diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,64468 +1,162884 @@
 {
-  "best_metric": 0.16217102110385895,
-  "best_model_checkpoint": "wyluilipe/deabuse/checkpoint-9000",
-  "epoch": 2.742032471437162,
-  "eval_steps": 120,
-  "global_step": 9120,
+  "best_metric": 0.07404118776321411,
+  "best_model_checkpoint": "wyluilipe/deabuse/checkpoint-22000",
+  "epoch": 2.9115552348382643,
+  "eval_steps": 400,
+  "global_step": 23200,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.0,
-      "grad_norm": 13.701467514038086,
+      "grad_norm": 728.5859375,
       "learning_rate": 2e-05,
-      "loss": 6.6206,
+      "loss": 15.4752,
       "step": 1
     },
     {
       "epoch": 0.0,
-      "grad_norm": 39.5434684753418,
-      "learning_rate": 1.9997995389395612e-05,
-      "loss": 4.445,
+      "grad_norm": 352.8587341308594,
+      "learning_rate": 1.9999163284943315e-05,
+      "loss": 15.063,
       "step": 2
     },
     {
       "epoch": 0.0,
-      "grad_norm": 83.92737579345703,
-      "learning_rate": 1.9995990778791222e-05,
-      "loss": 4.277,
+      "grad_norm": 518.958251953125,
+      "learning_rate": 1.9998326569886626e-05,
+      "loss": 11.8864,
       "step": 3
     },
     {
       "epoch": 0.0,
-      "grad_norm": 19.87366485595703,
-      "learning_rate": 1.9993986168186832e-05,
-      "loss": 4.9229,
+      "grad_norm": 1024.3834228515625,
+      "learning_rate": 1.999748985482994e-05,
+      "loss": 11.9749,
       "step": 4
     },
     {
       "epoch": 0.0,
-      "grad_norm": 22.43378448486328,
-      "learning_rate": 1.9991981557582442e-05,
-      "loss": 4.7554,
+      "grad_norm": 325.7301330566406,
+      "learning_rate": 1.9996653139773253e-05,
+      "loss": 13.9641,
       "step": 5
     },
     {
       "epoch": 0.0,
-      "grad_norm": 35.35092544555664,
-      "learning_rate": 1.9989976946978052e-05,
-      "loss": 4.1126,
+      "grad_norm": 313.9541931152344,
+      "learning_rate": 1.9995816424716566e-05,
+      "loss": 10.2194,
       "step": 6
     },
     {
       "epoch": 0.0,
-      "grad_norm": 133.4622344970703,
-      "learning_rate": 1.998797233637366e-05,
-      "loss": 4.5409,
+      "grad_norm": 358.2284240722656,
+      "learning_rate": 1.9994979709659877e-05,
+      "loss": 11.5145,
       "step": 7
     },
     {
       "epoch": 0.0,
-      "grad_norm": 20.33665657043457,
-      "learning_rate": 1.9985967725769272e-05,
-      "loss": 4.0595,
+      "grad_norm": 365.29931640625,
+      "learning_rate": 1.999414299460319e-05,
+      "loss": 10.7059,
       "step": 8
     },
     {
       "epoch": 0.0,
-      "grad_norm": 85.67984008789062,
-      "learning_rate": 1.9983963115164882e-05,
-      "loss": 4.9032,
+      "grad_norm": 490.46905517578125,
+      "learning_rate": 1.9993306279546504e-05,
+      "loss": 11.62,
       "step": 9
     },
     {
       "epoch": 0.0,
-      "grad_norm": 34.769309997558594,
-      "learning_rate": 1.998195850456049e-05,
-      "loss": 5.5253,
+      "grad_norm": 13291.681640625,
+      "learning_rate": 1.9992469564489814e-05,
+      "loss": 12.8367,
       "step": 10
     },
     {
       "epoch": 0.0,
-      "grad_norm": 40.53438949584961,
-      "learning_rate": 1.9979953893956102e-05,
-      "loss": 4.3671,
+      "grad_norm": 1022.1242065429688,
+      "learning_rate": 1.9991632849433128e-05,
+      "loss": 10.0496,
       "step": 11
     },
     {
       "epoch": 0.0,
-      "grad_norm": 21.633525848388672,
-      "learning_rate": 1.997794928335171e-05,
-      "loss": 3.9831,
+      "grad_norm": 192.495361328125,
+      "learning_rate": 1.999079613437644e-05,
+      "loss": 9.0021,
       "step": 12
     },
     {
       "epoch": 0.0,
-      "grad_norm": 20.7197322845459,
-      "learning_rate": 1.997594467274732e-05,
-      "loss": 4.4466,
+      "grad_norm": 211.52745056152344,
+      "learning_rate": 1.9989959419319752e-05,
+      "loss": 9.4134,
       "step": 13
     },
     {
       "epoch": 0.0,
-      "grad_norm": 110.01347351074219,
-      "learning_rate": 1.9973940062142933e-05,
-      "loss": 3.8518,
+      "grad_norm": 213.8238525390625,
+      "learning_rate": 1.9989122704263066e-05,
+      "loss": 10.3802,
       "step": 14
     },
     {
       "epoch": 0.0,
-      "grad_norm": 19.917098999023438,
-      "learning_rate": 1.997193545153854e-05,
-      "loss": 3.7986,
+      "grad_norm": 220.8242645263672,
+      "learning_rate": 1.9988285989206376e-05,
+      "loss": 8.2106,
       "step": 15
     },
     {
       "epoch": 0.0,
-      "grad_norm": 31.238054275512695,
-      "learning_rate": 1.996993084093415e-05,
-      "loss": 3.9456,
+      "grad_norm": 288.3450622558594,
+      "learning_rate": 1.998744927414969e-05,
+      "loss": 9.2149,
       "step": 16
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 20.226728439331055,
-      "learning_rate": 1.996792623032976e-05,
-      "loss": 4.1858,
+      "epoch": 0.0,
+      "grad_norm": 99.93714141845703,
+      "learning_rate": 1.9986612559093003e-05,
+      "loss": 7.4131,
       "step": 17
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 28.916229248046875,
-      "learning_rate": 1.996592161972537e-05,
-      "loss": 4.2707,
+      "epoch": 0.0,
+      "grad_norm": 1925.28076171875,
+      "learning_rate": 1.9985775844036313e-05,
+      "loss": 9.6749,
       "step": 18
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 20.103069305419922,
-      "learning_rate": 1.996391700912098e-05,
-      "loss": 5.2733,
+      "epoch": 0.0,
+      "grad_norm": 159.0907440185547,
+      "learning_rate": 1.9984939128979627e-05,
+      "loss": 7.9085,
       "step": 19
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 30.481670379638672,
-      "learning_rate": 1.996191239851659e-05,
-      "loss": 3.6844,
+      "epoch": 0.0,
+      "grad_norm": 196.09454345703125,
+      "learning_rate": 1.998410241392294e-05,
+      "loss": 8.3921,
       "step": 20
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 27.793643951416016,
-      "learning_rate": 1.99599077879122e-05,
-      "loss": 4.3403,
+      "epoch": 0.0,
+      "grad_norm": 104.76849365234375,
+      "learning_rate": 1.998326569886625e-05,
+      "loss": 7.0599,
       "step": 21
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 32.221065521240234,
-      "learning_rate": 1.995790317730781e-05,
-      "loss": 3.9117,
+      "epoch": 0.0,
+      "grad_norm": 121.42261505126953,
+      "learning_rate": 1.9982428983809565e-05,
+      "loss": 9.2175,
       "step": 22
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 23.525253295898438,
-      "learning_rate": 1.995589856670342e-05,
-      "loss": 3.804,
+      "epoch": 0.0,
+      "grad_norm": 118.76298522949219,
+      "learning_rate": 1.9981592268752878e-05,
+      "loss": 7.1647,
       "step": 23
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 16.033878326416016,
-      "learning_rate": 1.995389395609903e-05,
-      "loss": 5.166,
+      "epoch": 0.0,
+      "grad_norm": 176.77235412597656,
+      "learning_rate": 1.998075555369619e-05,
+      "loss": 6.6496,
       "step": 24
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 22.915647506713867,
-      "learning_rate": 1.995188934549464e-05,
-      "loss": 3.5634,
+      "epoch": 0.0,
+      "grad_norm": 58.71125030517578,
+      "learning_rate": 1.9979918838639502e-05,
+      "loss": 6.4695,
       "step": 25
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 18.16558837890625,
-      "learning_rate": 1.994988473489025e-05,
-      "loss": 4.9719,
+      "epoch": 0.0,
+      "grad_norm": 664.5093383789062,
+      "learning_rate": 1.9979082123582816e-05,
+      "loss": 10.4393,
       "step": 26
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 31.655603408813477,
-      "learning_rate": 1.994788012428586e-05,
-      "loss": 5.1661,
+      "epoch": 0.0,
+      "grad_norm": 114.95386505126953,
+      "learning_rate": 1.997824540852613e-05,
+      "loss": 7.0759,
       "step": 27
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 18.09580421447754,
-      "learning_rate": 1.994587551368147e-05,
-      "loss": 4.4041,
+      "epoch": 0.0,
+      "grad_norm": 45.126991271972656,
+      "learning_rate": 1.997740869346944e-05,
+      "loss": 6.3129,
       "step": 28
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 32.707950592041016,
-      "learning_rate": 1.9943870903077077e-05,
-      "loss": 4.5414,
+      "epoch": 0.0,
+      "grad_norm": 374.5563659667969,
+      "learning_rate": 1.9976571978412753e-05,
+      "loss": 5.6168,
       "step": 29
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 53.21809005737305,
-      "learning_rate": 1.994186629247269e-05,
-      "loss": 4.2779,
+      "epoch": 0.0,
+      "grad_norm": 1181.4608154296875,
+      "learning_rate": 1.9975735263356067e-05,
+      "loss": 9.343,
       "step": 30
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 36.61603927612305,
-      "learning_rate": 1.9939861681868297e-05,
-      "loss": 4.0579,
+      "epoch": 0.0,
+      "grad_norm": 283.9852294921875,
+      "learning_rate": 1.9974898548299377e-05,
+      "loss": 7.2791,
       "step": 31
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 21.24146842956543,
-      "learning_rate": 1.9937857071263907e-05,
-      "loss": 4.4486,
+      "epoch": 0.0,
+      "grad_norm": 164.09121704101562,
+      "learning_rate": 1.997406183324269e-05,
+      "loss": 7.4634,
       "step": 32
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 45.33989715576172,
-      "learning_rate": 1.993585246065952e-05,
-      "loss": 4.7946,
+      "epoch": 0.0,
+      "grad_norm": 409.6220703125,
+      "learning_rate": 1.9973225118186005e-05,
+      "loss": 6.7892,
       "step": 33
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 36.568145751953125,
-      "learning_rate": 1.9933847850055128e-05,
-      "loss": 4.6392,
+      "epoch": 0.0,
+      "grad_norm": 1974.9354248046875,
+      "learning_rate": 1.9972388403129318e-05,
+      "loss": 8.4132,
       "step": 34
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 19.171186447143555,
-      "learning_rate": 1.9931843239450738e-05,
-      "loss": 4.0399,
+      "epoch": 0.0,
+      "grad_norm": 105.84317016601562,
+      "learning_rate": 1.997155168807263e-05,
+      "loss": 6.9896,
       "step": 35
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 28.191804885864258,
-      "learning_rate": 1.9929838628846348e-05,
-      "loss": 4.9894,
+      "epoch": 0.0,
+      "grad_norm": 74.51609802246094,
+      "learning_rate": 1.9970714973015942e-05,
+      "loss": 7.7109,
       "step": 36
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 56.99246597290039,
-      "learning_rate": 1.9927834018241958e-05,
-      "loss": 3.7354,
+      "epoch": 0.0,
+      "grad_norm": 491.32171630859375,
+      "learning_rate": 1.9969878257959256e-05,
+      "loss": 7.9236,
       "step": 37
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 25.940950393676758,
-      "learning_rate": 1.9925829407637568e-05,
-      "loss": 3.4817,
+      "epoch": 0.0,
+      "grad_norm": 97.33172607421875,
+      "learning_rate": 1.9969041542902566e-05,
+      "loss": 5.2768,
       "step": 38
     },
     {
-      "epoch": 0.01,
-      "grad_norm": 38.012001037597656,
-      "learning_rate": 1.9923824797033178e-05,
-      "loss": 3.5887,
+      "epoch": 0.0,
+      "grad_norm": 187.27645874023438,
+      "learning_rate": 1.996820482784588e-05,
+      "loss": 5.6723,
       "step": 39
     },
     {
       "epoch": 0.01,
-      "grad_norm": 23.125755310058594,
-      "learning_rate": 1.9921820186428788e-05,
-      "loss": 4.8167,
+      "grad_norm": 143.09710693359375,
+      "learning_rate": 1.9967368112789193e-05,
+      "loss": 7.2103,
       "step": 40
     },
     {
       "epoch": 0.01,
-      "grad_norm": 25.253414154052734,
-      "learning_rate": 1.9919815575824398e-05,
-      "loss": 3.8167,
+      "grad_norm": 127.5011978149414,
+      "learning_rate": 1.9966531397732504e-05,
+      "loss": 6.2016,
       "step": 41
     },
     {
       "epoch": 0.01,
-      "grad_norm": 18.642539978027344,
-      "learning_rate": 1.991781096522001e-05,
-      "loss": 3.5483,
+      "grad_norm": 258.0473937988281,
+      "learning_rate": 1.9965694682675817e-05,
+      "loss": 6.897,
       "step": 42
     },
     {
       "epoch": 0.01,
-      "grad_norm": 102.15142059326172,
-      "learning_rate": 1.9915806354615615e-05,
-      "loss": 5.1262,
+      "grad_norm": 223.92303466796875,
+      "learning_rate": 1.9964857967619127e-05,
+      "loss": 10.1903,
       "step": 43
     },
     {
       "epoch": 0.01,
-      "grad_norm": 24.280908584594727,
-      "learning_rate": 1.991380174401123e-05,
-      "loss": 4.2309,
+      "grad_norm": 69.66583251953125,
+      "learning_rate": 1.996402125256244e-05,
+      "loss": 6.5244,
       "step": 44
     },
     {
       "epoch": 0.01,
-      "grad_norm": 18.596893310546875,
-      "learning_rate": 1.991179713340684e-05,
-      "loss": 3.3356,
+      "grad_norm": 292.11639404296875,
+      "learning_rate": 1.9963184537505755e-05,
+      "loss": 7.3886,
       "step": 45
     },
     {
       "epoch": 0.01,
-      "grad_norm": 21.592117309570312,
-      "learning_rate": 1.9909792522802445e-05,
-      "loss": 4.8101,
+      "grad_norm": 113.60519409179688,
+      "learning_rate": 1.9962347822449065e-05,
+      "loss": 6.6999,
       "step": 46
     },
     {
       "epoch": 0.01,
-      "grad_norm": 19.77571678161621,
-      "learning_rate": 1.990778791219806e-05,
-      "loss": 3.4436,
+      "grad_norm": 213.7831268310547,
+      "learning_rate": 1.996151110739238e-05,
+      "loss": 8.2679,
       "step": 47
     },
     {
       "epoch": 0.01,
-      "grad_norm": 47.20704650878906,
-      "learning_rate": 1.9905783301593665e-05,
-      "loss": 4.077,
+      "grad_norm": 60.55680847167969,
+      "learning_rate": 1.9960674392335692e-05,
+      "loss": 7.006,
       "step": 48
     },
     {
       "epoch": 0.01,
-      "grad_norm": 21.739601135253906,
-      "learning_rate": 1.9903778690989276e-05,
-      "loss": 4.5314,
+      "grad_norm": 70.07861328125,
+      "learning_rate": 1.9959837677279003e-05,
+      "loss": 7.2405,
       "step": 49
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 22.48163604736328,
-      "learning_rate": 1.9901774080384886e-05,
-      "loss": 4.1253,
+      "epoch": 0.01,
+      "grad_norm": 77.42964172363281,
+      "learning_rate": 1.9959000962222316e-05,
+      "loss": 5.8274,
       "step": 50
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 26.019960403442383,
-      "learning_rate": 1.9899769469780496e-05,
-      "loss": 4.4319,
+      "epoch": 0.01,
+      "grad_norm": 97.56826782226562,
+      "learning_rate": 1.995816424716563e-05,
+      "loss": 6.531,
       "step": 51
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 21.49262046813965,
-      "learning_rate": 1.9897764859176106e-05,
-      "loss": 4.7591,
+      "epoch": 0.01,
+      "grad_norm": 378.0497131347656,
+      "learning_rate": 1.995732753210894e-05,
+      "loss": 6.7833,
       "step": 52
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 21.852209091186523,
-      "learning_rate": 1.9895760248571716e-05,
-      "loss": 4.5679,
+      "epoch": 0.01,
+      "grad_norm": 121.92767333984375,
+      "learning_rate": 1.9956490817052254e-05,
+      "loss": 6.7799,
       "step": 53
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 17.225950241088867,
-      "learning_rate": 1.9893755637967326e-05,
-      "loss": 3.5482,
+      "epoch": 0.01,
+      "grad_norm": 85.76437377929688,
+      "learning_rate": 1.9955654101995567e-05,
+      "loss": 5.3537,
       "step": 54
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 17.815263748168945,
-      "learning_rate": 1.9891751027362936e-05,
-      "loss": 3.6293,
+      "epoch": 0.01,
+      "grad_norm": 40.82673645019531,
+      "learning_rate": 1.995481738693888e-05,
+      "loss": 6.3451,
       "step": 55
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 19.188623428344727,
-      "learning_rate": 1.9889746416758546e-05,
-      "loss": 4.1118,
+      "epoch": 0.01,
+      "grad_norm": 82.32181549072266,
+      "learning_rate": 1.995398067188219e-05,
+      "loss": 6.1245,
       "step": 56
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 21.920475006103516,
-      "learning_rate": 1.9887741806154156e-05,
-      "loss": 3.8108,
+      "epoch": 0.01,
+      "grad_norm": 84.50050354003906,
+      "learning_rate": 1.9953143956825505e-05,
+      "loss": 6.2817,
       "step": 57
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 24.557518005371094,
-      "learning_rate": 1.9885737195549766e-05,
-      "loss": 3.8317,
+      "epoch": 0.01,
+      "grad_norm": 89.99356842041016,
+      "learning_rate": 1.995230724176882e-05,
+      "loss": 7.0404,
       "step": 58
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 90.77320098876953,
-      "learning_rate": 1.9883732584945376e-05,
-      "loss": 5.1687,
+      "epoch": 0.01,
+      "grad_norm": 40.455955505371094,
+      "learning_rate": 1.995147052671213e-05,
+      "loss": 5.1922,
       "step": 59
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 25.83659553527832,
-      "learning_rate": 1.9881727974340986e-05,
-      "loss": 5.0576,
+      "epoch": 0.01,
+      "grad_norm": 99.8443603515625,
+      "learning_rate": 1.9950633811655443e-05,
+      "loss": 7.4731,
       "step": 60
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 19.337127685546875,
-      "learning_rate": 1.9879723363736597e-05,
-      "loss": 3.364,
+      "epoch": 0.01,
+      "grad_norm": 67.06603240966797,
+      "learning_rate": 1.9949797096598756e-05,
+      "loss": 7.7414,
       "step": 61
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 19.83724594116211,
-      "learning_rate": 1.9877718753132203e-05,
-      "loss": 3.5426,
+      "epoch": 0.01,
+      "grad_norm": 35.99038314819336,
+      "learning_rate": 1.994896038154207e-05,
+      "loss": 5.406,
       "step": 62
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 14.76127815246582,
-      "learning_rate": 1.9875714142527817e-05,
-      "loss": 3.8213,
+      "epoch": 0.01,
+      "grad_norm": 74.65876770019531,
+      "learning_rate": 1.994812366648538e-05,
+      "loss": 6.6748,
       "step": 63
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 21.111013412475586,
-      "learning_rate": 1.9873709531923427e-05,
-      "loss": 3.4596,
+      "epoch": 0.01,
+      "grad_norm": 49.295806884765625,
+      "learning_rate": 1.9947286951428694e-05,
+      "loss": 5.6246,
       "step": 64
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 63.46014404296875,
-      "learning_rate": 1.9871704921319033e-05,
-      "loss": 4.4058,
+      "epoch": 0.01,
+      "grad_norm": 1596.51708984375,
+      "learning_rate": 1.9946450236372007e-05,
+      "loss": 6.601,
       "step": 65
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 19.21253776550293,
-      "learning_rate": 1.9869700310714647e-05,
-      "loss": 3.8774,
+      "epoch": 0.01,
+      "grad_norm": 43.59678649902344,
+      "learning_rate": 1.9945613521315318e-05,
+      "loss": 5.112,
       "step": 66
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 23.148107528686523,
-      "learning_rate": 1.9867695700110254e-05,
-      "loss": 3.9923,
+      "epoch": 0.01,
+      "grad_norm": 112.09934997558594,
+      "learning_rate": 1.994477680625863e-05,
+      "loss": 5.9346,
       "step": 67
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 25.991294860839844,
-      "learning_rate": 1.9865691089505864e-05,
-      "loss": 4.5205,
+      "epoch": 0.01,
+      "grad_norm": 44.990169525146484,
+      "learning_rate": 1.994394009120194e-05,
+      "loss": 3.9908,
       "step": 68
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 18.949478149414062,
-      "learning_rate": 1.9863686478901477e-05,
-      "loss": 4.678,
+      "epoch": 0.01,
+      "grad_norm": 43.052791595458984,
+      "learning_rate": 1.9943103376145255e-05,
+      "loss": 4.6972,
       "step": 69
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 17.614856719970703,
-      "learning_rate": 1.9861681868297084e-05,
-      "loss": 3.758,
+      "epoch": 0.01,
+      "grad_norm": 66.66222381591797,
+      "learning_rate": 1.994226666108857e-05,
+      "loss": 4.5402,
       "step": 70
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 20.783035278320312,
-      "learning_rate": 1.9859677257692694e-05,
-      "loss": 4.5238,
+      "epoch": 0.01,
+      "grad_norm": 433.1274108886719,
+      "learning_rate": 1.994142994603188e-05,
+      "loss": 6.3487,
       "step": 71
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 27.744810104370117,
-      "learning_rate": 1.9857672647088304e-05,
-      "loss": 4.7011,
+      "epoch": 0.01,
+      "grad_norm": 35.271053314208984,
+      "learning_rate": 1.9940593230975193e-05,
+      "loss": 3.7392,
       "step": 72
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 22.72287368774414,
-      "learning_rate": 1.9855668036483914e-05,
-      "loss": 3.6715,
+      "epoch": 0.01,
+      "grad_norm": 55.31417465209961,
+      "learning_rate": 1.9939756515918503e-05,
+      "loss": 5.8701,
       "step": 73
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 23.866741180419922,
-      "learning_rate": 1.9853663425879524e-05,
-      "loss": 5.4669,
+      "epoch": 0.01,
+      "grad_norm": 434.6205139160156,
+      "learning_rate": 1.9938919800861817e-05,
+      "loss": 6.7362,
       "step": 74
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 16.936256408691406,
-      "learning_rate": 1.9851658815275134e-05,
-      "loss": 3.3014,
+      "epoch": 0.01,
+      "grad_norm": 47.17160415649414,
+      "learning_rate": 1.993808308580513e-05,
+      "loss": 4.5174,
       "step": 75
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 22.439456939697266,
-      "learning_rate": 1.9849654204670744e-05,
-      "loss": 3.202,
+      "epoch": 0.01,
+      "grad_norm": 320.0890197753906,
+      "learning_rate": 1.9937246370748444e-05,
+      "loss": 6.1796,
       "step": 76
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 16.06572723388672,
-      "learning_rate": 1.9847649594066354e-05,
-      "loss": 3.3253,
+      "epoch": 0.01,
+      "grad_norm": 50.07407760620117,
+      "learning_rate": 1.9936409655691754e-05,
+      "loss": 3.736,
       "step": 77
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 29.42819595336914,
-      "learning_rate": 1.9845644983461965e-05,
-      "loss": 4.1956,
+      "epoch": 0.01,
+      "grad_norm": 53.458953857421875,
+      "learning_rate": 1.9935572940635068e-05,
+      "loss": 5.0562,
       "step": 78
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 20.668291091918945,
-      "learning_rate": 1.9843640372857575e-05,
-      "loss": 4.4557,
+      "epoch": 0.01,
+      "grad_norm": 67.62255096435547,
+      "learning_rate": 1.993473622557838e-05,
+      "loss": 6.4006,
       "step": 79
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 21.241540908813477,
-      "learning_rate": 1.9841635762253185e-05,
-      "loss": 2.7819,
+      "epoch": 0.01,
+      "grad_norm": 144.18785095214844,
+      "learning_rate": 1.9933899510521692e-05,
+      "loss": 4.4056,
       "step": 80
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 39.1732292175293,
-      "learning_rate": 1.9839631151648795e-05,
-      "loss": 4.044,
+      "epoch": 0.01,
+      "grad_norm": 207.25999450683594,
+      "learning_rate": 1.9933062795465005e-05,
+      "loss": 7.6115,
       "step": 81
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 17.961164474487305,
-      "learning_rate": 1.9837626541044405e-05,
-      "loss": 4.0641,
+      "epoch": 0.01,
+      "grad_norm": 152.46791076660156,
+      "learning_rate": 1.993222608040832e-05,
+      "loss": 6.7522,
       "step": 82
     },
     {
-      "epoch": 0.02,
-      "grad_norm": 23.27247428894043,
-      "learning_rate": 1.9835621930440015e-05,
-      "loss": 4.1457,
+      "epoch": 0.01,
+      "grad_norm": 50.106224060058594,
+      "learning_rate": 1.9931389365351633e-05,
+      "loss": 5.4387,
       "step": 83
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 33.87324523925781,
-      "learning_rate": 1.983361731983562e-05,
-      "loss": 4.4233,
+      "epoch": 0.01,
+      "grad_norm": 46.966224670410156,
+      "learning_rate": 1.9930552650294943e-05,
+      "loss": 4.5016,
       "step": 84
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 27.01508903503418,
-      "learning_rate": 1.9831612709231235e-05,
-      "loss": 3.5391,
+      "epoch": 0.01,
+      "grad_norm": 46.80181121826172,
+      "learning_rate": 1.9929715935238257e-05,
+      "loss": 6.9584,
       "step": 85
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 18.453147888183594,
-      "learning_rate": 1.9829608098626842e-05,
-      "loss": 3.8098,
+      "epoch": 0.01,
+      "grad_norm": 167.78543090820312,
+      "learning_rate": 1.992887922018157e-05,
+      "loss": 5.6751,
       "step": 86
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 551.41259765625,
-      "learning_rate": 1.9827603488022452e-05,
-      "loss": 3.3489,
+      "epoch": 0.01,
+      "grad_norm": 222.9422149658203,
+      "learning_rate": 1.992804250512488e-05,
+      "loss": 5.5146,
       "step": 87
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 33.0008544921875,
-      "learning_rate": 1.9825598877418065e-05,
-      "loss": 3.8738,
+      "epoch": 0.01,
+      "grad_norm": 46.92573165893555,
+      "learning_rate": 1.9927205790068194e-05,
+      "loss": 4.6874,
       "step": 88
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 80.79717254638672,
-      "learning_rate": 1.9823594266813672e-05,
-      "loss": 3.9937,
+      "epoch": 0.01,
+      "grad_norm": 246.22035217285156,
+      "learning_rate": 1.9926369075011508e-05,
+      "loss": 5.9676,
       "step": 89
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 20.0161075592041,
-      "learning_rate": 1.9821589656209282e-05,
-      "loss": 3.2797,
+      "epoch": 0.01,
+      "grad_norm": 57.03669738769531,
+      "learning_rate": 1.992553235995482e-05,
+      "loss": 5.1982,
       "step": 90
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 25.17633056640625,
-      "learning_rate": 1.9819585045604892e-05,
-      "loss": 3.3241,
+      "epoch": 0.01,
+      "grad_norm": 51.081966400146484,
+      "learning_rate": 1.9924695644898132e-05,
+      "loss": 3.6639,
       "step": 91
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 50.05910873413086,
-      "learning_rate": 1.9817580435000502e-05,
-      "loss": 5.4063,
+      "epoch": 0.01,
+      "grad_norm": 31.671546936035156,
+      "learning_rate": 1.9923858929841445e-05,
+      "loss": 3.5103,
       "step": 92
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 19.56489372253418,
-      "learning_rate": 1.9815575824396112e-05,
-      "loss": 3.3133,
+      "epoch": 0.01,
+      "grad_norm": 47.267677307128906,
+      "learning_rate": 1.992302221478476e-05,
+      "loss": 3.8276,
       "step": 93
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 66.2906265258789,
-      "learning_rate": 1.9813571213791723e-05,
-      "loss": 4.3678,
+      "epoch": 0.01,
+      "grad_norm": 74.63932800292969,
+      "learning_rate": 1.992218549972807e-05,
+      "loss": 6.8494,
       "step": 94
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 17.404077529907227,
-      "learning_rate": 1.9811566603187333e-05,
-      "loss": 3.7368,
+      "epoch": 0.01,
+      "grad_norm": 68.33775329589844,
+      "learning_rate": 1.9921348784671383e-05,
+      "loss": 5.1167,
       "step": 95
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 17.154563903808594,
-      "learning_rate": 1.9809561992582943e-05,
-      "loss": 3.6814,
+      "epoch": 0.01,
+      "grad_norm": 40.34266662597656,
+      "learning_rate": 1.9920512069614693e-05,
+      "loss": 4.5064,
       "step": 96
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 21.113162994384766,
-      "learning_rate": 1.9807557381978553e-05,
-      "loss": 4.3332,
+      "epoch": 0.01,
+      "grad_norm": 219.28070068359375,
+      "learning_rate": 1.9919675354558007e-05,
+      "loss": 3.9197,
       "step": 97
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 35.284725189208984,
-      "learning_rate": 1.9805552771374163e-05,
-      "loss": 3.2572,
+      "epoch": 0.01,
+      "grad_norm": 162.0477752685547,
+      "learning_rate": 1.991883863950132e-05,
+      "loss": 5.7777,
       "step": 98
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 19.46085548400879,
-      "learning_rate": 1.9803548160769773e-05,
-      "loss": 3.5422,
+      "epoch": 0.01,
+      "grad_norm": 35.64801788330078,
+      "learning_rate": 1.991800192444463e-05,
+      "loss": 5.6309,
       "step": 99
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 24.549236297607422,
-      "learning_rate": 1.9801543550165383e-05,
-      "loss": 3.071,
+      "epoch": 0.01,
+      "grad_norm": 164.5700225830078,
+      "learning_rate": 1.9917165209387944e-05,
+      "loss": 5.6857,
       "step": 100
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 24.21219253540039,
-      "learning_rate": 1.9799538939560993e-05,
-      "loss": 3.7529,
+      "epoch": 0.01,
+      "grad_norm": 42.44116973876953,
+      "learning_rate": 1.9916328494331255e-05,
+      "loss": 5.1935,
       "step": 101
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 19.04281234741211,
-      "learning_rate": 1.9797534328956603e-05,
-      "loss": 3.5407,
+      "epoch": 0.01,
+      "grad_norm": 30.379470825195312,
+      "learning_rate": 1.991549177927457e-05,
+      "loss": 4.7206,
       "step": 102
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 46.35712814331055,
-      "learning_rate": 1.979552971835221e-05,
-      "loss": 3.0938,
+      "epoch": 0.01,
+      "grad_norm": 56.13416290283203,
+      "learning_rate": 1.9914655064217882e-05,
+      "loss": 5.3201,
       "step": 103
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 32.1992073059082,
-      "learning_rate": 1.9793525107747823e-05,
-      "loss": 4.1555,
+      "epoch": 0.01,
+      "grad_norm": 57.6418571472168,
+      "learning_rate": 1.9913818349161196e-05,
+      "loss": 3.8999,
       "step": 104
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 18.388561248779297,
-      "learning_rate": 1.979152049714343e-05,
-      "loss": 3.25,
+      "epoch": 0.01,
+      "grad_norm": 70.00899505615234,
+      "learning_rate": 1.9912981634104506e-05,
+      "loss": 4.046,
       "step": 105
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 25.422645568847656,
-      "learning_rate": 1.978951588653904e-05,
-      "loss": 3.7947,
+      "epoch": 0.01,
+      "grad_norm": 41.35757064819336,
+      "learning_rate": 1.991214491904782e-05,
+      "loss": 6.1811,
       "step": 106
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 20.530521392822266,
-      "learning_rate": 1.9787511275934654e-05,
-      "loss": 3.3529,
+      "epoch": 0.01,
+      "grad_norm": 92.74214172363281,
+      "learning_rate": 1.9911308203991133e-05,
+      "loss": 4.667,
       "step": 107
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 40.206138610839844,
-      "learning_rate": 1.978550666533026e-05,
-      "loss": 4.6079,
+      "epoch": 0.01,
+      "grad_norm": 64.18307495117188,
+      "learning_rate": 1.9910471488934443e-05,
+      "loss": 6.3902,
       "step": 108
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 43.84805679321289,
-      "learning_rate": 1.978350205472587e-05,
-      "loss": 3.6345,
+      "epoch": 0.01,
+      "grad_norm": 38.1926383972168,
+      "learning_rate": 1.9909634773877757e-05,
+      "loss": 6.1301,
       "step": 109
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 26.29327392578125,
-      "learning_rate": 1.978149744412148e-05,
-      "loss": 4.9385,
+      "epoch": 0.01,
+      "grad_norm": 89.6794662475586,
+      "learning_rate": 1.990879805882107e-05,
+      "loss": 4.9284,
       "step": 110
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 24.576719284057617,
-      "learning_rate": 1.977949283351709e-05,
-      "loss": 3.7024,
+      "epoch": 0.01,
+      "grad_norm": 52.6729621887207,
+      "learning_rate": 1.9907961343764384e-05,
+      "loss": 5.0203,
       "step": 111
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 21.797935485839844,
-      "learning_rate": 1.97774882229127e-05,
-      "loss": 4.0716,
+      "epoch": 0.01,
+      "grad_norm": 96.23528289794922,
+      "learning_rate": 1.9907124628707695e-05,
+      "loss": 3.5231,
       "step": 112
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 51.30803298950195,
-      "learning_rate": 1.977548361230831e-05,
-      "loss": 4.4803,
+      "epoch": 0.01,
+      "grad_norm": 51.27971649169922,
+      "learning_rate": 1.990628791365101e-05,
+      "loss": 3.9009,
       "step": 113
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 17.133638381958008,
-      "learning_rate": 1.977347900170392e-05,
-      "loss": 2.996,
+      "epoch": 0.01,
+      "grad_norm": 55.0560302734375,
+      "learning_rate": 1.9905451198594322e-05,
+      "loss": 5.9776,
       "step": 114
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 23.30716323852539,
-      "learning_rate": 1.977147439109953e-05,
-      "loss": 3.6316,
+      "epoch": 0.01,
+      "grad_norm": 28.32353401184082,
+      "learning_rate": 1.9904614483537632e-05,
+      "loss": 5.6077,
       "step": 115
     },
     {
-      "epoch": 0.03,
-      "grad_norm": 21.08536720275879,
-      "learning_rate": 1.976946978049514e-05,
-      "loss": 3.2101,
+      "epoch": 0.01,
+      "grad_norm": 31.120115280151367,
+      "learning_rate": 1.9903777768480946e-05,
+      "loss": 4.6517,
       "step": 116
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 40.39435577392578,
-      "learning_rate": 1.9767465169890748e-05,
-      "loss": 3.9427,
+      "epoch": 0.01,
+      "grad_norm": 41.79515838623047,
+      "learning_rate": 1.990294105342426e-05,
+      "loss": 3.8836,
       "step": 117
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 20.827363967895508,
-      "learning_rate": 1.976546055928636e-05,
-      "loss": 3.7911,
+      "epoch": 0.01,
+      "grad_norm": 47.09407043457031,
+      "learning_rate": 1.9902104338367573e-05,
+      "loss": 5.2293,
       "step": 118
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 31.05581283569336,
-      "learning_rate": 1.976345594868197e-05,
-      "loss": 4.0686,
+      "epoch": 0.01,
+      "grad_norm": 56.67197799682617,
+      "learning_rate": 1.9901267623310883e-05,
+      "loss": 5.4069,
       "step": 119
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 75.66057586669922,
-      "learning_rate": 1.976145133807758e-05,
-      "loss": 3.659,
-      "step": 120
-    },
-    {
-      "epoch": 0.04,
-      "eval_loss": 1.6635291576385498,
-      "eval_runtime": 43.6414,
-      "eval_samples_per_second": 33.89,
-      "eval_steps_per_second": 33.89,
+      "epoch": 0.02,
+      "grad_norm": 58.246856689453125,
+      "learning_rate": 1.9900430908254197e-05,
+      "loss": 5.2024,
       "step": 120
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 14.022979736328125,
-      "learning_rate": 1.975944672747319e-05,
-      "loss": 3.0417,
+      "epoch": 0.02,
+      "grad_norm": 54.86061477661133,
+      "learning_rate": 1.9899594193197507e-05,
+      "loss": 3.8366,
       "step": 121
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 38.61516189575195,
-      "learning_rate": 1.9757442116868798e-05,
-      "loss": 4.3155,
+      "epoch": 0.02,
+      "grad_norm": 70.79507446289062,
+      "learning_rate": 1.989875747814082e-05,
+      "loss": 4.7176,
       "step": 122
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 19.140295028686523,
-      "learning_rate": 1.975543750626441e-05,
-      "loss": 4.6368,
+      "epoch": 0.02,
+      "grad_norm": 155.59671020507812,
+      "learning_rate": 1.9897920763084135e-05,
+      "loss": 4.8716,
       "step": 123
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 24.628950119018555,
-      "learning_rate": 1.975343289566002e-05,
-      "loss": 4.131,
+      "epoch": 0.02,
+      "grad_norm": 60.553348541259766,
+      "learning_rate": 1.9897084048027445e-05,
+      "loss": 6.3457,
       "step": 124
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 32.278045654296875,
-      "learning_rate": 1.975142828505563e-05,
-      "loss": 4.4181,
+      "epoch": 0.02,
+      "grad_norm": 29.441686630249023,
+      "learning_rate": 1.989624733297076e-05,
+      "loss": 7.1413,
       "step": 125
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 25.219432830810547,
-      "learning_rate": 1.9749423674451242e-05,
-      "loss": 3.8382,
+      "epoch": 0.02,
+      "grad_norm": 77.06675720214844,
+      "learning_rate": 1.989541061791407e-05,
+      "loss": 5.7867,
       "step": 126
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 20.332189559936523,
-      "learning_rate": 1.974741906384685e-05,
-      "loss": 2.7729,
+      "epoch": 0.02,
+      "grad_norm": 84.96646881103516,
+      "learning_rate": 1.9894573902857382e-05,
+      "loss": 5.201,
       "step": 127
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 24.03636932373047,
-      "learning_rate": 1.974541445324246e-05,
-      "loss": 3.4281,
+      "epoch": 0.02,
+      "grad_norm": 27.29684066772461,
+      "learning_rate": 1.9893737187800696e-05,
+      "loss": 4.5,
       "step": 128
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 48.14682388305664,
-      "learning_rate": 1.974340984263807e-05,
-      "loss": 3.3327,
+      "epoch": 0.02,
+      "grad_norm": 69.64246368408203,
+      "learning_rate": 1.9892900472744006e-05,
+      "loss": 5.7104,
       "step": 129
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 22.612560272216797,
-      "learning_rate": 1.974140523203368e-05,
-      "loss": 2.8426,
+      "epoch": 0.02,
+      "grad_norm": 26.5671443939209,
+      "learning_rate": 1.989206375768732e-05,
+      "loss": 7.3485,
       "step": 130
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 18.185649871826172,
-      "learning_rate": 1.973940062142929e-05,
-      "loss": 3.8089,
+      "epoch": 0.02,
+      "grad_norm": 28.373462677001953,
+      "learning_rate": 1.9891227042630634e-05,
+      "loss": 4.5569,
       "step": 131
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 39.714744567871094,
-      "learning_rate": 1.97373960108249e-05,
-      "loss": 3.5672,
+      "epoch": 0.02,
+      "grad_norm": 32.1534309387207,
+      "learning_rate": 1.9890390327573947e-05,
+      "loss": 4.7846,
       "step": 132
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 63.07310485839844,
-      "learning_rate": 1.973539140022051e-05,
-      "loss": 3.1297,
+      "epoch": 0.02,
+      "grad_norm": 51.85831832885742,
+      "learning_rate": 1.9889553612517258e-05,
+      "loss": 4.2173,
       "step": 133
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 22.210826873779297,
-      "learning_rate": 1.973338678961612e-05,
-      "loss": 5.2622,
+      "epoch": 0.02,
+      "grad_norm": 25.660507202148438,
+      "learning_rate": 1.988871689746057e-05,
+      "loss": 3.8976,
       "step": 134
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 15.51650619506836,
-      "learning_rate": 1.973138217901173e-05,
-      "loss": 3.0659,
+      "epoch": 0.02,
+      "grad_norm": 33.85279083251953,
+      "learning_rate": 1.9887880182403885e-05,
+      "loss": 4.4822,
       "step": 135
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 20.847930908203125,
-      "learning_rate": 1.9729377568407336e-05,
-      "loss": 3.1732,
+      "epoch": 0.02,
+      "grad_norm": 43.09807205200195,
+      "learning_rate": 1.9887043467347195e-05,
+      "loss": 3.6526,
       "step": 136
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 18.530948638916016,
-      "learning_rate": 1.972737295780295e-05,
-      "loss": 2.7667,
+      "epoch": 0.02,
+      "grad_norm": 36.18055725097656,
+      "learning_rate": 1.988620675229051e-05,
+      "loss": 4.2724,
       "step": 137
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 47.08240509033203,
-      "learning_rate": 1.972536834719856e-05,
-      "loss": 2.5596,
+      "epoch": 0.02,
+      "grad_norm": 82.61565399169922,
+      "learning_rate": 1.9885370037233822e-05,
+      "loss": 4.4369,
       "step": 138
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 26.97058868408203,
-      "learning_rate": 1.9723363736594166e-05,
-      "loss": 3.129,
+      "epoch": 0.02,
+      "grad_norm": 37.68891906738281,
+      "learning_rate": 1.9884533322177136e-05,
+      "loss": 5.8417,
       "step": 139
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 20.2777099609375,
-      "learning_rate": 1.972135912598978e-05,
-      "loss": 3.4919,
+      "epoch": 0.02,
+      "grad_norm": 28.7679443359375,
+      "learning_rate": 1.9883696607120446e-05,
+      "loss": 6.1488,
       "step": 140
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 21.72832679748535,
-      "learning_rate": 1.9719354515385386e-05,
-      "loss": 4.1341,
+      "epoch": 0.02,
+      "grad_norm": 55.896026611328125,
+      "learning_rate": 1.988285989206376e-05,
+      "loss": 5.2888,
       "step": 141
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 17.30972671508789,
-      "learning_rate": 1.9717349904780996e-05,
-      "loss": 3.2444,
+      "epoch": 0.02,
+      "grad_norm": 28.051551818847656,
+      "learning_rate": 1.9882023177007074e-05,
+      "loss": 3.9439,
       "step": 142
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 49.03208923339844,
-      "learning_rate": 1.971534529417661e-05,
-      "loss": 4.314,
+      "epoch": 0.02,
+      "grad_norm": 40.94978713989258,
+      "learning_rate": 1.9881186461950384e-05,
+      "loss": 3.9018,
       "step": 143
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 26.420650482177734,
-      "learning_rate": 1.9713340683572217e-05,
-      "loss": 3.5411,
+      "epoch": 0.02,
+      "grad_norm": 37.6737060546875,
+      "learning_rate": 1.9880349746893698e-05,
+      "loss": 4.2534,
       "step": 144
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 25.890357971191406,
-      "learning_rate": 1.9711336072967827e-05,
-      "loss": 3.741,
+      "epoch": 0.02,
+      "grad_norm": 31.931015014648438,
+      "learning_rate": 1.987951303183701e-05,
+      "loss": 5.6937,
       "step": 145
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 22.66683578491211,
-      "learning_rate": 1.9709331462363437e-05,
-      "loss": 3.116,
+      "epoch": 0.02,
+      "grad_norm": 30.532413482666016,
+      "learning_rate": 1.9878676316780325e-05,
+      "loss": 3.8673,
       "step": 146
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 31.934288024902344,
-      "learning_rate": 1.9707326851759047e-05,
-      "loss": 3.8739,
+      "epoch": 0.02,
+      "grad_norm": 29.377347946166992,
+      "learning_rate": 1.9877839601723635e-05,
+      "loss": 5.2521,
       "step": 147
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 29.55501937866211,
-      "learning_rate": 1.9705322241154657e-05,
-      "loss": 3.7167,
+      "epoch": 0.02,
+      "grad_norm": 38.87949752807617,
+      "learning_rate": 1.987700288666695e-05,
+      "loss": 3.4218,
       "step": 148
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 21.13062286376953,
-      "learning_rate": 1.9703317630550267e-05,
-      "loss": 3.0495,
+      "epoch": 0.02,
+      "grad_norm": 21.628009796142578,
+      "learning_rate": 1.987616617161026e-05,
+      "loss": 2.7079,
       "step": 149
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 27.444517135620117,
-      "learning_rate": 1.9701313019945877e-05,
-      "loss": 3.2852,
+      "epoch": 0.02,
+      "grad_norm": 31.379051208496094,
+      "learning_rate": 1.9875329456553573e-05,
+      "loss": 5.888,
       "step": 150
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 19.37506675720215,
-      "learning_rate": 1.9699308409341487e-05,
-      "loss": 5.2696,
+      "epoch": 0.02,
+      "grad_norm": 28.817869186401367,
+      "learning_rate": 1.9874492741496886e-05,
+      "loss": 3.8517,
       "step": 151
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 49.605247497558594,
-      "learning_rate": 1.9697303798737097e-05,
-      "loss": 3.7375,
+      "epoch": 0.02,
+      "grad_norm": 24.719228744506836,
+      "learning_rate": 1.9873656026440197e-05,
+      "loss": 4.9138,
       "step": 152
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 19.60191535949707,
-      "learning_rate": 1.9695299188132707e-05,
-      "loss": 3.6209,
+      "epoch": 0.02,
+      "grad_norm": 42.05404281616211,
+      "learning_rate": 1.987281931138351e-05,
+      "loss": 3.521,
       "step": 153
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 18.86824607849121,
-      "learning_rate": 1.9693294577528317e-05,
-      "loss": 3.1842,
+      "epoch": 0.02,
+      "grad_norm": 28.471454620361328,
+      "learning_rate": 1.987198259632682e-05,
+      "loss": 4.5766,
       "step": 154
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 30.93740463256836,
-      "learning_rate": 1.9691289966923928e-05,
-      "loss": 3.555,
+      "epoch": 0.02,
+      "grad_norm": 92.56484985351562,
+      "learning_rate": 1.9871145881270134e-05,
+      "loss": 6.5103,
       "step": 155
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 24.22440528869629,
-      "learning_rate": 1.9689285356319538e-05,
-      "loss": 3.5906,
+      "epoch": 0.02,
+      "grad_norm": 28.869182586669922,
+      "learning_rate": 1.9870309166213448e-05,
+      "loss": 4.8127,
       "step": 156
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 43.011539459228516,
-      "learning_rate": 1.9687280745715148e-05,
-      "loss": 4.5245,
+      "epoch": 0.02,
+      "grad_norm": 96.62332153320312,
+      "learning_rate": 1.9869472451156758e-05,
+      "loss": 5.0915,
       "step": 157
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 24.332199096679688,
-      "learning_rate": 1.9685276135110754e-05,
-      "loss": 3.9531,
+      "epoch": 0.02,
+      "grad_norm": 83.90554809570312,
+      "learning_rate": 1.9868635736100072e-05,
+      "loss": 4.9871,
       "step": 158
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 37.94855880737305,
-      "learning_rate": 1.9683271524506368e-05,
-      "loss": 4.7006,
+      "epoch": 0.02,
+      "grad_norm": 27.655275344848633,
+      "learning_rate": 1.9867799021043385e-05,
+      "loss": 1.8241,
       "step": 159
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 20.99969482421875,
-      "learning_rate": 1.9681266913901975e-05,
-      "loss": 4.7367,
+      "epoch": 0.02,
+      "grad_norm": 24.46050453186035,
+      "learning_rate": 1.98669623059867e-05,
+      "loss": 4.7742,
       "step": 160
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 34.952232360839844,
-      "learning_rate": 1.9679262303297585e-05,
-      "loss": 3.2787,
+      "epoch": 0.02,
+      "grad_norm": 27.145444869995117,
+      "learning_rate": 1.986612559093001e-05,
+      "loss": 5.0453,
       "step": 161
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 25.392833709716797,
-      "learning_rate": 1.9677257692693198e-05,
-      "loss": 5.474,
+      "epoch": 0.02,
+      "grad_norm": 22.919443130493164,
+      "learning_rate": 1.9865288875873323e-05,
+      "loss": 2.4198,
       "step": 162
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 20.959766387939453,
-      "learning_rate": 1.9675253082088805e-05,
-      "loss": 4.8595,
+      "epoch": 0.02,
+      "grad_norm": 24.94849395751953,
+      "learning_rate": 1.9864452160816637e-05,
+      "loss": 5.8624,
       "step": 163
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 49.37668228149414,
-      "learning_rate": 1.9673248471484415e-05,
-      "loss": 4.2558,
+      "epoch": 0.02,
+      "grad_norm": 20.590097427368164,
+      "learning_rate": 1.9863615445759947e-05,
+      "loss": 3.993,
       "step": 164
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 23.258516311645508,
-      "learning_rate": 1.9671243860880025e-05,
-      "loss": 4.362,
+      "epoch": 0.02,
+      "grad_norm": 15.112648963928223,
+      "learning_rate": 1.986277873070326e-05,
+      "loss": 3.5619,
       "step": 165
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 23.190126419067383,
-      "learning_rate": 1.9669239250275635e-05,
-      "loss": 3.2228,
+      "epoch": 0.02,
+      "grad_norm": 56.81850051879883,
+      "learning_rate": 1.9861942015646574e-05,
+      "loss": 2.6144,
       "step": 166
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 18.251544952392578,
-      "learning_rate": 1.9667234639671245e-05,
-      "loss": 4.7009,
+      "epoch": 0.02,
+      "grad_norm": 63.49656677246094,
+      "learning_rate": 1.9861105300589888e-05,
+      "loss": 3.4254,
       "step": 167
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 19.142702102661133,
-      "learning_rate": 1.9665230029066855e-05,
-      "loss": 3.9582,
+      "epoch": 0.02,
+      "grad_norm": 52.94969940185547,
+      "learning_rate": 1.9860268585533198e-05,
+      "loss": 4.6606,
       "step": 168
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 21.59228515625,
-      "learning_rate": 1.9663225418462465e-05,
-      "loss": 5.7372,
+      "epoch": 0.02,
+      "grad_norm": 48.12769317626953,
+      "learning_rate": 1.985943187047651e-05,
+      "loss": 4.2959,
       "step": 169
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 55.695281982421875,
-      "learning_rate": 1.9661220807858075e-05,
-      "loss": 3.9148,
+      "epoch": 0.02,
+      "grad_norm": 31.31679916381836,
+      "learning_rate": 1.9858595155419825e-05,
+      "loss": 4.3399,
       "step": 170
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 68.97881317138672,
-      "learning_rate": 1.9659216197253685e-05,
-      "loss": 4.3682,
+      "epoch": 0.02,
+      "grad_norm": 28.835838317871094,
+      "learning_rate": 1.9857758440363136e-05,
+      "loss": 6.7479,
       "step": 171
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 20.77494239807129,
-      "learning_rate": 1.9657211586649296e-05,
-      "loss": 4.0729,
+      "epoch": 0.02,
+      "grad_norm": 26.881940841674805,
+      "learning_rate": 1.985692172530645e-05,
+      "loss": 5.6704,
       "step": 172
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 48.90683364868164,
-      "learning_rate": 1.9655206976044906e-05,
-      "loss": 3.9279,
+      "epoch": 0.02,
+      "grad_norm": 29.435422897338867,
+      "learning_rate": 1.9856085010249763e-05,
+      "loss": 3.5854,
       "step": 173
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 27.227113723754883,
-      "learning_rate": 1.9653202365440516e-05,
-      "loss": 4.6083,
+      "epoch": 0.02,
+      "grad_norm": 28.923675537109375,
+      "learning_rate": 1.9855248295193073e-05,
+      "loss": 5.2268,
       "step": 174
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 25.539125442504883,
-      "learning_rate": 1.9651197754836126e-05,
-      "loss": 3.8157,
+      "epoch": 0.02,
+      "grad_norm": 50.01896667480469,
+      "learning_rate": 1.9854411580136387e-05,
+      "loss": 2.6085,
       "step": 175
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 21.638704299926758,
-      "learning_rate": 1.9649193144231736e-05,
-      "loss": 3.8649,
+      "epoch": 0.02,
+      "grad_norm": 34.73915100097656,
+      "learning_rate": 1.98535748650797e-05,
+      "loss": 4.8505,
       "step": 176
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 29.82769775390625,
-      "learning_rate": 1.9647188533627343e-05,
-      "loss": 4.5611,
+      "epoch": 0.02,
+      "grad_norm": 96.74727630615234,
+      "learning_rate": 1.985273815002301e-05,
+      "loss": 5.0014,
       "step": 177
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 22.91179656982422,
-      "learning_rate": 1.9645183923022956e-05,
-      "loss": 3.083,
+      "epoch": 0.02,
+      "grad_norm": 32.39076232910156,
+      "learning_rate": 1.9851901434966324e-05,
+      "loss": 6.2986,
       "step": 178
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 19.676715850830078,
-      "learning_rate": 1.9643179312418563e-05,
-      "loss": 3.8661,
+      "epoch": 0.02,
+      "grad_norm": 20.07256317138672,
+      "learning_rate": 1.9851064719909635e-05,
+      "loss": 4.4205,
       "step": 179
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 18.871702194213867,
-      "learning_rate": 1.9641174701814173e-05,
-      "loss": 3.8232,
+      "epoch": 0.02,
+      "grad_norm": 20.149194717407227,
+      "learning_rate": 1.9850228004852948e-05,
+      "loss": 4.715,
       "step": 180
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 33.633079528808594,
-      "learning_rate": 1.9639170091209786e-05,
-      "loss": 3.9483,
+      "epoch": 0.02,
+      "grad_norm": 34.583003997802734,
+      "learning_rate": 1.9849391289796262e-05,
+      "loss": 6.2908,
       "step": 181
     },
     {
-      "epoch": 0.05,
-      "grad_norm": 17.3801212310791,
-      "learning_rate": 1.9637165480605393e-05,
-      "loss": 3.4511,
+      "epoch": 0.02,
+      "grad_norm": 15.587189674377441,
+      "learning_rate": 1.9848554574739572e-05,
+      "loss": 2.9253,
       "step": 182
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 30.476238250732422,
-      "learning_rate": 1.9635160870001003e-05,
-      "loss": 4.5803,
+      "epoch": 0.02,
+      "grad_norm": 30.812002182006836,
+      "learning_rate": 1.9847717859682886e-05,
+      "loss": 4.1018,
       "step": 183
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 20.700347900390625,
-      "learning_rate": 1.9633156259396613e-05,
-      "loss": 3.6991,
+      "epoch": 0.02,
+      "grad_norm": 22.826852798461914,
+      "learning_rate": 1.98468811446262e-05,
+      "loss": 2.9077,
       "step": 184
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 16.463884353637695,
-      "learning_rate": 1.9631151648792223e-05,
-      "loss": 3.2284,
+      "epoch": 0.02,
+      "grad_norm": 92.27213287353516,
+      "learning_rate": 1.984604442956951e-05,
+      "loss": 4.9038,
       "step": 185
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 22.347118377685547,
-      "learning_rate": 1.9629147038187833e-05,
-      "loss": 4.603,
+      "epoch": 0.02,
+      "grad_norm": 34.23793029785156,
+      "learning_rate": 1.9845207714512823e-05,
+      "loss": 4.782,
       "step": 186
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 22.56680679321289,
-      "learning_rate": 1.9627142427583443e-05,
-      "loss": 4.332,
+      "epoch": 0.02,
+      "grad_norm": 40.07436752319336,
+      "learning_rate": 1.9844370999456137e-05,
+      "loss": 6.2297,
       "step": 187
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 31.80207061767578,
-      "learning_rate": 1.9625137816979054e-05,
-      "loss": 3.8008,
+      "epoch": 0.02,
+      "grad_norm": 38.16945266723633,
+      "learning_rate": 1.9843534284399447e-05,
+      "loss": 6.1175,
       "step": 188
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 22.534713745117188,
-      "learning_rate": 1.9623133206374664e-05,
-      "loss": 3.7199,
+      "epoch": 0.02,
+      "grad_norm": 38.583229064941406,
+      "learning_rate": 1.984269756934276e-05,
+      "loss": 3.8736,
       "step": 189
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 22.314205169677734,
-      "learning_rate": 1.9621128595770274e-05,
-      "loss": 4.3184,
+      "epoch": 0.02,
+      "grad_norm": 36.794189453125,
+      "learning_rate": 1.9841860854286075e-05,
+      "loss": 4.1338,
       "step": 190
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 24.710988998413086,
-      "learning_rate": 1.9619123985165884e-05,
-      "loss": 4.7067,
+      "epoch": 0.02,
+      "grad_norm": 37.98091125488281,
+      "learning_rate": 1.9841024139229388e-05,
+      "loss": 5.0907,
       "step": 191
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 34.565242767333984,
-      "learning_rate": 1.9617119374561494e-05,
-      "loss": 3.0013,
+      "epoch": 0.02,
+      "grad_norm": 21.34034538269043,
+      "learning_rate": 1.98401874241727e-05,
+      "loss": 3.6487,
       "step": 192
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 21.252920150756836,
-      "learning_rate": 1.9615114763957104e-05,
-      "loss": 5.5017,
+      "epoch": 0.02,
+      "grad_norm": 38.78040313720703,
+      "learning_rate": 1.9839350709116012e-05,
+      "loss": 3.5643,
       "step": 193
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 18.080705642700195,
-      "learning_rate": 1.9613110153352714e-05,
-      "loss": 3.4748,
+      "epoch": 0.02,
+      "grad_norm": 31.415233612060547,
+      "learning_rate": 1.9838513994059326e-05,
+      "loss": 2.7609,
       "step": 194
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 20.119901657104492,
-      "learning_rate": 1.9611105542748324e-05,
-      "loss": 4.6922,
+      "epoch": 0.02,
+      "grad_norm": 28.023122787475586,
+      "learning_rate": 1.9837677279002636e-05,
+      "loss": 4.1092,
       "step": 195
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 20.794160842895508,
-      "learning_rate": 1.960910093214393e-05,
-      "loss": 4.3524,
+      "epoch": 0.02,
+      "grad_norm": 22.933298110961914,
+      "learning_rate": 1.983684056394595e-05,
+      "loss": 3.9792,
       "step": 196
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 16.995960235595703,
-      "learning_rate": 1.9607096321539544e-05,
-      "loss": 5.311,
+      "epoch": 0.02,
+      "grad_norm": 25.238265991210938,
+      "learning_rate": 1.9836003848889263e-05,
+      "loss": 3.7678,
       "step": 197
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 28.07927131652832,
-      "learning_rate": 1.9605091710935154e-05,
-      "loss": 4.5386,
+      "epoch": 0.02,
+      "grad_norm": 34.65019989013672,
+      "learning_rate": 1.9835167133832577e-05,
+      "loss": 3.7924,
       "step": 198
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 20.054431915283203,
-      "learning_rate": 1.960308710033076e-05,
-      "loss": 3.8613,
+      "epoch": 0.02,
+      "grad_norm": 40.85706329345703,
+      "learning_rate": 1.9834330418775887e-05,
+      "loss": 4.5442,
       "step": 199
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 19.08523941040039,
-      "learning_rate": 1.9601082489726375e-05,
-      "loss": 4.5862,
+      "epoch": 0.03,
+      "grad_norm": 19.91678237915039,
+      "learning_rate": 1.98334937037192e-05,
+      "loss": 4.335,
       "step": 200
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 20.86153793334961,
-      "learning_rate": 1.959907787912198e-05,
-      "loss": 4.4943,
+      "epoch": 0.03,
+      "grad_norm": 27.40424156188965,
+      "learning_rate": 1.9832656988662515e-05,
+      "loss": 4.0261,
       "step": 201
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 17.117141723632812,
-      "learning_rate": 1.959707326851759e-05,
-      "loss": 3.1455,
+      "epoch": 0.03,
+      "grad_norm": 28.54695701599121,
+      "learning_rate": 1.9831820273605825e-05,
+      "loss": 5.5133,
       "step": 202
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 20.701730728149414,
-      "learning_rate": 1.95950686579132e-05,
-      "loss": 4.1613,
+      "epoch": 0.03,
+      "grad_norm": 21.289785385131836,
+      "learning_rate": 1.983098355854914e-05,
+      "loss": 2.3904,
       "step": 203
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 17.824867248535156,
-      "learning_rate": 1.959306404730881e-05,
-      "loss": 3.764,
+      "epoch": 0.03,
+      "grad_norm": 20.005878448486328,
+      "learning_rate": 1.9830146843492452e-05,
+      "loss": 4.9863,
       "step": 204
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 23.23500633239746,
-      "learning_rate": 1.959105943670442e-05,
-      "loss": 3.647,
+      "epoch": 0.03,
+      "grad_norm": 27.44087028503418,
+      "learning_rate": 1.9829310128435762e-05,
+      "loss": 4.5081,
       "step": 205
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 19.766569137573242,
-      "learning_rate": 1.958905482610003e-05,
-      "loss": 3.9145,
+      "epoch": 0.03,
+      "grad_norm": 42.43584442138672,
+      "learning_rate": 1.9828473413379076e-05,
+      "loss": 5.4452,
       "step": 206
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 22.38438606262207,
-      "learning_rate": 1.9587050215495642e-05,
-      "loss": 4.4528,
+      "epoch": 0.03,
+      "grad_norm": 25.16995620727539,
+      "learning_rate": 1.9827636698322386e-05,
+      "loss": 4.7488,
       "step": 207
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 17.628910064697266,
-      "learning_rate": 1.9585045604891252e-05,
-      "loss": 3.7731,
+      "epoch": 0.03,
+      "grad_norm": 13.992096900939941,
+      "learning_rate": 1.98267999832657e-05,
+      "loss": 5.4432,
       "step": 208
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 18.171613693237305,
-      "learning_rate": 1.9583040994286862e-05,
-      "loss": 4.1904,
+      "epoch": 0.03,
+      "grad_norm": 32.74681854248047,
+      "learning_rate": 1.9825963268209014e-05,
+      "loss": 4.6107,
       "step": 209
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 27.142881393432617,
-      "learning_rate": 1.958103638368247e-05,
-      "loss": 4.3347,
+      "epoch": 0.03,
+      "grad_norm": 29.164478302001953,
+      "learning_rate": 1.9825126553152324e-05,
+      "loss": 3.5511,
       "step": 210
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 13.47857666015625,
-      "learning_rate": 1.9579031773078082e-05,
-      "loss": 3.2086,
+      "epoch": 0.03,
+      "grad_norm": 25.634967803955078,
+      "learning_rate": 1.9824289838095637e-05,
+      "loss": 4.1461,
       "step": 211
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 18.275033950805664,
-      "learning_rate": 1.9577027162473692e-05,
-      "loss": 4.6453,
+      "epoch": 0.03,
+      "grad_norm": 26.063108444213867,
+      "learning_rate": 1.982345312303895e-05,
+      "loss": 2.5191,
       "step": 212
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 17.69455337524414,
-      "learning_rate": 1.95750225518693e-05,
-      "loss": 3.5041,
+      "epoch": 0.03,
+      "grad_norm": 30.398160934448242,
+      "learning_rate": 1.982261640798226e-05,
+      "loss": 4.1432,
       "step": 213
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 16.023883819580078,
-      "learning_rate": 1.9573017941264912e-05,
-      "loss": 4.2656,
+      "epoch": 0.03,
+      "grad_norm": 20.190427780151367,
+      "learning_rate": 1.9821779692925575e-05,
+      "loss": 2.7101,
       "step": 214
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 25.941577911376953,
-      "learning_rate": 1.957101333066052e-05,
-      "loss": 3.9683,
+      "epoch": 0.03,
+      "grad_norm": 29.440567016601562,
+      "learning_rate": 1.982094297786889e-05,
+      "loss": 4.5135,
       "step": 215
     },
     {
-      "epoch": 0.06,
-      "grad_norm": 28.56459617614746,
-      "learning_rate": 1.956900872005613e-05,
-      "loss": 4.1389,
+      "epoch": 0.03,
+      "grad_norm": 24.399301528930664,
+      "learning_rate": 1.98201062628122e-05,
+      "loss": 3.446,
       "step": 216
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 18.201026916503906,
-      "learning_rate": 1.9567004109451743e-05,
-      "loss": 4.2534,
+      "epoch": 0.03,
+      "grad_norm": 19.4206600189209,
+      "learning_rate": 1.9819269547755513e-05,
+      "loss": 3.162,
       "step": 217
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 17.178457260131836,
-      "learning_rate": 1.956499949884735e-05,
-      "loss": 3.8277,
+      "epoch": 0.03,
+      "grad_norm": 23.20094108581543,
+      "learning_rate": 1.9818432832698826e-05,
+      "loss": 5.614,
       "step": 218
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 38.801876068115234,
-      "learning_rate": 1.956299488824296e-05,
-      "loss": 4.1429,
+      "epoch": 0.03,
+      "grad_norm": 23.896318435668945,
+      "learning_rate": 1.981759611764214e-05,
+      "loss": 2.5011,
       "step": 219
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 23.956295013427734,
-      "learning_rate": 1.956099027763857e-05,
-      "loss": 5.4837,
+      "epoch": 0.03,
+      "grad_norm": 27.542938232421875,
+      "learning_rate": 1.981675940258545e-05,
+      "loss": 1.6989,
       "step": 220
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 16.060426712036133,
-      "learning_rate": 1.955898566703418e-05,
-      "loss": 3.7044,
+      "epoch": 0.03,
+      "grad_norm": 20.2366886138916,
+      "learning_rate": 1.9815922687528764e-05,
+      "loss": 2.322,
       "step": 221
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 18.859159469604492,
-      "learning_rate": 1.955698105642979e-05,
-      "loss": 4.168,
+      "epoch": 0.03,
+      "grad_norm": 19.03917694091797,
+      "learning_rate": 1.9815085972472077e-05,
+      "loss": 7.0657,
       "step": 222
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 21.508644104003906,
-      "learning_rate": 1.95549764458254e-05,
-      "loss": 4.5285,
+      "epoch": 0.03,
+      "grad_norm": 80.15726470947266,
+      "learning_rate": 1.9814249257415388e-05,
+      "loss": 4.5171,
       "step": 223
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 25.023488998413086,
-      "learning_rate": 1.955297183522101e-05,
-      "loss": 3.2762,
+      "epoch": 0.03,
+      "grad_norm": 24.671308517456055,
+      "learning_rate": 1.98134125423587e-05,
+      "loss": 5.1079,
       "step": 224
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 16.727890014648438,
-      "learning_rate": 1.955096722461662e-05,
-      "loss": 3.3409,
+      "epoch": 0.03,
+      "grad_norm": 17.61369514465332,
+      "learning_rate": 1.9812575827302015e-05,
+      "loss": 2.1667,
       "step": 225
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 20.456737518310547,
-      "learning_rate": 1.954896261401223e-05,
-      "loss": 4.1665,
+      "epoch": 0.03,
+      "grad_norm": 23.398298263549805,
+      "learning_rate": 1.981173911224533e-05,
+      "loss": 4.2372,
       "step": 226
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 17.329370498657227,
-      "learning_rate": 1.954695800340784e-05,
-      "loss": 3.0992,
+      "epoch": 0.03,
+      "grad_norm": 29.489824295043945,
+      "learning_rate": 1.981090239718864e-05,
+      "loss": 4.3447,
       "step": 227
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 39.43778610229492,
-      "learning_rate": 1.954495339280345e-05,
-      "loss": 4.5199,
+      "epoch": 0.03,
+      "grad_norm": 44.913883209228516,
+      "learning_rate": 1.9810065682131953e-05,
+      "loss": 3.274,
       "step": 228
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 20.120681762695312,
-      "learning_rate": 1.954294878219906e-05,
-      "loss": 4.6025,
+      "epoch": 0.03,
+      "grad_norm": 33.340850830078125,
+      "learning_rate": 1.9809228967075266e-05,
+      "loss": 3.2506,
       "step": 229
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 22.665719985961914,
-      "learning_rate": 1.954094417159467e-05,
-      "loss": 4.3815,
+      "epoch": 0.03,
+      "grad_norm": 19.065631866455078,
+      "learning_rate": 1.9808392252018576e-05,
+      "loss": 4.2629,
       "step": 230
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 24.487995147705078,
-      "learning_rate": 1.953893956099028e-05,
-      "loss": 4.2686,
+      "epoch": 0.03,
+      "grad_norm": 24.77507972717285,
+      "learning_rate": 1.980755553696189e-05,
+      "loss": 3.236,
       "step": 231
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 32.88031768798828,
-      "learning_rate": 1.9536934950385887e-05,
-      "loss": 4.3775,
+      "epoch": 0.03,
+      "grad_norm": 55.65793991088867,
+      "learning_rate": 1.98067188219052e-05,
+      "loss": 3.031,
       "step": 232
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 22.327272415161133,
-      "learning_rate": 1.95349303397815e-05,
-      "loss": 3.7784,
+      "epoch": 0.03,
+      "grad_norm": 14.242398262023926,
+      "learning_rate": 1.9805882106848514e-05,
+      "loss": 4.0164,
       "step": 233
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 33.93623352050781,
-      "learning_rate": 1.9532925729177107e-05,
-      "loss": 3.2231,
+      "epoch": 0.03,
+      "grad_norm": 25.252952575683594,
+      "learning_rate": 1.9805045391791828e-05,
+      "loss": 3.5733,
       "step": 234
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 24.04096794128418,
-      "learning_rate": 1.9530921118572717e-05,
-      "loss": 4.1805,
+      "epoch": 0.03,
+      "grad_norm": 26.969528198242188,
+      "learning_rate": 1.9804208676735138e-05,
+      "loss": 4.2902,
       "step": 235
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 18.985858917236328,
-      "learning_rate": 1.952891650796833e-05,
-      "loss": 4.1365,
+      "epoch": 0.03,
+      "grad_norm": 18.014312744140625,
+      "learning_rate": 1.980337196167845e-05,
+      "loss": 4.9391,
       "step": 236
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 21.616893768310547,
-      "learning_rate": 1.9526911897363937e-05,
-      "loss": 2.4831,
+      "epoch": 0.03,
+      "grad_norm": 67.04398345947266,
+      "learning_rate": 1.9802535246621762e-05,
+      "loss": 2.8289,
       "step": 237
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 26.768409729003906,
-      "learning_rate": 1.9524907286759548e-05,
-      "loss": 4.6538,
+      "epoch": 0.03,
+      "grad_norm": 24.836767196655273,
+      "learning_rate": 1.9801698531565076e-05,
+      "loss": 2.4735,
       "step": 238
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 21.572690963745117,
-      "learning_rate": 1.9522902676155158e-05,
-      "loss": 3.7805,
+      "epoch": 0.03,
+      "grad_norm": 36.715023040771484,
+      "learning_rate": 1.980086181650839e-05,
+      "loss": 4.0768,
       "step": 239
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 14.675576210021973,
-      "learning_rate": 1.9520898065550768e-05,
-      "loss": 3.3591,
-      "step": 240
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 1.399770975112915,
-      "eval_runtime": 43.6698,
-      "eval_samples_per_second": 33.868,
-      "eval_steps_per_second": 33.868,
+      "epoch": 0.03,
+      "grad_norm": 17.729856491088867,
+      "learning_rate": 1.9800025101451703e-05,
+      "loss": 5.27,
       "step": 240
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 39.946598052978516,
-      "learning_rate": 1.9518893454946378e-05,
-      "loss": 3.3287,
+      "epoch": 0.03,
+      "grad_norm": 45.96723556518555,
+      "learning_rate": 1.9799188386395013e-05,
+      "loss": 4.0394,
       "step": 241
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 22.521690368652344,
-      "learning_rate": 1.9516888844341988e-05,
-      "loss": 3.3806,
+      "epoch": 0.03,
+      "grad_norm": 92.66098022460938,
+      "learning_rate": 1.9798351671338327e-05,
+      "loss": 3.5547,
       "step": 242
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 18.76473045349121,
-      "learning_rate": 1.9514884233737598e-05,
-      "loss": 4.1907,
+      "epoch": 0.03,
+      "grad_norm": 19.052738189697266,
+      "learning_rate": 1.979751495628164e-05,
+      "loss": 3.59,
       "step": 243
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 17.154766082763672,
-      "learning_rate": 1.9512879623133208e-05,
-      "loss": 3.7032,
+      "epoch": 0.03,
+      "grad_norm": 27.923494338989258,
+      "learning_rate": 1.979667824122495e-05,
+      "loss": 3.4968,
       "step": 244
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 25.60683250427246,
-      "learning_rate": 1.9510875012528818e-05,
-      "loss": 4.4485,
+      "epoch": 0.03,
+      "grad_norm": 24.136812210083008,
+      "learning_rate": 1.9795841526168264e-05,
+      "loss": 4.4669,
       "step": 245
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 29.663965225219727,
-      "learning_rate": 1.9508870401924428e-05,
-      "loss": 4.5449,
+      "epoch": 0.03,
+      "grad_norm": 22.240131378173828,
+      "learning_rate": 1.9795004811111578e-05,
+      "loss": 4.423,
       "step": 246
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 24.08624267578125,
-      "learning_rate": 1.950686579132004e-05,
-      "loss": 4.6023,
+      "epoch": 0.03,
+      "grad_norm": 20.765474319458008,
+      "learning_rate": 1.979416809605489e-05,
+      "loss": 3.8466,
       "step": 247
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 17.730571746826172,
-      "learning_rate": 1.950486118071565e-05,
-      "loss": 4.7885,
+      "epoch": 0.03,
+      "grad_norm": 16.91594123840332,
+      "learning_rate": 1.9793331380998202e-05,
+      "loss": 3.9957,
       "step": 248
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 16.919239044189453,
-      "learning_rate": 1.950285657011126e-05,
-      "loss": 3.2188,
+      "epoch": 0.03,
+      "grad_norm": 46.77788543701172,
+      "learning_rate": 1.9792494665941515e-05,
+      "loss": 2.9274,
       "step": 249
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 16.421016693115234,
-      "learning_rate": 1.950085195950687e-05,
-      "loss": 4.6058,
+      "epoch": 0.03,
+      "grad_norm": 26.280046463012695,
+      "learning_rate": 1.979165795088483e-05,
+      "loss": 5.7432,
       "step": 250
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 26.03248405456543,
-      "learning_rate": 1.9498847348902475e-05,
-      "loss": 5.1777,
+      "epoch": 0.03,
+      "grad_norm": 32.9533576965332,
+      "learning_rate": 1.979082123582814e-05,
+      "loss": 3.7103,
       "step": 251
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 24.347097396850586,
-      "learning_rate": 1.949684273829809e-05,
-      "loss": 4.5525,
+      "epoch": 0.03,
+      "grad_norm": 21.88126564025879,
+      "learning_rate": 1.9789984520771453e-05,
+      "loss": 5.8115,
       "step": 252
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 16.834897994995117,
-      "learning_rate": 1.9494838127693695e-05,
-      "loss": 3.7237,
+      "epoch": 0.03,
+      "grad_norm": 25.590606689453125,
+      "learning_rate": 1.9789147805714767e-05,
+      "loss": 5.702,
       "step": 253
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 14.267677307128906,
-      "learning_rate": 1.9492833517089306e-05,
-      "loss": 4.2919,
+      "epoch": 0.03,
+      "grad_norm": 16.906322479248047,
+      "learning_rate": 1.978831109065808e-05,
+      "loss": 4.2226,
       "step": 254
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 14.566679954528809,
-      "learning_rate": 1.949082890648492e-05,
-      "loss": 4.2664,
+      "epoch": 0.03,
+      "grad_norm": 54.520263671875,
+      "learning_rate": 1.978747437560139e-05,
+      "loss": 3.7309,
       "step": 255
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 20.296588897705078,
-      "learning_rate": 1.9488824295880526e-05,
-      "loss": 4.1636,
+      "epoch": 0.03,
+      "grad_norm": 15.407824516296387,
+      "learning_rate": 1.9786637660544704e-05,
+      "loss": 1.4514,
       "step": 256
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 19.508743286132812,
-      "learning_rate": 1.9486819685276136e-05,
-      "loss": 3.5817,
+      "epoch": 0.03,
+      "grad_norm": 27.355886459350586,
+      "learning_rate": 1.9785800945488018e-05,
+      "loss": 6.2087,
       "step": 257
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 22.892982482910156,
-      "learning_rate": 1.9484815074671746e-05,
-      "loss": 4.1911,
+      "epoch": 0.03,
+      "grad_norm": 19.986955642700195,
+      "learning_rate": 1.9784964230431328e-05,
+      "loss": 4.4818,
       "step": 258
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 20.90638542175293,
-      "learning_rate": 1.9482810464067356e-05,
-      "loss": 3.3314,
+      "epoch": 0.03,
+      "grad_norm": 19.87293243408203,
+      "learning_rate": 1.9784127515374642e-05,
+      "loss": 3.9836,
       "step": 259
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 16.314008712768555,
-      "learning_rate": 1.9480805853462966e-05,
-      "loss": 3.5795,
+      "epoch": 0.03,
+      "grad_norm": 19.814401626586914,
+      "learning_rate": 1.9783290800317952e-05,
+      "loss": 4.6888,
       "step": 260
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 19.024072647094727,
-      "learning_rate": 1.9478801242858576e-05,
-      "loss": 3.4856,
+      "epoch": 0.03,
+      "grad_norm": 63.62940979003906,
+      "learning_rate": 1.9782454085261266e-05,
+      "loss": 4.4955,
       "step": 261
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 15.41294002532959,
-      "learning_rate": 1.9476796632254186e-05,
-      "loss": 3.9168,
+      "epoch": 0.03,
+      "grad_norm": 21.967994689941406,
+      "learning_rate": 1.978161737020458e-05,
+      "loss": 3.8234,
       "step": 262
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 27.279674530029297,
-      "learning_rate": 1.9474792021649796e-05,
-      "loss": 4.4373,
+      "epoch": 0.03,
+      "grad_norm": 24.47323989868164,
+      "learning_rate": 1.978078065514789e-05,
+      "loss": 3.2202,
       "step": 263
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 26.365346908569336,
-      "learning_rate": 1.9472787411045406e-05,
-      "loss": 4.3507,
+      "epoch": 0.03,
+      "grad_norm": 31.2077693939209,
+      "learning_rate": 1.9779943940091203e-05,
+      "loss": 4.4682,
       "step": 264
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 18.319107055664062,
-      "learning_rate": 1.9470782800441016e-05,
-      "loss": 4.1936,
+      "epoch": 0.03,
+      "grad_norm": 16.750263214111328,
+      "learning_rate": 1.9779107225034514e-05,
+      "loss": 3.5919,
       "step": 265
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 16.385744094848633,
-      "learning_rate": 1.9468778189836627e-05,
-      "loss": 4.3412,
+      "epoch": 0.03,
+      "grad_norm": 23.57237434387207,
+      "learning_rate": 1.9778270509977827e-05,
+      "loss": 1.6616,
       "step": 266
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 17.853548049926758,
-      "learning_rate": 1.9466773579232237e-05,
-      "loss": 4.4193,
+      "epoch": 0.03,
+      "grad_norm": 22.44601058959961,
+      "learning_rate": 1.977743379492114e-05,
+      "loss": 4.5676,
       "step": 267
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 23.52976417541504,
-      "learning_rate": 1.9464768968627847e-05,
-      "loss": 5.0003,
+      "epoch": 0.03,
+      "grad_norm": 24.28874969482422,
+      "learning_rate": 1.9776597079864454e-05,
+      "loss": 4.0052,
       "step": 268
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 20.375717163085938,
-      "learning_rate": 1.9462764358023457e-05,
-      "loss": 3.6193,
+      "epoch": 0.03,
+      "grad_norm": 15.053487777709961,
+      "learning_rate": 1.9775760364807765e-05,
+      "loss": 5.6047,
       "step": 269
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 19.429637908935547,
-      "learning_rate": 1.9460759747419063e-05,
-      "loss": 4.0949,
+      "epoch": 0.03,
+      "grad_norm": 16.717178344726562,
+      "learning_rate": 1.977492364975108e-05,
+      "loss": 1.4622,
       "step": 270
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 16.466793060302734,
-      "learning_rate": 1.9458755136814677e-05,
-      "loss": 3.3825,
+      "epoch": 0.03,
+      "grad_norm": 127.36611938476562,
+      "learning_rate": 1.9774086934694392e-05,
+      "loss": 3.3587,
       "step": 271
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 28.024784088134766,
-      "learning_rate": 1.9456750526210287e-05,
-      "loss": 4.1028,
+      "epoch": 0.03,
+      "grad_norm": 24.546730041503906,
+      "learning_rate": 1.9773250219637702e-05,
+      "loss": 2.8608,
       "step": 272
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 20.214069366455078,
-      "learning_rate": 1.9454745915605894e-05,
-      "loss": 3.611,
+      "epoch": 0.03,
+      "grad_norm": 27.449810028076172,
+      "learning_rate": 1.9772413504581016e-05,
+      "loss": 3.8879,
       "step": 273
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 17.36444854736328,
-      "learning_rate": 1.9452741305001507e-05,
-      "loss": 3.7358,
+      "epoch": 0.03,
+      "grad_norm": 23.76478385925293,
+      "learning_rate": 1.977157678952433e-05,
+      "loss": 3.8756,
       "step": 274
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 17.746870040893555,
-      "learning_rate": 1.9450736694397114e-05,
-      "loss": 3.527,
+      "epoch": 0.03,
+      "grad_norm": 28.613237380981445,
+      "learning_rate": 1.9770740074467643e-05,
+      "loss": 1.9286,
       "step": 275
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 15.168383598327637,
-      "learning_rate": 1.9448732083792724e-05,
-      "loss": 3.3192,
+      "epoch": 0.03,
+      "grad_norm": 47.288856506347656,
+      "learning_rate": 1.9769903359410954e-05,
+      "loss": 4.0929,
       "step": 276
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 24.655101776123047,
-      "learning_rate": 1.9446727473188334e-05,
-      "loss": 5.3757,
+      "epoch": 0.03,
+      "grad_norm": 24.35297966003418,
+      "learning_rate": 1.9769066644354267e-05,
+      "loss": 4.455,
       "step": 277
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 22.04741668701172,
-      "learning_rate": 1.9444722862583944e-05,
-      "loss": 3.5021,
+      "epoch": 0.03,
+      "grad_norm": 70.38229370117188,
+      "learning_rate": 1.976822992929758e-05,
+      "loss": 3.4509,
       "step": 278
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 15.345871925354004,
-      "learning_rate": 1.9442718251979554e-05,
-      "loss": 4.5815,
+      "epoch": 0.04,
+      "grad_norm": 15.766568183898926,
+      "learning_rate": 1.976739321424089e-05,
+      "loss": 3.3342,
       "step": 279
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 17.076583862304688,
-      "learning_rate": 1.9440713641375164e-05,
-      "loss": 4.7359,
+      "epoch": 0.04,
+      "grad_norm": 22.440858840942383,
+      "learning_rate": 1.9766556499184205e-05,
+      "loss": 5.897,
       "step": 280
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 17.223661422729492,
-      "learning_rate": 1.9438709030770774e-05,
-      "loss": 3.6285,
+      "epoch": 0.04,
+      "grad_norm": 15.124077796936035,
+      "learning_rate": 1.976571978412752e-05,
+      "loss": 3.1477,
       "step": 281
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 19.869312286376953,
-      "learning_rate": 1.9436704420166385e-05,
-      "loss": 3.9027,
+      "epoch": 0.04,
+      "grad_norm": 17.690269470214844,
+      "learning_rate": 1.9764883069070832e-05,
+      "loss": 4.4101,
       "step": 282
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 20.547405242919922,
-      "learning_rate": 1.9434699809561995e-05,
-      "loss": 3.6283,
+      "epoch": 0.04,
+      "grad_norm": 45.84641647338867,
+      "learning_rate": 1.9764046354014142e-05,
+      "loss": 6.0457,
       "step": 283
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 18.522144317626953,
-      "learning_rate": 1.9432695198957605e-05,
-      "loss": 4.2172,
+      "epoch": 0.04,
+      "grad_norm": 20.248172760009766,
+      "learning_rate": 1.9763209638957456e-05,
+      "loss": 3.7918,
       "step": 284
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 30.722158432006836,
-      "learning_rate": 1.9430690588353215e-05,
-      "loss": 3.6669,
+      "epoch": 0.04,
+      "grad_norm": 18.53105926513672,
+      "learning_rate": 1.9762372923900766e-05,
+      "loss": 3.0112,
       "step": 285
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 16.493730545043945,
-      "learning_rate": 1.9428685977748825e-05,
-      "loss": 4.1544,
+      "epoch": 0.04,
+      "grad_norm": 44.17702102661133,
+      "learning_rate": 1.976153620884408e-05,
+      "loss": 3.3778,
       "step": 286
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 15.535743713378906,
-      "learning_rate": 1.942668136714443e-05,
-      "loss": 4.1201,
+      "epoch": 0.04,
+      "grad_norm": 27.555721282958984,
+      "learning_rate": 1.9760699493787393e-05,
+      "loss": 3.6562,
       "step": 287
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 136.23187255859375,
-      "learning_rate": 1.9424676756540045e-05,
-      "loss": 4.5668,
+      "epoch": 0.04,
+      "grad_norm": 29.536550521850586,
+      "learning_rate": 1.9759862778730704e-05,
+      "loss": 5.412,
       "step": 288
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 15.681511878967285,
-      "learning_rate": 1.9422672145935652e-05,
-      "loss": 4.1992,
+      "epoch": 0.04,
+      "grad_norm": 24.517093658447266,
+      "learning_rate": 1.9759026063674017e-05,
+      "loss": 4.254,
       "step": 289
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 12.06187915802002,
-      "learning_rate": 1.9420667535331262e-05,
-      "loss": 2.6199,
+      "epoch": 0.04,
+      "grad_norm": 17.41961097717285,
+      "learning_rate": 1.9758189348617328e-05,
+      "loss": 3.973,
       "step": 290
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 14.176081657409668,
-      "learning_rate": 1.9418662924726875e-05,
-      "loss": 2.6663,
+      "epoch": 0.04,
+      "grad_norm": 14.359147071838379,
+      "learning_rate": 1.975735263356064e-05,
+      "loss": 1.5225,
       "step": 291
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 16.331350326538086,
-      "learning_rate": 1.9416658314122482e-05,
-      "loss": 4.1395,
+      "epoch": 0.04,
+      "grad_norm": 18.383588790893555,
+      "learning_rate": 1.9756515918503955e-05,
+      "loss": 2.5873,
       "step": 292
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 25.406064987182617,
-      "learning_rate": 1.9414653703518095e-05,
-      "loss": 3.2631,
+      "epoch": 0.04,
+      "grad_norm": 20.837644577026367,
+      "learning_rate": 1.9755679203447265e-05,
+      "loss": 4.5986,
       "step": 293
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 25.89323616027832,
-      "learning_rate": 1.9412649092913702e-05,
-      "loss": 4.4711,
+      "epoch": 0.04,
+      "grad_norm": 17.78486442565918,
+      "learning_rate": 1.975484248839058e-05,
+      "loss": 3.6508,
       "step": 294
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 12.358001708984375,
-      "learning_rate": 1.9410644482309312e-05,
-      "loss": 4.0686,
+      "epoch": 0.04,
+      "grad_norm": 23.529550552368164,
+      "learning_rate": 1.9754005773333893e-05,
+      "loss": 3.8641,
       "step": 295
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 17.452009201049805,
-      "learning_rate": 1.9408639871704922e-05,
-      "loss": 3.6253,
+      "epoch": 0.04,
+      "grad_norm": 28.262685775756836,
+      "learning_rate": 1.9753169058277206e-05,
+      "loss": 5.1546,
       "step": 296
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 28.617233276367188,
-      "learning_rate": 1.9406635261100532e-05,
-      "loss": 4.9357,
+      "epoch": 0.04,
+      "grad_norm": 45.46820831298828,
+      "learning_rate": 1.9752332343220516e-05,
+      "loss": 3.7566,
       "step": 297
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 17.00152587890625,
-      "learning_rate": 1.9404630650496142e-05,
-      "loss": 3.7579,
+      "epoch": 0.04,
+      "grad_norm": 18.303991317749023,
+      "learning_rate": 1.975149562816383e-05,
+      "loss": 2.5184,
       "step": 298
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 17.73528289794922,
-      "learning_rate": 1.9402626039891753e-05,
-      "loss": 3.7798,
+      "epoch": 0.04,
+      "grad_norm": 14.7951078414917,
+      "learning_rate": 1.9750658913107144e-05,
+      "loss": 3.0745,
       "step": 299
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 15.016806602478027,
-      "learning_rate": 1.9400621429287363e-05,
-      "loss": 4.3192,
+      "epoch": 0.04,
+      "grad_norm": 42.80633544921875,
+      "learning_rate": 1.9749822198050454e-05,
+      "loss": 3.0089,
       "step": 300
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 12.723051071166992,
-      "learning_rate": 1.9398616818682973e-05,
-      "loss": 3.8939,
+      "epoch": 0.04,
+      "grad_norm": 30.628992080688477,
+      "learning_rate": 1.9748985482993768e-05,
+      "loss": 4.6328,
       "step": 301
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 14.875953674316406,
-      "learning_rate": 1.9396612208078583e-05,
-      "loss": 2.8365,
+      "epoch": 0.04,
+      "grad_norm": 17.47206687927246,
+      "learning_rate": 1.974814876793708e-05,
+      "loss": 3.6176,
       "step": 302
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 16.63105583190918,
-      "learning_rate": 1.9394607597474193e-05,
-      "loss": 3.7121,
+      "epoch": 0.04,
+      "grad_norm": 12.888838768005371,
+      "learning_rate": 1.9747312052880395e-05,
+      "loss": 1.8506,
       "step": 303
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 22.02267074584961,
-      "learning_rate": 1.9392602986869803e-05,
-      "loss": 5.0273,
+      "epoch": 0.04,
+      "grad_norm": 14.139362335205078,
+      "learning_rate": 1.9746475337823705e-05,
+      "loss": 2.9678,
       "step": 304
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 44.309234619140625,
-      "learning_rate": 1.9390598376265413e-05,
-      "loss": 4.5806,
+      "epoch": 0.04,
+      "grad_norm": 22.3652400970459,
+      "learning_rate": 1.974563862276702e-05,
+      "loss": 3.513,
       "step": 305
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 20.403287887573242,
-      "learning_rate": 1.938859376566102e-05,
-      "loss": 3.2671,
+      "epoch": 0.04,
+      "grad_norm": 16.76512336730957,
+      "learning_rate": 1.9744801907710332e-05,
+      "loss": 3.6319,
       "step": 306
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 15.081746101379395,
-      "learning_rate": 1.9386589155056633e-05,
-      "loss": 4.2114,
+      "epoch": 0.04,
+      "grad_norm": 14.686978340148926,
+      "learning_rate": 1.9743965192653643e-05,
+      "loss": 2.93,
       "step": 307
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 16.417646408081055,
-      "learning_rate": 1.938458454445224e-05,
-      "loss": 3.878,
+      "epoch": 0.04,
+      "grad_norm": 17.279701232910156,
+      "learning_rate": 1.9743128477596956e-05,
+      "loss": 6.5399,
       "step": 308
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 20.749244689941406,
-      "learning_rate": 1.938257993384785e-05,
-      "loss": 3.3179,
+      "epoch": 0.04,
+      "grad_norm": 33.21736526489258,
+      "learning_rate": 1.974229176254027e-05,
+      "loss": 4.1478,
       "step": 309
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 21.518075942993164,
-      "learning_rate": 1.9380575323243463e-05,
-      "loss": 3.4908,
+      "epoch": 0.04,
+      "grad_norm": 15.007304191589355,
+      "learning_rate": 1.974145504748358e-05,
+      "loss": 4.3891,
       "step": 310
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 17.475610733032227,
-      "learning_rate": 1.937857071263907e-05,
-      "loss": 4.1835,
+      "epoch": 0.04,
+      "grad_norm": 14.931962966918945,
+      "learning_rate": 1.9740618332426894e-05,
+      "loss": 4.9052,
       "step": 311
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 20.958995819091797,
-      "learning_rate": 1.937656610203468e-05,
-      "loss": 4.5692,
+      "epoch": 0.04,
+      "grad_norm": 16.65130615234375,
+      "learning_rate": 1.9739781617370208e-05,
+      "loss": 2.7062,
       "step": 312
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 16.16900062561035,
-      "learning_rate": 1.937456149143029e-05,
-      "loss": 2.9379,
+      "epoch": 0.04,
+      "grad_norm": 23.00065040588379,
+      "learning_rate": 1.9738944902313518e-05,
+      "loss": 3.7142,
       "step": 313
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 29.76364517211914,
-      "learning_rate": 1.93725568808259e-05,
-      "loss": 4.0285,
+      "epoch": 0.04,
+      "grad_norm": 19.257022857666016,
+      "learning_rate": 1.973810818725683e-05,
+      "loss": 3.6447,
       "step": 314
     },
     {
-      "epoch": 0.09,
-      "grad_norm": 14.153215408325195,
-      "learning_rate": 1.937055227022151e-05,
-      "loss": 3.6143,
+      "epoch": 0.04,
+      "grad_norm": 18.844392776489258,
+      "learning_rate": 1.9737271472200145e-05,
+      "loss": 3.9087,
       "step": 315
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 25.9129581451416,
-      "learning_rate": 1.936854765961712e-05,
-      "loss": 4.9313,
+      "epoch": 0.04,
+      "grad_norm": 21.749853134155273,
+      "learning_rate": 1.9736434757143455e-05,
+      "loss": 2.765,
       "step": 316
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 12.36880874633789,
-      "learning_rate": 1.936654304901273e-05,
-      "loss": 3.5271,
+      "epoch": 0.04,
+      "grad_norm": 16.286396026611328,
+      "learning_rate": 1.973559804208677e-05,
+      "loss": 3.5264,
       "step": 317
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 20.27505874633789,
-      "learning_rate": 1.936453843840834e-05,
-      "loss": 3.8401,
+      "epoch": 0.04,
+      "grad_norm": 21.13903045654297,
+      "learning_rate": 1.973476132703008e-05,
+      "loss": 4.9465,
       "step": 318
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 20.562847137451172,
-      "learning_rate": 1.936253382780395e-05,
-      "loss": 4.0286,
+      "epoch": 0.04,
+      "grad_norm": 18.01095199584961,
+      "learning_rate": 1.9733924611973393e-05,
+      "loss": 4.4266,
       "step": 319
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 23.03611183166504,
-      "learning_rate": 1.936052921719956e-05,
-      "loss": 3.1755,
+      "epoch": 0.04,
+      "grad_norm": 57.30530548095703,
+      "learning_rate": 1.9733087896916707e-05,
+      "loss": 4.1182,
       "step": 320
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 24.045217514038086,
-      "learning_rate": 1.935852460659517e-05,
-      "loss": 4.1921,
+      "epoch": 0.04,
+      "grad_norm": 34.54814529418945,
+      "learning_rate": 1.9732251181860017e-05,
+      "loss": 6.332,
       "step": 321
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 15.012109756469727,
-      "learning_rate": 1.935651999599078e-05,
-      "loss": 3.9538,
+      "epoch": 0.04,
+      "grad_norm": 15.039722442626953,
+      "learning_rate": 1.973141446680333e-05,
+      "loss": 2.0336,
       "step": 322
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 16.769580841064453,
-      "learning_rate": 1.935451538538639e-05,
-      "loss": 4.1792,
+      "epoch": 0.04,
+      "grad_norm": 16.953218460083008,
+      "learning_rate": 1.9730577751746644e-05,
+      "loss": 4.2034,
       "step": 323
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 15.990212440490723,
-      "learning_rate": 1.9352510774782e-05,
-      "loss": 3.3792,
+      "epoch": 0.04,
+      "grad_norm": 12.90640926361084,
+      "learning_rate": 1.9729741036689958e-05,
+      "loss": 2.2464,
       "step": 324
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 17.586837768554688,
-      "learning_rate": 1.9350506164177608e-05,
-      "loss": 3.8851,
+      "epoch": 0.04,
+      "grad_norm": 28.966737747192383,
+      "learning_rate": 1.9728904321633268e-05,
+      "loss": 2.8233,
       "step": 325
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 28.66331672668457,
-      "learning_rate": 1.934850155357322e-05,
-      "loss": 4.4007,
+      "epoch": 0.04,
+      "grad_norm": 21.538097381591797,
+      "learning_rate": 1.9728067606576582e-05,
+      "loss": 2.7435,
       "step": 326
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 19.1878604888916,
-      "learning_rate": 1.934649694296883e-05,
-      "loss": 3.7238,
+      "epoch": 0.04,
+      "grad_norm": 20.6904354095459,
+      "learning_rate": 1.9727230891519895e-05,
+      "loss": 5.3992,
       "step": 327
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 27.458173751831055,
-      "learning_rate": 1.9344492332364438e-05,
-      "loss": 3.6198,
+      "epoch": 0.04,
+      "grad_norm": 15.666752815246582,
+      "learning_rate": 1.9726394176463206e-05,
+      "loss": 3.7641,
       "step": 328
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 23.124168395996094,
-      "learning_rate": 1.934248772176005e-05,
-      "loss": 3.9671,
+      "epoch": 0.04,
+      "grad_norm": 43.591861724853516,
+      "learning_rate": 1.972555746140652e-05,
+      "loss": 2.9297,
       "step": 329
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 22.697099685668945,
-      "learning_rate": 1.934048311115566e-05,
-      "loss": 4.6046,
+      "epoch": 0.04,
+      "grad_norm": 23.864118576049805,
+      "learning_rate": 1.9724720746349833e-05,
+      "loss": 3.9387,
       "step": 330
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 16.66628074645996,
-      "learning_rate": 1.933847850055127e-05,
-      "loss": 3.4079,
+      "epoch": 0.04,
+      "grad_norm": 19.16819953918457,
+      "learning_rate": 1.9723884031293147e-05,
+      "loss": 3.6225,
       "step": 331
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 21.602243423461914,
-      "learning_rate": 1.933647388994688e-05,
-      "loss": 2.6947,
+      "epoch": 0.04,
+      "grad_norm": 25.3775634765625,
+      "learning_rate": 1.9723047316236457e-05,
+      "loss": 4.0582,
       "step": 332
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 19.987550735473633,
-      "learning_rate": 1.933446927934249e-05,
-      "loss": 4.0264,
+      "epoch": 0.04,
+      "grad_norm": 17.970884323120117,
+      "learning_rate": 1.972221060117977e-05,
+      "loss": 5.6782,
       "step": 333
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 27.548095703125,
-      "learning_rate": 1.93324646687381e-05,
-      "loss": 3.6234,
+      "epoch": 0.04,
+      "grad_norm": 18.719879150390625,
+      "learning_rate": 1.9721373886123084e-05,
+      "loss": 5.1692,
       "step": 334
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 16.985288619995117,
-      "learning_rate": 1.933046005813371e-05,
-      "loss": 3.5273,
+      "epoch": 0.04,
+      "grad_norm": 24.480466842651367,
+      "learning_rate": 1.9720537171066394e-05,
+      "loss": 4.19,
       "step": 335
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 249.91415405273438,
-      "learning_rate": 1.932845544752932e-05,
-      "loss": 3.0628,
+      "epoch": 0.04,
+      "grad_norm": 34.20682907104492,
+      "learning_rate": 1.9719700456009708e-05,
+      "loss": 1.6398,
       "step": 336
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 22.47792625427246,
-      "learning_rate": 1.932645083692493e-05,
-      "loss": 3.7402,
+      "epoch": 0.04,
+      "grad_norm": 21.29566192626953,
+      "learning_rate": 1.9718863740953022e-05,
+      "loss": 2.8134,
       "step": 337
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 28.15104866027832,
-      "learning_rate": 1.932444622632054e-05,
-      "loss": 3.5902,
+      "epoch": 0.04,
+      "grad_norm": 32.886497497558594,
+      "learning_rate": 1.9718027025896332e-05,
+      "loss": 5.3904,
       "step": 338
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 15.687856674194336,
-      "learning_rate": 1.932244161571615e-05,
-      "loss": 3.4816,
+      "epoch": 0.04,
+      "grad_norm": 20.40854835510254,
+      "learning_rate": 1.9717190310839646e-05,
+      "loss": 4.7706,
       "step": 339
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 15.60968017578125,
-      "learning_rate": 1.932043700511176e-05,
-      "loss": 3.8098,
+      "epoch": 0.04,
+      "grad_norm": 23.373363494873047,
+      "learning_rate": 1.971635359578296e-05,
+      "loss": 3.6134,
       "step": 340
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 21.058176040649414,
-      "learning_rate": 1.931843239450737e-05,
-      "loss": 4.0355,
+      "epoch": 0.04,
+      "grad_norm": 23.77679443359375,
+      "learning_rate": 1.971551688072627e-05,
+      "loss": 4.1803,
       "step": 341
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 52.26857376098633,
-      "learning_rate": 1.931642778390298e-05,
-      "loss": 5.9303,
+      "epoch": 0.04,
+      "grad_norm": 24.839336395263672,
+      "learning_rate": 1.9714680165669583e-05,
+      "loss": 3.7854,
       "step": 342
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 14.664571762084961,
-      "learning_rate": 1.931442317329859e-05,
-      "loss": 3.5722,
+      "epoch": 0.04,
+      "grad_norm": 13.832033157348633,
+      "learning_rate": 1.9713843450612893e-05,
+      "loss": 3.402,
       "step": 343
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 11.48156452178955,
-      "learning_rate": 1.9312418562694196e-05,
-      "loss": 3.8408,
+      "epoch": 0.04,
+      "grad_norm": 19.192140579223633,
+      "learning_rate": 1.9713006735556207e-05,
+      "loss": 2.4615,
       "step": 344
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 18.76517677307129,
-      "learning_rate": 1.931041395208981e-05,
-      "loss": 3.5406,
+      "epoch": 0.04,
+      "grad_norm": 11.908378601074219,
+      "learning_rate": 1.971217002049952e-05,
+      "loss": 3.0252,
       "step": 345
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 15.111738204956055,
-      "learning_rate": 1.930840934148542e-05,
-      "loss": 3.6611,
+      "epoch": 0.04,
+      "grad_norm": 49.56410598754883,
+      "learning_rate": 1.971133330544283e-05,
+      "loss": 2.499,
       "step": 346
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 11.203923225402832,
-      "learning_rate": 1.9306404730881026e-05,
-      "loss": 3.0199,
+      "epoch": 0.04,
+      "grad_norm": 19.634252548217773,
+      "learning_rate": 1.9710496590386145e-05,
+      "loss": 3.4309,
       "step": 347
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 21.51028060913086,
-      "learning_rate": 1.930440012027664e-05,
-      "loss": 3.1991,
+      "epoch": 0.04,
+      "grad_norm": 25.267864227294922,
+      "learning_rate": 1.970965987532946e-05,
+      "loss": 3.8254,
       "step": 348
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 21.56675910949707,
-      "learning_rate": 1.9302395509672247e-05,
-      "loss": 3.6963,
+      "epoch": 0.04,
+      "grad_norm": 27.672697067260742,
+      "learning_rate": 1.970882316027277e-05,
+      "loss": 4.0687,
       "step": 349
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 13.743776321411133,
-      "learning_rate": 1.9300390899067857e-05,
-      "loss": 3.6142,
+      "epoch": 0.04,
+      "grad_norm": 29.05243492126465,
+      "learning_rate": 1.9707986445216082e-05,
+      "loss": 2.8896,
       "step": 350
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 15.742630004882812,
-      "learning_rate": 1.9298386288463467e-05,
-      "loss": 2.8555,
+      "epoch": 0.04,
+      "grad_norm": 13.088189125061035,
+      "learning_rate": 1.9707149730159396e-05,
+      "loss": 2.2714,
       "step": 351
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 14.802862167358398,
-      "learning_rate": 1.9296381677859077e-05,
-      "loss": 2.9698,
+      "epoch": 0.04,
+      "grad_norm": 23.043155670166016,
+      "learning_rate": 1.970631301510271e-05,
+      "loss": 4.1696,
       "step": 352
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 16.21837615966797,
-      "learning_rate": 1.9294377067254687e-05,
-      "loss": 5.3529,
+      "epoch": 0.04,
+      "grad_norm": 47.78111267089844,
+      "learning_rate": 1.970547630004602e-05,
+      "loss": 4.2777,
       "step": 353
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 14.180800437927246,
-      "learning_rate": 1.9292372456650297e-05,
-      "loss": 3.7083,
+      "epoch": 0.04,
+      "grad_norm": 28.614927291870117,
+      "learning_rate": 1.9704639584989333e-05,
+      "loss": 4.6397,
       "step": 354
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 16.493022918701172,
-      "learning_rate": 1.9290367846045907e-05,
-      "loss": 2.6302,
+      "epoch": 0.04,
+      "grad_norm": 25.07124137878418,
+      "learning_rate": 1.9703802869932647e-05,
+      "loss": 3.9658,
       "step": 355
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 19.73296356201172,
-      "learning_rate": 1.9288363235441517e-05,
-      "loss": 3.4556,
+      "epoch": 0.04,
+      "grad_norm": 18.310688018798828,
+      "learning_rate": 1.9702966154875957e-05,
+      "loss": 3.4839,
       "step": 356
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 20.31797218322754,
-      "learning_rate": 1.9286358624837127e-05,
-      "loss": 4.2031,
+      "epoch": 0.04,
+      "grad_norm": 21.59184455871582,
+      "learning_rate": 1.970212943981927e-05,
+      "loss": 2.9543,
       "step": 357
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 24.127195358276367,
-      "learning_rate": 1.9284354014232737e-05,
-      "loss": 4.2493,
+      "epoch": 0.04,
+      "grad_norm": 25.248348236083984,
+      "learning_rate": 1.9701292724762585e-05,
+      "loss": 3.7935,
       "step": 358
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 33.403282165527344,
-      "learning_rate": 1.9282349403628347e-05,
-      "loss": 4.6944,
+      "epoch": 0.05,
+      "grad_norm": 18.58963394165039,
+      "learning_rate": 1.9700456009705898e-05,
+      "loss": 3.1486,
       "step": 359
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 52.27415466308594,
-      "learning_rate": 1.9280344793023958e-05,
-      "loss": 3.5661,
-      "step": 360
-    },
-    {
-      "epoch": 0.11,
-      "eval_loss": 1.2527186870574951,
-      "eval_runtime": 43.7248,
-      "eval_samples_per_second": 33.825,
-      "eval_steps_per_second": 33.825,
+      "epoch": 0.05,
+      "grad_norm": 14.266563415527344,
+      "learning_rate": 1.969961929464921e-05,
+      "loss": 3.7676,
       "step": 360
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 18.096208572387695,
-      "learning_rate": 1.9278340182419568e-05,
-      "loss": 4.3089,
+      "epoch": 0.05,
+      "grad_norm": 18.322986602783203,
+      "learning_rate": 1.9698782579592522e-05,
+      "loss": 3.0848,
       "step": 361
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 12.513582229614258,
-      "learning_rate": 1.9276335571815178e-05,
-      "loss": 3.7448,
+      "epoch": 0.05,
+      "grad_norm": 14.636101722717285,
+      "learning_rate": 1.9697945864535836e-05,
+      "loss": 4.5866,
       "step": 362
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 18.849184036254883,
-      "learning_rate": 1.9274330961210784e-05,
-      "loss": 4.8774,
+      "epoch": 0.05,
+      "grad_norm": 41.74934005737305,
+      "learning_rate": 1.9697109149479146e-05,
+      "loss": 2.5856,
       "step": 363
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 14.817676544189453,
-      "learning_rate": 1.9272326350606398e-05,
-      "loss": 4.2751,
+      "epoch": 0.05,
+      "grad_norm": 14.788008689880371,
+      "learning_rate": 1.969627243442246e-05,
+      "loss": 3.0474,
       "step": 364
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 24.063844680786133,
-      "learning_rate": 1.9270321740002008e-05,
-      "loss": 3.8558,
+      "epoch": 0.05,
+      "grad_norm": 18.032474517822266,
+      "learning_rate": 1.9695435719365773e-05,
+      "loss": 6.5002,
       "step": 365
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 18.589746475219727,
-      "learning_rate": 1.9268317129397615e-05,
-      "loss": 2.9172,
+      "epoch": 0.05,
+      "grad_norm": 22.44284439086914,
+      "learning_rate": 1.9694599004309084e-05,
+      "loss": 2.2314,
       "step": 366
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 51.5893669128418,
-      "learning_rate": 1.9266312518793228e-05,
-      "loss": 3.2441,
+      "epoch": 0.05,
+      "grad_norm": 13.548026084899902,
+      "learning_rate": 1.9693762289252397e-05,
+      "loss": 3.7871,
       "step": 367
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 18.138118743896484,
-      "learning_rate": 1.9264307908188835e-05,
-      "loss": 2.9894,
+      "epoch": 0.05,
+      "grad_norm": 24.17680549621582,
+      "learning_rate": 1.969292557419571e-05,
+      "loss": 4.6725,
       "step": 368
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 14.857394218444824,
-      "learning_rate": 1.9262303297584445e-05,
-      "loss": 3.1436,
+      "epoch": 0.05,
+      "grad_norm": 22.945974349975586,
+      "learning_rate": 1.969208885913902e-05,
+      "loss": 3.5503,
       "step": 369
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 17.468170166015625,
-      "learning_rate": 1.926029868698006e-05,
-      "loss": 3.2683,
+      "epoch": 0.05,
+      "grad_norm": 14.245816230773926,
+      "learning_rate": 1.9691252144082335e-05,
+      "loss": 4.1735,
       "step": 370
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 17.41706657409668,
-      "learning_rate": 1.9258294076375665e-05,
-      "loss": 3.6597,
+      "epoch": 0.05,
+      "grad_norm": 16.24998664855957,
+      "learning_rate": 1.9690415429025645e-05,
+      "loss": 1.5692,
       "step": 371
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 15.607809066772461,
-      "learning_rate": 1.9256289465771275e-05,
-      "loss": 3.5317,
+      "epoch": 0.05,
+      "grad_norm": 18.322080612182617,
+      "learning_rate": 1.968957871396896e-05,
+      "loss": 3.6277,
       "step": 372
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 25.48333168029785,
-      "learning_rate": 1.9254284855166885e-05,
-      "loss": 4.017,
+      "epoch": 0.05,
+      "grad_norm": 19.254554748535156,
+      "learning_rate": 1.9688741998912272e-05,
+      "loss": 3.3539,
       "step": 373
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 26.36655044555664,
-      "learning_rate": 1.9252280244562495e-05,
-      "loss": 3.1871,
+      "epoch": 0.05,
+      "grad_norm": 49.72288131713867,
+      "learning_rate": 1.9687905283855583e-05,
+      "loss": 3.7778,
       "step": 374
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 29.703283309936523,
-      "learning_rate": 1.9250275633958105e-05,
-      "loss": 3.56,
+      "epoch": 0.05,
+      "grad_norm": 18.277828216552734,
+      "learning_rate": 1.9687068568798896e-05,
+      "loss": 3.7181,
       "step": 375
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 12.352496147155762,
-      "learning_rate": 1.9248271023353715e-05,
-      "loss": 3.2961,
+      "epoch": 0.05,
+      "grad_norm": 10.964229583740234,
+      "learning_rate": 1.968623185374221e-05,
+      "loss": 2.8094,
       "step": 376
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 24.971912384033203,
-      "learning_rate": 1.9246266412749326e-05,
-      "loss": 3.8872,
+      "epoch": 0.05,
+      "grad_norm": 18.096059799194336,
+      "learning_rate": 1.968539513868552e-05,
+      "loss": 3.2346,
       "step": 377
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 20.017274856567383,
-      "learning_rate": 1.9244261802144936e-05,
-      "loss": 4.3486,
+      "epoch": 0.05,
+      "grad_norm": 14.133166313171387,
+      "learning_rate": 1.9684558423628834e-05,
+      "loss": 3.6737,
       "step": 378
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 16.611631393432617,
-      "learning_rate": 1.9242257191540546e-05,
-      "loss": 4.1083,
+      "epoch": 0.05,
+      "grad_norm": 13.820868492126465,
+      "learning_rate": 1.9683721708572148e-05,
+      "loss": 2.8704,
       "step": 379
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 94.53077697753906,
-      "learning_rate": 1.9240252580936152e-05,
-      "loss": 3.7256,
+      "epoch": 0.05,
+      "grad_norm": 28.055957794189453,
+      "learning_rate": 1.968288499351546e-05,
+      "loss": 4.1075,
       "step": 380
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 14.142595291137695,
-      "learning_rate": 1.9238247970331766e-05,
-      "loss": 3.9881,
+      "epoch": 0.05,
+      "grad_norm": 18.781705856323242,
+      "learning_rate": 1.968204827845877e-05,
+      "loss": 3.0917,
       "step": 381
     },
     {
-      "epoch": 0.11,
-      "grad_norm": 15.951706886291504,
-      "learning_rate": 1.9236243359727373e-05,
-      "loss": 3.6539,
+      "epoch": 0.05,
+      "grad_norm": 14.126927375793457,
+      "learning_rate": 1.9681211563402085e-05,
+      "loss": 3.5304,
       "step": 382
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 16.317209243774414,
-      "learning_rate": 1.9234238749122983e-05,
-      "loss": 4.028,
+      "epoch": 0.05,
+      "grad_norm": 13.951945304870605,
+      "learning_rate": 1.96803748483454e-05,
+      "loss": 4.9655,
       "step": 383
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 15.425968170166016,
-      "learning_rate": 1.9232234138518596e-05,
-      "loss": 2.9222,
+      "epoch": 0.05,
+      "grad_norm": 21.58669090270996,
+      "learning_rate": 1.967953813328871e-05,
+      "loss": 2.5144,
       "step": 384
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 15.79262924194336,
-      "learning_rate": 1.9230229527914203e-05,
-      "loss": 3.1179,
+      "epoch": 0.05,
+      "grad_norm": 9.454935073852539,
+      "learning_rate": 1.9678701418232023e-05,
+      "loss": 4.0318,
       "step": 385
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 19.95915985107422,
-      "learning_rate": 1.9228224917309813e-05,
-      "loss": 3.9919,
+      "epoch": 0.05,
+      "grad_norm": 22.182817459106445,
+      "learning_rate": 1.9677864703175336e-05,
+      "loss": 2.4768,
       "step": 386
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 14.439689636230469,
-      "learning_rate": 1.9226220306705423e-05,
-      "loss": 3.257,
+      "epoch": 0.05,
+      "grad_norm": 30.83213996887207,
+      "learning_rate": 1.967702798811865e-05,
+      "loss": 5.7192,
       "step": 387
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 16.699209213256836,
-      "learning_rate": 1.9224215696101033e-05,
-      "loss": 3.8755,
+      "epoch": 0.05,
+      "grad_norm": 36.244850158691406,
+      "learning_rate": 1.967619127306196e-05,
+      "loss": 4.4631,
       "step": 388
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 24.074472427368164,
-      "learning_rate": 1.9222211085496643e-05,
-      "loss": 4.0977,
+      "epoch": 0.05,
+      "grad_norm": 18.907621383666992,
+      "learning_rate": 1.9675354558005274e-05,
+      "loss": 3.8661,
       "step": 389
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 15.57771110534668,
-      "learning_rate": 1.9220206474892253e-05,
-      "loss": 4.4361,
+      "epoch": 0.05,
+      "grad_norm": 22.727462768554688,
+      "learning_rate": 1.9674517842948588e-05,
+      "loss": 5.2718,
       "step": 390
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 13.52245807647705,
-      "learning_rate": 1.9218201864287863e-05,
-      "loss": 3.7686,
+      "epoch": 0.05,
+      "grad_norm": 23.575908660888672,
+      "learning_rate": 1.9673681127891898e-05,
+      "loss": 6.1416,
       "step": 391
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 19.746614456176758,
-      "learning_rate": 1.9216197253683473e-05,
-      "loss": 3.819,
+      "epoch": 0.05,
+      "grad_norm": 28.313125610351562,
+      "learning_rate": 1.967284441283521e-05,
+      "loss": 4.5796,
       "step": 392
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 14.887005805969238,
-      "learning_rate": 1.9214192643079084e-05,
-      "loss": 3.8956,
+      "epoch": 0.05,
+      "grad_norm": 16.64825439453125,
+      "learning_rate": 1.9672007697778525e-05,
+      "loss": 1.9543,
       "step": 393
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 18.658023834228516,
-      "learning_rate": 1.9212188032474694e-05,
-      "loss": 3.9219,
+      "epoch": 0.05,
+      "grad_norm": 28.147789001464844,
+      "learning_rate": 1.9671170982721835e-05,
+      "loss": 2.8313,
       "step": 394
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 14.493846893310547,
-      "learning_rate": 1.9210183421870304e-05,
-      "loss": 4.26,
+      "epoch": 0.05,
+      "grad_norm": 27.15233039855957,
+      "learning_rate": 1.967033426766515e-05,
+      "loss": 3.023,
       "step": 395
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 16.01069450378418,
-      "learning_rate": 1.9208178811265914e-05,
-      "loss": 2.9605,
+      "epoch": 0.05,
+      "grad_norm": 16.548398971557617,
+      "learning_rate": 1.966949755260846e-05,
+      "loss": 2.378,
       "step": 396
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 24.826866149902344,
-      "learning_rate": 1.9206174200661524e-05,
-      "loss": 4.3274,
+      "epoch": 0.05,
+      "grad_norm": 23.343488693237305,
+      "learning_rate": 1.9668660837551773e-05,
+      "loss": 5.2585,
       "step": 397
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 10.156407356262207,
-      "learning_rate": 1.9204169590057134e-05,
-      "loss": 3.2431,
+      "epoch": 0.05,
+      "grad_norm": 15.054876327514648,
+      "learning_rate": 1.9667824122495087e-05,
+      "loss": 4.0137,
       "step": 398
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 21.510562896728516,
-      "learning_rate": 1.920216497945274e-05,
-      "loss": 4.4685,
+      "epoch": 0.05,
+      "grad_norm": 27.556150436401367,
+      "learning_rate": 1.9666987407438397e-05,
+      "loss": 5.8062,
       "step": 399
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 25.705717086791992,
-      "learning_rate": 1.9200160368848354e-05,
-      "loss": 4.3723,
+      "epoch": 0.05,
+      "grad_norm": 20.49494171142578,
+      "learning_rate": 1.966615069238171e-05,
+      "loss": 5.7858,
       "step": 400
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 15.125773429870605,
-      "learning_rate": 1.9198155758243964e-05,
-      "loss": 3.9991,
+      "epoch": 0.05,
+      "eval_loss": 0.7426895499229431,
+      "eval_runtime": 268.2339,
+      "eval_samples_per_second": 13.205,
+      "eval_steps_per_second": 13.205,
+      "step": 400
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.434072494506836,
+      "learning_rate": 1.9665313977325024e-05,
+      "loss": 4.5162,
       "step": 401
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 11.93122386932373,
-      "learning_rate": 1.919615114763957e-05,
-      "loss": 3.5661,
+      "epoch": 0.05,
+      "grad_norm": 22.502578735351562,
+      "learning_rate": 1.9664477262268334e-05,
+      "loss": 4.2813,
       "step": 402
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 9.818219184875488,
-      "learning_rate": 1.9194146537035184e-05,
-      "loss": 3.4324,
+      "epoch": 0.05,
+      "grad_norm": 17.459829330444336,
+      "learning_rate": 1.9663640547211648e-05,
+      "loss": 2.355,
       "step": 403
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 23.427623748779297,
-      "learning_rate": 1.919214192643079e-05,
-      "loss": 4.2537,
+      "epoch": 0.05,
+      "grad_norm": 21.548038482666016,
+      "learning_rate": 1.966280383215496e-05,
+      "loss": 6.0367,
       "step": 404
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 18.164058685302734,
-      "learning_rate": 1.91901373158264e-05,
-      "loss": 3.2003,
+      "epoch": 0.05,
+      "grad_norm": 16.913127899169922,
+      "learning_rate": 1.9661967117098272e-05,
+      "loss": 3.0731,
       "step": 405
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 20.42031478881836,
-      "learning_rate": 1.918813270522201e-05,
-      "loss": 2.4072,
+      "epoch": 0.05,
+      "grad_norm": 23.094449996948242,
+      "learning_rate": 1.9661130402041586e-05,
+      "loss": 3.1689,
       "step": 406
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 21.685937881469727,
-      "learning_rate": 1.918612809461762e-05,
-      "loss": 3.5641,
+      "epoch": 0.05,
+      "grad_norm": 19.218231201171875,
+      "learning_rate": 1.96602936869849e-05,
+      "loss": 2.5354,
       "step": 407
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 19.072772979736328,
-      "learning_rate": 1.918412348401323e-05,
-      "loss": 3.8314,
+      "epoch": 0.05,
+      "grad_norm": 14.398783683776855,
+      "learning_rate": 1.9659456971928213e-05,
+      "loss": 2.5509,
       "step": 408
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 11.509551048278809,
-      "learning_rate": 1.918211887340884e-05,
-      "loss": 2.837,
+      "epoch": 0.05,
+      "grad_norm": 28.161535263061523,
+      "learning_rate": 1.9658620256871523e-05,
+      "loss": 4.797,
       "step": 409
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 14.759239196777344,
-      "learning_rate": 1.918011426280445e-05,
-      "loss": 3.501,
+      "epoch": 0.05,
+      "grad_norm": 17.712581634521484,
+      "learning_rate": 1.9657783541814837e-05,
+      "loss": 3.4253,
       "step": 410
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 25.07742691040039,
-      "learning_rate": 1.917810965220006e-05,
-      "loss": 3.0863,
+      "epoch": 0.05,
+      "grad_norm": 27.91653823852539,
+      "learning_rate": 1.965694682675815e-05,
+      "loss": 4.9134,
       "step": 411
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 17.365577697753906,
-      "learning_rate": 1.9176105041595672e-05,
-      "loss": 3.6191,
+      "epoch": 0.05,
+      "grad_norm": 19.58594512939453,
+      "learning_rate": 1.965611011170146e-05,
+      "loss": 2.3899,
       "step": 412
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 15.148959159851074,
-      "learning_rate": 1.9174100430991282e-05,
-      "loss": 3.6065,
+      "epoch": 0.05,
+      "grad_norm": 37.195430755615234,
+      "learning_rate": 1.9655273396644774e-05,
+      "loss": 4.6033,
       "step": 413
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 22.2659854888916,
-      "learning_rate": 1.9172095820386892e-05,
-      "loss": 3.7687,
+      "epoch": 0.05,
+      "grad_norm": 21.229907989501953,
+      "learning_rate": 1.9654436681588088e-05,
+      "loss": 2.5836,
       "step": 414
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 10.9700927734375,
-      "learning_rate": 1.9170091209782502e-05,
-      "loss": 3.6995,
+      "epoch": 0.05,
+      "grad_norm": 13.44748592376709,
+      "learning_rate": 1.96535999665314e-05,
+      "loss": 4.605,
       "step": 415
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 16.873733520507812,
-      "learning_rate": 1.9168086599178112e-05,
-      "loss": 2.3494,
+      "epoch": 0.05,
+      "grad_norm": 16.857240676879883,
+      "learning_rate": 1.9652763251474712e-05,
+      "loss": 2.4381,
       "step": 416
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 22.604717254638672,
-      "learning_rate": 1.9166081988573722e-05,
-      "loss": 3.417,
+      "epoch": 0.05,
+      "grad_norm": 17.680007934570312,
+      "learning_rate": 1.9651926536418026e-05,
+      "loss": 2.3925,
       "step": 417
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 16.342870712280273,
-      "learning_rate": 1.916407737796933e-05,
-      "loss": 3.7907,
+      "epoch": 0.05,
+      "grad_norm": 17.41832733154297,
+      "learning_rate": 1.965108982136134e-05,
+      "loss": 5.1594,
       "step": 418
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 43.08720016479492,
-      "learning_rate": 1.9162072767364942e-05,
-      "loss": 2.4359,
+      "epoch": 0.05,
+      "grad_norm": 10.49217414855957,
+      "learning_rate": 1.965025310630465e-05,
+      "loss": 3.4874,
       "step": 419
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 18.63433074951172,
-      "learning_rate": 1.9160068156760552e-05,
-      "loss": 4.5936,
+      "epoch": 0.05,
+      "grad_norm": 18.354557037353516,
+      "learning_rate": 1.9649416391247963e-05,
+      "loss": 3.7841,
       "step": 420
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 20.13007164001465,
-      "learning_rate": 1.915806354615616e-05,
-      "loss": 4.5972,
+      "epoch": 0.05,
+      "grad_norm": 14.493927955627441,
+      "learning_rate": 1.9648579676191273e-05,
+      "loss": 2.6667,
       "step": 421
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 17.850637435913086,
-      "learning_rate": 1.9156058935551773e-05,
-      "loss": 3.1744,
+      "epoch": 0.05,
+      "grad_norm": 12.498456954956055,
+      "learning_rate": 1.9647742961134587e-05,
+      "loss": 4.7955,
       "step": 422
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 15.922115325927734,
-      "learning_rate": 1.915405432494738e-05,
-      "loss": 3.862,
+      "epoch": 0.05,
+      "grad_norm": 29.352039337158203,
+      "learning_rate": 1.96469062460779e-05,
+      "loss": 2.5031,
       "step": 423
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 20.92375946044922,
-      "learning_rate": 1.915204971434299e-05,
-      "loss": 2.907,
+      "epoch": 0.05,
+      "grad_norm": 30.123577117919922,
+      "learning_rate": 1.964606953102121e-05,
+      "loss": 3.5351,
       "step": 424
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 18.767166137695312,
-      "learning_rate": 1.91500451037386e-05,
-      "loss": 4.3281,
+      "epoch": 0.05,
+      "grad_norm": 13.854634284973145,
+      "learning_rate": 1.9645232815964525e-05,
+      "loss": 4.9223,
       "step": 425
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 16.2836856842041,
-      "learning_rate": 1.914804049313421e-05,
-      "loss": 3.5196,
+      "epoch": 0.05,
+      "grad_norm": 13.630191802978516,
+      "learning_rate": 1.9644396100907838e-05,
+      "loss": 4.2064,
       "step": 426
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 13.612652778625488,
-      "learning_rate": 1.914603588252982e-05,
-      "loss": 4.0671,
+      "epoch": 0.05,
+      "grad_norm": 17.01755714416504,
+      "learning_rate": 1.964355938585115e-05,
+      "loss": 2.1371,
       "step": 427
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 20.497167587280273,
-      "learning_rate": 1.914403127192543e-05,
-      "loss": 4.5793,
+      "epoch": 0.05,
+      "grad_norm": 16.535648345947266,
+      "learning_rate": 1.9642722670794462e-05,
+      "loss": 3.343,
       "step": 428
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 16.230653762817383,
-      "learning_rate": 1.914202666132104e-05,
-      "loss": 4.3463,
+      "epoch": 0.05,
+      "grad_norm": 17.35195541381836,
+      "learning_rate": 1.9641885955737776e-05,
+      "loss": 3.838,
       "step": 429
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 17.90238380432129,
-      "learning_rate": 1.914002205071665e-05,
-      "loss": 3.6869,
+      "epoch": 0.05,
+      "grad_norm": 12.293807983398438,
+      "learning_rate": 1.9641049240681086e-05,
+      "loss": 3.1724,
       "step": 430
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 23.716754913330078,
-      "learning_rate": 1.913801744011226e-05,
-      "loss": 3.8698,
+      "epoch": 0.05,
+      "grad_norm": 66.72679138183594,
+      "learning_rate": 1.96402125256244e-05,
+      "loss": 4.5378,
       "step": 431
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 17.085433959960938,
-      "learning_rate": 1.913601282950787e-05,
-      "loss": 2.4927,
+      "epoch": 0.05,
+      "grad_norm": 24.131763458251953,
+      "learning_rate": 1.9639375810567713e-05,
+      "loss": 3.6992,
       "step": 432
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 19.192747116088867,
-      "learning_rate": 1.913400821890348e-05,
-      "loss": 3.9132,
+      "epoch": 0.05,
+      "grad_norm": 17.67955780029297,
+      "learning_rate": 1.9638539095511024e-05,
+      "loss": 2.1648,
       "step": 433
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 13.243489265441895,
-      "learning_rate": 1.913200360829909e-05,
-      "loss": 2.9527,
+      "epoch": 0.05,
+      "grad_norm": 14.534732818603516,
+      "learning_rate": 1.9637702380454337e-05,
+      "loss": 3.1824,
       "step": 434
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 15.914332389831543,
-      "learning_rate": 1.91299989976947e-05,
-      "loss": 3.9814,
+      "epoch": 0.05,
+      "grad_norm": 24.638200759887695,
+      "learning_rate": 1.963686566539765e-05,
+      "loss": 3.8397,
       "step": 435
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 16.82046127319336,
-      "learning_rate": 1.912799438709031e-05,
-      "loss": 3.2976,
+      "epoch": 0.05,
+      "grad_norm": 11.531396865844727,
+      "learning_rate": 1.9636028950340965e-05,
+      "loss": 3.1445,
       "step": 436
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 30.610519409179688,
-      "learning_rate": 1.9125989776485917e-05,
-      "loss": 2.8746,
+      "epoch": 0.05,
+      "grad_norm": 31.642684936523438,
+      "learning_rate": 1.9635192235284275e-05,
+      "loss": 4.803,
       "step": 437
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 14.973118782043457,
-      "learning_rate": 1.912398516588153e-05,
-      "loss": 3.4231,
+      "epoch": 0.05,
+      "grad_norm": 12.691168785095215,
+      "learning_rate": 1.963435552022759e-05,
+      "loss": 3.1471,
       "step": 438
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 17.690595626831055,
-      "learning_rate": 1.912198055527714e-05,
-      "loss": 2.8452,
+      "epoch": 0.06,
+      "grad_norm": 17.246782302856445,
+      "learning_rate": 1.9633518805170902e-05,
+      "loss": 3.2402,
       "step": 439
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 13.494382858276367,
-      "learning_rate": 1.9119975944672747e-05,
-      "loss": 3.4656,
+      "epoch": 0.06,
+      "grad_norm": 22.924819946289062,
+      "learning_rate": 1.9632682090114212e-05,
+      "loss": 3.015,
       "step": 440
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 15.88420581817627,
-      "learning_rate": 1.911797133406836e-05,
-      "loss": 3.4494,
+      "epoch": 0.06,
+      "grad_norm": 13.039619445800781,
+      "learning_rate": 1.9631845375057526e-05,
+      "loss": 3.6388,
       "step": 441
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 15.013689041137695,
-      "learning_rate": 1.9115966723463967e-05,
-      "loss": 3.7962,
+      "epoch": 0.06,
+      "grad_norm": 15.010822296142578,
+      "learning_rate": 1.963100866000084e-05,
+      "loss": 6.2867,
       "step": 442
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 19.49163246154785,
-      "learning_rate": 1.9113962112859578e-05,
-      "loss": 3.5861,
+      "epoch": 0.06,
+      "grad_norm": 14.180012702941895,
+      "learning_rate": 1.9630171944944153e-05,
+      "loss": 5.0498,
       "step": 443
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 14.526626586914062,
-      "learning_rate": 1.911195750225519e-05,
-      "loss": 4.3179,
+      "epoch": 0.06,
+      "grad_norm": 14.439057350158691,
+      "learning_rate": 1.9629335229887464e-05,
+      "loss": 5.7437,
       "step": 444
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 14.432934761047363,
-      "learning_rate": 1.9109952891650798e-05,
-      "loss": 3.6713,
+      "epoch": 0.06,
+      "grad_norm": 25.79570198059082,
+      "learning_rate": 1.9628498514830777e-05,
+      "loss": 3.5513,
       "step": 445
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 17.720027923583984,
-      "learning_rate": 1.9107948281046408e-05,
-      "loss": 2.9935,
+      "epoch": 0.06,
+      "grad_norm": 13.489237785339355,
+      "learning_rate": 1.962766179977409e-05,
+      "loss": 4.9429,
       "step": 446
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 30.397794723510742,
-      "learning_rate": 1.9105943670442018e-05,
-      "loss": 3.4623,
+      "epoch": 0.06,
+      "grad_norm": 18.599224090576172,
+      "learning_rate": 1.96268250847174e-05,
+      "loss": 3.3341,
       "step": 447
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 14.584136009216309,
-      "learning_rate": 1.9103939059837628e-05,
-      "loss": 3.4608,
+      "epoch": 0.06,
+      "grad_norm": 11.782310485839844,
+      "learning_rate": 1.9625988369660715e-05,
+      "loss": 2.9087,
       "step": 448
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 10.301027297973633,
-      "learning_rate": 1.9101934449233238e-05,
-      "loss": 3.1884,
+      "epoch": 0.06,
+      "grad_norm": 12.502921104431152,
+      "learning_rate": 1.9625151654604025e-05,
+      "loss": 2.8944,
       "step": 449
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 17.015613555908203,
-      "learning_rate": 1.9099929838628848e-05,
-      "loss": 3.9679,
+      "epoch": 0.06,
+      "grad_norm": 14.942875862121582,
+      "learning_rate": 1.962431493954734e-05,
+      "loss": 3.2534,
       "step": 450
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 13.941701889038086,
-      "learning_rate": 1.9097925228024458e-05,
-      "loss": 3.3073,
+      "epoch": 0.06,
+      "grad_norm": 12.024401664733887,
+      "learning_rate": 1.9623478224490652e-05,
+      "loss": 5.8004,
       "step": 451
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 16.194982528686523,
-      "learning_rate": 1.909592061742007e-05,
-      "loss": 3.6541,
+      "epoch": 0.06,
+      "grad_norm": 14.10036563873291,
+      "learning_rate": 1.9622641509433963e-05,
+      "loss": 3.5902,
       "step": 452
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 19.734464645385742,
-      "learning_rate": 1.909391600681568e-05,
-      "loss": 4.4298,
+      "epoch": 0.06,
+      "grad_norm": 21.69317054748535,
+      "learning_rate": 1.9621804794377276e-05,
+      "loss": 6.0422,
       "step": 453
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 12.222415924072266,
-      "learning_rate": 1.9091911396211285e-05,
-      "loss": 2.5833,
+      "epoch": 0.06,
+      "grad_norm": 18.428606033325195,
+      "learning_rate": 1.9620968079320587e-05,
+      "loss": 3.6164,
       "step": 454
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 21.09163475036621,
-      "learning_rate": 1.90899067856069e-05,
-      "loss": 3.8562,
+      "epoch": 0.06,
+      "grad_norm": 15.561418533325195,
+      "learning_rate": 1.96201313642639e-05,
+      "loss": 3.0053,
       "step": 455
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 14.021400451660156,
-      "learning_rate": 1.908790217500251e-05,
-      "loss": 3.9132,
+      "epoch": 0.06,
+      "grad_norm": 16.941303253173828,
+      "learning_rate": 1.9619294649207214e-05,
+      "loss": 3.5116,
       "step": 456
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 19.492759704589844,
-      "learning_rate": 1.9085897564398115e-05,
-      "loss": 4.7376,
+      "epoch": 0.06,
+      "grad_norm": 16.592493057250977,
+      "learning_rate": 1.9618457934150527e-05,
+      "loss": 4.3639,
       "step": 457
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 13.180451393127441,
-      "learning_rate": 1.908389295379373e-05,
-      "loss": 2.8013,
+      "epoch": 0.06,
+      "grad_norm": 16.290321350097656,
+      "learning_rate": 1.9617621219093838e-05,
+      "loss": 1.1296,
       "step": 458
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 19.046419143676758,
-      "learning_rate": 1.9081888343189336e-05,
-      "loss": 3.2353,
+      "epoch": 0.06,
+      "grad_norm": 36.56865692138672,
+      "learning_rate": 1.961678450403715e-05,
+      "loss": 4.6438,
       "step": 459
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 24.7363338470459,
-      "learning_rate": 1.9079883732584946e-05,
-      "loss": 4.2447,
+      "epoch": 0.06,
+      "grad_norm": 20.538652420043945,
+      "learning_rate": 1.9615947788980465e-05,
+      "loss": 3.0496,
       "step": 460
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 12.855481147766113,
-      "learning_rate": 1.9077879121980556e-05,
-      "loss": 3.3459,
+      "epoch": 0.06,
+      "grad_norm": 14.338871955871582,
+      "learning_rate": 1.9615111073923775e-05,
+      "loss": 4.9444,
       "step": 461
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 13.02658462524414,
-      "learning_rate": 1.9075874511376166e-05,
-      "loss": 4.0793,
+      "epoch": 0.06,
+      "grad_norm": 20.616947174072266,
+      "learning_rate": 1.961427435886709e-05,
+      "loss": 2.8608,
       "step": 462
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 17.221975326538086,
-      "learning_rate": 1.907386990077178e-05,
-      "loss": 3.3399,
+      "epoch": 0.06,
+      "grad_norm": 22.124589920043945,
+      "learning_rate": 1.9613437643810403e-05,
+      "loss": 2.2952,
       "step": 463
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 22.799137115478516,
-      "learning_rate": 1.9071865290167386e-05,
-      "loss": 3.7892,
+      "epoch": 0.06,
+      "grad_norm": 16.3970947265625,
+      "learning_rate": 1.9612600928753716e-05,
+      "loss": 4.7596,
       "step": 464
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 21.327260971069336,
-      "learning_rate": 1.9069860679562996e-05,
-      "loss": 2.5584,
+      "epoch": 0.06,
+      "grad_norm": 15.964753150939941,
+      "learning_rate": 1.9611764213697026e-05,
+      "loss": 3.176,
       "step": 465
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 26.7298583984375,
-      "learning_rate": 1.9067856068958606e-05,
-      "loss": 4.5134,
+      "epoch": 0.06,
+      "grad_norm": 16.239675521850586,
+      "learning_rate": 1.961092749864034e-05,
+      "loss": 4.4223,
       "step": 466
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 25.994600296020508,
-      "learning_rate": 1.9065851458354216e-05,
-      "loss": 2.8459,
+      "epoch": 0.06,
+      "grad_norm": 17.314661026000977,
+      "learning_rate": 1.9610090783583654e-05,
+      "loss": 3.0885,
       "step": 467
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 22.67238998413086,
-      "learning_rate": 1.9063846847749826e-05,
-      "loss": 4.0606,
+      "epoch": 0.06,
+      "grad_norm": 9.08109188079834,
+      "learning_rate": 1.9609254068526964e-05,
+      "loss": 3.6177,
       "step": 468
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 16.200130462646484,
-      "learning_rate": 1.9061842237145436e-05,
-      "loss": 4.0715,
+      "epoch": 0.06,
+      "grad_norm": 13.201598167419434,
+      "learning_rate": 1.9608417353470278e-05,
+      "loss": 5.4196,
       "step": 469
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 12.668680191040039,
-      "learning_rate": 1.9059837626541046e-05,
-      "loss": 2.9427,
+      "epoch": 0.06,
+      "grad_norm": 12.478797912597656,
+      "learning_rate": 1.960758063841359e-05,
+      "loss": 2.9956,
       "step": 470
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 24.67599868774414,
-      "learning_rate": 1.9057833015936657e-05,
-      "loss": 3.3925,
+      "epoch": 0.06,
+      "grad_norm": 16.85390281677246,
+      "learning_rate": 1.9606743923356905e-05,
+      "loss": 3.9526,
       "step": 471
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 26.510074615478516,
-      "learning_rate": 1.9055828405332267e-05,
-      "loss": 2.774,
+      "epoch": 0.06,
+      "grad_norm": 14.20464038848877,
+      "learning_rate": 1.9605907208300215e-05,
+      "loss": 4.3399,
       "step": 472
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 19.766721725463867,
-      "learning_rate": 1.9053823794727873e-05,
-      "loss": 3.1132,
+      "epoch": 0.06,
+      "grad_norm": 20.508764266967773,
+      "learning_rate": 1.960507049324353e-05,
+      "loss": 4.9114,
       "step": 473
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 14.419114112854004,
-      "learning_rate": 1.9051819184123487e-05,
-      "loss": 2.8139,
+      "epoch": 0.06,
+      "grad_norm": 20.2303409576416,
+      "learning_rate": 1.960423377818684e-05,
+      "loss": 3.7521,
       "step": 474
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 11.945268630981445,
-      "learning_rate": 1.9049814573519097e-05,
-      "loss": 3.4746,
+      "epoch": 0.06,
+      "grad_norm": 14.391060829162598,
+      "learning_rate": 1.9603397063130153e-05,
+      "loss": 5.0591,
       "step": 475
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 19.036060333251953,
-      "learning_rate": 1.9047809962914704e-05,
-      "loss": 4.0888,
+      "epoch": 0.06,
+      "grad_norm": 23.435667037963867,
+      "learning_rate": 1.9602560348073466e-05,
+      "loss": 3.3647,
       "step": 476
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 18.684444427490234,
-      "learning_rate": 1.9045805352310317e-05,
-      "loss": 3.8154,
+      "epoch": 0.06,
+      "grad_norm": 11.666590690612793,
+      "learning_rate": 1.9601723633016777e-05,
+      "loss": 5.2999,
       "step": 477
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 10.9110689163208,
-      "learning_rate": 1.9043800741705924e-05,
-      "loss": 3.7957,
+      "epoch": 0.06,
+      "grad_norm": 14.70217514038086,
+      "learning_rate": 1.960088691796009e-05,
+      "loss": 3.5773,
       "step": 478
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 22.50055503845215,
-      "learning_rate": 1.9041796131101534e-05,
-      "loss": 4.0542,
+      "epoch": 0.06,
+      "grad_norm": 10.582998275756836,
+      "learning_rate": 1.9600050202903404e-05,
+      "loss": 3.3664,
       "step": 479
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 14.225284576416016,
-      "learning_rate": 1.9039791520497144e-05,
-      "loss": 2.9864,
-      "step": 480
-    },
-    {
-      "epoch": 0.14,
-      "eval_loss": 0.9622519612312317,
-      "eval_runtime": 43.4567,
-      "eval_samples_per_second": 34.034,
-      "eval_steps_per_second": 34.034,
+      "epoch": 0.06,
+      "grad_norm": 15.844144821166992,
+      "learning_rate": 1.9599213487846714e-05,
+      "loss": 1.8579,
       "step": 480
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 14.51484489440918,
-      "learning_rate": 1.9037786909892754e-05,
-      "loss": 3.9772,
+      "epoch": 0.06,
+      "grad_norm": 11.509811401367188,
+      "learning_rate": 1.9598376772790028e-05,
+      "loss": 1.7489,
       "step": 481
     },
     {
-      "epoch": 0.14,
-      "grad_norm": 51.21427536010742,
-      "learning_rate": 1.9035782299288364e-05,
-      "loss": 3.8034,
+      "epoch": 0.06,
+      "grad_norm": 14.378329277038574,
+      "learning_rate": 1.9597540057733338e-05,
+      "loss": 3.6322,
       "step": 482
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 26.110877990722656,
-      "learning_rate": 1.9033777688683974e-05,
-      "loss": 3.6146,
+      "epoch": 0.06,
+      "grad_norm": 43.20583724975586,
+      "learning_rate": 1.9596703342676652e-05,
+      "loss": 5.0654,
       "step": 483
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 16.056385040283203,
-      "learning_rate": 1.9031773078079584e-05,
-      "loss": 3.8006,
+      "epoch": 0.06,
+      "grad_norm": 36.17166519165039,
+      "learning_rate": 1.9595866627619965e-05,
+      "loss": 3.6418,
       "step": 484
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 15.42993450164795,
-      "learning_rate": 1.9029768467475194e-05,
-      "loss": 3.0251,
+      "epoch": 0.06,
+      "grad_norm": 14.509894371032715,
+      "learning_rate": 1.959502991256328e-05,
+      "loss": 4.5501,
       "step": 485
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 12.17597484588623,
-      "learning_rate": 1.9027763856870804e-05,
-      "loss": 3.4756,
+      "epoch": 0.06,
+      "grad_norm": 13.088136672973633,
+      "learning_rate": 1.959419319750659e-05,
+      "loss": 5.2244,
       "step": 486
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 11.121559143066406,
-      "learning_rate": 1.9025759246266415e-05,
-      "loss": 2.7291,
+      "epoch": 0.06,
+      "grad_norm": 27.84554100036621,
+      "learning_rate": 1.9593356482449903e-05,
+      "loss": 3.7659,
       "step": 487
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 17.4909725189209,
-      "learning_rate": 1.9023754635662025e-05,
-      "loss": 3.6708,
+      "epoch": 0.06,
+      "grad_norm": 16.524335861206055,
+      "learning_rate": 1.9592519767393217e-05,
+      "loss": 5.8718,
       "step": 488
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 23.629941940307617,
-      "learning_rate": 1.9021750025057635e-05,
-      "loss": 3.0275,
+      "epoch": 0.06,
+      "grad_norm": 12.565080642700195,
+      "learning_rate": 1.9591683052336527e-05,
+      "loss": 3.1432,
       "step": 489
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 12.820405960083008,
-      "learning_rate": 1.9019745414453245e-05,
-      "loss": 3.0267,
+      "epoch": 0.06,
+      "grad_norm": 26.694808959960938,
+      "learning_rate": 1.959084633727984e-05,
+      "loss": 3.5153,
       "step": 490
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 22.951032638549805,
-      "learning_rate": 1.9017740803848855e-05,
-      "loss": 2.393,
+      "epoch": 0.06,
+      "grad_norm": 19.632619857788086,
+      "learning_rate": 1.9590009622223154e-05,
+      "loss": 5.4893,
       "step": 491
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 11.71376895904541,
-      "learning_rate": 1.901573619324446e-05,
-      "loss": 2.9705,
+      "epoch": 0.06,
+      "grad_norm": 13.182287216186523,
+      "learning_rate": 1.9589172907166468e-05,
+      "loss": 2.4552,
       "step": 492
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 14.28131103515625,
-      "learning_rate": 1.9013731582640075e-05,
-      "loss": 2.727,
+      "epoch": 0.06,
+      "grad_norm": 25.17426872253418,
+      "learning_rate": 1.9588336192109778e-05,
+      "loss": 3.209,
       "step": 493
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 16.59454345703125,
-      "learning_rate": 1.9011726972035685e-05,
-      "loss": 3.5118,
+      "epoch": 0.06,
+      "grad_norm": 17.61745834350586,
+      "learning_rate": 1.9587499477053092e-05,
+      "loss": 4.6118,
       "step": 494
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 20.26374053955078,
-      "learning_rate": 1.9009722361431292e-05,
-      "loss": 4.6275,
+      "epoch": 0.06,
+      "grad_norm": 12.008716583251953,
+      "learning_rate": 1.9586662761996405e-05,
+      "loss": 3.6998,
       "step": 495
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 13.29510498046875,
-      "learning_rate": 1.9007717750826905e-05,
-      "loss": 3.8611,
+      "epoch": 0.06,
+      "grad_norm": 13.607529640197754,
+      "learning_rate": 1.9585826046939716e-05,
+      "loss": 2.8954,
       "step": 496
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 14.79573917388916,
-      "learning_rate": 1.9005713140222512e-05,
-      "loss": 4.632,
+      "epoch": 0.06,
+      "grad_norm": 13.992676734924316,
+      "learning_rate": 1.958498933188303e-05,
+      "loss": 1.3716,
       "step": 497
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 19.15884017944336,
-      "learning_rate": 1.9003708529618122e-05,
-      "loss": 3.5274,
+      "epoch": 0.06,
+      "grad_norm": 15.480781555175781,
+      "learning_rate": 1.9584152616826343e-05,
+      "loss": 3.2069,
       "step": 498
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 11.399006843566895,
-      "learning_rate": 1.9001703919013736e-05,
-      "loss": 4.0488,
+      "epoch": 0.06,
+      "grad_norm": 16.785871505737305,
+      "learning_rate": 1.9583315901769657e-05,
+      "loss": 4.2125,
       "step": 499
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 12.227916717529297,
-      "learning_rate": 1.8999699308409342e-05,
-      "loss": 3.8005,
+      "epoch": 0.06,
+      "grad_norm": 43.2510871887207,
+      "learning_rate": 1.9582479186712967e-05,
+      "loss": 2.4049,
       "step": 500
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 15.450165748596191,
-      "learning_rate": 1.8997694697804952e-05,
-      "loss": 2.9546,
+      "epoch": 0.06,
+      "grad_norm": 20.25304412841797,
+      "learning_rate": 1.958164247165628e-05,
+      "loss": 1.7508,
       "step": 501
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 19.479894638061523,
-      "learning_rate": 1.8995690087200562e-05,
-      "loss": 3.056,
+      "epoch": 0.06,
+      "grad_norm": 10.431807518005371,
+      "learning_rate": 1.958080575659959e-05,
+      "loss": 2.7182,
       "step": 502
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 18.609848022460938,
-      "learning_rate": 1.8993685476596172e-05,
-      "loss": 3.8573,
+      "epoch": 0.06,
+      "grad_norm": 14.986677169799805,
+      "learning_rate": 1.9579969041542904e-05,
+      "loss": 3.2185,
       "step": 503
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 19.30628204345703,
-      "learning_rate": 1.8991680865991783e-05,
-      "loss": 2.9196,
+      "epoch": 0.06,
+      "grad_norm": 17.70635986328125,
+      "learning_rate": 1.9579132326486218e-05,
+      "loss": 4.4745,
       "step": 504
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 15.707794189453125,
-      "learning_rate": 1.8989676255387393e-05,
-      "loss": 3.6089,
+      "epoch": 0.06,
+      "grad_norm": 13.41077995300293,
+      "learning_rate": 1.957829561142953e-05,
+      "loss": 2.5006,
       "step": 505
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 18.78173065185547,
-      "learning_rate": 1.8987671644783003e-05,
-      "loss": 2.8028,
+      "epoch": 0.06,
+      "grad_norm": 20.12737464904785,
+      "learning_rate": 1.9577458896372842e-05,
+      "loss": 3.9931,
       "step": 506
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 17.811344146728516,
-      "learning_rate": 1.8985667034178613e-05,
-      "loss": 3.645,
+      "epoch": 0.06,
+      "grad_norm": 21.820350646972656,
+      "learning_rate": 1.9576622181316152e-05,
+      "loss": 2.7736,
       "step": 507
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 15.66889476776123,
-      "learning_rate": 1.8983662423574223e-05,
-      "loss": 3.5172,
+      "epoch": 0.06,
+      "grad_norm": 18.02536964416504,
+      "learning_rate": 1.9575785466259466e-05,
+      "loss": 2.1139,
       "step": 508
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 13.496955871582031,
-      "learning_rate": 1.8981657812969833e-05,
-      "loss": 3.2022,
+      "epoch": 0.06,
+      "grad_norm": 14.28238582611084,
+      "learning_rate": 1.957494875120278e-05,
+      "loss": 4.572,
       "step": 509
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 14.416739463806152,
-      "learning_rate": 1.8979653202365443e-05,
-      "loss": 3.7212,
+      "epoch": 0.06,
+      "grad_norm": 17.39335060119629,
+      "learning_rate": 1.957411203614609e-05,
+      "loss": 4.1144,
       "step": 510
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 13.47453784942627,
-      "learning_rate": 1.897764859176105e-05,
-      "loss": 4.1187,
+      "epoch": 0.06,
+      "grad_norm": 16.540075302124023,
+      "learning_rate": 1.9573275321089403e-05,
+      "loss": 3.5938,
       "step": 511
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 18.787689208984375,
-      "learning_rate": 1.8975643981156663e-05,
-      "loss": 3.6262,
+      "epoch": 0.06,
+      "grad_norm": 12.823972702026367,
+      "learning_rate": 1.9572438606032717e-05,
+      "loss": 2.467,
       "step": 512
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 10.850929260253906,
-      "learning_rate": 1.8973639370552273e-05,
-      "loss": 2.937,
+      "epoch": 0.06,
+      "grad_norm": 16.05743408203125,
+      "learning_rate": 1.957160189097603e-05,
+      "loss": 4.8134,
       "step": 513
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 14.07578182220459,
-      "learning_rate": 1.897163475994788e-05,
-      "loss": 4.5924,
+      "epoch": 0.06,
+      "grad_norm": 16.572154998779297,
+      "learning_rate": 1.957076517591934e-05,
+      "loss": 5.0406,
       "step": 514
     },
     {
-      "epoch": 0.15,
-      "grad_norm": 13.420841217041016,
-      "learning_rate": 1.8969630149343493e-05,
-      "loss": 2.5988,
+      "epoch": 0.06,
+      "grad_norm": 10.041598320007324,
+      "learning_rate": 1.9569928460862655e-05,
+      "loss": 3.0591,
       "step": 515
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 16.54571533203125,
-      "learning_rate": 1.89676255387391e-05,
-      "loss": 3.835,
+      "epoch": 0.06,
+      "grad_norm": 16.29025650024414,
+      "learning_rate": 1.956909174580597e-05,
+      "loss": 2.9466,
       "step": 516
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 68.53701782226562,
-      "learning_rate": 1.896562092813471e-05,
-      "loss": 5.1961,
+      "epoch": 0.06,
+      "grad_norm": 16.037534713745117,
+      "learning_rate": 1.956825503074928e-05,
+      "loss": 4.3558,
       "step": 517
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 9.695450782775879,
-      "learning_rate": 1.8963616317530324e-05,
-      "loss": 3.4607,
+      "epoch": 0.07,
+      "grad_norm": 25.85873031616211,
+      "learning_rate": 1.9567418315692592e-05,
+      "loss": 2.2824,
       "step": 518
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 15.207024574279785,
-      "learning_rate": 1.896161170692593e-05,
-      "loss": 3.3355,
+      "epoch": 0.07,
+      "grad_norm": 30.12692642211914,
+      "learning_rate": 1.9566581600635906e-05,
+      "loss": 2.4197,
       "step": 519
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 14.207443237304688,
-      "learning_rate": 1.895960709632154e-05,
-      "loss": 3.5345,
+      "epoch": 0.07,
+      "grad_norm": 15.433709144592285,
+      "learning_rate": 1.956574488557922e-05,
+      "loss": 3.2912,
       "step": 520
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 18.703203201293945,
-      "learning_rate": 1.895760248571715e-05,
-      "loss": 4.1013,
+      "epoch": 0.07,
+      "grad_norm": 13.080670356750488,
+      "learning_rate": 1.956490817052253e-05,
+      "loss": 5.63,
       "step": 521
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 22.39604949951172,
-      "learning_rate": 1.895559787511276e-05,
-      "loss": 4.8708,
+      "epoch": 0.07,
+      "grad_norm": 16.858592987060547,
+      "learning_rate": 1.9564071455465843e-05,
+      "loss": 3.2096,
       "step": 522
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 19.35679817199707,
-      "learning_rate": 1.895359326450837e-05,
-      "loss": 2.9551,
+      "epoch": 0.07,
+      "grad_norm": 7.967859268188477,
+      "learning_rate": 1.9563234740409157e-05,
+      "loss": 6.2087,
       "step": 523
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 13.426315307617188,
-      "learning_rate": 1.895158865390398e-05,
-      "loss": 2.9543,
+      "epoch": 0.07,
+      "grad_norm": 9.962607383728027,
+      "learning_rate": 1.9562398025352467e-05,
+      "loss": 4.5942,
       "step": 524
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 18.244346618652344,
-      "learning_rate": 1.894958404329959e-05,
-      "loss": 3.8422,
+      "epoch": 0.07,
+      "grad_norm": 24.848005294799805,
+      "learning_rate": 1.956156131029578e-05,
+      "loss": 4.6929,
       "step": 525
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 16.004077911376953,
-      "learning_rate": 1.89475794326952e-05,
-      "loss": 2.7931,
+      "epoch": 0.07,
+      "grad_norm": 15.170857429504395,
+      "learning_rate": 1.9560724595239095e-05,
+      "loss": 3.1667,
       "step": 526
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 23.495067596435547,
-      "learning_rate": 1.894557482209081e-05,
-      "loss": 4.6575,
+      "epoch": 0.07,
+      "grad_norm": 15.806061744689941,
+      "learning_rate": 1.9559887880182405e-05,
+      "loss": 4.0896,
       "step": 527
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 16.585546493530273,
-      "learning_rate": 1.894357021148642e-05,
-      "loss": 3.2343,
+      "epoch": 0.07,
+      "grad_norm": 14.205904006958008,
+      "learning_rate": 1.955905116512572e-05,
+      "loss": 3.6082,
       "step": 528
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 25.565746307373047,
-      "learning_rate": 1.894156560088203e-05,
-      "loss": 3.6168,
+      "epoch": 0.07,
+      "grad_norm": 16.476051330566406,
+      "learning_rate": 1.9558214450069032e-05,
+      "loss": 4.0229,
       "step": 529
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 17.877038955688477,
-      "learning_rate": 1.893956099027764e-05,
-      "loss": 4.1159,
+      "epoch": 0.07,
+      "grad_norm": 18.11623764038086,
+      "learning_rate": 1.9557377735012342e-05,
+      "loss": 3.9217,
       "step": 530
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 25.356380462646484,
-      "learning_rate": 1.893755637967325e-05,
-      "loss": 4.1038,
+      "epoch": 0.07,
+      "grad_norm": 13.680058479309082,
+      "learning_rate": 1.9556541019955656e-05,
+      "loss": 3.6137,
       "step": 531
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 28.306737899780273,
-      "learning_rate": 1.893555176906886e-05,
-      "loss": 3.7835,
+      "epoch": 0.07,
+      "grad_norm": 11.90891170501709,
+      "learning_rate": 1.9555704304898966e-05,
+      "loss": 4.1743,
       "step": 532
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 16.962968826293945,
-      "learning_rate": 1.8933547158464468e-05,
-      "loss": 2.9008,
+      "epoch": 0.07,
+      "grad_norm": 25.42872428894043,
+      "learning_rate": 1.955486758984228e-05,
+      "loss": 5.0914,
       "step": 533
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 20.51753044128418,
-      "learning_rate": 1.8931542547860082e-05,
-      "loss": 4.6783,
+      "epoch": 0.07,
+      "grad_norm": 12.091551780700684,
+      "learning_rate": 1.9554030874785594e-05,
+      "loss": 4.2007,
       "step": 534
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 17.836177825927734,
-      "learning_rate": 1.892953793725569e-05,
-      "loss": 2.8979,
+      "epoch": 0.07,
+      "grad_norm": 11.26998233795166,
+      "learning_rate": 1.9553194159728904e-05,
+      "loss": 3.4855,
       "step": 535
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 14.355405807495117,
-      "learning_rate": 1.89275333266513e-05,
-      "loss": 3.6143,
+      "epoch": 0.07,
+      "grad_norm": 11.434247016906738,
+      "learning_rate": 1.9552357444672218e-05,
+      "loss": 4.8023,
       "step": 536
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 17.681434631347656,
-      "learning_rate": 1.8925528716046912e-05,
-      "loss": 2.4729,
+      "epoch": 0.07,
+      "grad_norm": 14.716187477111816,
+      "learning_rate": 1.955152072961553e-05,
+      "loss": 1.4756,
       "step": 537
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 16.02050018310547,
-      "learning_rate": 1.892352410544252e-05,
-      "loss": 2.3059,
+      "epoch": 0.07,
+      "grad_norm": 11.850278854370117,
+      "learning_rate": 1.955068401455884e-05,
+      "loss": 5.4931,
       "step": 538
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 12.686012268066406,
-      "learning_rate": 1.892151949483813e-05,
-      "loss": 3.307,
+      "epoch": 0.07,
+      "grad_norm": 13.068528175354004,
+      "learning_rate": 1.9549847299502155e-05,
+      "loss": 4.0506,
       "step": 539
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 13.235198974609375,
-      "learning_rate": 1.891951488423374e-05,
-      "loss": 4.249,
+      "epoch": 0.07,
+      "grad_norm": 13.046282768249512,
+      "learning_rate": 1.954901058444547e-05,
+      "loss": 3.9209,
       "step": 540
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 16.329849243164062,
-      "learning_rate": 1.891751027362935e-05,
-      "loss": 3.9762,
+      "epoch": 0.07,
+      "grad_norm": 18.51168441772461,
+      "learning_rate": 1.954817386938878e-05,
+      "loss": 2.3712,
       "step": 541
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 18.02645492553711,
-      "learning_rate": 1.891550566302496e-05,
-      "loss": 3.3566,
+      "epoch": 0.07,
+      "grad_norm": 12.119020462036133,
+      "learning_rate": 1.9547337154332093e-05,
+      "loss": 1.7651,
       "step": 542
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 15.097955703735352,
-      "learning_rate": 1.891350105242057e-05,
-      "loss": 3.683,
+      "epoch": 0.07,
+      "grad_norm": 10.600531578063965,
+      "learning_rate": 1.9546500439275406e-05,
+      "loss": 4.215,
       "step": 543
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 14.13657283782959,
-      "learning_rate": 1.891149644181618e-05,
-      "loss": 2.5315,
+      "epoch": 0.07,
+      "grad_norm": 18.362884521484375,
+      "learning_rate": 1.954566372421872e-05,
+      "loss": 3.8651,
       "step": 544
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 13.064208984375,
-      "learning_rate": 1.890949183121179e-05,
-      "loss": 3.7249,
+      "epoch": 0.07,
+      "grad_norm": 10.03396987915039,
+      "learning_rate": 1.954482700916203e-05,
+      "loss": 3.7257,
       "step": 545
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 12.626958847045898,
-      "learning_rate": 1.89074872206074e-05,
-      "loss": 3.4621,
+      "epoch": 0.07,
+      "grad_norm": 17.79392433166504,
+      "learning_rate": 1.9543990294105344e-05,
+      "loss": 2.2889,
       "step": 546
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 15.529770851135254,
-      "learning_rate": 1.8905482610003006e-05,
-      "loss": 3.6776,
+      "epoch": 0.07,
+      "grad_norm": 15.041823387145996,
+      "learning_rate": 1.9543153579048658e-05,
+      "loss": 3.0014,
       "step": 547
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 16.237417221069336,
-      "learning_rate": 1.890347799939862e-05,
-      "loss": 5.0018,
+      "epoch": 0.07,
+      "grad_norm": 19.62635612487793,
+      "learning_rate": 1.9542316863991968e-05,
+      "loss": 5.0205,
       "step": 548
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 24.94883155822754,
-      "learning_rate": 1.890147338879423e-05,
-      "loss": 3.5904,
+      "epoch": 0.07,
+      "grad_norm": 23.438556671142578,
+      "learning_rate": 1.954148014893528e-05,
+      "loss": 3.028,
       "step": 549
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 18.61686134338379,
-      "learning_rate": 1.8899468778189836e-05,
-      "loss": 3.47,
+      "epoch": 0.07,
+      "grad_norm": 19.551422119140625,
+      "learning_rate": 1.9540643433878595e-05,
+      "loss": 2.8019,
       "step": 550
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 23.455402374267578,
-      "learning_rate": 1.889746416758545e-05,
-      "loss": 3.8038,
+      "epoch": 0.07,
+      "grad_norm": 17.06743049621582,
+      "learning_rate": 1.953980671882191e-05,
+      "loss": 3.2838,
       "step": 551
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 14.99221420288086,
-      "learning_rate": 1.8895459556981056e-05,
-      "loss": 3.4294,
+      "epoch": 0.07,
+      "grad_norm": 16.77594566345215,
+      "learning_rate": 1.953897000376522e-05,
+      "loss": 3.2405,
       "step": 552
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 15.767837524414062,
-      "learning_rate": 1.8893454946376667e-05,
-      "loss": 3.2229,
+      "epoch": 0.07,
+      "grad_norm": 15.835091590881348,
+      "learning_rate": 1.9538133288708533e-05,
+      "loss": 4.1767,
       "step": 553
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 12.863175392150879,
-      "learning_rate": 1.8891450335772277e-05,
-      "loss": 3.6184,
+      "epoch": 0.07,
+      "grad_norm": 35.76700973510742,
+      "learning_rate": 1.9537296573651846e-05,
+      "loss": 1.8702,
       "step": 554
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 14.412956237792969,
-      "learning_rate": 1.8889445725167887e-05,
-      "loss": 4.1913,
+      "epoch": 0.07,
+      "grad_norm": 21.611915588378906,
+      "learning_rate": 1.9536459858595157e-05,
+      "loss": 4.1775,
       "step": 555
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 24.851470947265625,
-      "learning_rate": 1.8887441114563497e-05,
-      "loss": 3.9418,
+      "epoch": 0.07,
+      "grad_norm": 14.899382591247559,
+      "learning_rate": 1.953562314353847e-05,
+      "loss": 3.198,
       "step": 556
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 47.26935958862305,
-      "learning_rate": 1.8885436503959107e-05,
-      "loss": 5.1826,
+      "epoch": 0.07,
+      "grad_norm": 15.35878849029541,
+      "learning_rate": 1.9534786428481784e-05,
+      "loss": 2.8263,
       "step": 557
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 15.844667434692383,
-      "learning_rate": 1.8883431893354717e-05,
-      "loss": 3.7824,
+      "epoch": 0.07,
+      "grad_norm": 14.225360870361328,
+      "learning_rate": 1.9533949713425094e-05,
+      "loss": 3.9339,
       "step": 558
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 11.842270851135254,
-      "learning_rate": 1.8881427282750327e-05,
-      "loss": 2.9349,
+      "epoch": 0.07,
+      "grad_norm": 28.126028060913086,
+      "learning_rate": 1.9533112998368408e-05,
+      "loss": 1.6965,
       "step": 559
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 21.0179386138916,
-      "learning_rate": 1.8879422672145937e-05,
-      "loss": 3.2554,
+      "epoch": 0.07,
+      "grad_norm": 33.49496078491211,
+      "learning_rate": 1.9532276283311718e-05,
+      "loss": 3.8451,
       "step": 560
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 19.371559143066406,
-      "learning_rate": 1.8877418061541547e-05,
-      "loss": 3.6965,
+      "epoch": 0.07,
+      "grad_norm": 22.24785041809082,
+      "learning_rate": 1.9531439568255032e-05,
+      "loss": 3.5929,
       "step": 561
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 17.8704833984375,
-      "learning_rate": 1.8875413450937157e-05,
-      "loss": 3.6738,
+      "epoch": 0.07,
+      "grad_norm": 36.69216537475586,
+      "learning_rate": 1.9530602853198345e-05,
+      "loss": 3.6625,
       "step": 562
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 11.61600112915039,
-      "learning_rate": 1.8873408840332767e-05,
-      "loss": 2.5997,
+      "epoch": 0.07,
+      "grad_norm": 13.007461547851562,
+      "learning_rate": 1.9529766138141656e-05,
+      "loss": 3.823,
       "step": 563
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 17.82415199279785,
-      "learning_rate": 1.8871404229728377e-05,
-      "loss": 3.402,
+      "epoch": 0.07,
+      "grad_norm": 19.145416259765625,
+      "learning_rate": 1.952892942308497e-05,
+      "loss": 3.2177,
       "step": 564
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 13.841615676879883,
-      "learning_rate": 1.8869399619123988e-05,
-      "loss": 2.5574,
+      "epoch": 0.07,
+      "grad_norm": 28.419158935546875,
+      "learning_rate": 1.9528092708028283e-05,
+      "loss": 1.5909,
       "step": 565
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 17.705219268798828,
-      "learning_rate": 1.8867395008519594e-05,
-      "loss": 3.4216,
+      "epoch": 0.07,
+      "grad_norm": 14.794642448425293,
+      "learning_rate": 1.9527255992971593e-05,
+      "loss": 1.8523,
       "step": 566
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 15.329522132873535,
-      "learning_rate": 1.8865390397915208e-05,
-      "loss": 2.5929,
+      "epoch": 0.07,
+      "grad_norm": 19.6970157623291,
+      "learning_rate": 1.9526419277914907e-05,
+      "loss": 5.4043,
       "step": 567
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 14.040759086608887,
-      "learning_rate": 1.8863385787310818e-05,
-      "loss": 3.7188,
+      "epoch": 0.07,
+      "grad_norm": 20.794288635253906,
+      "learning_rate": 1.952558256285822e-05,
+      "loss": 2.3032,
       "step": 568
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 20.66543960571289,
-      "learning_rate": 1.8861381176706424e-05,
-      "loss": 3.3664,
+      "epoch": 0.07,
+      "grad_norm": 18.208457946777344,
+      "learning_rate": 1.952474584780153e-05,
+      "loss": 3.9328,
       "step": 569
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 10.702879905700684,
-      "learning_rate": 1.8859376566102038e-05,
-      "loss": 3.0084,
+      "epoch": 0.07,
+      "grad_norm": 18.695064544677734,
+      "learning_rate": 1.9523909132744844e-05,
+      "loss": 2.7843,
       "step": 570
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 15.897429466247559,
-      "learning_rate": 1.8857371955497645e-05,
-      "loss": 3.7669,
+      "epoch": 0.07,
+      "grad_norm": 10.989561080932617,
+      "learning_rate": 1.9523072417688158e-05,
+      "loss": 2.9185,
       "step": 571
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 11.95020866394043,
-      "learning_rate": 1.8855367344893255e-05,
-      "loss": 3.9482,
+      "epoch": 0.07,
+      "grad_norm": 13.847416877746582,
+      "learning_rate": 1.9522235702631472e-05,
+      "loss": 3.4503,
       "step": 572
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 27.125240325927734,
-      "learning_rate": 1.8853362734288868e-05,
-      "loss": 2.7438,
+      "epoch": 0.07,
+      "grad_norm": 140.5792999267578,
+      "learning_rate": 1.9521398987574782e-05,
+      "loss": 3.8577,
       "step": 573
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 13.781254768371582,
-      "learning_rate": 1.8851358123684475e-05,
-      "loss": 3.3872,
+      "epoch": 0.07,
+      "grad_norm": 11.583138465881348,
+      "learning_rate": 1.9520562272518096e-05,
+      "loss": 3.792,
       "step": 574
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 29.12127113342285,
-      "learning_rate": 1.8849353513080085e-05,
-      "loss": 4.5567,
+      "epoch": 0.07,
+      "grad_norm": 27.13224983215332,
+      "learning_rate": 1.951972555746141e-05,
+      "loss": 6.0304,
       "step": 575
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 15.671697616577148,
-      "learning_rate": 1.8847348902475695e-05,
-      "loss": 2.5302,
+      "epoch": 0.07,
+      "grad_norm": 21.271699905395508,
+      "learning_rate": 1.951888884240472e-05,
+      "loss": 4.7579,
       "step": 576
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 17.996654510498047,
-      "learning_rate": 1.8845344291871305e-05,
-      "loss": 3.1327,
+      "epoch": 0.07,
+      "grad_norm": 13.934300422668457,
+      "learning_rate": 1.9518052127348033e-05,
+      "loss": 2.274,
       "step": 577
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 16.846689224243164,
-      "learning_rate": 1.8843339681266915e-05,
-      "loss": 2.5453,
+      "epoch": 0.07,
+      "grad_norm": 14.28530216217041,
+      "learning_rate": 1.9517215412291347e-05,
+      "loss": 3.3553,
       "step": 578
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 21.583017349243164,
-      "learning_rate": 1.8841335070662525e-05,
-      "loss": 3.5481,
+      "epoch": 0.07,
+      "grad_norm": 21.369596481323242,
+      "learning_rate": 1.951637869723466e-05,
+      "loss": 3.0815,
       "step": 579
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 14.08506965637207,
-      "learning_rate": 1.8839330460058135e-05,
-      "loss": 3.8354,
+      "epoch": 0.07,
+      "grad_norm": 70.39269256591797,
+      "learning_rate": 1.951554198217797e-05,
+      "loss": 1.4678,
       "step": 580
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 12.634489059448242,
-      "learning_rate": 1.8837325849453745e-05,
-      "loss": 3.1067,
+      "epoch": 0.07,
+      "grad_norm": 10.346799850463867,
+      "learning_rate": 1.9514705267121284e-05,
+      "loss": 3.7247,
       "step": 581
     },
     {
-      "epoch": 0.17,
-      "grad_norm": 11.801117897033691,
-      "learning_rate": 1.8835321238849356e-05,
-      "loss": 3.4585,
+      "epoch": 0.07,
+      "grad_norm": 16.728269577026367,
+      "learning_rate": 1.9513868552064598e-05,
+      "loss": 4.006,
       "step": 582
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 15.204583168029785,
-      "learning_rate": 1.8833316628244966e-05,
-      "loss": 3.3097,
+      "epoch": 0.07,
+      "grad_norm": 11.536478996276855,
+      "learning_rate": 1.9513031837007908e-05,
+      "loss": 3.0046,
       "step": 583
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 15.794861793518066,
-      "learning_rate": 1.8831312017640576e-05,
-      "loss": 3.0038,
+      "epoch": 0.07,
+      "grad_norm": 13.823684692382812,
+      "learning_rate": 1.9512195121951222e-05,
+      "loss": 3.4075,
       "step": 584
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 13.39342212677002,
-      "learning_rate": 1.8829307407036186e-05,
-      "loss": 3.4626,
+      "epoch": 0.07,
+      "grad_norm": 11.703648567199707,
+      "learning_rate": 1.9511358406894532e-05,
+      "loss": 3.6496,
       "step": 585
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 14.78641414642334,
-      "learning_rate": 1.8827302796431796e-05,
-      "loss": 3.4037,
+      "epoch": 0.07,
+      "grad_norm": 13.754158020019531,
+      "learning_rate": 1.9510521691837846e-05,
+      "loss": 2.1938,
       "step": 586
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 16.814594268798828,
-      "learning_rate": 1.8825298185827406e-05,
-      "loss": 3.7882,
+      "epoch": 0.07,
+      "grad_norm": 14.499614715576172,
+      "learning_rate": 1.950968497678116e-05,
+      "loss": 2.7492,
       "step": 587
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 27.2209415435791,
-      "learning_rate": 1.8823293575223013e-05,
-      "loss": 3.2455,
+      "epoch": 0.07,
+      "grad_norm": 15.740714073181152,
+      "learning_rate": 1.950884826172447e-05,
+      "loss": 2.9027,
       "step": 588
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 16.686363220214844,
-      "learning_rate": 1.8821288964618626e-05,
-      "loss": 3.932,
+      "epoch": 0.07,
+      "grad_norm": 15.932597160339355,
+      "learning_rate": 1.9508011546667783e-05,
+      "loss": 3.1953,
       "step": 589
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 17.841867446899414,
-      "learning_rate": 1.8819284354014233e-05,
-      "loss": 2.3729,
+      "epoch": 0.07,
+      "grad_norm": 21.080705642700195,
+      "learning_rate": 1.9507174831611097e-05,
+      "loss": 2.7916,
       "step": 590
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 13.670007705688477,
-      "learning_rate": 1.8817279743409843e-05,
-      "loss": 3.9139,
+      "epoch": 0.07,
+      "grad_norm": 19.372400283813477,
+      "learning_rate": 1.9506338116554407e-05,
+      "loss": 4.3448,
       "step": 591
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 12.825089454650879,
-      "learning_rate": 1.8815275132805456e-05,
-      "loss": 3.4298,
+      "epoch": 0.07,
+      "grad_norm": 18.336488723754883,
+      "learning_rate": 1.950550140149772e-05,
+      "loss": 4.4037,
       "step": 592
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 21.140291213989258,
-      "learning_rate": 1.8813270522201063e-05,
-      "loss": 3.6672,
+      "epoch": 0.07,
+      "grad_norm": 12.87525749206543,
+      "learning_rate": 1.9504664686441035e-05,
+      "loss": 1.776,
       "step": 593
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 19.407217025756836,
-      "learning_rate": 1.8811265911596673e-05,
-      "loss": 3.853,
+      "epoch": 0.07,
+      "grad_norm": 31.6224308013916,
+      "learning_rate": 1.9503827971384345e-05,
+      "loss": 2.0704,
       "step": 594
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 17.949951171875,
-      "learning_rate": 1.8809261300992283e-05,
-      "loss": 4.0853,
+      "epoch": 0.07,
+      "grad_norm": 21.20492172241211,
+      "learning_rate": 1.950299125632766e-05,
+      "loss": 2.3727,
       "step": 595
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 14.42286205291748,
-      "learning_rate": 1.8807256690387893e-05,
-      "loss": 3.8624,
+      "epoch": 0.07,
+      "grad_norm": 24.187694549560547,
+      "learning_rate": 1.9502154541270972e-05,
+      "loss": 2.6064,
       "step": 596
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 13.524324417114258,
-      "learning_rate": 1.8805252079783503e-05,
-      "loss": 3.4537,
+      "epoch": 0.07,
+      "grad_norm": 17.776546478271484,
+      "learning_rate": 1.9501317826214282e-05,
+      "loss": 3.9096,
       "step": 597
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 16.790464401245117,
-      "learning_rate": 1.8803247469179114e-05,
-      "loss": 3.3715,
+      "epoch": 0.08,
+      "grad_norm": 9.369264602661133,
+      "learning_rate": 1.9500481111157596e-05,
+      "loss": 2.1012,
       "step": 598
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 28.930021286010742,
-      "learning_rate": 1.8801242858574724e-05,
-      "loss": 4.5601,
+      "epoch": 0.08,
+      "grad_norm": 15.953502655029297,
+      "learning_rate": 1.949964439610091e-05,
+      "loss": 6.1173,
       "step": 599
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 17.764270782470703,
-      "learning_rate": 1.8799238247970334e-05,
-      "loss": 3.0025,
-      "step": 600
-    },
-    {
-      "epoch": 0.18,
-      "eval_loss": 0.8108684420585632,
-      "eval_runtime": 43.7575,
-      "eval_samples_per_second": 33.8,
-      "eval_steps_per_second": 33.8,
+      "epoch": 0.08,
+      "grad_norm": 17.869020462036133,
+      "learning_rate": 1.9498807681044223e-05,
+      "loss": 2.898,
       "step": 600
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 12.37165641784668,
-      "learning_rate": 1.8797233637365944e-05,
-      "loss": 3.7558,
+      "epoch": 0.08,
+      "grad_norm": 13.049903869628906,
+      "learning_rate": 1.9497970965987534e-05,
+      "loss": 2.7439,
       "step": 601
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 31.21900177001953,
-      "learning_rate": 1.8795229026761554e-05,
-      "loss": 3.6418,
+      "epoch": 0.08,
+      "grad_norm": 10.503606796264648,
+      "learning_rate": 1.9497134250930847e-05,
+      "loss": 1.334,
       "step": 602
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 12.556445121765137,
-      "learning_rate": 1.8793224416157164e-05,
-      "loss": 2.7222,
+      "epoch": 0.08,
+      "grad_norm": 20.092920303344727,
+      "learning_rate": 1.949629753587416e-05,
+      "loss": 3.6018,
       "step": 603
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 12.976155281066895,
-      "learning_rate": 1.8791219805552774e-05,
-      "loss": 3.1556,
+      "epoch": 0.08,
+      "grad_norm": 106.8263168334961,
+      "learning_rate": 1.949546082081747e-05,
+      "loss": 3.2016,
       "step": 604
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 14.281742095947266,
-      "learning_rate": 1.8789215194948384e-05,
-      "loss": 3.5987,
+      "epoch": 0.08,
+      "grad_norm": 26.08072280883789,
+      "learning_rate": 1.9494624105760785e-05,
+      "loss": 4.7636,
       "step": 605
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 17.95466423034668,
-      "learning_rate": 1.8787210584343994e-05,
-      "loss": 2.8724,
+      "epoch": 0.08,
+      "grad_norm": 18.82193374633789,
+      "learning_rate": 1.94937873907041e-05,
+      "loss": 2.8383,
       "step": 606
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 11.985798835754395,
-      "learning_rate": 1.87852059737396e-05,
-      "loss": 3.3135,
+      "epoch": 0.08,
+      "grad_norm": 20.548948287963867,
+      "learning_rate": 1.9492950675647412e-05,
+      "loss": 2.4919,
       "step": 607
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 11.839190483093262,
-      "learning_rate": 1.8783201363135214e-05,
-      "loss": 2.7356,
+      "epoch": 0.08,
+      "grad_norm": 14.64707088470459,
+      "learning_rate": 1.9492113960590722e-05,
+      "loss": 3.689,
       "step": 608
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 19.593158721923828,
-      "learning_rate": 1.878119675253082e-05,
-      "loss": 2.48,
+      "epoch": 0.08,
+      "grad_norm": 16.494598388671875,
+      "learning_rate": 1.9491277245534036e-05,
+      "loss": 1.9832,
       "step": 609
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 11.183398246765137,
-      "learning_rate": 1.877919214192643e-05,
-      "loss": 2.3999,
+      "epoch": 0.08,
+      "grad_norm": 19.26803207397461,
+      "learning_rate": 1.949044053047735e-05,
+      "loss": 3.0708,
       "step": 610
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 16.24125862121582,
-      "learning_rate": 1.8777187531322045e-05,
-      "loss": 3.8464,
+      "epoch": 0.08,
+      "grad_norm": 20.39113426208496,
+      "learning_rate": 1.948960381542066e-05,
+      "loss": 2.8375,
       "step": 611
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 17.56856346130371,
-      "learning_rate": 1.877518292071765e-05,
-      "loss": 3.5414,
+      "epoch": 0.08,
+      "grad_norm": 13.161185264587402,
+      "learning_rate": 1.9488767100363974e-05,
+      "loss": 2.4338,
       "step": 612
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 15.031785011291504,
-      "learning_rate": 1.877317831011326e-05,
-      "loss": 3.4706,
+      "epoch": 0.08,
+      "grad_norm": 22.47385025024414,
+      "learning_rate": 1.9487930385307284e-05,
+      "loss": 5.3107,
       "step": 613
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 14.207649230957031,
-      "learning_rate": 1.877117369950887e-05,
-      "loss": 3.0532,
+      "epoch": 0.08,
+      "grad_norm": 9.89582633972168,
+      "learning_rate": 1.9487093670250598e-05,
+      "loss": 2.2135,
       "step": 614
     },
     {
-      "epoch": 0.18,
-      "grad_norm": 15.561878204345703,
-      "learning_rate": 1.876916908890448e-05,
-      "loss": 3.5544,
+      "epoch": 0.08,
+      "grad_norm": 14.444812774658203,
+      "learning_rate": 1.948625695519391e-05,
+      "loss": 2.9484,
       "step": 615
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 15.425860404968262,
-      "learning_rate": 1.876716447830009e-05,
-      "loss": 3.2101,
+      "epoch": 0.08,
+      "grad_norm": 17.883405685424805,
+      "learning_rate": 1.948542024013722e-05,
+      "loss": 1.481,
       "step": 616
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 17.44911766052246,
-      "learning_rate": 1.8765159867695702e-05,
-      "loss": 4.0486,
+      "epoch": 0.08,
+      "grad_norm": 15.512957572937012,
+      "learning_rate": 1.9484583525080535e-05,
+      "loss": 3.8052,
       "step": 617
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 16.89653205871582,
-      "learning_rate": 1.8763155257091312e-05,
-      "loss": 3.1268,
+      "epoch": 0.08,
+      "grad_norm": 9.79510498046875,
+      "learning_rate": 1.9483746810023845e-05,
+      "loss": 1.2129,
       "step": 618
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 17.181591033935547,
-      "learning_rate": 1.8761150646486922e-05,
-      "loss": 3.0687,
+      "epoch": 0.08,
+      "grad_norm": 12.645746231079102,
+      "learning_rate": 1.948291009496716e-05,
+      "loss": 1.5934,
       "step": 619
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 14.057995796203613,
-      "learning_rate": 1.8759146035882532e-05,
-      "loss": 2.7292,
+      "epoch": 0.08,
+      "grad_norm": 29.721193313598633,
+      "learning_rate": 1.9482073379910473e-05,
+      "loss": 5.759,
       "step": 620
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 62.06224822998047,
-      "learning_rate": 1.875714142527814e-05,
-      "loss": 2.8716,
+      "epoch": 0.08,
+      "grad_norm": 21.66473388671875,
+      "learning_rate": 1.9481236664853786e-05,
+      "loss": 3.0079,
       "step": 621
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 15.26970100402832,
-      "learning_rate": 1.8755136814673752e-05,
-      "loss": 3.4212,
+      "epoch": 0.08,
+      "grad_norm": 10.992793083190918,
+      "learning_rate": 1.9480399949797097e-05,
+      "loss": 3.4051,
       "step": 622
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 12.904154777526855,
-      "learning_rate": 1.8753132204069362e-05,
-      "loss": 2.5465,
+      "epoch": 0.08,
+      "grad_norm": 14.940546989440918,
+      "learning_rate": 1.947956323474041e-05,
+      "loss": 4.0389,
       "step": 623
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 16.669986724853516,
-      "learning_rate": 1.875112759346497e-05,
-      "loss": 2.8482,
+      "epoch": 0.08,
+      "grad_norm": 14.911699295043945,
+      "learning_rate": 1.9478726519683724e-05,
+      "loss": 3.7682,
       "step": 624
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 10.211058616638184,
-      "learning_rate": 1.8749122982860582e-05,
-      "loss": 2.5476,
+      "epoch": 0.08,
+      "grad_norm": 13.005440711975098,
+      "learning_rate": 1.9477889804627034e-05,
+      "loss": 1.5709,
       "step": 625
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 21.396862030029297,
-      "learning_rate": 1.874711837225619e-05,
-      "loss": 3.7639,
+      "epoch": 0.08,
+      "grad_norm": 15.729487419128418,
+      "learning_rate": 1.9477053089570348e-05,
+      "loss": 1.1742,
       "step": 626
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 20.889707565307617,
-      "learning_rate": 1.87451137616518e-05,
-      "loss": 3.3019,
+      "epoch": 0.08,
+      "grad_norm": 20.291719436645508,
+      "learning_rate": 1.947621637451366e-05,
+      "loss": 4.8735,
       "step": 627
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 25.096691131591797,
-      "learning_rate": 1.874310915104741e-05,
-      "loss": 4.246,
+      "epoch": 0.08,
+      "grad_norm": 13.93189525604248,
+      "learning_rate": 1.9475379659456975e-05,
+      "loss": 2.2031,
       "step": 628
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 13.878023147583008,
-      "learning_rate": 1.874110454044302e-05,
-      "loss": 2.9438,
+      "epoch": 0.08,
+      "grad_norm": 11.237191200256348,
+      "learning_rate": 1.9474542944400285e-05,
+      "loss": 4.0729,
       "step": 629
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 18.34180450439453,
-      "learning_rate": 1.873909992983863e-05,
-      "loss": 4.3573,
+      "epoch": 0.08,
+      "grad_norm": 14.943650245666504,
+      "learning_rate": 1.94737062293436e-05,
+      "loss": 1.8902,
       "step": 630
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 18.738731384277344,
-      "learning_rate": 1.873709531923424e-05,
-      "loss": 2.7748,
+      "epoch": 0.08,
+      "grad_norm": 14.849425315856934,
+      "learning_rate": 1.9472869514286913e-05,
+      "loss": 3.2545,
       "step": 631
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 15.163655281066895,
-      "learning_rate": 1.873509070862985e-05,
-      "loss": 3.4938,
+      "epoch": 0.08,
+      "grad_norm": 11.831103324890137,
+      "learning_rate": 1.9472032799230223e-05,
+      "loss": 2.6383,
       "step": 632
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 16.00517463684082,
-      "learning_rate": 1.873308609802546e-05,
-      "loss": 4.1243,
+      "epoch": 0.08,
+      "grad_norm": 12.109277725219727,
+      "learning_rate": 1.9471196084173537e-05,
+      "loss": 4.2325,
       "step": 633
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 15.933029174804688,
-      "learning_rate": 1.873108148742107e-05,
-      "loss": 3.8793,
+      "epoch": 0.08,
+      "grad_norm": 14.199204444885254,
+      "learning_rate": 1.947035936911685e-05,
+      "loss": 3.5085,
       "step": 634
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 17.073196411132812,
-      "learning_rate": 1.872907687681668e-05,
-      "loss": 3.3813,
+      "epoch": 0.08,
+      "grad_norm": 17.686521530151367,
+      "learning_rate": 1.9469522654060164e-05,
+      "loss": 4.728,
       "step": 635
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 16.49427604675293,
-      "learning_rate": 1.872707226621229e-05,
-      "loss": 3.9439,
+      "epoch": 0.08,
+      "grad_norm": 11.224983215332031,
+      "learning_rate": 1.9468685939003474e-05,
+      "loss": 1.8126,
       "step": 636
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 14.988619804382324,
-      "learning_rate": 1.87250676556079e-05,
-      "loss": 3.4706,
+      "epoch": 0.08,
+      "grad_norm": 11.27236270904541,
+      "learning_rate": 1.9467849223946788e-05,
+      "loss": 2.9768,
       "step": 637
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 14.796496391296387,
-      "learning_rate": 1.872306304500351e-05,
-      "loss": 2.572,
+      "epoch": 0.08,
+      "grad_norm": 17.28792953491211,
+      "learning_rate": 1.9467012508890098e-05,
+      "loss": 3.3681,
       "step": 638
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 14.737870216369629,
-      "learning_rate": 1.872105843439912e-05,
-      "loss": 4.6644,
+      "epoch": 0.08,
+      "grad_norm": 134.4596405029297,
+      "learning_rate": 1.946617579383341e-05,
+      "loss": 2.3149,
       "step": 639
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 14.593988418579102,
-      "learning_rate": 1.8719053823794727e-05,
-      "loss": 2.9476,
+      "epoch": 0.08,
+      "grad_norm": 15.561910629272461,
+      "learning_rate": 1.9465339078776725e-05,
+      "loss": 3.1515,
       "step": 640
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 24.125364303588867,
-      "learning_rate": 1.871704921319034e-05,
-      "loss": 3.0569,
+      "epoch": 0.08,
+      "grad_norm": 23.368066787719727,
+      "learning_rate": 1.9464502363720036e-05,
+      "loss": 4.4627,
       "step": 641
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 12.783120155334473,
-      "learning_rate": 1.871504460258595e-05,
-      "loss": 2.7435,
+      "epoch": 0.08,
+      "grad_norm": 18.397462844848633,
+      "learning_rate": 1.946366564866335e-05,
+      "loss": 3.3775,
       "step": 642
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 20.349702835083008,
-      "learning_rate": 1.8713039991981557e-05,
-      "loss": 3.924,
+      "epoch": 0.08,
+      "grad_norm": 13.091896057128906,
+      "learning_rate": 1.946282893360666e-05,
+      "loss": 5.6346,
       "step": 643
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 18.848377227783203,
-      "learning_rate": 1.871103538137717e-05,
-      "loss": 3.6072,
+      "epoch": 0.08,
+      "grad_norm": 12.951624870300293,
+      "learning_rate": 1.9461992218549973e-05,
+      "loss": 5.0077,
       "step": 644
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 16.555147171020508,
-      "learning_rate": 1.8709030770772777e-05,
-      "loss": 2.6862,
+      "epoch": 0.08,
+      "grad_norm": 9.754615783691406,
+      "learning_rate": 1.9461155503493287e-05,
+      "loss": 3.2439,
       "step": 645
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 93.45834350585938,
-      "learning_rate": 1.8707026160168387e-05,
-      "loss": 3.1187,
+      "epoch": 0.08,
+      "grad_norm": 11.347737312316895,
+      "learning_rate": 1.9460318788436597e-05,
+      "loss": 2.7666,
       "step": 646
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 12.666962623596191,
-      "learning_rate": 1.8705021549564e-05,
-      "loss": 3.3528,
+      "epoch": 0.08,
+      "grad_norm": 45.7727165222168,
+      "learning_rate": 1.945948207337991e-05,
+      "loss": 4.1038,
       "step": 647
     },
     {
-      "epoch": 0.19,
-      "grad_norm": 13.548149108886719,
-      "learning_rate": 1.8703016938959608e-05,
-      "loss": 2.8839,
+      "epoch": 0.08,
+      "grad_norm": 12.478466033935547,
+      "learning_rate": 1.9458645358323224e-05,
+      "loss": 4.6921,
       "step": 648
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 20.26407241821289,
-      "learning_rate": 1.8701012328355218e-05,
-      "loss": 3.2999,
+      "epoch": 0.08,
+      "grad_norm": 21.725603103637695,
+      "learning_rate": 1.9457808643266538e-05,
+      "loss": 3.9947,
       "step": 649
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 16.36279296875,
-      "learning_rate": 1.8699007717750828e-05,
-      "loss": 2.7201,
+      "epoch": 0.08,
+      "grad_norm": 19.537290573120117,
+      "learning_rate": 1.9456971928209848e-05,
+      "loss": 5.4066,
       "step": 650
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 15.859626770019531,
-      "learning_rate": 1.8697003107146438e-05,
-      "loss": 3.3978,
+      "epoch": 0.08,
+      "grad_norm": 16.776832580566406,
+      "learning_rate": 1.9456135213153162e-05,
+      "loss": 4.4481,
       "step": 651
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 20.308744430541992,
-      "learning_rate": 1.8694998496542048e-05,
-      "loss": 3.7883,
+      "epoch": 0.08,
+      "grad_norm": 12.141162872314453,
+      "learning_rate": 1.9455298498096476e-05,
+      "loss": 2.9702,
       "step": 652
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 14.849808692932129,
-      "learning_rate": 1.8692993885937658e-05,
-      "loss": 3.0277,
+      "epoch": 0.08,
+      "grad_norm": 14.00488567352295,
+      "learning_rate": 1.9454461783039786e-05,
+      "loss": 3.5068,
       "step": 653
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 10.114745140075684,
-      "learning_rate": 1.8690989275333268e-05,
-      "loss": 2.4441,
+      "epoch": 0.08,
+      "grad_norm": 21.130966186523438,
+      "learning_rate": 1.94536250679831e-05,
+      "loss": 4.976,
       "step": 654
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 21.375455856323242,
-      "learning_rate": 1.8688984664728878e-05,
-      "loss": 2.9501,
+      "epoch": 0.08,
+      "grad_norm": 79.34236907958984,
+      "learning_rate": 1.9452788352926413e-05,
+      "loss": 2.7563,
       "step": 655
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 21.958810806274414,
-      "learning_rate": 1.8686980054124488e-05,
-      "loss": 2.8862,
+      "epoch": 0.08,
+      "grad_norm": 22.180267333984375,
+      "learning_rate": 1.9451951637869727e-05,
+      "loss": 3.4531,
       "step": 656
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 14.946088790893555,
-      "learning_rate": 1.86849754435201e-05,
-      "loss": 3.4693,
+      "epoch": 0.08,
+      "grad_norm": 26.059926986694336,
+      "learning_rate": 1.9451114922813037e-05,
+      "loss": 3.8926,
       "step": 657
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 13.369003295898438,
-      "learning_rate": 1.868297083291571e-05,
-      "loss": 4.4842,
+      "epoch": 0.08,
+      "grad_norm": 16.83011245727539,
+      "learning_rate": 1.945027820775635e-05,
+      "loss": 3.0436,
       "step": 658
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 13.760272979736328,
-      "learning_rate": 1.868096622231132e-05,
-      "loss": 4.0259,
+      "epoch": 0.08,
+      "grad_norm": 13.718125343322754,
+      "learning_rate": 1.9449441492699664e-05,
+      "loss": 3.8004,
       "step": 659
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 18.290048599243164,
-      "learning_rate": 1.867896161170693e-05,
-      "loss": 2.4139,
+      "epoch": 0.08,
+      "grad_norm": 12.119871139526367,
+      "learning_rate": 1.9448604777642975e-05,
+      "loss": 3.8753,
       "step": 660
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 20.004146575927734,
-      "learning_rate": 1.867695700110254e-05,
-      "loss": 3.8439,
+      "epoch": 0.08,
+      "grad_norm": 22.682151794433594,
+      "learning_rate": 1.9447768062586288e-05,
+      "loss": 4.6795,
       "step": 661
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 15.349987983703613,
-      "learning_rate": 1.8674952390498145e-05,
-      "loss": 2.9132,
+      "epoch": 0.08,
+      "grad_norm": 17.96092987060547,
+      "learning_rate": 1.9446931347529602e-05,
+      "loss": 1.6681,
       "step": 662
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 14.10637092590332,
-      "learning_rate": 1.867294777989376e-05,
-      "loss": 2.9145,
+      "epoch": 0.08,
+      "grad_norm": 13.491999626159668,
+      "learning_rate": 1.9446094632472915e-05,
+      "loss": 3.6529,
       "step": 663
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 15.24035358428955,
-      "learning_rate": 1.8670943169289366e-05,
-      "loss": 2.748,
+      "epoch": 0.08,
+      "grad_norm": 12.081392288208008,
+      "learning_rate": 1.9445257917416226e-05,
+      "loss": 3.4167,
       "step": 664
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 23.505083084106445,
-      "learning_rate": 1.8668938558684976e-05,
-      "loss": 2.8452,
+      "epoch": 0.08,
+      "grad_norm": 16.117155075073242,
+      "learning_rate": 1.944442120235954e-05,
+      "loss": 4.0835,
       "step": 665
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 82.25447845458984,
-      "learning_rate": 1.866693394808059e-05,
-      "loss": 2.8634,
+      "epoch": 0.08,
+      "grad_norm": 12.962418556213379,
+      "learning_rate": 1.944358448730285e-05,
+      "loss": 3.6016,
       "step": 666
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 24.59209632873535,
-      "learning_rate": 1.8664929337476196e-05,
-      "loss": 2.9632,
+      "epoch": 0.08,
+      "grad_norm": 11.88910961151123,
+      "learning_rate": 1.9442747772246163e-05,
+      "loss": 1.6991,
       "step": 667
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 18.12932014465332,
-      "learning_rate": 1.8662924726871806e-05,
-      "loss": 3.9964,
+      "epoch": 0.08,
+      "grad_norm": 13.218146324157715,
+      "learning_rate": 1.9441911057189477e-05,
+      "loss": 3.126,
       "step": 668
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 13.9307222366333,
-      "learning_rate": 1.8660920116267416e-05,
-      "loss": 3.37,
+      "epoch": 0.08,
+      "grad_norm": 11.371163368225098,
+      "learning_rate": 1.9441074342132787e-05,
+      "loss": 2.4902,
       "step": 669
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 15.804916381835938,
-      "learning_rate": 1.8658915505663026e-05,
-      "loss": 3.7911,
+      "epoch": 0.08,
+      "grad_norm": 23.65850257873535,
+      "learning_rate": 1.94402376270761e-05,
+      "loss": 2.7793,
       "step": 670
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 12.954413414001465,
-      "learning_rate": 1.8656910895058636e-05,
-      "loss": 3.4076,
+      "epoch": 0.08,
+      "grad_norm": 10.849726676940918,
+      "learning_rate": 1.943940091201941e-05,
+      "loss": 1.7024,
       "step": 671
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 19.96038055419922,
-      "learning_rate": 1.8654906284454246e-05,
-      "loss": 2.6706,
+      "epoch": 0.08,
+      "grad_norm": 19.88506317138672,
+      "learning_rate": 1.9438564196962725e-05,
+      "loss": 2.9413,
       "step": 672
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 14.459244728088379,
-      "learning_rate": 1.8652901673849856e-05,
-      "loss": 2.6447,
+      "epoch": 0.08,
+      "grad_norm": 43.79939270019531,
+      "learning_rate": 1.943772748190604e-05,
+      "loss": 2.4117,
       "step": 673
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 14.168243408203125,
-      "learning_rate": 1.8650897063245466e-05,
-      "loss": 2.8916,
+      "epoch": 0.08,
+      "grad_norm": 17.82375144958496,
+      "learning_rate": 1.943689076684935e-05,
+      "loss": 2.5445,
       "step": 674
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 12.294631004333496,
-      "learning_rate": 1.8648892452641076e-05,
-      "loss": 3.911,
+      "epoch": 0.08,
+      "grad_norm": 14.555415153503418,
+      "learning_rate": 1.9436054051792662e-05,
+      "loss": 3.7538,
       "step": 675
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 16.20514488220215,
-      "learning_rate": 1.8646887842036687e-05,
-      "loss": 4.1275,
+      "epoch": 0.08,
+      "grad_norm": 12.511565208435059,
+      "learning_rate": 1.9435217336735976e-05,
+      "loss": 2.6467,
       "step": 676
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 13.80794906616211,
-      "learning_rate": 1.8644883231432297e-05,
-      "loss": 2.5423,
+      "epoch": 0.08,
+      "grad_norm": 8.727785110473633,
+      "learning_rate": 1.943438062167929e-05,
+      "loss": 2.6967,
       "step": 677
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 13.199609756469727,
-      "learning_rate": 1.8642878620827907e-05,
-      "loss": 3.119,
+      "epoch": 0.09,
+      "grad_norm": 9.602141380310059,
+      "learning_rate": 1.94335439066226e-05,
+      "loss": 2.9577,
       "step": 678
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 17.830242156982422,
-      "learning_rate": 1.8640874010223517e-05,
-      "loss": 3.9047,
+      "epoch": 0.09,
+      "grad_norm": 33.51832962036133,
+      "learning_rate": 1.9432707191565914e-05,
+      "loss": 5.2958,
       "step": 679
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 17.04703140258789,
-      "learning_rate": 1.8638869399619127e-05,
-      "loss": 2.4903,
+      "epoch": 0.09,
+      "grad_norm": 14.140719413757324,
+      "learning_rate": 1.9431870476509227e-05,
+      "loss": 4.4345,
       "step": 680
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 16.005630493164062,
-      "learning_rate": 1.8636864789014734e-05,
-      "loss": 3.5141,
+      "epoch": 0.09,
+      "grad_norm": 18.002798080444336,
+      "learning_rate": 1.9431033761452537e-05,
+      "loss": 4.8873,
       "step": 681
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 12.92525577545166,
-      "learning_rate": 1.8634860178410347e-05,
-      "loss": 3.5435,
+      "epoch": 0.09,
+      "grad_norm": 18.150636672973633,
+      "learning_rate": 1.943019704639585e-05,
+      "loss": 2.9637,
       "step": 682
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 13.173433303833008,
-      "learning_rate": 1.8632855567805954e-05,
-      "loss": 3.4652,
+      "epoch": 0.09,
+      "grad_norm": 35.52592086791992,
+      "learning_rate": 1.9429360331339165e-05,
+      "loss": 3.5024,
       "step": 683
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 15.04622745513916,
-      "learning_rate": 1.8630850957201564e-05,
-      "loss": 3.2746,
+      "epoch": 0.09,
+      "grad_norm": 16.548412322998047,
+      "learning_rate": 1.942852361628248e-05,
+      "loss": 3.9262,
       "step": 684
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 16.044187545776367,
-      "learning_rate": 1.8628846346597177e-05,
-      "loss": 3.5208,
+      "epoch": 0.09,
+      "grad_norm": 12.454635620117188,
+      "learning_rate": 1.942768690122579e-05,
+      "loss": 3.7851,
       "step": 685
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 12.709644317626953,
-      "learning_rate": 1.8626841735992784e-05,
-      "loss": 2.9898,
+      "epoch": 0.09,
+      "grad_norm": 35.238014221191406,
+      "learning_rate": 1.9426850186169102e-05,
+      "loss": 4.1137,
       "step": 686
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 62.63556671142578,
-      "learning_rate": 1.8624837125388394e-05,
-      "loss": 4.3599,
+      "epoch": 0.09,
+      "grad_norm": 14.9273099899292,
+      "learning_rate": 1.9426013471112416e-05,
+      "loss": 3.4451,
       "step": 687
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 14.936622619628906,
-      "learning_rate": 1.8622832514784004e-05,
-      "loss": 3.7374,
+      "epoch": 0.09,
+      "grad_norm": 12.957579612731934,
+      "learning_rate": 1.9425176756055726e-05,
+      "loss": 2.667,
       "step": 688
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 18.37685203552246,
-      "learning_rate": 1.8620827904179614e-05,
-      "loss": 2.8972,
+      "epoch": 0.09,
+      "grad_norm": 22.085285186767578,
+      "learning_rate": 1.942434004099904e-05,
+      "loss": 4.9414,
       "step": 689
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 18.769283294677734,
-      "learning_rate": 1.8618823293575224e-05,
-      "loss": 2.7669,
+      "epoch": 0.09,
+      "grad_norm": 12.428492546081543,
+      "learning_rate": 1.9423503325942354e-05,
+      "loss": 2.9308,
       "step": 690
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 14.026557922363281,
-      "learning_rate": 1.8616818682970834e-05,
-      "loss": 3.4005,
+      "epoch": 0.09,
+      "grad_norm": 21.980300903320312,
+      "learning_rate": 1.9422666610885664e-05,
+      "loss": 3.3054,
       "step": 691
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 15.341630935668945,
-      "learning_rate": 1.8614814072366445e-05,
-      "loss": 3.8892,
+      "epoch": 0.09,
+      "grad_norm": 10.78618335723877,
+      "learning_rate": 1.9421829895828977e-05,
+      "loss": 0.7016,
       "step": 692
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 20.44842529296875,
-      "learning_rate": 1.8612809461762055e-05,
-      "loss": 4.4409,
+      "epoch": 0.09,
+      "grad_norm": 14.255931854248047,
+      "learning_rate": 1.942099318077229e-05,
+      "loss": 4.1619,
       "step": 693
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 23.54950523376465,
-      "learning_rate": 1.8610804851157665e-05,
-      "loss": 3.9646,
+      "epoch": 0.09,
+      "grad_norm": 11.36681842803955,
+      "learning_rate": 1.94201564657156e-05,
+      "loss": 1.7237,
       "step": 694
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 13.78188419342041,
-      "learning_rate": 1.860880024055327e-05,
-      "loss": 4.5521,
+      "epoch": 0.09,
+      "grad_norm": 14.176616668701172,
+      "learning_rate": 1.9419319750658915e-05,
+      "loss": 3.3943,
       "step": 695
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 21.157424926757812,
-      "learning_rate": 1.8606795629948885e-05,
-      "loss": 3.6927,
+      "epoch": 0.09,
+      "grad_norm": 35.78500747680664,
+      "learning_rate": 1.9418483035602225e-05,
+      "loss": 3.7633,
       "step": 696
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 12.406957626342773,
-      "learning_rate": 1.8604791019344495e-05,
-      "loss": 3.2458,
+      "epoch": 0.09,
+      "grad_norm": 11.32670783996582,
+      "learning_rate": 1.941764632054554e-05,
+      "loss": 1.5977,
       "step": 697
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 18.12921142578125,
-      "learning_rate": 1.86027864087401e-05,
-      "loss": 3.7537,
+      "epoch": 0.09,
+      "grad_norm": 9.296537399291992,
+      "learning_rate": 1.9416809605488853e-05,
+      "loss": 1.8491,
       "step": 698
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 21.271060943603516,
-      "learning_rate": 1.8600781798135715e-05,
-      "loss": 4.3055,
+      "epoch": 0.09,
+      "grad_norm": 12.421931266784668,
+      "learning_rate": 1.9415972890432163e-05,
+      "loss": 2.4607,
       "step": 699
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 13.396303176879883,
-      "learning_rate": 1.8598777187531322e-05,
-      "loss": 3.5275,
+      "epoch": 0.09,
+      "grad_norm": 11.723788261413574,
+      "learning_rate": 1.9415136175375476e-05,
+      "loss": 3.9441,
       "step": 700
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 11.93570613861084,
-      "learning_rate": 1.8596772576926935e-05,
-      "loss": 3.0972,
+      "epoch": 0.09,
+      "grad_norm": 120.77423095703125,
+      "learning_rate": 1.941429946031879e-05,
+      "loss": 2.6284,
       "step": 701
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 14.72830581665039,
-      "learning_rate": 1.8594767966322545e-05,
-      "loss": 4.0018,
+      "epoch": 0.09,
+      "grad_norm": 13.209537506103516,
+      "learning_rate": 1.94134627452621e-05,
+      "loss": 2.1441,
       "step": 702
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 23.694110870361328,
-      "learning_rate": 1.8592763355718152e-05,
-      "loss": 3.2218,
+      "epoch": 0.09,
+      "grad_norm": 10.035263061523438,
+      "learning_rate": 1.9412626030205414e-05,
+      "loss": 2.6173,
       "step": 703
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 17.545129776000977,
-      "learning_rate": 1.8590758745113766e-05,
-      "loss": 3.2768,
+      "epoch": 0.09,
+      "grad_norm": 15.197259902954102,
+      "learning_rate": 1.9411789315148728e-05,
+      "loss": 4.7402,
       "step": 704
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 16.36639976501465,
-      "learning_rate": 1.8588754134509372e-05,
-      "loss": 3.1514,
+      "epoch": 0.09,
+      "grad_norm": 15.673565864562988,
+      "learning_rate": 1.941095260009204e-05,
+      "loss": 3.5522,
       "step": 705
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 11.386521339416504,
-      "learning_rate": 1.8586749523904982e-05,
-      "loss": 2.3357,
+      "epoch": 0.09,
+      "grad_norm": 13.20335578918457,
+      "learning_rate": 1.941011588503535e-05,
+      "loss": 4.3,
       "step": 706
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 10.972953796386719,
-      "learning_rate": 1.8584744913300592e-05,
-      "loss": 2.8289,
+      "epoch": 0.09,
+      "grad_norm": 25.170381546020508,
+      "learning_rate": 1.9409279169978665e-05,
+      "loss": 5.6345,
       "step": 707
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 16.927831649780273,
-      "learning_rate": 1.8582740302696202e-05,
-      "loss": 3.086,
+      "epoch": 0.09,
+      "grad_norm": 20.28534507751465,
+      "learning_rate": 1.940844245492198e-05,
+      "loss": 4.7721,
       "step": 708
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 14.967674255371094,
-      "learning_rate": 1.8580735692091813e-05,
-      "loss": 3.6209,
+      "epoch": 0.09,
+      "grad_norm": 22.23134422302246,
+      "learning_rate": 1.940760573986529e-05,
+      "loss": 3.1,
       "step": 709
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 16.293773651123047,
-      "learning_rate": 1.8578731081487423e-05,
-      "loss": 4.2359,
+      "epoch": 0.09,
+      "grad_norm": 11.009767532348633,
+      "learning_rate": 1.9406769024808603e-05,
+      "loss": 0.5144,
       "step": 710
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 14.0774564743042,
-      "learning_rate": 1.8576726470883033e-05,
-      "loss": 3.7756,
+      "epoch": 0.09,
+      "grad_norm": 15.88630199432373,
+      "learning_rate": 1.9405932309751916e-05,
+      "loss": 4.2522,
       "step": 711
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 26.526830673217773,
-      "learning_rate": 1.8574721860278643e-05,
-      "loss": 3.1339,
+      "epoch": 0.09,
+      "grad_norm": 13.1906156539917,
+      "learning_rate": 1.940509559469523e-05,
+      "loss": 2.494,
       "step": 712
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 13.169925689697266,
-      "learning_rate": 1.8572717249674253e-05,
-      "loss": 2.6322,
+      "epoch": 0.09,
+      "grad_norm": 21.883312225341797,
+      "learning_rate": 1.940425887963854e-05,
+      "loss": 2.4317,
       "step": 713
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 13.879220962524414,
-      "learning_rate": 1.857071263906986e-05,
-      "loss": 2.5253,
+      "epoch": 0.09,
+      "grad_norm": 11.292802810668945,
+      "learning_rate": 1.9403422164581854e-05,
+      "loss": 2.2441,
       "step": 714
     },
     {
-      "epoch": 0.21,
-      "grad_norm": 15.33066177368164,
-      "learning_rate": 1.8568708028465473e-05,
-      "loss": 3.3604,
+      "epoch": 0.09,
+      "grad_norm": 13.797185897827148,
+      "learning_rate": 1.9402585449525168e-05,
+      "loss": 4.7378,
       "step": 715
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 16.04281997680664,
-      "learning_rate": 1.8566703417861083e-05,
-      "loss": 2.824,
+      "epoch": 0.09,
+      "grad_norm": 22.5351505279541,
+      "learning_rate": 1.9401748734468478e-05,
+      "loss": 4.5674,
       "step": 716
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 13.022286415100098,
-      "learning_rate": 1.856469880725669e-05,
-      "loss": 2.6692,
+      "epoch": 0.09,
+      "grad_norm": 8.449362754821777,
+      "learning_rate": 1.940091201941179e-05,
+      "loss": 1.1736,
       "step": 717
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 14.34041976928711,
-      "learning_rate": 1.8562694196652303e-05,
-      "loss": 3.0903,
+      "epoch": 0.09,
+      "grad_norm": 13.076533317565918,
+      "learning_rate": 1.9400075304355105e-05,
+      "loss": 3.7764,
       "step": 718
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 17.68020248413086,
-      "learning_rate": 1.856068958604791e-05,
-      "loss": 3.4143,
+      "epoch": 0.09,
+      "grad_norm": 14.160146713256836,
+      "learning_rate": 1.9399238589298415e-05,
+      "loss": 2.7199,
       "step": 719
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 13.933920860290527,
-      "learning_rate": 1.855868497544352e-05,
-      "loss": 3.2975,
-      "step": 720
-    },
-    {
-      "epoch": 0.22,
-      "eval_loss": 0.7311321496963501,
-      "eval_runtime": 43.9321,
-      "eval_samples_per_second": 33.666,
-      "eval_steps_per_second": 33.666,
+      "epoch": 0.09,
+      "grad_norm": 11.873703956604004,
+      "learning_rate": 1.939840187424173e-05,
+      "loss": 2.8223,
       "step": 720
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 14.72537612915039,
-      "learning_rate": 1.8556680364839134e-05,
-      "loss": 3.9612,
+      "epoch": 0.09,
+      "grad_norm": 12.563404083251953,
+      "learning_rate": 1.9397565159185043e-05,
+      "loss": 2.5746,
       "step": 721
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 26.851642608642578,
-      "learning_rate": 1.855467575423474e-05,
-      "loss": 3.4687,
+      "epoch": 0.09,
+      "grad_norm": 10.250195503234863,
+      "learning_rate": 1.9396728444128353e-05,
+      "loss": 2.0382,
       "step": 722
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 16.372344970703125,
-      "learning_rate": 1.855267114363035e-05,
-      "loss": 3.7605,
+      "epoch": 0.09,
+      "grad_norm": 7.887467384338379,
+      "learning_rate": 1.9395891729071667e-05,
+      "loss": 0.775,
       "step": 723
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 17.88177490234375,
-      "learning_rate": 1.855066653302596e-05,
-      "loss": 3.4436,
+      "epoch": 0.09,
+      "grad_norm": 16.83657455444336,
+      "learning_rate": 1.9395055014014977e-05,
+      "loss": 5.009,
       "step": 724
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 16.162473678588867,
-      "learning_rate": 1.854866192242157e-05,
-      "loss": 3.4066,
+      "epoch": 0.09,
+      "grad_norm": 11.711037635803223,
+      "learning_rate": 1.939421829895829e-05,
+      "loss": 3.1593,
       "step": 725
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 12.335894584655762,
-      "learning_rate": 1.854665731181718e-05,
-      "loss": 2.9006,
+      "epoch": 0.09,
+      "grad_norm": 17.11760711669922,
+      "learning_rate": 1.9393381583901604e-05,
+      "loss": 4.1511,
       "step": 726
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 19.66415786743164,
-      "learning_rate": 1.854465270121279e-05,
-      "loss": 3.5752,
+      "epoch": 0.09,
+      "grad_norm": 18.830835342407227,
+      "learning_rate": 1.9392544868844914e-05,
+      "loss": 4.5674,
       "step": 727
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 11.808516502380371,
-      "learning_rate": 1.85426480906084e-05,
-      "loss": 1.8741,
+      "epoch": 0.09,
+      "grad_norm": 13.822317123413086,
+      "learning_rate": 1.9391708153788228e-05,
+      "loss": 2.7104,
       "step": 728
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 15.641440391540527,
-      "learning_rate": 1.854064348000401e-05,
-      "loss": 3.3391,
+      "epoch": 0.09,
+      "grad_norm": 11.05170726776123,
+      "learning_rate": 1.9390871438731542e-05,
+      "loss": 2.7689,
       "step": 729
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 13.590608596801758,
-      "learning_rate": 1.853863886939962e-05,
-      "loss": 2.9265,
+      "epoch": 0.09,
+      "grad_norm": 13.445638656616211,
+      "learning_rate": 1.9390034723674852e-05,
+      "loss": 2.9387,
       "step": 730
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 18.237276077270508,
-      "learning_rate": 1.853663425879523e-05,
-      "loss": 3.6717,
+      "epoch": 0.09,
+      "grad_norm": 22.369338989257812,
+      "learning_rate": 1.9389198008618166e-05,
+      "loss": 3.1684,
       "step": 731
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 41.96309280395508,
-      "learning_rate": 1.853462964819084e-05,
-      "loss": 2.857,
+      "epoch": 0.09,
+      "grad_norm": 11.012389183044434,
+      "learning_rate": 1.938836129356148e-05,
+      "loss": 2.211,
       "step": 732
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 15.107841491699219,
-      "learning_rate": 1.853262503758645e-05,
-      "loss": 3.9187,
+      "epoch": 0.09,
+      "grad_norm": 11.709720611572266,
+      "learning_rate": 1.9387524578504793e-05,
+      "loss": 2.822,
       "step": 733
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 15.574593544006348,
-      "learning_rate": 1.853062042698206e-05,
-      "loss": 3.1606,
+      "epoch": 0.09,
+      "grad_norm": 13.748392105102539,
+      "learning_rate": 1.9386687863448103e-05,
+      "loss": 2.3774,
       "step": 734
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 21.812761306762695,
-      "learning_rate": 1.852861581637767e-05,
-      "loss": 2.8953,
+      "epoch": 0.09,
+      "grad_norm": 17.709924697875977,
+      "learning_rate": 1.9385851148391417e-05,
+      "loss": 2.5742,
       "step": 735
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 21.730745315551758,
-      "learning_rate": 1.8526611205773278e-05,
-      "loss": 3.3289,
+      "epoch": 0.09,
+      "grad_norm": 35.1605110168457,
+      "learning_rate": 1.938501443333473e-05,
+      "loss": 1.6553,
       "step": 736
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 33.00816345214844,
-      "learning_rate": 1.852460659516889e-05,
-      "loss": 2.98,
+      "epoch": 0.09,
+      "grad_norm": 17.295238494873047,
+      "learning_rate": 1.938417771827804e-05,
+      "loss": 2.2971,
       "step": 737
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 16.756553649902344,
-      "learning_rate": 1.8522601984564498e-05,
-      "loss": 3.0438,
+      "epoch": 0.09,
+      "grad_norm": 14.092886924743652,
+      "learning_rate": 1.9383341003221354e-05,
+      "loss": 2.3941,
       "step": 738
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 24.259780883789062,
-      "learning_rate": 1.852059737396011e-05,
-      "loss": 3.9168,
+      "epoch": 0.09,
+      "grad_norm": 20.23340606689453,
+      "learning_rate": 1.9382504288164668e-05,
+      "loss": 1.5338,
       "step": 739
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 10.961197853088379,
-      "learning_rate": 1.8518592763355722e-05,
-      "loss": 2.521,
+      "epoch": 0.09,
+      "grad_norm": 15.133199691772461,
+      "learning_rate": 1.9381667573107982e-05,
+      "loss": 1.0674,
       "step": 740
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 11.985860824584961,
-      "learning_rate": 1.851658815275133e-05,
-      "loss": 3.6082,
+      "epoch": 0.09,
+      "grad_norm": 14.169252395629883,
+      "learning_rate": 1.9380830858051292e-05,
+      "loss": 2.9105,
       "step": 741
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 13.038666725158691,
-      "learning_rate": 1.851458354214694e-05,
-      "loss": 2.8915,
+      "epoch": 0.09,
+      "grad_norm": 153.477783203125,
+      "learning_rate": 1.9379994142994606e-05,
+      "loss": 2.8067,
       "step": 742
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 15.792445182800293,
-      "learning_rate": 1.851257893154255e-05,
-      "loss": 3.0136,
+      "epoch": 0.09,
+      "grad_norm": 7.675416946411133,
+      "learning_rate": 1.937915742793792e-05,
+      "loss": 2.2885,
       "step": 743
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 11.698639869689941,
-      "learning_rate": 1.851057432093816e-05,
-      "loss": 2.9538,
+      "epoch": 0.09,
+      "grad_norm": 16.272003173828125,
+      "learning_rate": 1.937832071288123e-05,
+      "loss": 2.3131,
       "step": 744
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 14.640687942504883,
-      "learning_rate": 1.850856971033377e-05,
-      "loss": 2.1114,
+      "epoch": 0.09,
+      "grad_norm": 104.13090515136719,
+      "learning_rate": 1.9377483997824543e-05,
+      "loss": 3.6301,
       "step": 745
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 21.3159122467041,
-      "learning_rate": 1.850656509972938e-05,
-      "loss": 3.1929,
+      "epoch": 0.09,
+      "grad_norm": 19.843830108642578,
+      "learning_rate": 1.9376647282767857e-05,
+      "loss": 1.6103,
       "step": 746
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 15.516459465026855,
-      "learning_rate": 1.850456048912499e-05,
-      "loss": 3.4558,
+      "epoch": 0.09,
+      "grad_norm": 33.996437072753906,
+      "learning_rate": 1.9375810567711167e-05,
+      "loss": 5.4643,
       "step": 747
     },
     {
-      "epoch": 0.22,
-      "grad_norm": 12.31661605834961,
-      "learning_rate": 1.85025558785206e-05,
-      "loss": 3.5311,
+      "epoch": 0.09,
+      "grad_norm": 9.253028869628906,
+      "learning_rate": 1.937497385265448e-05,
+      "loss": 1.7174,
       "step": 748
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 17.94566535949707,
-      "learning_rate": 1.850055126791621e-05,
-      "loss": 3.6713,
+      "epoch": 0.09,
+      "grad_norm": 19.09623908996582,
+      "learning_rate": 1.937413713759779e-05,
+      "loss": 2.9651,
       "step": 749
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 13.146416664123535,
-      "learning_rate": 1.849854665731182e-05,
-      "loss": 2.0929,
+      "epoch": 0.09,
+      "grad_norm": 16.510766983032227,
+      "learning_rate": 1.9373300422541105e-05,
+      "loss": 5.6302,
       "step": 750
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 15.036728858947754,
-      "learning_rate": 1.849654204670743e-05,
-      "loss": 3.2656,
+      "epoch": 0.09,
+      "grad_norm": 13.26391315460205,
+      "learning_rate": 1.937246370748442e-05,
+      "loss": 2.3617,
       "step": 751
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 12.377967834472656,
-      "learning_rate": 1.849453743610304e-05,
-      "loss": 2.811,
+      "epoch": 0.09,
+      "grad_norm": 16.245691299438477,
+      "learning_rate": 1.937162699242773e-05,
+      "loss": 2.2042,
       "step": 752
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 48.185482025146484,
-      "learning_rate": 1.849253282549865e-05,
-      "loss": 4.3531,
+      "epoch": 0.09,
+      "grad_norm": 14.730291366577148,
+      "learning_rate": 1.9370790277371042e-05,
+      "loss": 4.1242,
       "step": 753
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 16.886388778686523,
-      "learning_rate": 1.849052821489426e-05,
-      "loss": 3.3402,
+      "epoch": 0.09,
+      "grad_norm": 15.235692977905273,
+      "learning_rate": 1.9369953562314356e-05,
+      "loss": 2.8035,
       "step": 754
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 10.66983699798584,
-      "learning_rate": 1.8488523604289866e-05,
-      "loss": 2.5462,
+      "epoch": 0.09,
+      "grad_norm": 8.804192543029785,
+      "learning_rate": 1.9369116847257666e-05,
+      "loss": 3.199,
       "step": 755
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 13.7207670211792,
-      "learning_rate": 1.848651899368548e-05,
-      "loss": 4.2536,
+      "epoch": 0.09,
+      "grad_norm": 14.533276557922363,
+      "learning_rate": 1.936828013220098e-05,
+      "loss": 4.2,
       "step": 756
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 11.786395072937012,
-      "learning_rate": 1.8484514383081086e-05,
-      "loss": 3.3007,
+      "epoch": 0.1,
+      "grad_norm": 12.746406555175781,
+      "learning_rate": 1.9367443417144293e-05,
+      "loss": 2.9343,
       "step": 757
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 19.73274803161621,
-      "learning_rate": 1.8482509772476697e-05,
-      "loss": 3.2301,
+      "epoch": 0.1,
+      "grad_norm": 9.367931365966797,
+      "learning_rate": 1.9366606702087604e-05,
+      "loss": 1.7168,
       "step": 758
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 15.213275909423828,
-      "learning_rate": 1.848050516187231e-05,
-      "loss": 3.0059,
+      "epoch": 0.1,
+      "grad_norm": 17.672361373901367,
+      "learning_rate": 1.9365769987030917e-05,
+      "loss": 4.2855,
       "step": 759
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 19.37066078186035,
-      "learning_rate": 1.8478500551267917e-05,
-      "loss": 2.8112,
+      "epoch": 0.1,
+      "grad_norm": 24.997514724731445,
+      "learning_rate": 1.936493327197423e-05,
+      "loss": 4.1359,
       "step": 760
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 48.94891357421875,
-      "learning_rate": 1.8476495940663527e-05,
-      "loss": 4.0552,
+      "epoch": 0.1,
+      "grad_norm": 13.02668285369873,
+      "learning_rate": 1.9364096556917545e-05,
+      "loss": 2.802,
       "step": 761
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 11.666791915893555,
-      "learning_rate": 1.8474491330059137e-05,
-      "loss": 3.0329,
+      "epoch": 0.1,
+      "grad_norm": 16.64696502685547,
+      "learning_rate": 1.9363259841860855e-05,
+      "loss": 2.6909,
       "step": 762
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 14.530978202819824,
-      "learning_rate": 1.8472486719454747e-05,
-      "loss": 2.9601,
+      "epoch": 0.1,
+      "grad_norm": 10.787168502807617,
+      "learning_rate": 1.936242312680417e-05,
+      "loss": 1.8418,
       "step": 763
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 15.476301193237305,
-      "learning_rate": 1.8470482108850357e-05,
-      "loss": 3.8685,
+      "epoch": 0.1,
+      "grad_norm": 24.767377853393555,
+      "learning_rate": 1.9361586411747482e-05,
+      "loss": 5.6238,
       "step": 764
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 34.36149597167969,
-      "learning_rate": 1.8468477498245967e-05,
-      "loss": 3.6309,
+      "epoch": 0.1,
+      "grad_norm": 12.260977745056152,
+      "learning_rate": 1.9360749696690792e-05,
+      "loss": 3.3651,
       "step": 765
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 13.753215789794922,
-      "learning_rate": 1.8466472887641577e-05,
-      "loss": 2.5185,
+      "epoch": 0.1,
+      "grad_norm": 15.056289672851562,
+      "learning_rate": 1.9359912981634106e-05,
+      "loss": 2.7135,
       "step": 766
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 14.692520141601562,
-      "learning_rate": 1.8464468277037187e-05,
-      "loss": 3.4938,
+      "epoch": 0.1,
+      "grad_norm": 27.636878967285156,
+      "learning_rate": 1.935907626657742e-05,
+      "loss": 3.2168,
       "step": 767
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 19.021665573120117,
-      "learning_rate": 1.8462463666432797e-05,
-      "loss": 2.4541,
+      "epoch": 0.1,
+      "grad_norm": 13.80523681640625,
+      "learning_rate": 1.9358239551520733e-05,
+      "loss": 4.6771,
       "step": 768
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 15.532613754272461,
-      "learning_rate": 1.8460459055828407e-05,
-      "loss": 3.5129,
+      "epoch": 0.1,
+      "grad_norm": 12.087339401245117,
+      "learning_rate": 1.9357402836464044e-05,
+      "loss": 3.8014,
       "step": 769
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 19.325273513793945,
-      "learning_rate": 1.8458454445224018e-05,
-      "loss": 3.8114,
+      "epoch": 0.1,
+      "grad_norm": 13.019579887390137,
+      "learning_rate": 1.9356566121407357e-05,
+      "loss": 2.8844,
       "step": 770
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 19.3370361328125,
-      "learning_rate": 1.8456449834619628e-05,
-      "loss": 3.7453,
+      "epoch": 0.1,
+      "grad_norm": 10.89457893371582,
+      "learning_rate": 1.935572940635067e-05,
+      "loss": 2.0939,
       "step": 771
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 17.998004913330078,
-      "learning_rate": 1.8454445224015238e-05,
-      "loss": 4.1352,
+      "epoch": 0.1,
+      "grad_norm": 13.409041404724121,
+      "learning_rate": 1.935489269129398e-05,
+      "loss": 3.9463,
       "step": 772
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 14.474387168884277,
-      "learning_rate": 1.8452440613410848e-05,
-      "loss": 3.0304,
+      "epoch": 0.1,
+      "grad_norm": 10.823927879333496,
+      "learning_rate": 1.9354055976237295e-05,
+      "loss": 3.0251,
       "step": 773
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 12.42952823638916,
-      "learning_rate": 1.8450436002806454e-05,
-      "loss": 2.8432,
+      "epoch": 0.1,
+      "grad_norm": 14.99421501159668,
+      "learning_rate": 1.935321926118061e-05,
+      "loss": 1.5831,
       "step": 774
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 15.06114387512207,
-      "learning_rate": 1.8448431392202068e-05,
-      "loss": 2.9204,
+      "epoch": 0.1,
+      "grad_norm": 14.503152847290039,
+      "learning_rate": 1.935238254612392e-05,
+      "loss": 3.4288,
       "step": 775
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 19.3477840423584,
-      "learning_rate": 1.8446426781597678e-05,
-      "loss": 2.6387,
+      "epoch": 0.1,
+      "grad_norm": 12.275667190551758,
+      "learning_rate": 1.9351545831067232e-05,
+      "loss": 3.406,
       "step": 776
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 19.3367862701416,
-      "learning_rate": 1.8444422170993285e-05,
-      "loss": 3.7189,
+      "epoch": 0.1,
+      "grad_norm": 16.560117721557617,
+      "learning_rate": 1.9350709116010543e-05,
+      "loss": 4.3817,
       "step": 777
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 18.679487228393555,
-      "learning_rate": 1.8442417560388898e-05,
-      "loss": 2.7699,
+      "epoch": 0.1,
+      "grad_norm": 16.34395980834961,
+      "learning_rate": 1.9349872400953856e-05,
+      "loss": 1.9061,
       "step": 778
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 17.268980026245117,
-      "learning_rate": 1.8440412949784505e-05,
-      "loss": 3.1304,
+      "epoch": 0.1,
+      "grad_norm": 11.401074409484863,
+      "learning_rate": 1.934903568589717e-05,
+      "loss": 2.8721,
       "step": 779
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 12.782285690307617,
-      "learning_rate": 1.8438408339180115e-05,
-      "loss": 2.1415,
+      "epoch": 0.1,
+      "grad_norm": 18.35641098022461,
+      "learning_rate": 1.934819897084048e-05,
+      "loss": 4.7659,
       "step": 780
     },
     {
-      "epoch": 0.23,
-      "grad_norm": 13.401495933532715,
-      "learning_rate": 1.8436403728575725e-05,
-      "loss": 2.8394,
+      "epoch": 0.1,
+      "grad_norm": 14.767510414123535,
+      "learning_rate": 1.9347362255783794e-05,
+      "loss": 4.2284,
       "step": 781
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 17.430240631103516,
-      "learning_rate": 1.8434399117971335e-05,
-      "loss": 3.3336,
+      "epoch": 0.1,
+      "grad_norm": 42.032958984375,
+      "learning_rate": 1.9346525540727108e-05,
+      "loss": 2.738,
       "step": 782
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 50.752410888671875,
-      "learning_rate": 1.8432394507366945e-05,
-      "loss": 3.1897,
+      "epoch": 0.1,
+      "grad_norm": 15.29719352722168,
+      "learning_rate": 1.9345688825670418e-05,
+      "loss": 3.3126,
       "step": 783
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 14.428933143615723,
-      "learning_rate": 1.8430389896762555e-05,
-      "loss": 2.8947,
+      "epoch": 0.1,
+      "grad_norm": 29.157716751098633,
+      "learning_rate": 1.934485211061373e-05,
+      "loss": 5.0679,
       "step": 784
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 60.45037078857422,
-      "learning_rate": 1.8428385286158165e-05,
-      "loss": 2.7407,
+      "epoch": 0.1,
+      "grad_norm": 11.626741409301758,
+      "learning_rate": 1.9344015395557045e-05,
+      "loss": 2.2906,
       "step": 785
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 15.153972625732422,
-      "learning_rate": 1.8426380675553776e-05,
-      "loss": 3.9118,
+      "epoch": 0.1,
+      "grad_norm": 14.68950366973877,
+      "learning_rate": 1.9343178680500355e-05,
+      "loss": 4.9856,
       "step": 786
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 13.527739524841309,
-      "learning_rate": 1.8424376064949386e-05,
-      "loss": 2.8687,
+      "epoch": 0.1,
+      "grad_norm": 9.994388580322266,
+      "learning_rate": 1.934234196544367e-05,
+      "loss": 4.3106,
       "step": 787
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 13.59441089630127,
-      "learning_rate": 1.8422371454344996e-05,
-      "loss": 3.4082,
+      "epoch": 0.1,
+      "grad_norm": 9.545372009277344,
+      "learning_rate": 1.9341505250386983e-05,
+      "loss": 2.5157,
       "step": 788
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 18.013368606567383,
-      "learning_rate": 1.8420366843740606e-05,
-      "loss": 3.1523,
+      "epoch": 0.1,
+      "grad_norm": 12.77479362487793,
+      "learning_rate": 1.9340668535330296e-05,
+      "loss": 1.9969,
       "step": 789
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 23.031423568725586,
-      "learning_rate": 1.8418362233136216e-05,
-      "loss": 3.8803,
+      "epoch": 0.1,
+      "grad_norm": 16.627382278442383,
+      "learning_rate": 1.9339831820273607e-05,
+      "loss": 3.4785,
       "step": 790
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 12.914963722229004,
-      "learning_rate": 1.8416357622531823e-05,
-      "loss": 3.0994,
+      "epoch": 0.1,
+      "grad_norm": 27.862689971923828,
+      "learning_rate": 1.933899510521692e-05,
+      "loss": 4.5648,
       "step": 791
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 25.980188369750977,
-      "learning_rate": 1.8414353011927436e-05,
-      "loss": 1.9718,
+      "epoch": 0.1,
+      "grad_norm": 15.007376670837402,
+      "learning_rate": 1.9338158390160234e-05,
+      "loss": 1.4494,
       "step": 792
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 14.725564002990723,
-      "learning_rate": 1.8412348401323043e-05,
-      "loss": 2.472,
+      "epoch": 0.1,
+      "grad_norm": 12.32243537902832,
+      "learning_rate": 1.9337321675103544e-05,
+      "loss": 4.0946,
       "step": 793
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 13.297785758972168,
-      "learning_rate": 1.8410343790718653e-05,
-      "loss": 4.0097,
+      "epoch": 0.1,
+      "grad_norm": 13.405266761779785,
+      "learning_rate": 1.9336484960046858e-05,
+      "loss": 3.2625,
       "step": 794
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 15.004578590393066,
-      "learning_rate": 1.8408339180114266e-05,
-      "loss": 2.8653,
+      "epoch": 0.1,
+      "grad_norm": 23.589693069458008,
+      "learning_rate": 1.933564824499017e-05,
+      "loss": 5.7331,
       "step": 795
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 19.27330780029297,
-      "learning_rate": 1.8406334569509873e-05,
-      "loss": 3.1684,
+      "epoch": 0.1,
+      "grad_norm": 35.675025939941406,
+      "learning_rate": 1.9334811529933485e-05,
+      "loss": 2.3257,
       "step": 796
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 25.846576690673828,
-      "learning_rate": 1.8404329958905483e-05,
-      "loss": 2.6346,
+      "epoch": 0.1,
+      "grad_norm": 12.68811321258545,
+      "learning_rate": 1.9333974814876795e-05,
+      "loss": 4.5812,
       "step": 797
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 9.999025344848633,
-      "learning_rate": 1.8402325348301093e-05,
-      "loss": 1.9912,
+      "epoch": 0.1,
+      "grad_norm": 22.3652286529541,
+      "learning_rate": 1.933313809982011e-05,
+      "loss": 3.5698,
       "step": 798
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 11.801385879516602,
-      "learning_rate": 1.8400320737696703e-05,
-      "loss": 3.3957,
+      "epoch": 0.1,
+      "grad_norm": 20.88276481628418,
+      "learning_rate": 1.9332301384763423e-05,
+      "loss": 5.2997,
       "step": 799
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 12.033352851867676,
-      "learning_rate": 1.8398316127092313e-05,
-      "loss": 2.62,
+      "epoch": 0.1,
+      "grad_norm": 7.848814010620117,
+      "learning_rate": 1.9331464669706733e-05,
+      "loss": 4.3608,
       "step": 800
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 12.087538719177246,
-      "learning_rate": 1.8396311516487923e-05,
-      "loss": 3.2451,
+      "epoch": 0.1,
+      "eval_loss": 0.4046749472618103,
+      "eval_runtime": 95.7577,
+      "eval_samples_per_second": 36.989,
+      "eval_steps_per_second": 36.989,
+      "step": 800
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.814807891845703,
+      "learning_rate": 1.9330627954650047e-05,
+      "loss": 1.7813,
       "step": 801
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 9.926271438598633,
-      "learning_rate": 1.8394306905883533e-05,
-      "loss": 2.5499,
+      "epoch": 0.1,
+      "grad_norm": 10.413105010986328,
+      "learning_rate": 1.9329791239593357e-05,
+      "loss": 3.4774,
       "step": 802
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 15.592676162719727,
-      "learning_rate": 1.8392302295279144e-05,
-      "loss": 3.7471,
+      "epoch": 0.1,
+      "grad_norm": 21.069353103637695,
+      "learning_rate": 1.932895452453667e-05,
+      "loss": 2.5785,
       "step": 803
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 16.60494613647461,
-      "learning_rate": 1.8390297684674754e-05,
-      "loss": 3.218,
+      "epoch": 0.1,
+      "grad_norm": 12.616759300231934,
+      "learning_rate": 1.9328117809479984e-05,
+      "loss": 2.8462,
       "step": 804
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 20.184545516967773,
-      "learning_rate": 1.8388293074070364e-05,
-      "loss": 2.9737,
+      "epoch": 0.1,
+      "grad_norm": 23.985736846923828,
+      "learning_rate": 1.9327281094423294e-05,
+      "loss": 4.0714,
       "step": 805
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 12.5513334274292,
-      "learning_rate": 1.8386288463465974e-05,
-      "loss": 2.3381,
+      "epoch": 0.1,
+      "grad_norm": 21.885786056518555,
+      "learning_rate": 1.9326444379366608e-05,
+      "loss": 3.4187,
       "step": 806
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 19.2335262298584,
-      "learning_rate": 1.8384283852861584e-05,
-      "loss": 3.5287,
+      "epoch": 0.1,
+      "grad_norm": 15.95827865600586,
+      "learning_rate": 1.9325607664309918e-05,
+      "loss": 3.7852,
       "step": 807
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 16.872840881347656,
-      "learning_rate": 1.8382279242257194e-05,
-      "loss": 3.1892,
+      "epoch": 0.1,
+      "grad_norm": 11.442026138305664,
+      "learning_rate": 1.9324770949253232e-05,
+      "loss": 4.8199,
       "step": 808
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 22.987356185913086,
-      "learning_rate": 1.8380274631652804e-05,
-      "loss": 2.534,
+      "epoch": 0.1,
+      "grad_norm": 12.986435890197754,
+      "learning_rate": 1.9323934234196546e-05,
+      "loss": 2.5731,
       "step": 809
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 18.01645851135254,
-      "learning_rate": 1.837827002104841e-05,
-      "loss": 3.5278,
+      "epoch": 0.1,
+      "grad_norm": 9.10637092590332,
+      "learning_rate": 1.932309751913986e-05,
+      "loss": 2.5007,
       "step": 810
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 23.722457885742188,
-      "learning_rate": 1.8376265410444024e-05,
-      "loss": 3.9305,
+      "epoch": 0.1,
+      "grad_norm": 15.682181358337402,
+      "learning_rate": 1.932226080408317e-05,
+      "loss": 3.6707,
       "step": 811
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 11.13327693939209,
-      "learning_rate": 1.837426079983963e-05,
-      "loss": 3.9406,
+      "epoch": 0.1,
+      "grad_norm": 9.2945556640625,
+      "learning_rate": 1.9321424089026483e-05,
+      "loss": 2.9492,
       "step": 812
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 18.1392765045166,
-      "learning_rate": 1.837225618923524e-05,
-      "loss": 2.3532,
+      "epoch": 0.1,
+      "grad_norm": 19.119644165039062,
+      "learning_rate": 1.9320587373969797e-05,
+      "loss": 5.1159,
       "step": 813
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 17.197660446166992,
-      "learning_rate": 1.8370251578630854e-05,
-      "loss": 3.8571,
+      "epoch": 0.1,
+      "grad_norm": 12.646649360656738,
+      "learning_rate": 1.9319750658913107e-05,
+      "loss": 3.6129,
       "step": 814
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 16.233530044555664,
-      "learning_rate": 1.836824696802646e-05,
-      "loss": 2.8763,
+      "epoch": 0.1,
+      "grad_norm": 14.362308502197266,
+      "learning_rate": 1.931891394385642e-05,
+      "loss": 3.9929,
       "step": 815
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 104.82502746582031,
-      "learning_rate": 1.836624235742207e-05,
-      "loss": 3.3302,
+      "epoch": 0.1,
+      "grad_norm": 13.062503814697266,
+      "learning_rate": 1.9318077228799734e-05,
+      "loss": 3.6153,
       "step": 816
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 14.149030685424805,
-      "learning_rate": 1.836423774681768e-05,
-      "loss": 3.2956,
+      "epoch": 0.1,
+      "grad_norm": 17.66871452331543,
+      "learning_rate": 1.9317240513743048e-05,
+      "loss": 2.5809,
       "step": 817
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 13.704061508178711,
-      "learning_rate": 1.836223313621329e-05,
-      "loss": 2.7446,
+      "epoch": 0.1,
+      "grad_norm": 39.23408508300781,
+      "learning_rate": 1.9316403798686358e-05,
+      "loss": 2.9452,
       "step": 818
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 14.284625053405762,
-      "learning_rate": 1.83602285256089e-05,
-      "loss": 2.9887,
+      "epoch": 0.1,
+      "grad_norm": 20.960430145263672,
+      "learning_rate": 1.9315567083629672e-05,
+      "loss": 4.0357,
       "step": 819
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 20.733549118041992,
-      "learning_rate": 1.835822391500451e-05,
-      "loss": 3.3273,
+      "epoch": 0.1,
+      "grad_norm": 11.633240699768066,
+      "learning_rate": 1.9314730368572986e-05,
+      "loss": 2.0731,
       "step": 820
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 27.305700302124023,
-      "learning_rate": 1.835621930440012e-05,
-      "loss": 3.061,
+      "epoch": 0.1,
+      "grad_norm": 10.769817352294922,
+      "learning_rate": 1.9313893653516296e-05,
+      "loss": 1.9501,
       "step": 821
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 11.420647621154785,
-      "learning_rate": 1.8354214693795732e-05,
-      "loss": 3.1825,
+      "epoch": 0.1,
+      "grad_norm": 16.530324935913086,
+      "learning_rate": 1.931305693845961e-05,
+      "loss": 3.8412,
       "step": 822
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 13.486950874328613,
-      "learning_rate": 1.8352210083191342e-05,
-      "loss": 3.1205,
+      "epoch": 0.1,
+      "grad_norm": 10.702472686767578,
+      "learning_rate": 1.9312220223402923e-05,
+      "loss": 2.675,
       "step": 823
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 21.656282424926758,
-      "learning_rate": 1.8350205472586952e-05,
-      "loss": 2.7608,
+      "epoch": 0.1,
+      "grad_norm": 26.523433685302734,
+      "learning_rate": 1.9311383508346237e-05,
+      "loss": 1.1108,
       "step": 824
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 31.164608001708984,
-      "learning_rate": 1.8348200861982562e-05,
-      "loss": 3.3097,
+      "epoch": 0.1,
+      "grad_norm": 21.036243438720703,
+      "learning_rate": 1.9310546793289547e-05,
+      "loss": 3.4756,
       "step": 825
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 36.02324676513672,
-      "learning_rate": 1.8346196251378172e-05,
-      "loss": 3.2008,
+      "epoch": 0.1,
+      "grad_norm": 18.37494468688965,
+      "learning_rate": 1.930971007823286e-05,
+      "loss": 4.5667,
       "step": 826
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 15.944084167480469,
-      "learning_rate": 1.8344191640773782e-05,
-      "loss": 2.9221,
+      "epoch": 0.1,
+      "grad_norm": 12.04539966583252,
+      "learning_rate": 1.9308873363176174e-05,
+      "loss": 1.546,
       "step": 827
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 14.965399742126465,
-      "learning_rate": 1.8342187030169392e-05,
-      "loss": 3.1973,
+      "epoch": 0.1,
+      "grad_norm": 12.4345064163208,
+      "learning_rate": 1.9308036648119485e-05,
+      "loss": 3.3334,
       "step": 828
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 20.134925842285156,
-      "learning_rate": 1.8340182419565e-05,
-      "loss": 3.521,
+      "epoch": 0.1,
+      "grad_norm": 69.78355407714844,
+      "learning_rate": 1.9307199933062798e-05,
+      "loss": 5.0178,
       "step": 829
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 17.916906356811523,
-      "learning_rate": 1.8338177808960612e-05,
-      "loss": 3.8329,
+      "epoch": 0.1,
+      "grad_norm": 15.295698165893555,
+      "learning_rate": 1.930636321800611e-05,
+      "loss": 2.6952,
       "step": 830
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 22.273197174072266,
-      "learning_rate": 1.8336173198356223e-05,
-      "loss": 3.0719,
+      "epoch": 0.1,
+      "grad_norm": 11.989553451538086,
+      "learning_rate": 1.9305526502949422e-05,
+      "loss": 2.7478,
       "step": 831
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 18.33119773864746,
-      "learning_rate": 1.833416858775183e-05,
-      "loss": 3.1312,
+      "epoch": 0.1,
+      "grad_norm": 14.243673324584961,
+      "learning_rate": 1.9304689787892736e-05,
+      "loss": 4.1795,
       "step": 832
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 17.575429916381836,
-      "learning_rate": 1.8332163977147443e-05,
-      "loss": 2.7529,
+      "epoch": 0.1,
+      "grad_norm": 15.110532760620117,
+      "learning_rate": 1.9303853072836046e-05,
+      "loss": 2.8877,
       "step": 833
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 16.5457820892334,
-      "learning_rate": 1.833015936654305e-05,
-      "loss": 2.3905,
+      "epoch": 0.1,
+      "grad_norm": 12.30280590057373,
+      "learning_rate": 1.930301635777936e-05,
+      "loss": 3.3959,
       "step": 834
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 19.157955169677734,
-      "learning_rate": 1.832815475593866e-05,
-      "loss": 3.6365,
+      "epoch": 0.1,
+      "grad_norm": 10.67061996459961,
+      "learning_rate": 1.930217964272267e-05,
+      "loss": 3.561,
       "step": 835
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 16.239450454711914,
-      "learning_rate": 1.832615014533427e-05,
-      "loss": 2.864,
+      "epoch": 0.1,
+      "grad_norm": 16.788816452026367,
+      "learning_rate": 1.9301342927665984e-05,
+      "loss": 3.3843,
       "step": 836
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 24.656652450561523,
-      "learning_rate": 1.832414553472988e-05,
-      "loss": 3.4563,
+      "epoch": 0.11,
+      "grad_norm": 14.812087059020996,
+      "learning_rate": 1.9300506212609297e-05,
+      "loss": 1.2697,
       "step": 837
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 21.09084129333496,
-      "learning_rate": 1.832214092412549e-05,
-      "loss": 3.3509,
+      "epoch": 0.11,
+      "grad_norm": 17.998449325561523,
+      "learning_rate": 1.929966949755261e-05,
+      "loss": 3.3954,
       "step": 838
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 14.556492805480957,
-      "learning_rate": 1.83201363135211e-05,
-      "loss": 3.0753,
+      "epoch": 0.11,
+      "grad_norm": 10.879467010498047,
+      "learning_rate": 1.929883278249592e-05,
+      "loss": 3.1961,
       "step": 839
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 19.413835525512695,
-      "learning_rate": 1.831813170291671e-05,
-      "loss": 2.9689,
-      "step": 840
-    },
-    {
-      "epoch": 0.25,
-      "eval_loss": 0.5996547937393188,
-      "eval_runtime": 43.6208,
-      "eval_samples_per_second": 33.906,
-      "eval_steps_per_second": 33.906,
+      "epoch": 0.11,
+      "grad_norm": 13.773147583007812,
+      "learning_rate": 1.9297996067439235e-05,
+      "loss": 5.1689,
       "step": 840
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 17.024858474731445,
-      "learning_rate": 1.831612709231232e-05,
-      "loss": 4.6989,
+      "epoch": 0.11,
+      "grad_norm": 19.973291397094727,
+      "learning_rate": 1.929715935238255e-05,
+      "loss": 4.2266,
       "step": 841
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 12.914769172668457,
-      "learning_rate": 1.831412248170793e-05,
-      "loss": 3.6312,
+      "epoch": 0.11,
+      "grad_norm": 14.159008979797363,
+      "learning_rate": 1.929632263732586e-05,
+      "loss": 5.0419,
       "step": 842
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 89.747802734375,
-      "learning_rate": 1.831211787110354e-05,
-      "loss": 2.7291,
+      "epoch": 0.11,
+      "grad_norm": 9.060222625732422,
+      "learning_rate": 1.9295485922269172e-05,
+      "loss": 1.2281,
       "step": 843
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 15.173033714294434,
-      "learning_rate": 1.831011326049915e-05,
-      "loss": 3.1805,
+      "epoch": 0.11,
+      "grad_norm": 18.025102615356445,
+      "learning_rate": 1.9294649207212486e-05,
+      "loss": 1.9924,
       "step": 844
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 40.06729507446289,
-      "learning_rate": 1.830810864989476e-05,
-      "loss": 4.5534,
+      "epoch": 0.11,
+      "grad_norm": 20.9349308013916,
+      "learning_rate": 1.92938124921558e-05,
+      "loss": 2.8242,
       "step": 845
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 13.402242660522461,
-      "learning_rate": 1.830610403929037e-05,
-      "loss": 2.946,
+      "epoch": 0.11,
+      "grad_norm": 12.945772171020508,
+      "learning_rate": 1.929297577709911e-05,
+      "loss": 5.5559,
       "step": 846
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 18.814979553222656,
-      "learning_rate": 1.830409942868598e-05,
-      "loss": 3.3728,
+      "epoch": 0.11,
+      "grad_norm": 37.96195983886719,
+      "learning_rate": 1.9292139062042424e-05,
+      "loss": 2.7299,
       "step": 847
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 12.975325584411621,
-      "learning_rate": 1.8302094818081587e-05,
-      "loss": 2.8497,
+      "epoch": 0.11,
+      "grad_norm": 10.461493492126465,
+      "learning_rate": 1.9291302346985737e-05,
+      "loss": 3.2226,
       "step": 848
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 9.28498363494873,
-      "learning_rate": 1.83000902074772e-05,
-      "loss": 2.4016,
+      "epoch": 0.11,
+      "grad_norm": 17.57033920288086,
+      "learning_rate": 1.9290465631929047e-05,
+      "loss": 3.6556,
       "step": 849
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 15.675614356994629,
-      "learning_rate": 1.829808559687281e-05,
-      "loss": 2.7848,
+      "epoch": 0.11,
+      "grad_norm": 10.791287422180176,
+      "learning_rate": 1.928962891687236e-05,
+      "loss": 2.4405,
       "step": 850
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 13.038680076599121,
-      "learning_rate": 1.8296080986268417e-05,
-      "loss": 3.4155,
+      "epoch": 0.11,
+      "grad_norm": 14.49622631072998,
+      "learning_rate": 1.9288792201815675e-05,
+      "loss": 5.1708,
       "step": 851
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 11.621955871582031,
-      "learning_rate": 1.829407637566403e-05,
-      "loss": 2.8026,
+      "epoch": 0.11,
+      "grad_norm": 14.331541061401367,
+      "learning_rate": 1.928795548675899e-05,
+      "loss": 3.636,
       "step": 852
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 11.502103805541992,
-      "learning_rate": 1.8292071765059638e-05,
-      "loss": 2.2732,
+      "epoch": 0.11,
+      "grad_norm": 11.339052200317383,
+      "learning_rate": 1.92871187717023e-05,
+      "loss": 2.8692,
       "step": 853
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 27.41511344909668,
-      "learning_rate": 1.8290067154455248e-05,
-      "loss": 4.0694,
+      "epoch": 0.11,
+      "grad_norm": 26.334575653076172,
+      "learning_rate": 1.9286282056645612e-05,
+      "loss": 3.4792,
       "step": 854
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 16.437061309814453,
-      "learning_rate": 1.8288062543850858e-05,
-      "loss": 3.0165,
+      "epoch": 0.11,
+      "grad_norm": 13.978888511657715,
+      "learning_rate": 1.9285445341588923e-05,
+      "loss": 2.6604,
       "step": 855
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 14.086379051208496,
-      "learning_rate": 1.8286057933246468e-05,
-      "loss": 2.6271,
+      "epoch": 0.11,
+      "grad_norm": 12.099700927734375,
+      "learning_rate": 1.9284608626532236e-05,
+      "loss": 2.2098,
       "step": 856
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 18.89293670654297,
-      "learning_rate": 1.8284053322642078e-05,
-      "loss": 3.2895,
+      "epoch": 0.11,
+      "grad_norm": 13.464815139770508,
+      "learning_rate": 1.928377191147555e-05,
+      "loss": 1.981,
       "step": 857
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 14.886481285095215,
-      "learning_rate": 1.8282048712037688e-05,
-      "loss": 3.3004,
+      "epoch": 0.11,
+      "grad_norm": 40.965389251708984,
+      "learning_rate": 1.928293519641886e-05,
+      "loss": 4.6309,
       "step": 858
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 21.090091705322266,
-      "learning_rate": 1.8280044101433298e-05,
-      "loss": 3.0523,
+      "epoch": 0.11,
+      "grad_norm": 13.609042167663574,
+      "learning_rate": 1.9282098481362174e-05,
+      "loss": 1.7419,
       "step": 859
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 13.093547821044922,
-      "learning_rate": 1.8278039490828908e-05,
-      "loss": 2.3511,
+      "epoch": 0.11,
+      "grad_norm": 12.86327075958252,
+      "learning_rate": 1.9281261766305484e-05,
+      "loss": 3.5797,
       "step": 860
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 13.202871322631836,
-      "learning_rate": 1.8276034880224518e-05,
-      "loss": 2.7941,
+      "epoch": 0.11,
+      "grad_norm": 11.691978454589844,
+      "learning_rate": 1.9280425051248798e-05,
+      "loss": 2.4073,
       "step": 861
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 10.028319358825684,
-      "learning_rate": 1.827403026962013e-05,
-      "loss": 2.6855,
+      "epoch": 0.11,
+      "grad_norm": 14.143892288208008,
+      "learning_rate": 1.927958833619211e-05,
+      "loss": 3.1444,
       "step": 862
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 14.725035667419434,
-      "learning_rate": 1.827202565901574e-05,
-      "loss": 3.3114,
+      "epoch": 0.11,
+      "grad_norm": 13.53709888458252,
+      "learning_rate": 1.927875162113542e-05,
+      "loss": 2.2037,
       "step": 863
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 20.055673599243164,
-      "learning_rate": 1.827002104841135e-05,
-      "loss": 4.0679,
+      "epoch": 0.11,
+      "grad_norm": 18.455381393432617,
+      "learning_rate": 1.9277914906078735e-05,
+      "loss": 3.7846,
       "step": 864
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 15.065644264221191,
-      "learning_rate": 1.8268016437806955e-05,
-      "loss": 2.8739,
+      "epoch": 0.11,
+      "grad_norm": 18.374515533447266,
+      "learning_rate": 1.927707819102205e-05,
+      "loss": 2.4646,
       "step": 865
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 11.230770111083984,
-      "learning_rate": 1.826601182720257e-05,
-      "loss": 3.2987,
+      "epoch": 0.11,
+      "grad_norm": 11.123920440673828,
+      "learning_rate": 1.9276241475965363e-05,
+      "loss": 3.8492,
       "step": 866
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 27.67386245727539,
-      "learning_rate": 1.8264007216598175e-05,
-      "loss": 3.0173,
+      "epoch": 0.11,
+      "grad_norm": 10.278346061706543,
+      "learning_rate": 1.9275404760908673e-05,
+      "loss": 1.2724,
       "step": 867
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 18.033613204956055,
-      "learning_rate": 1.8262002605993785e-05,
-      "loss": 3.4757,
+      "epoch": 0.11,
+      "grad_norm": 15.350423812866211,
+      "learning_rate": 1.9274568045851986e-05,
+      "loss": 2.0491,
       "step": 868
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 12.066390037536621,
-      "learning_rate": 1.82599979953894e-05,
-      "loss": 2.815,
+      "epoch": 0.11,
+      "grad_norm": 11.977092742919922,
+      "learning_rate": 1.92737313307953e-05,
+      "loss": 3.7265,
       "step": 869
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 18.521556854248047,
-      "learning_rate": 1.8257993384785006e-05,
-      "loss": 4.0135,
+      "epoch": 0.11,
+      "grad_norm": 18.873859405517578,
+      "learning_rate": 1.927289461573861e-05,
+      "loss": 3.1926,
       "step": 870
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 19.63759422302246,
-      "learning_rate": 1.825598877418062e-05,
-      "loss": 4.0507,
+      "epoch": 0.11,
+      "grad_norm": 15.339083671569824,
+      "learning_rate": 1.9272057900681924e-05,
+      "loss": 3.3854,
       "step": 871
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 34.88851547241211,
-      "learning_rate": 1.8253984163576226e-05,
-      "loss": 4.7139,
+      "epoch": 0.11,
+      "grad_norm": 23.82866668701172,
+      "learning_rate": 1.9271221185625238e-05,
+      "loss": 4.5427,
       "step": 872
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 15.407425880432129,
-      "learning_rate": 1.8251979552971836e-05,
-      "loss": 3.7188,
+      "epoch": 0.11,
+      "grad_norm": 16.461936950683594,
+      "learning_rate": 1.927038447056855e-05,
+      "loss": 3.9018,
       "step": 873
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 12.915898323059082,
-      "learning_rate": 1.824997494236745e-05,
-      "loss": 2.5801,
+      "epoch": 0.11,
+      "grad_norm": 19.81767463684082,
+      "learning_rate": 1.926954775551186e-05,
+      "loss": 4.0158,
       "step": 874
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 12.690547943115234,
-      "learning_rate": 1.8247970331763056e-05,
-      "loss": 2.3076,
+      "epoch": 0.11,
+      "grad_norm": 21.873647689819336,
+      "learning_rate": 1.9268711040455175e-05,
+      "loss": 3.0256,
       "step": 875
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 31.50592041015625,
-      "learning_rate": 1.8245965721158666e-05,
-      "loss": 2.8265,
+      "epoch": 0.11,
+      "grad_norm": 20.212047576904297,
+      "learning_rate": 1.926787432539849e-05,
+      "loss": 3.4756,
       "step": 876
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 13.253562927246094,
-      "learning_rate": 1.8243961110554276e-05,
-      "loss": 2.612,
+      "epoch": 0.11,
+      "grad_norm": 13.910633087158203,
+      "learning_rate": 1.92670376103418e-05,
+      "loss": 3.4597,
       "step": 877
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 21.80856704711914,
-      "learning_rate": 1.8241956499949886e-05,
-      "loss": 3.0365,
+      "epoch": 0.11,
+      "grad_norm": 13.977246284484863,
+      "learning_rate": 1.9266200895285113e-05,
+      "loss": 2.26,
       "step": 878
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 17.941879272460938,
-      "learning_rate": 1.8239951889345496e-05,
-      "loss": 2.9886,
+      "epoch": 0.11,
+      "grad_norm": 16.114044189453125,
+      "learning_rate": 1.9265364180228426e-05,
+      "loss": 5.6168,
       "step": 879
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 14.684907913208008,
-      "learning_rate": 1.8237947278741106e-05,
-      "loss": 2.6941,
+      "epoch": 0.11,
+      "grad_norm": 21.8970890045166,
+      "learning_rate": 1.926452746517174e-05,
+      "loss": 2.5029,
       "step": 880
     },
     {
-      "epoch": 0.26,
-      "grad_norm": 19.08102798461914,
-      "learning_rate": 1.8235942668136717e-05,
-      "loss": 3.4722,
+      "epoch": 0.11,
+      "grad_norm": 21.701828002929688,
+      "learning_rate": 1.926369075011505e-05,
+      "loss": 5.1717,
       "step": 881
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 14.615809440612793,
-      "learning_rate": 1.8233938057532327e-05,
-      "loss": 2.9938,
+      "epoch": 0.11,
+      "grad_norm": 13.911944389343262,
+      "learning_rate": 1.9262854035058364e-05,
+      "loss": 3.6187,
       "step": 882
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 22.291290283203125,
-      "learning_rate": 1.8231933446927937e-05,
-      "loss": 4.3829,
+      "epoch": 0.11,
+      "grad_norm": 15.300573348999023,
+      "learning_rate": 1.9262017320001674e-05,
+      "loss": 4.3989,
       "step": 883
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 16.102741241455078,
-      "learning_rate": 1.8229928836323543e-05,
-      "loss": 3.5552,
+      "epoch": 0.11,
+      "grad_norm": 22.996200561523438,
+      "learning_rate": 1.9261180604944988e-05,
+      "loss": 3.8584,
       "step": 884
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 20.495725631713867,
-      "learning_rate": 1.8227924225719157e-05,
-      "loss": 3.3252,
+      "epoch": 0.11,
+      "grad_norm": 9.07385540008545,
+      "learning_rate": 1.92603438898883e-05,
+      "loss": 4.0722,
       "step": 885
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 25.180877685546875,
-      "learning_rate": 1.8225919615114764e-05,
-      "loss": 2.6339,
+      "epoch": 0.11,
+      "grad_norm": 20.0780086517334,
+      "learning_rate": 1.9259507174831612e-05,
+      "loss": 3.5784,
       "step": 886
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 14.054975509643555,
-      "learning_rate": 1.8223915004510374e-05,
-      "loss": 3.2697,
+      "epoch": 0.11,
+      "grad_norm": 15.519294738769531,
+      "learning_rate": 1.9258670459774925e-05,
+      "loss": 4.7674,
       "step": 887
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 14.858386993408203,
-      "learning_rate": 1.8221910393905987e-05,
-      "loss": 2.3764,
+      "epoch": 0.11,
+      "grad_norm": 11.765948295593262,
+      "learning_rate": 1.9257833744718236e-05,
+      "loss": 1.0629,
       "step": 888
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 14.020156860351562,
-      "learning_rate": 1.8219905783301594e-05,
-      "loss": 2.4207,
+      "epoch": 0.11,
+      "grad_norm": 10.282224655151367,
+      "learning_rate": 1.925699702966155e-05,
+      "loss": 0.4704,
       "step": 889
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 20.844745635986328,
-      "learning_rate": 1.8217901172697204e-05,
-      "loss": 2.8965,
+      "epoch": 0.11,
+      "grad_norm": 11.294329643249512,
+      "learning_rate": 1.9256160314604863e-05,
+      "loss": 3.0352,
       "step": 890
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 12.82536792755127,
-      "learning_rate": 1.8215896562092814e-05,
-      "loss": 2.1605,
+      "epoch": 0.11,
+      "grad_norm": 21.708587646484375,
+      "learning_rate": 1.9255323599548173e-05,
+      "loss": 3.151,
       "step": 891
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 11.74587345123291,
-      "learning_rate": 1.8213891951488424e-05,
-      "loss": 3.171,
+      "epoch": 0.11,
+      "grad_norm": 14.7886323928833,
+      "learning_rate": 1.9254486884491487e-05,
+      "loss": 3.4023,
       "step": 892
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 21.482084274291992,
-      "learning_rate": 1.8211887340884034e-05,
-      "loss": 3.142,
+      "epoch": 0.11,
+      "grad_norm": 12.128108978271484,
+      "learning_rate": 1.92536501694348e-05,
+      "loss": 1.6436,
       "step": 893
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 28.953731536865234,
-      "learning_rate": 1.8209882730279644e-05,
-      "loss": 3.2382,
+      "epoch": 0.11,
+      "grad_norm": 27.2979736328125,
+      "learning_rate": 1.925281345437811e-05,
+      "loss": 3.554,
       "step": 894
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 17.727638244628906,
-      "learning_rate": 1.8207878119675254e-05,
-      "loss": 3.2516,
+      "epoch": 0.11,
+      "grad_norm": 12.120756149291992,
+      "learning_rate": 1.9251976739321425e-05,
+      "loss": 3.8833,
       "step": 895
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 24.647932052612305,
-      "learning_rate": 1.8205873509070864e-05,
-      "loss": 3.1515,
+      "epoch": 0.11,
+      "grad_norm": 15.177332878112793,
+      "learning_rate": 1.9251140024264738e-05,
+      "loss": 2.3427,
       "step": 896
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 15.397368431091309,
-      "learning_rate": 1.8203868898466475e-05,
-      "loss": 3.8315,
+      "epoch": 0.11,
+      "grad_norm": 14.57490062713623,
+      "learning_rate": 1.9250303309208052e-05,
+      "loss": 2.8223,
       "step": 897
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 10.54749584197998,
-      "learning_rate": 1.8201864287862085e-05,
-      "loss": 1.3177,
+      "epoch": 0.11,
+      "grad_norm": 14.658095359802246,
+      "learning_rate": 1.9249466594151362e-05,
+      "loss": 4.0262,
       "step": 898
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 22.014423370361328,
-      "learning_rate": 1.8199859677257695e-05,
-      "loss": 3.8405,
+      "epoch": 0.11,
+      "grad_norm": 25.267988204956055,
+      "learning_rate": 1.9248629879094676e-05,
+      "loss": 3.9113,
       "step": 899
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 18.356632232666016,
-      "learning_rate": 1.8197855066653305e-05,
-      "loss": 3.0761,
+      "epoch": 0.11,
+      "grad_norm": 11.966208457946777,
+      "learning_rate": 1.924779316403799e-05,
+      "loss": 4.6879,
       "step": 900
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 19.23956298828125,
-      "learning_rate": 1.8195850456048915e-05,
-      "loss": 3.8153,
+      "epoch": 0.11,
+      "grad_norm": 20.84747314453125,
+      "learning_rate": 1.92469564489813e-05,
+      "loss": 3.4313,
       "step": 901
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 21.569475173950195,
-      "learning_rate": 1.8193845845444525e-05,
-      "loss": 2.913,
+      "epoch": 0.11,
+      "grad_norm": 23.972240447998047,
+      "learning_rate": 1.9246119733924613e-05,
+      "loss": 3.6033,
       "step": 902
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 10.557537078857422,
-      "learning_rate": 1.819184123484013e-05,
-      "loss": 2.8867,
+      "epoch": 0.11,
+      "grad_norm": 14.662901878356934,
+      "learning_rate": 1.9245283018867927e-05,
+      "loss": 1.7671,
       "step": 903
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 14.843564987182617,
-      "learning_rate": 1.8189836624235745e-05,
-      "loss": 2.9857,
+      "epoch": 0.11,
+      "grad_norm": 17.601343154907227,
+      "learning_rate": 1.924444630381124e-05,
+      "loss": 3.5829,
       "step": 904
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 10.520699501037598,
-      "learning_rate": 1.8187832013631355e-05,
-      "loss": 3.0075,
+      "epoch": 0.11,
+      "grad_norm": 13.459490776062012,
+      "learning_rate": 1.924360958875455e-05,
+      "loss": 3.098,
       "step": 905
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 18.192188262939453,
-      "learning_rate": 1.8185827403026962e-05,
-      "loss": 4.4008,
+      "epoch": 0.11,
+      "grad_norm": 16.708049774169922,
+      "learning_rate": 1.9242772873697864e-05,
+      "loss": 4.3347,
       "step": 906
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 13.926733016967773,
-      "learning_rate": 1.8183822792422575e-05,
-      "loss": 3.1211,
+      "epoch": 0.11,
+      "grad_norm": 15.664067268371582,
+      "learning_rate": 1.9241936158641178e-05,
+      "loss": 2.6975,
       "step": 907
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 17.12418556213379,
-      "learning_rate": 1.8181818181818182e-05,
-      "loss": 3.4901,
+      "epoch": 0.11,
+      "grad_norm": 20.520610809326172,
+      "learning_rate": 1.924109944358449e-05,
+      "loss": 3.8704,
       "step": 908
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 16.681673049926758,
-      "learning_rate": 1.8179813571213792e-05,
-      "loss": 3.2524,
+      "epoch": 0.11,
+      "grad_norm": 7.673875331878662,
+      "learning_rate": 1.9240262728527802e-05,
+      "loss": 1.7854,
       "step": 909
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 18.500484466552734,
-      "learning_rate": 1.8177808960609402e-05,
-      "loss": 3.7737,
+      "epoch": 0.11,
+      "grad_norm": 12.935189247131348,
+      "learning_rate": 1.9239426013471116e-05,
+      "loss": 3.7663,
       "step": 910
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 13.746018409729004,
-      "learning_rate": 1.8175804350005012e-05,
-      "loss": 2.4568,
+      "epoch": 0.11,
+      "grad_norm": 11.801560401916504,
+      "learning_rate": 1.9238589298414426e-05,
+      "loss": 3.6699,
       "step": 911
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 12.869297981262207,
-      "learning_rate": 1.8173799739400622e-05,
-      "loss": 2.6976,
+      "epoch": 0.11,
+      "grad_norm": 16.79932403564453,
+      "learning_rate": 1.923775258335774e-05,
+      "loss": 5.206,
       "step": 912
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 17.630205154418945,
-      "learning_rate": 1.8171795128796232e-05,
-      "loss": 3.2424,
+      "epoch": 0.11,
+      "grad_norm": 10.527442932128906,
+      "learning_rate": 1.923691586830105e-05,
+      "loss": 2.3717,
       "step": 913
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 11.685369491577148,
-      "learning_rate": 1.8169790518191843e-05,
-      "loss": 2.3216,
+      "epoch": 0.11,
+      "grad_norm": 10.9661865234375,
+      "learning_rate": 1.9236079153244364e-05,
+      "loss": 2.4747,
       "step": 914
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 23.479877471923828,
-      "learning_rate": 1.8167785907587453e-05,
-      "loss": 3.3934,
+      "epoch": 0.11,
+      "grad_norm": 10.854809761047363,
+      "learning_rate": 1.9235242438187677e-05,
+      "loss": 4.4707,
       "step": 915
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 11.527375221252441,
-      "learning_rate": 1.8165781296983063e-05,
-      "loss": 2.9907,
+      "epoch": 0.11,
+      "grad_norm": 13.150390625,
+      "learning_rate": 1.9234405723130987e-05,
+      "loss": 3.8095,
       "step": 916
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 11.39271354675293,
-      "learning_rate": 1.8163776686378673e-05,
-      "loss": 2.8909,
+      "epoch": 0.12,
+      "grad_norm": 15.951024055480957,
+      "learning_rate": 1.92335690080743e-05,
+      "loss": 3.7779,
       "step": 917
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 18.55435562133789,
-      "learning_rate": 1.8161772075774283e-05,
-      "loss": 2.9944,
+      "epoch": 0.12,
+      "grad_norm": 14.399053573608398,
+      "learning_rate": 1.9232732293017615e-05,
+      "loss": 2.7399,
       "step": 918
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 13.561935424804688,
-      "learning_rate": 1.8159767465169893e-05,
-      "loss": 2.385,
+      "epoch": 0.12,
+      "grad_norm": 15.080406188964844,
+      "learning_rate": 1.9231895577960925e-05,
+      "loss": 4.3843,
       "step": 919
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 16.513639450073242,
-      "learning_rate": 1.8157762854565503e-05,
-      "loss": 3.0903,
+      "epoch": 0.12,
+      "grad_norm": 9.267796516418457,
+      "learning_rate": 1.923105886290424e-05,
+      "loss": 1.9494,
       "step": 920
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 19.51789093017578,
-      "learning_rate": 1.8155758243961113e-05,
-      "loss": 4.0376,
+      "epoch": 0.12,
+      "grad_norm": 13.215700149536133,
+      "learning_rate": 1.9230222147847552e-05,
+      "loss": 3.3097,
       "step": 921
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 16.57154655456543,
-      "learning_rate": 1.815375363335672e-05,
-      "loss": 3.3258,
+      "epoch": 0.12,
+      "grad_norm": 14.999835014343262,
+      "learning_rate": 1.9229385432790863e-05,
+      "loss": 0.6485,
       "step": 922
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 12.257933616638184,
-      "learning_rate": 1.8151749022752333e-05,
-      "loss": 2.85,
+      "epoch": 0.12,
+      "grad_norm": 20.776092529296875,
+      "learning_rate": 1.9228548717734176e-05,
+      "loss": 4.4746,
       "step": 923
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 19.822603225708008,
-      "learning_rate": 1.8149744412147943e-05,
-      "loss": 2.7183,
+      "epoch": 0.12,
+      "grad_norm": 13.484353065490723,
+      "learning_rate": 1.922771200267749e-05,
+      "loss": 3.3389,
       "step": 924
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 13.144048690795898,
-      "learning_rate": 1.814773980154355e-05,
-      "loss": 2.9005,
+      "epoch": 0.12,
+      "grad_norm": 17.923723220825195,
+      "learning_rate": 1.9226875287620803e-05,
+      "loss": 2.1387,
       "step": 925
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 13.302081108093262,
-      "learning_rate": 1.8145735190939164e-05,
-      "loss": 3.5627,
+      "epoch": 0.12,
+      "grad_norm": 90.81083679199219,
+      "learning_rate": 1.9226038572564114e-05,
+      "loss": 1.8454,
       "step": 926
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 18.423614501953125,
-      "learning_rate": 1.814373058033477e-05,
-      "loss": 3.7219,
+      "epoch": 0.12,
+      "grad_norm": 22.24589729309082,
+      "learning_rate": 1.9225201857507427e-05,
+      "loss": 2.0114,
       "step": 927
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 16.71343231201172,
-      "learning_rate": 1.814172596973038e-05,
-      "loss": 2.3099,
+      "epoch": 0.12,
+      "grad_norm": 15.252634048461914,
+      "learning_rate": 1.922436514245074e-05,
+      "loss": 4.191,
       "step": 928
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 12.817221641540527,
-      "learning_rate": 1.813972135912599e-05,
-      "loss": 3.1187,
+      "epoch": 0.12,
+      "grad_norm": 23.059764862060547,
+      "learning_rate": 1.922352842739405e-05,
+      "loss": 2.4768,
       "step": 929
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 23.68539047241211,
-      "learning_rate": 1.81377167485216e-05,
-      "loss": 2.8866,
+      "epoch": 0.12,
+      "grad_norm": 18.246435165405273,
+      "learning_rate": 1.9222691712337365e-05,
+      "loss": 4.5082,
       "step": 930
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 19.9428768157959,
-      "learning_rate": 1.813571213791721e-05,
-      "loss": 3.3457,
+      "epoch": 0.12,
+      "grad_norm": 13.866202354431152,
+      "learning_rate": 1.922185499728068e-05,
+      "loss": 4.214,
       "step": 931
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 15.457541465759277,
-      "learning_rate": 1.813370752731282e-05,
-      "loss": 2.5672,
+      "epoch": 0.12,
+      "grad_norm": 17.0211238861084,
+      "learning_rate": 1.9221018282223992e-05,
+      "loss": 2.7061,
       "step": 932
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 14.215147972106934,
-      "learning_rate": 1.813170291670843e-05,
-      "loss": 3.2077,
+      "epoch": 0.12,
+      "grad_norm": 9.928666114807129,
+      "learning_rate": 1.9220181567167303e-05,
+      "loss": 2.093,
       "step": 933
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 23.65352439880371,
-      "learning_rate": 1.812969830610404e-05,
-      "loss": 4.0376,
+      "epoch": 0.12,
+      "grad_norm": 12.430704116821289,
+      "learning_rate": 1.9219344852110616e-05,
+      "loss": 1.8956,
       "step": 934
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 15.811484336853027,
-      "learning_rate": 1.812769369549965e-05,
-      "loss": 3.4806,
+      "epoch": 0.12,
+      "grad_norm": 25.73609161376953,
+      "learning_rate": 1.921850813705393e-05,
+      "loss": 3.4497,
       "step": 935
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 22.097625732421875,
-      "learning_rate": 1.812568908489526e-05,
-      "loss": 2.9302,
+      "epoch": 0.12,
+      "grad_norm": 10.325471878051758,
+      "learning_rate": 1.921767142199724e-05,
+      "loss": 3.3559,
       "step": 936
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 21.592880249023438,
-      "learning_rate": 1.812368447429087e-05,
-      "loss": 3.5155,
+      "epoch": 0.12,
+      "grad_norm": 14.831570625305176,
+      "learning_rate": 1.9216834706940554e-05,
+      "loss": 2.6587,
       "step": 937
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 13.204874992370605,
-      "learning_rate": 1.812167986368648e-05,
-      "loss": 3.172,
+      "epoch": 0.12,
+      "grad_norm": 13.535395622253418,
+      "learning_rate": 1.9215997991883867e-05,
+      "loss": 2.3793,
       "step": 938
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 19.575117111206055,
-      "learning_rate": 1.811967525308209e-05,
-      "loss": 3.1032,
+      "epoch": 0.12,
+      "grad_norm": 27.883773803710938,
+      "learning_rate": 1.9215161276827178e-05,
+      "loss": 3.3018,
       "step": 939
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 19.439678192138672,
-      "learning_rate": 1.81176706424777e-05,
-      "loss": 3.1572,
+      "epoch": 0.12,
+      "grad_norm": 12.214103698730469,
+      "learning_rate": 1.921432456177049e-05,
+      "loss": 4.0168,
       "step": 940
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 18.003063201904297,
-      "learning_rate": 1.8115666031873308e-05,
-      "loss": 3.6915,
+      "epoch": 0.12,
+      "grad_norm": 13.346578598022461,
+      "learning_rate": 1.92134878467138e-05,
+      "loss": 2.6091,
       "step": 941
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 14.612069129943848,
-      "learning_rate": 1.811366142126892e-05,
-      "loss": 2.6144,
+      "epoch": 0.12,
+      "grad_norm": 13.722108840942383,
+      "learning_rate": 1.9212651131657115e-05,
+      "loss": 3.6267,
       "step": 942
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 26.311756134033203,
-      "learning_rate": 1.811165681066453e-05,
-      "loss": 4.1978,
+      "epoch": 0.12,
+      "grad_norm": 25.921424865722656,
+      "learning_rate": 1.921181441660043e-05,
+      "loss": 3.3123,
       "step": 943
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 12.142837524414062,
-      "learning_rate": 1.810965220006014e-05,
-      "loss": 1.9791,
+      "epoch": 0.12,
+      "grad_norm": 11.782764434814453,
+      "learning_rate": 1.921097770154374e-05,
+      "loss": 2.578,
       "step": 944
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 16.259824752807617,
-      "learning_rate": 1.8107647589455752e-05,
-      "loss": 3.0366,
+      "epoch": 0.12,
+      "grad_norm": 23.230119705200195,
+      "learning_rate": 1.9210140986487053e-05,
+      "loss": 2.4671,
       "step": 945
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 15.532299041748047,
-      "learning_rate": 1.810564297885136e-05,
-      "loss": 2.9808,
+      "epoch": 0.12,
+      "grad_norm": 17.044708251953125,
+      "learning_rate": 1.9209304271430366e-05,
+      "loss": 3.4549,
       "step": 946
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 31.902145385742188,
-      "learning_rate": 1.810363836824697e-05,
-      "loss": 4.2394,
+      "epoch": 0.12,
+      "grad_norm": 12.964016914367676,
+      "learning_rate": 1.9208467556373677e-05,
+      "loss": 3.3802,
       "step": 947
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 20.337215423583984,
-      "learning_rate": 1.8101633757642582e-05,
-      "loss": 3.8077,
+      "epoch": 0.12,
+      "grad_norm": 10.344799041748047,
+      "learning_rate": 1.920763084131699e-05,
+      "loss": 2.1356,
       "step": 948
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 15.51062297821045,
-      "learning_rate": 1.809962914703819e-05,
-      "loss": 1.9024,
+      "epoch": 0.12,
+      "grad_norm": 15.343401908874512,
+      "learning_rate": 1.9206794126260304e-05,
+      "loss": 3.9159,
       "step": 949
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 21.733741760253906,
-      "learning_rate": 1.80976245364338e-05,
-      "loss": 3.1563,
+      "epoch": 0.12,
+      "grad_norm": 18.93767738342285,
+      "learning_rate": 1.9205957411203614e-05,
+      "loss": 3.2314,
       "step": 950
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 27.472270965576172,
-      "learning_rate": 1.809561992582941e-05,
-      "loss": 3.7646,
+      "epoch": 0.12,
+      "grad_norm": 16.269479751586914,
+      "learning_rate": 1.9205120696146928e-05,
+      "loss": 2.7732,
       "step": 951
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 33.5841178894043,
-      "learning_rate": 1.809361531522502e-05,
-      "loss": 3.6314,
+      "epoch": 0.12,
+      "grad_norm": 11.748882293701172,
+      "learning_rate": 1.920428398109024e-05,
+      "loss": 2.2064,
       "step": 952
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 15.488965034484863,
-      "learning_rate": 1.809161070462063e-05,
-      "loss": 3.8817,
+      "epoch": 0.12,
+      "grad_norm": 19.187292098999023,
+      "learning_rate": 1.9203447266033555e-05,
+      "loss": 2.2557,
       "step": 953
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 11.746240615844727,
-      "learning_rate": 1.808960609401624e-05,
-      "loss": 2.1518,
+      "epoch": 0.12,
+      "grad_norm": 15.426502227783203,
+      "learning_rate": 1.9202610550976865e-05,
+      "loss": 3.5879,
       "step": 954
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 9.70715618133545,
-      "learning_rate": 1.808760148341185e-05,
-      "loss": 2.5758,
+      "epoch": 0.12,
+      "grad_norm": 18.423494338989258,
+      "learning_rate": 1.920177383592018e-05,
+      "loss": 2.3315,
       "step": 955
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 15.384037017822266,
-      "learning_rate": 1.808559687280746e-05,
-      "loss": 3.2143,
+      "epoch": 0.12,
+      "grad_norm": 21.4342041015625,
+      "learning_rate": 1.9200937120863493e-05,
+      "loss": 3.2196,
       "step": 956
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 18.4316349029541,
-      "learning_rate": 1.808359226220307e-05,
-      "loss": 4.8752,
+      "epoch": 0.12,
+      "grad_norm": 27.630748748779297,
+      "learning_rate": 1.9200100405806803e-05,
+      "loss": 5.156,
       "step": 957
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 13.498906135559082,
-      "learning_rate": 1.8081587651598676e-05,
-      "loss": 3.2658,
+      "epoch": 0.12,
+      "grad_norm": 11.966787338256836,
+      "learning_rate": 1.9199263690750117e-05,
+      "loss": 1.4733,
       "step": 958
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 26.12436866760254,
-      "learning_rate": 1.807958304099429e-05,
-      "loss": 1.9117,
+      "epoch": 0.12,
+      "grad_norm": 16.689373016357422,
+      "learning_rate": 1.919842697569343e-05,
+      "loss": 2.2288,
       "step": 959
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 14.987210273742676,
-      "learning_rate": 1.80775784303899e-05,
-      "loss": 2.1122,
-      "step": 960
-    },
-    {
-      "epoch": 0.29,
-      "eval_loss": 0.5726422667503357,
-      "eval_runtime": 43.6835,
-      "eval_samples_per_second": 33.857,
-      "eval_steps_per_second": 33.857,
+      "epoch": 0.12,
+      "grad_norm": 36.253570556640625,
+      "learning_rate": 1.9197590260636744e-05,
+      "loss": 3.6702,
       "step": 960
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 11.189974784851074,
-      "learning_rate": 1.8075573819785506e-05,
-      "loss": 2.7843,
+      "epoch": 0.12,
+      "grad_norm": 22.745553970336914,
+      "learning_rate": 1.9196753545580054e-05,
+      "loss": 2.2173,
       "step": 961
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 13.831470489501953,
-      "learning_rate": 1.807356920918112e-05,
-      "loss": 3.8859,
+      "epoch": 0.12,
+      "grad_norm": 8.689067840576172,
+      "learning_rate": 1.9195916830523368e-05,
+      "loss": 1.2961,
       "step": 962
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 14.756105422973633,
-      "learning_rate": 1.8071564598576727e-05,
-      "loss": 2.6933,
+      "epoch": 0.12,
+      "grad_norm": 15.15867805480957,
+      "learning_rate": 1.919508011546668e-05,
+      "loss": 3.4019,
       "step": 963
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 13.506275177001953,
-      "learning_rate": 1.8069559987972337e-05,
-      "loss": 2.7927,
+      "epoch": 0.12,
+      "grad_norm": 15.15098762512207,
+      "learning_rate": 1.9194243400409992e-05,
+      "loss": 3.566,
       "step": 964
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 27.999460220336914,
-      "learning_rate": 1.8067555377367947e-05,
-      "loss": 3.3839,
+      "epoch": 0.12,
+      "grad_norm": 17.65106201171875,
+      "learning_rate": 1.9193406685353305e-05,
+      "loss": 2.8859,
       "step": 965
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 13.807269096374512,
-      "learning_rate": 1.8065550766763557e-05,
-      "loss": 2.0444,
+      "epoch": 0.12,
+      "grad_norm": 15.059209823608398,
+      "learning_rate": 1.9192569970296616e-05,
+      "loss": 2.474,
       "step": 966
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 15.027080535888672,
-      "learning_rate": 1.8063546156159167e-05,
-      "loss": 2.5851,
+      "epoch": 0.12,
+      "grad_norm": 13.547049522399902,
+      "learning_rate": 1.919173325523993e-05,
+      "loss": 2.8793,
       "step": 967
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 46.652530670166016,
-      "learning_rate": 1.8061541545554777e-05,
-      "loss": 2.5573,
+      "epoch": 0.12,
+      "grad_norm": 15.953766822814941,
+      "learning_rate": 1.9190896540183243e-05,
+      "loss": 2.6781,
       "step": 968
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 17.231346130371094,
-      "learning_rate": 1.8059536934950387e-05,
-      "loss": 2.3659,
+      "epoch": 0.12,
+      "grad_norm": 19.517332077026367,
+      "learning_rate": 1.9190059825126553e-05,
+      "loss": 4.631,
       "step": 969
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 17.74903678894043,
-      "learning_rate": 1.8057532324345997e-05,
-      "loss": 2.8746,
+      "epoch": 0.12,
+      "grad_norm": 13.69670295715332,
+      "learning_rate": 1.9189223110069867e-05,
+      "loss": 3.5081,
       "step": 970
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 9.483039855957031,
-      "learning_rate": 1.8055527713741607e-05,
-      "loss": 2.6829,
+      "epoch": 0.12,
+      "grad_norm": 26.607080459594727,
+      "learning_rate": 1.9188386395013177e-05,
+      "loss": 3.4868,
       "step": 971
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 10.787821769714355,
-      "learning_rate": 1.8053523103137217e-05,
-      "loss": 2.395,
+      "epoch": 0.12,
+      "grad_norm": 11.74852466583252,
+      "learning_rate": 1.918754967995649e-05,
+      "loss": 1.8016,
       "step": 972
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 16.904396057128906,
-      "learning_rate": 1.8051518492532827e-05,
-      "loss": 2.8498,
+      "epoch": 0.12,
+      "grad_norm": 14.13051986694336,
+      "learning_rate": 1.9186712964899804e-05,
+      "loss": 3.496,
       "step": 973
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 18.882564544677734,
-      "learning_rate": 1.8049513881928437e-05,
-      "loss": 3.3251,
+      "epoch": 0.12,
+      "grad_norm": 13.406763076782227,
+      "learning_rate": 1.9185876249843118e-05,
+      "loss": 4.1372,
       "step": 974
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 18.128250122070312,
-      "learning_rate": 1.8047509271324048e-05,
-      "loss": 3.1162,
+      "epoch": 0.12,
+      "grad_norm": 12.86937141418457,
+      "learning_rate": 1.918503953478643e-05,
+      "loss": 4.6107,
       "step": 975
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 19.998552322387695,
-      "learning_rate": 1.8045504660719658e-05,
-      "loss": 3.183,
+      "epoch": 0.12,
+      "grad_norm": 17.223657608032227,
+      "learning_rate": 1.9184202819729742e-05,
+      "loss": 3.7678,
       "step": 976
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 19.723297119140625,
-      "learning_rate": 1.8043500050115264e-05,
-      "loss": 3.0132,
+      "epoch": 0.12,
+      "grad_norm": 13.638578414916992,
+      "learning_rate": 1.9183366104673056e-05,
+      "loss": 3.242,
       "step": 977
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 18.414817810058594,
-      "learning_rate": 1.8041495439510878e-05,
-      "loss": 2.6902,
+      "epoch": 0.12,
+      "grad_norm": 14.202733993530273,
+      "learning_rate": 1.9182529389616366e-05,
+      "loss": 2.0694,
       "step": 978
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 15.55300521850586,
-      "learning_rate": 1.8039490828906488e-05,
-      "loss": 2.0389,
+      "epoch": 0.12,
+      "grad_norm": 20.721847534179688,
+      "learning_rate": 1.918169267455968e-05,
+      "loss": 4.5076,
       "step": 979
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 19.29793357849121,
-      "learning_rate": 1.8037486218302095e-05,
-      "loss": 3.276,
+      "epoch": 0.12,
+      "grad_norm": 14.14547348022461,
+      "learning_rate": 1.9180855959502993e-05,
+      "loss": 4.7961,
       "step": 980
     },
     {
-      "epoch": 0.29,
-      "grad_norm": 18.001596450805664,
-      "learning_rate": 1.8035481607697708e-05,
-      "loss": 2.2783,
+      "epoch": 0.12,
+      "grad_norm": 13.157936096191406,
+      "learning_rate": 1.9180019244446307e-05,
+      "loss": 1.7192,
       "step": 981
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 12.103147506713867,
-      "learning_rate": 1.8033476997093315e-05,
-      "loss": 3.8083,
+      "epoch": 0.12,
+      "grad_norm": 15.036370277404785,
+      "learning_rate": 1.9179182529389617e-05,
+      "loss": 2.2217,
       "step": 982
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 15.265912055969238,
-      "learning_rate": 1.8031472386488925e-05,
-      "loss": 2.1125,
+      "epoch": 0.12,
+      "grad_norm": 33.67008590698242,
+      "learning_rate": 1.917834581433293e-05,
+      "loss": 2.4785,
       "step": 983
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 20.835962295532227,
-      "learning_rate": 1.8029467775884535e-05,
-      "loss": 2.128,
+      "epoch": 0.12,
+      "grad_norm": 14.870255470275879,
+      "learning_rate": 1.9177509099276244e-05,
+      "loss": 2.7684,
       "step": 984
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 27.07541847229004,
-      "learning_rate": 1.8027463165280145e-05,
-      "loss": 2.9855,
+      "epoch": 0.12,
+      "grad_norm": 25.22587013244629,
+      "learning_rate": 1.9176672384219555e-05,
+      "loss": 2.7044,
       "step": 985
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 17.84712791442871,
-      "learning_rate": 1.8025458554675755e-05,
-      "loss": 2.7988,
+      "epoch": 0.12,
+      "grad_norm": 9.532674789428711,
+      "learning_rate": 1.917583566916287e-05,
+      "loss": 2.7126,
       "step": 986
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 15.661683082580566,
-      "learning_rate": 1.8023453944071365e-05,
-      "loss": 3.1825,
+      "epoch": 0.12,
+      "grad_norm": 12.704583168029785,
+      "learning_rate": 1.9174998954106182e-05,
+      "loss": 2.1161,
       "step": 987
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 21.294462203979492,
-      "learning_rate": 1.8021449333466975e-05,
-      "loss": 3.2597,
+      "epoch": 0.12,
+      "grad_norm": 25.719287872314453,
+      "learning_rate": 1.9174162239049496e-05,
+      "loss": 4.9047,
       "step": 988
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 14.863018035888672,
-      "learning_rate": 1.8019444722862585e-05,
-      "loss": 2.7602,
+      "epoch": 0.12,
+      "grad_norm": 12.979522705078125,
+      "learning_rate": 1.9173325523992806e-05,
+      "loss": 2.8941,
       "step": 989
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 14.493101119995117,
-      "learning_rate": 1.8017440112258195e-05,
-      "loss": 3.5392,
+      "epoch": 0.12,
+      "grad_norm": 10.867280960083008,
+      "learning_rate": 1.917248880893612e-05,
+      "loss": 1.9709,
       "step": 990
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 16.697307586669922,
-      "learning_rate": 1.8015435501653806e-05,
-      "loss": 3.0317,
+      "epoch": 0.12,
+      "grad_norm": 11.747751235961914,
+      "learning_rate": 1.9171652093879433e-05,
+      "loss": 2.9887,
       "step": 991
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 21.106035232543945,
-      "learning_rate": 1.8013430891049416e-05,
-      "loss": 2.6286,
+      "epoch": 0.12,
+      "grad_norm": 18.359580993652344,
+      "learning_rate": 1.9170815378822743e-05,
+      "loss": 4.3102,
       "step": 992
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 14.834815979003906,
-      "learning_rate": 1.8011426280445026e-05,
-      "loss": 2.2224,
+      "epoch": 0.12,
+      "grad_norm": 19.873567581176758,
+      "learning_rate": 1.9169978663766057e-05,
+      "loss": 3.9871,
       "step": 993
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 16.66942596435547,
-      "learning_rate": 1.8009421669840636e-05,
-      "loss": 3.2398,
+      "epoch": 0.12,
+      "grad_norm": 18.60597038269043,
+      "learning_rate": 1.9169141948709367e-05,
+      "loss": 3.5496,
       "step": 994
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 20.955211639404297,
-      "learning_rate": 1.8007417059236246e-05,
-      "loss": 3.2694,
+      "epoch": 0.12,
+      "grad_norm": 17.307985305786133,
+      "learning_rate": 1.916830523365268e-05,
+      "loss": 3.6558,
       "step": 995
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 16.78022575378418,
-      "learning_rate": 1.8005412448631853e-05,
-      "loss": 3.3401,
+      "epoch": 0.12,
+      "grad_norm": 16.86280059814453,
+      "learning_rate": 1.9167468518595995e-05,
+      "loss": 2.0322,
       "step": 996
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 10.238494873046875,
-      "learning_rate": 1.8003407838027466e-05,
-      "loss": 2.686,
+      "epoch": 0.13,
+      "grad_norm": 13.198040962219238,
+      "learning_rate": 1.9166631803539305e-05,
+      "loss": 2.366,
       "step": 997
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 25.67633628845215,
-      "learning_rate": 1.8001403227423076e-05,
-      "loss": 4.3305,
+      "epoch": 0.13,
+      "grad_norm": 14.232464790344238,
+      "learning_rate": 1.916579508848262e-05,
+      "loss": 2.8017,
       "step": 998
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 18.46412467956543,
-      "learning_rate": 1.7999398616818683e-05,
-      "loss": 4.0384,
+      "epoch": 0.13,
+      "grad_norm": 15.965517044067383,
+      "learning_rate": 1.916495837342593e-05,
+      "loss": 4.077,
       "step": 999
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 16.967388153076172,
-      "learning_rate": 1.7997394006214296e-05,
-      "loss": 2.752,
+      "epoch": 0.13,
+      "grad_norm": 10.01248550415039,
+      "learning_rate": 1.9164121658369242e-05,
+      "loss": 0.4904,
       "step": 1000
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 16.72841453552246,
-      "learning_rate": 1.7995389395609903e-05,
-      "loss": 1.9186,
+      "epoch": 0.13,
+      "grad_norm": 36.76567077636719,
+      "learning_rate": 1.9163284943312556e-05,
+      "loss": 1.9846,
       "step": 1001
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 17.65038299560547,
-      "learning_rate": 1.7993384785005513e-05,
-      "loss": 2.8156,
+      "epoch": 0.13,
+      "grad_norm": 14.137666702270508,
+      "learning_rate": 1.916244822825587e-05,
+      "loss": 1.5229,
       "step": 1002
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 15.546381950378418,
-      "learning_rate": 1.7991380174401123e-05,
-      "loss": 2.9373,
+      "epoch": 0.13,
+      "grad_norm": 21.321977615356445,
+      "learning_rate": 1.916161151319918e-05,
+      "loss": 5.1578,
       "step": 1003
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 13.01362133026123,
-      "learning_rate": 1.7989375563796733e-05,
-      "loss": 2.4807,
+      "epoch": 0.13,
+      "grad_norm": 11.998373985290527,
+      "learning_rate": 1.9160774798142494e-05,
+      "loss": 2.938,
       "step": 1004
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 18.31947135925293,
-      "learning_rate": 1.7987370953192343e-05,
-      "loss": 2.6384,
+      "epoch": 0.13,
+      "grad_norm": 15.488204956054688,
+      "learning_rate": 1.9159938083085807e-05,
+      "loss": 5.3228,
       "step": 1005
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 19.710500717163086,
-      "learning_rate": 1.7985366342587953e-05,
-      "loss": 2.9314,
+      "epoch": 0.13,
+      "grad_norm": 24.641403198242188,
+      "learning_rate": 1.9159101368029118e-05,
+      "loss": 2.3875,
       "step": 1006
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 12.663653373718262,
-      "learning_rate": 1.7983361731983563e-05,
-      "loss": 1.9977,
+      "epoch": 0.13,
+      "grad_norm": 12.331722259521484,
+      "learning_rate": 1.915826465297243e-05,
+      "loss": 1.4247,
       "step": 1007
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 23.557924270629883,
-      "learning_rate": 1.7981357121379174e-05,
-      "loss": 3.6326,
+      "epoch": 0.13,
+      "grad_norm": 10.565719604492188,
+      "learning_rate": 1.9157427937915745e-05,
+      "loss": 2.8828,
       "step": 1008
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 12.505498886108398,
-      "learning_rate": 1.7979352510774784e-05,
-      "loss": 2.3971,
+      "epoch": 0.13,
+      "grad_norm": 15.967330932617188,
+      "learning_rate": 1.915659122285906e-05,
+      "loss": 4.3271,
       "step": 1009
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 21.709392547607422,
-      "learning_rate": 1.7977347900170394e-05,
-      "loss": 2.4877,
+      "epoch": 0.13,
+      "grad_norm": 9.954456329345703,
+      "learning_rate": 1.915575450780237e-05,
+      "loss": 3.6863,
       "step": 1010
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 12.365106582641602,
-      "learning_rate": 1.7975343289566004e-05,
-      "loss": 3.1574,
+      "epoch": 0.13,
+      "grad_norm": 18.42413902282715,
+      "learning_rate": 1.9154917792745682e-05,
+      "loss": 3.9867,
       "step": 1011
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 18.052745819091797,
-      "learning_rate": 1.7973338678961614e-05,
-      "loss": 3.6156,
+      "epoch": 0.13,
+      "grad_norm": 31.44066047668457,
+      "learning_rate": 1.9154081077688996e-05,
+      "loss": 5.7757,
       "step": 1012
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 16.1375732421875,
-      "learning_rate": 1.7971334068357224e-05,
-      "loss": 3.0042,
+      "epoch": 0.13,
+      "grad_norm": 22.84227752685547,
+      "learning_rate": 1.9153244362632306e-05,
+      "loss": 2.1757,
       "step": 1013
     },
     {
-      "epoch": 0.3,
-      "grad_norm": 15.52522087097168,
-      "learning_rate": 1.7969329457752834e-05,
-      "loss": 3.3669,
+      "epoch": 0.13,
+      "grad_norm": 16.95429801940918,
+      "learning_rate": 1.915240764757562e-05,
+      "loss": 3.5305,
       "step": 1014
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 18.78600311279297,
-      "learning_rate": 1.796732484714844e-05,
-      "loss": 4.9568,
+      "epoch": 0.13,
+      "grad_norm": 14.703499794006348,
+      "learning_rate": 1.9151570932518934e-05,
+      "loss": 2.3546,
       "step": 1015
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 32.19879150390625,
-      "learning_rate": 1.7965320236544054e-05,
-      "loss": 3.4862,
+      "epoch": 0.13,
+      "grad_norm": 14.213774681091309,
+      "learning_rate": 1.9150734217462247e-05,
+      "loss": 3.4051,
       "step": 1016
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 22.817644119262695,
-      "learning_rate": 1.7963315625939664e-05,
-      "loss": 3.6447,
+      "epoch": 0.13,
+      "grad_norm": 14.093331336975098,
+      "learning_rate": 1.9149897502405558e-05,
+      "loss": 3.1891,
       "step": 1017
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 13.087103843688965,
-      "learning_rate": 1.796131101533527e-05,
-      "loss": 3.1701,
+      "epoch": 0.13,
+      "grad_norm": 14.156906127929688,
+      "learning_rate": 1.914906078734887e-05,
+      "loss": 2.7102,
       "step": 1018
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 15.170973777770996,
-      "learning_rate": 1.7959306404730884e-05,
-      "loss": 2.907,
+      "epoch": 0.13,
+      "grad_norm": 17.22342872619629,
+      "learning_rate": 1.914822407229218e-05,
+      "loss": 3.1369,
       "step": 1019
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 18.237186431884766,
-      "learning_rate": 1.795730179412649e-05,
-      "loss": 3.34,
+      "epoch": 0.13,
+      "grad_norm": 16.823516845703125,
+      "learning_rate": 1.9147387357235495e-05,
+      "loss": 3.5317,
       "step": 1020
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 21.131467819213867,
-      "learning_rate": 1.79552971835221e-05,
-      "loss": 3.0868,
+      "epoch": 0.13,
+      "grad_norm": 10.506781578063965,
+      "learning_rate": 1.914655064217881e-05,
+      "loss": 2.212,
       "step": 1021
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 21.808650970458984,
-      "learning_rate": 1.7953292572917715e-05,
-      "loss": 2.4194,
+      "epoch": 0.13,
+      "grad_norm": 10.249368667602539,
+      "learning_rate": 1.914571392712212e-05,
+      "loss": 2.169,
       "step": 1022
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 8.451123237609863,
-      "learning_rate": 1.795128796231332e-05,
-      "loss": 2.0371,
+      "epoch": 0.13,
+      "grad_norm": 11.512591361999512,
+      "learning_rate": 1.9144877212065433e-05,
+      "loss": 2.7708,
       "step": 1023
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 13.38468074798584,
-      "learning_rate": 1.794928335170893e-05,
-      "loss": 2.1737,
+      "epoch": 0.13,
+      "grad_norm": 17.4201602935791,
+      "learning_rate": 1.9144040497008743e-05,
+      "loss": 4.4981,
       "step": 1024
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 11.502458572387695,
-      "learning_rate": 1.794727874110454e-05,
-      "loss": 2.5593,
+      "epoch": 0.13,
+      "grad_norm": 8.540460586547852,
+      "learning_rate": 1.9143203781952057e-05,
+      "loss": 1.1926,
       "step": 1025
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 8.056184768676758,
-      "learning_rate": 1.794527413050015e-05,
-      "loss": 2.0176,
+      "epoch": 0.13,
+      "grad_norm": 10.993705749511719,
+      "learning_rate": 1.914236706689537e-05,
+      "loss": 1.6827,
       "step": 1026
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 12.703288078308105,
-      "learning_rate": 1.7943269519895762e-05,
-      "loss": 3.1376,
+      "epoch": 0.13,
+      "grad_norm": 14.419856071472168,
+      "learning_rate": 1.914153035183868e-05,
+      "loss": 3.2344,
       "step": 1027
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 18.933555603027344,
-      "learning_rate": 1.7941264909291372e-05,
-      "loss": 2.6452,
+      "epoch": 0.13,
+      "grad_norm": 15.001279830932617,
+      "learning_rate": 1.9140693636781994e-05,
+      "loss": 3.0493,
       "step": 1028
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 19.963056564331055,
-      "learning_rate": 1.7939260298686982e-05,
-      "loss": 3.3245,
+      "epoch": 0.13,
+      "grad_norm": 30.563142776489258,
+      "learning_rate": 1.9139856921725308e-05,
+      "loss": 3.8982,
       "step": 1029
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 11.356467247009277,
-      "learning_rate": 1.7937255688082592e-05,
-      "loss": 2.8729,
+      "epoch": 0.13,
+      "grad_norm": 12.270652770996094,
+      "learning_rate": 1.913902020666862e-05,
+      "loss": 3.1659,
       "step": 1030
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 22.96855926513672,
-      "learning_rate": 1.7935251077478202e-05,
-      "loss": 3.4278,
+      "epoch": 0.13,
+      "grad_norm": 11.87707805633545,
+      "learning_rate": 1.913818349161193e-05,
+      "loss": 1.5629,
       "step": 1031
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 9.410229682922363,
-      "learning_rate": 1.793324646687381e-05,
-      "loss": 2.4024,
+      "epoch": 0.13,
+      "grad_norm": 14.46357250213623,
+      "learning_rate": 1.9137346776555245e-05,
+      "loss": 2.2168,
       "step": 1032
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 15.325308799743652,
-      "learning_rate": 1.7931241856269422e-05,
-      "loss": 3.0026,
+      "epoch": 0.13,
+      "grad_norm": 12.322527885437012,
+      "learning_rate": 1.913651006149856e-05,
+      "loss": 2.4828,
       "step": 1033
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 17.43353271484375,
-      "learning_rate": 1.7929237245665032e-05,
-      "loss": 3.4274,
+      "epoch": 0.13,
+      "grad_norm": 11.011430740356445,
+      "learning_rate": 1.913567334644187e-05,
+      "loss": 3.1574,
       "step": 1034
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 18.91429328918457,
-      "learning_rate": 1.792723263506064e-05,
-      "loss": 2.8543,
+      "epoch": 0.13,
+      "grad_norm": 9.569558143615723,
+      "learning_rate": 1.9134836631385183e-05,
+      "loss": 3.0719,
       "step": 1035
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 13.564133644104004,
-      "learning_rate": 1.7925228024456253e-05,
-      "loss": 1.9484,
+      "epoch": 0.13,
+      "grad_norm": 18.550840377807617,
+      "learning_rate": 1.9133999916328497e-05,
+      "loss": 2.4526,
       "step": 1036
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 21.08160972595215,
-      "learning_rate": 1.792322341385186e-05,
-      "loss": 3.0899,
+      "epoch": 0.13,
+      "grad_norm": 33.769222259521484,
+      "learning_rate": 1.913316320127181e-05,
+      "loss": 2.9887,
       "step": 1037
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 11.906721115112305,
-      "learning_rate": 1.792121880324747e-05,
-      "loss": 2.3356,
+      "epoch": 0.13,
+      "grad_norm": 24.496849060058594,
+      "learning_rate": 1.913232648621512e-05,
+      "loss": 2.9969,
       "step": 1038
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 12.753665924072266,
-      "learning_rate": 1.791921419264308e-05,
-      "loss": 1.7421,
+      "epoch": 0.13,
+      "grad_norm": 9.507521629333496,
+      "learning_rate": 1.9131489771158434e-05,
+      "loss": 2.0447,
       "step": 1039
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 13.137770652770996,
-      "learning_rate": 1.791720958203869e-05,
-      "loss": 3.0557,
+      "epoch": 0.13,
+      "grad_norm": 14.459228515625,
+      "learning_rate": 1.9130653056101748e-05,
+      "loss": 2.9356,
       "step": 1040
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 18.3424129486084,
-      "learning_rate": 1.7915204971434303e-05,
-      "loss": 2.9359,
+      "epoch": 0.13,
+      "grad_norm": 7.550471782684326,
+      "learning_rate": 1.9129816341045058e-05,
+      "loss": 1.9061,
       "step": 1041
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 14.686424255371094,
-      "learning_rate": 1.791320036082991e-05,
-      "loss": 2.947,
+      "epoch": 0.13,
+      "grad_norm": 17.238767623901367,
+      "learning_rate": 1.912897962598837e-05,
+      "loss": 4.294,
       "step": 1042
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 18.53430938720703,
-      "learning_rate": 1.791119575022552e-05,
-      "loss": 3.6603,
+      "epoch": 0.13,
+      "grad_norm": 19.979581832885742,
+      "learning_rate": 1.9128142910931685e-05,
+      "loss": 3.699,
       "step": 1043
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 12.61064338684082,
-      "learning_rate": 1.790919113962113e-05,
-      "loss": 3.071,
+      "epoch": 0.13,
+      "grad_norm": 22.493305206298828,
+      "learning_rate": 1.9127306195875e-05,
+      "loss": 3.2351,
       "step": 1044
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 16.695178985595703,
-      "learning_rate": 1.790718652901674e-05,
-      "loss": 3.6908,
+      "epoch": 0.13,
+      "grad_norm": 11.701733589172363,
+      "learning_rate": 1.912646948081831e-05,
+      "loss": 2.3459,
       "step": 1045
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 21.377464294433594,
-      "learning_rate": 1.790518191841235e-05,
-      "loss": 2.0132,
+      "epoch": 0.13,
+      "grad_norm": 19.428638458251953,
+      "learning_rate": 1.9125632765761623e-05,
+      "loss": 1.2258,
       "step": 1046
     },
     {
-      "epoch": 0.31,
-      "grad_norm": 12.307204246520996,
-      "learning_rate": 1.790317730780796e-05,
-      "loss": 3.3536,
+      "epoch": 0.13,
+      "grad_norm": 11.169611930847168,
+      "learning_rate": 1.9124796050704933e-05,
+      "loss": 0.8787,
       "step": 1047
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 18.136398315429688,
-      "learning_rate": 1.790117269720357e-05,
-      "loss": 3.3327,
+      "epoch": 0.13,
+      "grad_norm": 10.714649200439453,
+      "learning_rate": 1.9123959335648247e-05,
+      "loss": 2.9222,
       "step": 1048
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 14.812959671020508,
-      "learning_rate": 1.789916808659918e-05,
-      "loss": 2.203,
+      "epoch": 0.13,
+      "grad_norm": 17.899307250976562,
+      "learning_rate": 1.912312262059156e-05,
+      "loss": 2.7829,
       "step": 1049
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 16.73038101196289,
-      "learning_rate": 1.789716347599479e-05,
-      "loss": 2.5606,
+      "epoch": 0.13,
+      "grad_norm": 21.250642776489258,
+      "learning_rate": 1.912228590553487e-05,
+      "loss": 3.9042,
       "step": 1050
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 28.89952278137207,
-      "learning_rate": 1.7895158865390397e-05,
-      "loss": 3.7834,
+      "epoch": 0.13,
+      "grad_norm": 12.698055267333984,
+      "learning_rate": 1.9121449190478184e-05,
+      "loss": 3.1538,
       "step": 1051
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 16.880115509033203,
-      "learning_rate": 1.789315425478601e-05,
-      "loss": 2.5196,
+      "epoch": 0.13,
+      "grad_norm": 63.32461166381836,
+      "learning_rate": 1.9120612475421495e-05,
+      "loss": 2.2006,
       "step": 1052
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 13.364277839660645,
-      "learning_rate": 1.789114964418162e-05,
-      "loss": 2.9332,
+      "epoch": 0.13,
+      "grad_norm": 14.6691312789917,
+      "learning_rate": 1.9119775760364808e-05,
+      "loss": 3.7852,
       "step": 1053
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 34.66802978515625,
-      "learning_rate": 1.7889145033577227e-05,
-      "loss": 3.8191,
+      "epoch": 0.13,
+      "grad_norm": 12.2632474899292,
+      "learning_rate": 1.9118939045308122e-05,
+      "loss": 1.982,
       "step": 1054
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 31.339599609375,
-      "learning_rate": 1.788714042297284e-05,
-      "loss": 2.9259,
+      "epoch": 0.13,
+      "grad_norm": 20.339496612548828,
+      "learning_rate": 1.9118102330251432e-05,
+      "loss": 3.9451,
       "step": 1055
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 20.67884063720703,
-      "learning_rate": 1.7885135812368447e-05,
-      "loss": 2.7581,
+      "epoch": 0.13,
+      "grad_norm": 11.595175743103027,
+      "learning_rate": 1.9117265615194746e-05,
+      "loss": 1.8827,
       "step": 1056
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 15.182465553283691,
-      "learning_rate": 1.7883131201764058e-05,
-      "loss": 2.8311,
+      "epoch": 0.13,
+      "grad_norm": 17.194581985473633,
+      "learning_rate": 1.911642890013806e-05,
+      "loss": 0.9844,
       "step": 1057
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 18.39655303955078,
-      "learning_rate": 1.7881126591159668e-05,
-      "loss": 2.7088,
+      "epoch": 0.13,
+      "grad_norm": 12.588428497314453,
+      "learning_rate": 1.9115592185081373e-05,
+      "loss": 3.0674,
       "step": 1058
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 15.220178604125977,
-      "learning_rate": 1.7879121980555278e-05,
-      "loss": 3.1082,
+      "epoch": 0.13,
+      "grad_norm": 14.255656242370605,
+      "learning_rate": 1.9114755470024683e-05,
+      "loss": 3.8112,
       "step": 1059
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 13.494277954101562,
-      "learning_rate": 1.7877117369950888e-05,
-      "loss": 2.8682,
+      "epoch": 0.13,
+      "grad_norm": 18.233028411865234,
+      "learning_rate": 1.9113918754967997e-05,
+      "loss": 3.1386,
       "step": 1060
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 19.83290672302246,
-      "learning_rate": 1.7875112759346498e-05,
-      "loss": 3.3298,
+      "epoch": 0.13,
+      "grad_norm": 10.391460418701172,
+      "learning_rate": 1.911308203991131e-05,
+      "loss": 1.8707,
       "step": 1061
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 22.487585067749023,
-      "learning_rate": 1.7873108148742108e-05,
-      "loss": 3.2538,
+      "epoch": 0.13,
+      "grad_norm": 17.744470596313477,
+      "learning_rate": 1.911224532485462e-05,
+      "loss": 2.09,
       "step": 1062
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 24.89381217956543,
-      "learning_rate": 1.7871103538137718e-05,
-      "loss": 2.5621,
+      "epoch": 0.13,
+      "grad_norm": 13.269696235656738,
+      "learning_rate": 1.9111408609797935e-05,
+      "loss": 3.8912,
       "step": 1063
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 17.16600227355957,
-      "learning_rate": 1.7869098927533328e-05,
-      "loss": 2.8225,
+      "epoch": 0.13,
+      "grad_norm": 52.91433334350586,
+      "learning_rate": 1.9110571894741248e-05,
+      "loss": 2.4628,
       "step": 1064
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 10.632941246032715,
-      "learning_rate": 1.7867094316928938e-05,
-      "loss": 3.0174,
+      "epoch": 0.13,
+      "grad_norm": 10.043838500976562,
+      "learning_rate": 1.9109735179684562e-05,
+      "loss": 1.085,
       "step": 1065
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 20.951936721801758,
-      "learning_rate": 1.7865089706324548e-05,
-      "loss": 3.5367,
+      "epoch": 0.13,
+      "grad_norm": 11.903457641601562,
+      "learning_rate": 1.9108898464627872e-05,
+      "loss": 1.5179,
       "step": 1066
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 15.389318466186523,
-      "learning_rate": 1.786308509572016e-05,
-      "loss": 2.2259,
+      "epoch": 0.13,
+      "grad_norm": 11.420485496520996,
+      "learning_rate": 1.9108061749571186e-05,
+      "loss": 2.3194,
       "step": 1067
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 44.66872787475586,
-      "learning_rate": 1.786108048511577e-05,
-      "loss": 4.754,
+      "epoch": 0.13,
+      "grad_norm": 14.506755828857422,
+      "learning_rate": 1.91072250345145e-05,
+      "loss": 3.0316,
       "step": 1068
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 12.777280807495117,
-      "learning_rate": 1.785907587451138e-05,
-      "loss": 3.2503,
+      "epoch": 0.13,
+      "grad_norm": 8.62365436553955,
+      "learning_rate": 1.910638831945781e-05,
+      "loss": 3.8688,
       "step": 1069
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 130.45664978027344,
-      "learning_rate": 1.7857071263906985e-05,
-      "loss": 2.8328,
+      "epoch": 0.13,
+      "grad_norm": 15.223709106445312,
+      "learning_rate": 1.9105551604401123e-05,
+      "loss": 3.7763,
       "step": 1070
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 25.45984649658203,
-      "learning_rate": 1.78550666533026e-05,
-      "loss": 2.7516,
+      "epoch": 0.13,
+      "grad_norm": 18.46766471862793,
+      "learning_rate": 1.9104714889344437e-05,
+      "loss": 2.5704,
       "step": 1071
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 15.323507308959961,
-      "learning_rate": 1.785306204269821e-05,
-      "loss": 2.0025,
+      "epoch": 0.13,
+      "grad_norm": 10.691957473754883,
+      "learning_rate": 1.9103878174287747e-05,
+      "loss": 2.6815,
       "step": 1072
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 12.247516632080078,
-      "learning_rate": 1.7851057432093815e-05,
-      "loss": 2.7504,
+      "epoch": 0.13,
+      "grad_norm": 9.40165901184082,
+      "learning_rate": 1.910304145923106e-05,
+      "loss": 2.6142,
       "step": 1073
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 33.093528747558594,
-      "learning_rate": 1.784905282148943e-05,
-      "loss": 3.8759,
+      "epoch": 0.13,
+      "grad_norm": 19.226882934570312,
+      "learning_rate": 1.9102204744174375e-05,
+      "loss": 4.4681,
       "step": 1074
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 19.78754234313965,
-      "learning_rate": 1.7847048210885036e-05,
-      "loss": 3.0278,
+      "epoch": 0.13,
+      "grad_norm": 16.283008575439453,
+      "learning_rate": 1.9101368029117685e-05,
+      "loss": 3.1899,
       "step": 1075
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 15.436639785766602,
-      "learning_rate": 1.7845043600280646e-05,
-      "loss": 3.5511,
+      "epoch": 0.14,
+      "grad_norm": 23.305496215820312,
+      "learning_rate": 1.9100531314061e-05,
+      "loss": 4.6672,
       "step": 1076
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 12.693981170654297,
-      "learning_rate": 1.784303898967626e-05,
-      "loss": 3.4346,
+      "epoch": 0.14,
+      "grad_norm": 36.387516021728516,
+      "learning_rate": 1.909969459900431e-05,
+      "loss": 2.1056,
       "step": 1077
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 17.650697708129883,
-      "learning_rate": 1.7841034379071866e-05,
-      "loss": 2.3388,
+      "epoch": 0.14,
+      "grad_norm": 40.60980224609375,
+      "learning_rate": 1.9098857883947622e-05,
+      "loss": 3.5833,
       "step": 1078
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 18.479936599731445,
-      "learning_rate": 1.7839029768467476e-05,
-      "loss": 3.4169,
+      "epoch": 0.14,
+      "grad_norm": 13.623090744018555,
+      "learning_rate": 1.9098021168890936e-05,
+      "loss": 3.3983,
       "step": 1079
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 14.039420127868652,
-      "learning_rate": 1.7837025157863086e-05,
-      "loss": 3.0089,
-      "step": 1080
-    },
-    {
-      "epoch": 0.32,
-      "eval_loss": 0.5645097494125366,
-      "eval_runtime": 43.9047,
-      "eval_samples_per_second": 33.687,
-      "eval_steps_per_second": 33.687,
+      "epoch": 0.14,
+      "grad_norm": 16.30078125,
+      "learning_rate": 1.9097184453834246e-05,
+      "loss": 2.0246,
       "step": 1080
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 18.384023666381836,
-      "learning_rate": 1.7835020547258696e-05,
-      "loss": 5.0499,
+      "epoch": 0.14,
+      "grad_norm": 26.719799041748047,
+      "learning_rate": 1.909634773877756e-05,
+      "loss": 2.9841,
       "step": 1081
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 14.556102752685547,
-      "learning_rate": 1.7833015936654306e-05,
-      "loss": 2.7198,
+      "epoch": 0.14,
+      "grad_norm": 15.77614974975586,
+      "learning_rate": 1.9095511023720874e-05,
+      "loss": 2.7169,
       "step": 1082
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 16.748477935791016,
-      "learning_rate": 1.7831011326049916e-05,
-      "loss": 2.7809,
+      "epoch": 0.14,
+      "grad_norm": 20.72202491760254,
+      "learning_rate": 1.9094674308664184e-05,
+      "loss": 3.6718,
       "step": 1083
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 16.843488693237305,
-      "learning_rate": 1.7829006715445526e-05,
-      "loss": 3.1988,
+      "epoch": 0.14,
+      "grad_norm": 19.182992935180664,
+      "learning_rate": 1.9093837593607497e-05,
+      "loss": 4.8896,
       "step": 1084
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 60.20880889892578,
-      "learning_rate": 1.7827002104841137e-05,
-      "loss": 4.0739,
+      "epoch": 0.14,
+      "grad_norm": 9.58013916015625,
+      "learning_rate": 1.909300087855081e-05,
+      "loss": 2.4248,
       "step": 1085
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 22.219043731689453,
-      "learning_rate": 1.7824997494236747e-05,
-      "loss": 3.1217,
+      "epoch": 0.14,
+      "grad_norm": 16.581295013427734,
+      "learning_rate": 1.9092164163494125e-05,
+      "loss": 4.0646,
       "step": 1086
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 15.742715835571289,
-      "learning_rate": 1.7822992883632357e-05,
-      "loss": 2.1444,
+      "epoch": 0.14,
+      "grad_norm": 18.470375061035156,
+      "learning_rate": 1.9091327448437435e-05,
+      "loss": 4.0107,
       "step": 1087
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 13.159707069396973,
-      "learning_rate": 1.7820988273027967e-05,
-      "loss": 3.7656,
+      "epoch": 0.14,
+      "grad_norm": 13.326085090637207,
+      "learning_rate": 1.909049073338075e-05,
+      "loss": 3.0869,
       "step": 1088
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 18.79963493347168,
-      "learning_rate": 1.7818983662423573e-05,
-      "loss": 3.2334,
+      "epoch": 0.14,
+      "grad_norm": 16.086368560791016,
+      "learning_rate": 1.9089654018324062e-05,
+      "loss": 3.2626,
       "step": 1089
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 12.51858139038086,
-      "learning_rate": 1.7816979051819187e-05,
-      "loss": 1.6161,
+      "epoch": 0.14,
+      "grad_norm": 26.71560287475586,
+      "learning_rate": 1.9088817303267373e-05,
+      "loss": 3.6289,
       "step": 1090
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 18.549959182739258,
-      "learning_rate": 1.7814974441214797e-05,
-      "loss": 2.2857,
+      "epoch": 0.14,
+      "grad_norm": 11.671117782592773,
+      "learning_rate": 1.9087980588210686e-05,
+      "loss": 2.8107,
       "step": 1091
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 17.99864959716797,
-      "learning_rate": 1.7812969830610404e-05,
-      "loss": 2.2779,
+      "epoch": 0.14,
+      "grad_norm": 17.49782371520996,
+      "learning_rate": 1.9087143873154e-05,
+      "loss": 4.3494,
       "step": 1092
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 16.206727981567383,
-      "learning_rate": 1.7810965220006017e-05,
-      "loss": 2.1039,
+      "epoch": 0.14,
+      "grad_norm": 13.312760353088379,
+      "learning_rate": 1.9086307158097314e-05,
+      "loss": 1.8361,
       "step": 1093
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 14.30031681060791,
-      "learning_rate": 1.7808960609401624e-05,
-      "loss": 3.3124,
+      "epoch": 0.14,
+      "grad_norm": 13.810235977172852,
+      "learning_rate": 1.9085470443040624e-05,
+      "loss": 3.8031,
       "step": 1094
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 48.405094146728516,
-      "learning_rate": 1.7806955998797234e-05,
-      "loss": 3.8335,
+      "epoch": 0.14,
+      "grad_norm": 41.52159881591797,
+      "learning_rate": 1.9084633727983937e-05,
+      "loss": 4.49,
       "step": 1095
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 17.39105224609375,
-      "learning_rate": 1.7804951388192847e-05,
-      "loss": 2.8733,
+      "epoch": 0.14,
+      "grad_norm": 13.56728458404541,
+      "learning_rate": 1.908379701292725e-05,
+      "loss": 2.4451,
       "step": 1096
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 23.60736656188965,
-      "learning_rate": 1.7802946777588454e-05,
-      "loss": 2.9195,
+      "epoch": 0.14,
+      "grad_norm": 16.606473922729492,
+      "learning_rate": 1.908296029787056e-05,
+      "loss": 2.3632,
       "step": 1097
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 18.69573402404785,
-      "learning_rate": 1.7800942166984064e-05,
-      "loss": 3.4923,
+      "epoch": 0.14,
+      "grad_norm": 15.629426956176758,
+      "learning_rate": 1.9082123582813875e-05,
+      "loss": 3.2804,
       "step": 1098
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 31.806943893432617,
-      "learning_rate": 1.7798937556379674e-05,
-      "loss": 2.9115,
+      "epoch": 0.14,
+      "grad_norm": 14.610381126403809,
+      "learning_rate": 1.908128686775719e-05,
+      "loss": 3.0428,
       "step": 1099
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 13.861745834350586,
-      "learning_rate": 1.7796932945775284e-05,
-      "loss": 1.3565,
+      "epoch": 0.14,
+      "grad_norm": 31.57737159729004,
+      "learning_rate": 1.90804501527005e-05,
+      "loss": 4.185,
       "step": 1100
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 23.198631286621094,
-      "learning_rate": 1.7794928335170894e-05,
-      "loss": 3.0973,
+      "epoch": 0.14,
+      "grad_norm": 11.78358268737793,
+      "learning_rate": 1.9079613437643813e-05,
+      "loss": 3.4115,
       "step": 1101
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 16.94276237487793,
-      "learning_rate": 1.7792923724566505e-05,
-      "loss": 3.5986,
+      "epoch": 0.14,
+      "grad_norm": 12.632600784301758,
+      "learning_rate": 1.9078776722587126e-05,
+      "loss": 3.2309,
       "step": 1102
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 9.507773399353027,
-      "learning_rate": 1.7790919113962115e-05,
-      "loss": 1.7677,
+      "epoch": 0.14,
+      "grad_norm": 13.208819389343262,
+      "learning_rate": 1.9077940007530436e-05,
+      "loss": 2.3418,
       "step": 1103
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 61.569091796875,
-      "learning_rate": 1.7788914503357725e-05,
-      "loss": 2.5007,
+      "epoch": 0.14,
+      "grad_norm": 6.310235977172852,
+      "learning_rate": 1.907710329247375e-05,
+      "loss": 0.2954,
       "step": 1104
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 16.81398582458496,
-      "learning_rate": 1.7786909892753335e-05,
-      "loss": 3.5078,
+      "epoch": 0.14,
+      "grad_norm": 16.388343811035156,
+      "learning_rate": 1.907626657741706e-05,
+      "loss": 2.5308,
       "step": 1105
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 15.168441772460938,
-      "learning_rate": 1.778490528214894e-05,
-      "loss": 2.3151,
+      "epoch": 0.14,
+      "grad_norm": 10.25705623626709,
+      "learning_rate": 1.9075429862360374e-05,
+      "loss": 2.5711,
       "step": 1106
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 16.73635482788086,
-      "learning_rate": 1.7782900671544555e-05,
-      "loss": 3.0094,
+      "epoch": 0.14,
+      "grad_norm": 9.988468170166016,
+      "learning_rate": 1.9074593147303688e-05,
+      "loss": 1.9492,
       "step": 1107
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 16.253849029541016,
-      "learning_rate": 1.7780896060940165e-05,
-      "loss": 2.7682,
+      "epoch": 0.14,
+      "grad_norm": 20.494890213012695,
+      "learning_rate": 1.9073756432246998e-05,
+      "loss": 2.5956,
       "step": 1108
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 14.130172729492188,
-      "learning_rate": 1.7778891450335775e-05,
-      "loss": 3.0624,
+      "epoch": 0.14,
+      "grad_norm": 15.59097671508789,
+      "learning_rate": 1.907291971719031e-05,
+      "loss": 5.2766,
       "step": 1109
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 17.684720993041992,
-      "learning_rate": 1.7776886839731385e-05,
-      "loss": 3.3596,
+      "epoch": 0.14,
+      "grad_norm": 15.705918312072754,
+      "learning_rate": 1.9072083002133625e-05,
+      "loss": 3.8288,
       "step": 1110
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 29.59230613708496,
-      "learning_rate": 1.7774882229126992e-05,
-      "loss": 3.1807,
+      "epoch": 0.14,
+      "grad_norm": 11.32144832611084,
+      "learning_rate": 1.9071246287076935e-05,
+      "loss": 1.4441,
       "step": 1111
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 11.93481159210205,
-      "learning_rate": 1.7772877618522605e-05,
-      "loss": 2.8107,
+      "epoch": 0.14,
+      "grad_norm": 13.291257858276367,
+      "learning_rate": 1.907040957202025e-05,
+      "loss": 2.6686,
       "step": 1112
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 16.962127685546875,
-      "learning_rate": 1.7770873007918212e-05,
-      "loss": 2.9108,
+      "epoch": 0.14,
+      "grad_norm": 19.998872756958008,
+      "learning_rate": 1.9069572856963563e-05,
+      "loss": 2.3493,
       "step": 1113
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 13.559270858764648,
-      "learning_rate": 1.7768868397313822e-05,
-      "loss": 3.5231,
+      "epoch": 0.14,
+      "grad_norm": 18.568021774291992,
+      "learning_rate": 1.9068736141906876e-05,
+      "loss": 2.7368,
       "step": 1114
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 12.885640144348145,
-      "learning_rate": 1.7766863786709436e-05,
-      "loss": 2.7105,
+      "epoch": 0.14,
+      "grad_norm": 20.428361892700195,
+      "learning_rate": 1.9067899426850187e-05,
+      "loss": 3.5724,
       "step": 1115
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 29.005674362182617,
-      "learning_rate": 1.7764859176105042e-05,
-      "loss": 3.413,
+      "epoch": 0.14,
+      "grad_norm": 28.480815887451172,
+      "learning_rate": 1.90670627117935e-05,
+      "loss": 4.4699,
       "step": 1116
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 13.02987289428711,
-      "learning_rate": 1.7762854565500652e-05,
-      "loss": 1.8334,
+      "epoch": 0.14,
+      "grad_norm": 11.410590171813965,
+      "learning_rate": 1.9066225996736814e-05,
+      "loss": 4.0421,
       "step": 1117
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 14.34976863861084,
-      "learning_rate": 1.7760849954896263e-05,
-      "loss": 1.5514,
+      "epoch": 0.14,
+      "grad_norm": 17.83932113647461,
+      "learning_rate": 1.9065389281680124e-05,
+      "loss": 5.0705,
       "step": 1118
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 22.86460304260254,
-      "learning_rate": 1.7758845344291873e-05,
-      "loss": 3.4641,
+      "epoch": 0.14,
+      "grad_norm": 16.42825698852539,
+      "learning_rate": 1.9064552566623438e-05,
+      "loss": 2.6597,
       "step": 1119
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 28.6113224029541,
-      "learning_rate": 1.7756840733687483e-05,
-      "loss": 3.761,
+      "epoch": 0.14,
+      "grad_norm": 18.98484230041504,
+      "learning_rate": 1.906371585156675e-05,
+      "loss": 3.6132,
       "step": 1120
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 16.072839736938477,
-      "learning_rate": 1.7754836123083093e-05,
-      "loss": 1.5352,
+      "epoch": 0.14,
+      "grad_norm": 17.578466415405273,
+      "learning_rate": 1.9062879136510065e-05,
+      "loss": 3.7327,
       "step": 1121
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 18.935298919677734,
-      "learning_rate": 1.7752831512478703e-05,
-      "loss": 2.8795,
+      "epoch": 0.14,
+      "grad_norm": 12.911381721496582,
+      "learning_rate": 1.9062042421453375e-05,
+      "loss": 2.7183,
       "step": 1122
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 22.4312686920166,
-      "learning_rate": 1.7750826901874313e-05,
-      "loss": 4.1038,
+      "epoch": 0.14,
+      "grad_norm": 14.096404075622559,
+      "learning_rate": 1.906120570639669e-05,
+      "loss": 1.5354,
       "step": 1123
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 13.36023998260498,
-      "learning_rate": 1.7748822291269923e-05,
-      "loss": 2.9175,
+      "epoch": 0.14,
+      "grad_norm": 16.299760818481445,
+      "learning_rate": 1.9060368991340003e-05,
+      "loss": 2.2153,
       "step": 1124
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 10.283577919006348,
-      "learning_rate": 1.774681768066553e-05,
-      "loss": 1.6504,
+      "epoch": 0.14,
+      "grad_norm": 30.454317092895508,
+      "learning_rate": 1.9059532276283313e-05,
+      "loss": 2.3884,
       "step": 1125
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 16.4921817779541,
-      "learning_rate": 1.7744813070061143e-05,
-      "loss": 3.4366,
+      "epoch": 0.14,
+      "grad_norm": 16.80135726928711,
+      "learning_rate": 1.9058695561226627e-05,
+      "loss": 2.7318,
       "step": 1126
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 24.69314193725586,
-      "learning_rate": 1.7742808459456753e-05,
-      "loss": 3.5184,
+      "epoch": 0.14,
+      "grad_norm": 17.622236251831055,
+      "learning_rate": 1.905785884616994e-05,
+      "loss": 1.9797,
       "step": 1127
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 15.775534629821777,
-      "learning_rate": 1.774080384885236e-05,
-      "loss": 3.0247,
+      "epoch": 0.14,
+      "grad_norm": 18.480009078979492,
+      "learning_rate": 1.905702213111325e-05,
+      "loss": 4.2706,
       "step": 1128
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 35.886539459228516,
-      "learning_rate": 1.7738799238247973e-05,
-      "loss": 3.0789,
+      "epoch": 0.14,
+      "grad_norm": 24.628833770751953,
+      "learning_rate": 1.9056185416056564e-05,
+      "loss": 2.9627,
       "step": 1129
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 13.481932640075684,
-      "learning_rate": 1.773679462764358e-05,
-      "loss": 2.6866,
+      "epoch": 0.14,
+      "grad_norm": 13.772994041442871,
+      "learning_rate": 1.9055348700999874e-05,
+      "loss": 2.6898,
       "step": 1130
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 13.571722030639648,
-      "learning_rate": 1.773479001703919e-05,
-      "loss": 3.6435,
+      "epoch": 0.14,
+      "grad_norm": 14.796730041503906,
+      "learning_rate": 1.9054511985943188e-05,
+      "loss": 4.5945,
       "step": 1131
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 12.800562858581543,
-      "learning_rate": 1.77327854064348e-05,
-      "loss": 2.6728,
+      "epoch": 0.14,
+      "grad_norm": 16.073083877563477,
+      "learning_rate": 1.9053675270886502e-05,
+      "loss": 3.6796,
       "step": 1132
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 28.864028930664062,
-      "learning_rate": 1.773078079583041e-05,
-      "loss": 1.8873,
+      "epoch": 0.14,
+      "grad_norm": 12.354791641235352,
+      "learning_rate": 1.9052838555829812e-05,
+      "loss": 2.6927,
       "step": 1133
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 23.177940368652344,
-      "learning_rate": 1.772877618522602e-05,
-      "loss": 3.7982,
+      "epoch": 0.14,
+      "grad_norm": 32.54914093017578,
+      "learning_rate": 1.9052001840773126e-05,
+      "loss": 3.332,
       "step": 1134
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 18.07777214050293,
-      "learning_rate": 1.772677157462163e-05,
-      "loss": 3.1977,
+      "epoch": 0.14,
+      "grad_norm": 11.939435005187988,
+      "learning_rate": 1.905116512571644e-05,
+      "loss": 4.0946,
       "step": 1135
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 25.78776741027832,
-      "learning_rate": 1.772476696401724e-05,
-      "loss": 2.4591,
+      "epoch": 0.14,
+      "grad_norm": 19.629676818847656,
+      "learning_rate": 1.905032841065975e-05,
+      "loss": 2.8144,
       "step": 1136
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 14.545743942260742,
-      "learning_rate": 1.772276235341285e-05,
-      "loss": 2.5043,
+      "epoch": 0.14,
+      "grad_norm": 15.602178573608398,
+      "learning_rate": 1.9049491695603063e-05,
+      "loss": 1.3764,
       "step": 1137
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 12.708026885986328,
-      "learning_rate": 1.772075774280846e-05,
-      "loss": 3.234,
+      "epoch": 0.14,
+      "grad_norm": 13.58865737915039,
+      "learning_rate": 1.9048654980546377e-05,
+      "loss": 2.1669,
       "step": 1138
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 15.227660179138184,
-      "learning_rate": 1.771875313220407e-05,
-      "loss": 3.9899,
+      "epoch": 0.14,
+      "grad_norm": 10.807489395141602,
+      "learning_rate": 1.9047818265489687e-05,
+      "loss": 3.0976,
       "step": 1139
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 23.074007034301758,
-      "learning_rate": 1.771674852159968e-05,
-      "loss": 1.94,
+      "epoch": 0.14,
+      "grad_norm": 19.740665435791016,
+      "learning_rate": 1.9046981550433e-05,
+      "loss": 2.1226,
       "step": 1140
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 20.678417205810547,
-      "learning_rate": 1.771474391099529e-05,
-      "loss": 3.0762,
+      "epoch": 0.14,
+      "grad_norm": 27.950868606567383,
+      "learning_rate": 1.9046144835376314e-05,
+      "loss": 4.2818,
       "step": 1141
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 10.947648048400879,
-      "learning_rate": 1.77127393003909e-05,
-      "loss": 3.0063,
+      "epoch": 0.14,
+      "grad_norm": 12.425821304321289,
+      "learning_rate": 1.9045308120319628e-05,
+      "loss": 3.3513,
       "step": 1142
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 12.576761245727539,
-      "learning_rate": 1.771073468978651e-05,
-      "loss": 2.5958,
+      "epoch": 0.14,
+      "grad_norm": 12.599128723144531,
+      "learning_rate": 1.904447140526294e-05,
+      "loss": 2.0619,
       "step": 1143
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 25.190330505371094,
-      "learning_rate": 1.7708730079182118e-05,
-      "loss": 2.9465,
+      "epoch": 0.14,
+      "grad_norm": 9.102954864501953,
+      "learning_rate": 1.9043634690206252e-05,
+      "loss": 2.0338,
       "step": 1144
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 14.625896453857422,
-      "learning_rate": 1.770672546857773e-05,
-      "loss": 2.337,
+      "epoch": 0.14,
+      "grad_norm": 16.797597885131836,
+      "learning_rate": 1.9042797975149566e-05,
+      "loss": 3.4905,
       "step": 1145
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 13.233895301818848,
-      "learning_rate": 1.770472085797334e-05,
-      "loss": 2.7883,
+      "epoch": 0.14,
+      "grad_norm": 12.28773307800293,
+      "learning_rate": 1.9041961260092876e-05,
+      "loss": 2.2691,
       "step": 1146
     },
     {
-      "epoch": 0.34,
-      "grad_norm": 11.731785774230957,
-      "learning_rate": 1.7702716247368948e-05,
-      "loss": 2.4522,
+      "epoch": 0.14,
+      "grad_norm": 14.749690055847168,
+      "learning_rate": 1.904112454503619e-05,
+      "loss": 4.179,
       "step": 1147
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 13.378613471984863,
-      "learning_rate": 1.770071163676456e-05,
-      "loss": 2.6688,
+      "epoch": 0.14,
+      "grad_norm": 14.806706428527832,
+      "learning_rate": 1.9040287829979503e-05,
+      "loss": 3.3297,
       "step": 1148
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 34.357383728027344,
-      "learning_rate": 1.769870702616017e-05,
-      "loss": 3.0263,
+      "epoch": 0.14,
+      "grad_norm": 26.46034049987793,
+      "learning_rate": 1.9039451114922817e-05,
+      "loss": 3.7049,
       "step": 1149
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 13.445833206176758,
-      "learning_rate": 1.769670241555578e-05,
-      "loss": 1.7473,
+      "epoch": 0.14,
+      "grad_norm": 14.55521297454834,
+      "learning_rate": 1.9038614399866127e-05,
+      "loss": 3.3959,
       "step": 1150
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 13.28929328918457,
-      "learning_rate": 1.7694697804951392e-05,
-      "loss": 2.3902,
+      "epoch": 0.14,
+      "grad_norm": 15.360191345214844,
+      "learning_rate": 1.903777768480944e-05,
+      "loss": 3.5524,
       "step": 1151
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 19.005510330200195,
-      "learning_rate": 1.7692693194347e-05,
-      "loss": 2.506,
+      "epoch": 0.14,
+      "grad_norm": 17.11458969116211,
+      "learning_rate": 1.9036940969752754e-05,
+      "loss": 3.3466,
       "step": 1152
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 18.774620056152344,
-      "learning_rate": 1.769068858374261e-05,
-      "loss": 2.8765,
+      "epoch": 0.14,
+      "grad_norm": 10.738466262817383,
+      "learning_rate": 1.9036104254696065e-05,
+      "loss": 2.8425,
       "step": 1153
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 21.18247413635254,
-      "learning_rate": 1.768868397313822e-05,
-      "loss": 2.4275,
+      "epoch": 0.14,
+      "grad_norm": 9.902814865112305,
+      "learning_rate": 1.903526753963938e-05,
+      "loss": 2.3491,
       "step": 1154
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 22.34467124938965,
-      "learning_rate": 1.768667936253383e-05,
-      "loss": 2.6174,
+      "epoch": 0.14,
+      "grad_norm": 14.706478118896484,
+      "learning_rate": 1.9034430824582692e-05,
+      "loss": 1.6602,
       "step": 1155
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 28.30898094177246,
-      "learning_rate": 1.768467475192944e-05,
-      "loss": 3.1029,
+      "epoch": 0.15,
+      "grad_norm": 6.456505298614502,
+      "learning_rate": 1.9033594109526002e-05,
+      "loss": 1.7329,
       "step": 1156
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 12.033406257629395,
-      "learning_rate": 1.768267014132505e-05,
-      "loss": 3.4031,
+      "epoch": 0.15,
+      "grad_norm": 11.623553276062012,
+      "learning_rate": 1.9032757394469316e-05,
+      "loss": 1.9712,
       "step": 1157
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 16.54393768310547,
-      "learning_rate": 1.768066553072066e-05,
-      "loss": 2.6641,
+      "epoch": 0.15,
+      "grad_norm": 13.264073371887207,
+      "learning_rate": 1.9031920679412626e-05,
+      "loss": 5.3528,
       "step": 1158
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 29.7037410736084,
-      "learning_rate": 1.767866092011627e-05,
-      "loss": 2.8023,
+      "epoch": 0.15,
+      "grad_norm": 20.779993057250977,
+      "learning_rate": 1.903108396435594e-05,
+      "loss": 3.1254,
       "step": 1159
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 17.149572372436523,
-      "learning_rate": 1.767665630951188e-05,
-      "loss": 3.3883,
+      "epoch": 0.15,
+      "grad_norm": 11.741122245788574,
+      "learning_rate": 1.9030247249299253e-05,
+      "loss": 1.6191,
       "step": 1160
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 32.76160430908203,
-      "learning_rate": 1.767465169890749e-05,
-      "loss": 3.1329,
+      "epoch": 0.15,
+      "grad_norm": 12.015575408935547,
+      "learning_rate": 1.9029410534242564e-05,
+      "loss": 3.5642,
       "step": 1161
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 13.081724166870117,
-      "learning_rate": 1.76726470883031e-05,
-      "loss": 3.1289,
+      "epoch": 0.15,
+      "grad_norm": 10.663491249084473,
+      "learning_rate": 1.9028573819185877e-05,
+      "loss": 0.9802,
       "step": 1162
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 29.665246963500977,
-      "learning_rate": 1.767064247769871e-05,
-      "loss": 3.6106,
+      "epoch": 0.15,
+      "grad_norm": 8.336151123046875,
+      "learning_rate": 1.902773710412919e-05,
+      "loss": 2.5828,
       "step": 1163
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 19.989797592163086,
-      "learning_rate": 1.766863786709432e-05,
-      "loss": 3.4249,
+      "epoch": 0.15,
+      "grad_norm": 18.085256576538086,
+      "learning_rate": 1.90269003890725e-05,
+      "loss": 3.497,
       "step": 1164
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 11.273635864257812,
-      "learning_rate": 1.766663325648993e-05,
-      "loss": 3.0172,
+      "epoch": 0.15,
+      "grad_norm": 16.504741668701172,
+      "learning_rate": 1.9026063674015815e-05,
+      "loss": 3.6238,
       "step": 1165
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 25.350624084472656,
-      "learning_rate": 1.7664628645885536e-05,
-      "loss": 2.8628,
+      "epoch": 0.15,
+      "grad_norm": 13.649950981140137,
+      "learning_rate": 1.902522695895913e-05,
+      "loss": 3.8243,
       "step": 1166
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 14.034561157226562,
-      "learning_rate": 1.766262403528115e-05,
-      "loss": 2.7608,
+      "epoch": 0.15,
+      "grad_norm": 16.09895133972168,
+      "learning_rate": 1.902439024390244e-05,
+      "loss": 2.7386,
       "step": 1167
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 20.618732452392578,
-      "learning_rate": 1.7660619424676757e-05,
-      "loss": 2.6732,
+      "epoch": 0.15,
+      "grad_norm": 12.028850555419922,
+      "learning_rate": 1.9023553528845752e-05,
+      "loss": 2.5787,
       "step": 1168
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 18.557788848876953,
-      "learning_rate": 1.7658614814072367e-05,
-      "loss": 4.2311,
+      "epoch": 0.15,
+      "grad_norm": 12.9425630569458,
+      "learning_rate": 1.9022716813789066e-05,
+      "loss": 2.921,
       "step": 1169
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 22.748462677001953,
-      "learning_rate": 1.765661020346798e-05,
-      "loss": 3.0337,
+      "epoch": 0.15,
+      "grad_norm": 12.585747718811035,
+      "learning_rate": 1.902188009873238e-05,
+      "loss": 5.5789,
       "step": 1170
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 15.054831504821777,
-      "learning_rate": 1.7654605592863587e-05,
-      "loss": 2.5534,
+      "epoch": 0.15,
+      "grad_norm": 9.341103553771973,
+      "learning_rate": 1.902104338367569e-05,
+      "loss": 1.5048,
       "step": 1171
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 12.077056884765625,
-      "learning_rate": 1.7652600982259197e-05,
-      "loss": 1.9834,
+      "epoch": 0.15,
+      "grad_norm": 9.025690078735352,
+      "learning_rate": 1.9020206668619004e-05,
+      "loss": 0.6358,
       "step": 1172
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 17.17857551574707,
-      "learning_rate": 1.7650596371654807e-05,
-      "loss": 3.1654,
+      "epoch": 0.15,
+      "grad_norm": 17.043132781982422,
+      "learning_rate": 1.9019369953562317e-05,
+      "loss": 2.6911,
       "step": 1173
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 13.104513168334961,
-      "learning_rate": 1.7648591761050417e-05,
-      "loss": 2.999,
+      "epoch": 0.15,
+      "grad_norm": 7.509103775024414,
+      "learning_rate": 1.9018533238505628e-05,
+      "loss": 1.7465,
       "step": 1174
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 14.851997375488281,
-      "learning_rate": 1.7646587150446027e-05,
-      "loss": 3.3295,
+      "epoch": 0.15,
+      "grad_norm": 9.966673851013184,
+      "learning_rate": 1.901769652344894e-05,
+      "loss": 1.9566,
       "step": 1175
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 12.090683937072754,
-      "learning_rate": 1.7644582539841637e-05,
-      "loss": 2.4796,
+      "epoch": 0.15,
+      "grad_norm": 16.340253829956055,
+      "learning_rate": 1.9016859808392255e-05,
+      "loss": 3.069,
       "step": 1176
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 13.724775314331055,
-      "learning_rate": 1.7642577929237247e-05,
-      "loss": 2.9697,
+      "epoch": 0.15,
+      "grad_norm": 26.096837997436523,
+      "learning_rate": 1.901602309333557e-05,
+      "loss": 4.3065,
       "step": 1177
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 26.366907119750977,
-      "learning_rate": 1.7640573318632857e-05,
-      "loss": 3.2335,
+      "epoch": 0.15,
+      "grad_norm": 13.248631477355957,
+      "learning_rate": 1.901518637827888e-05,
+      "loss": 3.2177,
       "step": 1178
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 10.779017448425293,
-      "learning_rate": 1.7638568708028467e-05,
-      "loss": 2.4018,
+      "epoch": 0.15,
+      "grad_norm": 14.375895500183105,
+      "learning_rate": 1.9014349663222192e-05,
+      "loss": 2.944,
       "step": 1179
     },
     {
-      "epoch": 0.35,
-      "grad_norm": 22.231727600097656,
-      "learning_rate": 1.7636564097424078e-05,
-      "loss": 3.2377,
+      "epoch": 0.15,
+      "grad_norm": 19.621803283691406,
+      "learning_rate": 1.9013512948165506e-05,
+      "loss": 2.5023,
       "step": 1180
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 18.77667236328125,
-      "learning_rate": 1.7634559486819688e-05,
-      "loss": 3.0199,
+      "epoch": 0.15,
+      "grad_norm": 15.87476634979248,
+      "learning_rate": 1.9012676233108816e-05,
+      "loss": 4.5875,
       "step": 1181
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 16.207422256469727,
-      "learning_rate": 1.7632554876215298e-05,
-      "loss": 2.7055,
+      "epoch": 0.15,
+      "grad_norm": 10.495246887207031,
+      "learning_rate": 1.901183951805213e-05,
+      "loss": 3.1892,
       "step": 1182
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 16.809715270996094,
-      "learning_rate": 1.7630550265610908e-05,
-      "loss": 2.9821,
+      "epoch": 0.15,
+      "grad_norm": 23.24948501586914,
+      "learning_rate": 1.901100280299544e-05,
+      "loss": 3.406,
       "step": 1183
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 16.3651065826416,
-      "learning_rate": 1.7628545655006518e-05,
-      "loss": 2.1699,
+      "epoch": 0.15,
+      "grad_norm": 11.072179794311523,
+      "learning_rate": 1.9010166087938754e-05,
+      "loss": 2.5812,
       "step": 1184
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 11.468558311462402,
-      "learning_rate": 1.7626541044402125e-05,
-      "loss": 2.6325,
+      "epoch": 0.15,
+      "grad_norm": 12.650174140930176,
+      "learning_rate": 1.9009329372882068e-05,
+      "loss": 3.1434,
       "step": 1185
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 16.051881790161133,
-      "learning_rate": 1.7624536433797738e-05,
-      "loss": 2.3443,
+      "epoch": 0.15,
+      "grad_norm": 15.346962928771973,
+      "learning_rate": 1.9008492657825378e-05,
+      "loss": 3.4276,
       "step": 1186
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 12.057586669921875,
-      "learning_rate": 1.7622531823193345e-05,
-      "loss": 2.7623,
+      "epoch": 0.15,
+      "grad_norm": 13.713033676147461,
+      "learning_rate": 1.900765594276869e-05,
+      "loss": 2.194,
       "step": 1187
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 17.24924659729004,
-      "learning_rate": 1.7620527212588955e-05,
-      "loss": 3.3553,
+      "epoch": 0.15,
+      "grad_norm": 33.21901321411133,
+      "learning_rate": 1.9006819227712002e-05,
+      "loss": 3.0766,
       "step": 1188
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 12.06871223449707,
-      "learning_rate": 1.761852260198457e-05,
-      "loss": 1.9598,
+      "epoch": 0.15,
+      "grad_norm": 13.535443305969238,
+      "learning_rate": 1.9005982512655315e-05,
+      "loss": 3.4499,
       "step": 1189
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 20.307418823242188,
-      "learning_rate": 1.7616517991380175e-05,
-      "loss": 3.1637,
+      "epoch": 0.15,
+      "grad_norm": 10.614287376403809,
+      "learning_rate": 1.900514579759863e-05,
+      "loss": 1.821,
       "step": 1190
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 11.176687240600586,
-      "learning_rate": 1.7614513380775785e-05,
-      "loss": 2.5982,
+      "epoch": 0.15,
+      "grad_norm": 11.230908393859863,
+      "learning_rate": 1.9004309082541943e-05,
+      "loss": 1.0304,
       "step": 1191
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 14.596364974975586,
-      "learning_rate": 1.7612508770171395e-05,
-      "loss": 2.5782,
+      "epoch": 0.15,
+      "grad_norm": 7.948458194732666,
+      "learning_rate": 1.9003472367485253e-05,
+      "loss": 2.5713,
       "step": 1192
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 16.369272232055664,
-      "learning_rate": 1.7610504159567005e-05,
-      "loss": 2.7876,
+      "epoch": 0.15,
+      "grad_norm": 19.880294799804688,
+      "learning_rate": 1.9002635652428567e-05,
+      "loss": 3.7412,
       "step": 1193
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 16.84731101989746,
-      "learning_rate": 1.7608499548962615e-05,
-      "loss": 2.9056,
+      "epoch": 0.15,
+      "grad_norm": 21.333202362060547,
+      "learning_rate": 1.900179893737188e-05,
+      "loss": 1.8671,
       "step": 1194
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 51.60824203491211,
-      "learning_rate": 1.7606494938358225e-05,
-      "loss": 2.3254,
+      "epoch": 0.15,
+      "grad_norm": 10.239437103271484,
+      "learning_rate": 1.900096222231519e-05,
+      "loss": 1.6627,
       "step": 1195
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 18.47427749633789,
-      "learning_rate": 1.7604490327753836e-05,
-      "loss": 2.9355,
+      "epoch": 0.15,
+      "grad_norm": 9.137307167053223,
+      "learning_rate": 1.9000125507258504e-05,
+      "loss": 2.6344,
       "step": 1196
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 15.123434066772461,
-      "learning_rate": 1.7602485717149446e-05,
-      "loss": 2.588,
+      "epoch": 0.15,
+      "grad_norm": 13.801033020019531,
+      "learning_rate": 1.8999288792201818e-05,
+      "loss": 3.7894,
       "step": 1197
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 19.959136962890625,
-      "learning_rate": 1.7600481106545056e-05,
-      "loss": 2.5238,
+      "epoch": 0.15,
+      "grad_norm": 22.311199188232422,
+      "learning_rate": 1.899845207714513e-05,
+      "loss": 2.3578,
       "step": 1198
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 19.62584114074707,
-      "learning_rate": 1.7598476495940662e-05,
-      "loss": 3.2818,
+      "epoch": 0.15,
+      "grad_norm": 13.077570915222168,
+      "learning_rate": 1.8997615362088442e-05,
+      "loss": 2.8371,
       "step": 1199
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 19.437580108642578,
-      "learning_rate": 1.7596471885336276e-05,
-      "loss": 2.8019,
+      "epoch": 0.15,
+      "grad_norm": 20.19798469543457,
+      "learning_rate": 1.8996778647031755e-05,
+      "loss": 3.2919,
       "step": 1200
     },
     {
-      "epoch": 0.36,
-      "eval_loss": 0.5124673843383789,
-      "eval_runtime": 43.4829,
-      "eval_samples_per_second": 34.013,
-      "eval_steps_per_second": 34.013,
+      "epoch": 0.15,
+      "eval_loss": 0.2365642935037613,
+      "eval_runtime": 93.584,
+      "eval_samples_per_second": 37.848,
+      "eval_steps_per_second": 37.848,
       "step": 1200
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 41.26815414428711,
-      "learning_rate": 1.7594467274731886e-05,
-      "loss": 3.0088,
+      "epoch": 0.15,
+      "grad_norm": 13.819417953491211,
+      "learning_rate": 1.899594193197507e-05,
+      "loss": 3.3004,
       "step": 1201
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 23.42255973815918,
-      "learning_rate": 1.7592462664127493e-05,
-      "loss": 3.0811,
+      "epoch": 0.15,
+      "grad_norm": 22.852203369140625,
+      "learning_rate": 1.899510521691838e-05,
+      "loss": 3.276,
       "step": 1202
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 24.135406494140625,
-      "learning_rate": 1.7590458053523106e-05,
-      "loss": 3.4351,
+      "epoch": 0.15,
+      "grad_norm": 16.19291877746582,
+      "learning_rate": 1.8994268501861693e-05,
+      "loss": 3.67,
       "step": 1203
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 11.820728302001953,
-      "learning_rate": 1.7588453442918713e-05,
-      "loss": 2.2131,
+      "epoch": 0.15,
+      "grad_norm": 16.76343536376953,
+      "learning_rate": 1.8993431786805007e-05,
+      "loss": 1.8594,
       "step": 1204
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 12.589012145996094,
-      "learning_rate": 1.7586448832314323e-05,
-      "loss": 2.5815,
+      "epoch": 0.15,
+      "grad_norm": 6.684631824493408,
+      "learning_rate": 1.899259507174832e-05,
+      "loss": 0.4694,
       "step": 1205
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 14.668851852416992,
-      "learning_rate": 1.7584444221709936e-05,
-      "loss": 2.9765,
+      "epoch": 0.15,
+      "grad_norm": 12.731599807739258,
+      "learning_rate": 1.899175835669163e-05,
+      "loss": 2.2129,
       "step": 1206
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 13.122230529785156,
-      "learning_rate": 1.7582439611105543e-05,
-      "loss": 2.3476,
+      "epoch": 0.15,
+      "grad_norm": 42.011207580566406,
+      "learning_rate": 1.8990921641634944e-05,
+      "loss": 4.4189,
       "step": 1207
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 17.157947540283203,
-      "learning_rate": 1.7580435000501153e-05,
-      "loss": 2.8408,
+      "epoch": 0.15,
+      "grad_norm": 67.61883544921875,
+      "learning_rate": 1.8990084926578254e-05,
+      "loss": 3.1967,
       "step": 1208
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 21.18256950378418,
-      "learning_rate": 1.7578430389896763e-05,
-      "loss": 2.5619,
+      "epoch": 0.15,
+      "grad_norm": 15.998286247253418,
+      "learning_rate": 1.8989248211521568e-05,
+      "loss": 3.8409,
       "step": 1209
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 19.1740779876709,
-      "learning_rate": 1.7576425779292373e-05,
-      "loss": 4.0597,
+      "epoch": 0.15,
+      "grad_norm": 9.8434419631958,
+      "learning_rate": 1.8988411496464882e-05,
+      "loss": 2.0819,
       "step": 1210
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 20.59099769592285,
-      "learning_rate": 1.7574421168687983e-05,
-      "loss": 1.8192,
+      "epoch": 0.15,
+      "grad_norm": 148.34994506835938,
+      "learning_rate": 1.8987574781408192e-05,
+      "loss": 2.4031,
       "step": 1211
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 31.291309356689453,
-      "learning_rate": 1.7572416558083593e-05,
-      "loss": 3.6861,
+      "epoch": 0.15,
+      "grad_norm": 44.0875358581543,
+      "learning_rate": 1.8986738066351506e-05,
+      "loss": 2.2609,
       "step": 1212
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 11.655035018920898,
-      "learning_rate": 1.7570411947479204e-05,
-      "loss": 2.3163,
+      "epoch": 0.15,
+      "grad_norm": 13.564834594726562,
+      "learning_rate": 1.898590135129482e-05,
+      "loss": 3.4062,
       "step": 1213
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 27.356048583984375,
-      "learning_rate": 1.7568407336874814e-05,
-      "loss": 3.9249,
+      "epoch": 0.15,
+      "grad_norm": 13.418842315673828,
+      "learning_rate": 1.898506463623813e-05,
+      "loss": 2.1331,
       "step": 1214
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 22.89837646484375,
-      "learning_rate": 1.7566402726270424e-05,
-      "loss": 2.6824,
+      "epoch": 0.15,
+      "grad_norm": 11.270611763000488,
+      "learning_rate": 1.8984227921181443e-05,
+      "loss": 2.806,
       "step": 1215
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 16.46586799621582,
-      "learning_rate": 1.7564398115666034e-05,
-      "loss": 2.5388,
+      "epoch": 0.15,
+      "grad_norm": 34.41801452636719,
+      "learning_rate": 1.8983391206124753e-05,
+      "loss": 4.1672,
       "step": 1216
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 18.528961181640625,
-      "learning_rate": 1.7562393505061644e-05,
-      "loss": 2.435,
+      "epoch": 0.15,
+      "grad_norm": 17.68085289001465,
+      "learning_rate": 1.8982554491068067e-05,
+      "loss": 3.34,
       "step": 1217
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 10.361041069030762,
-      "learning_rate": 1.756038889445725e-05,
-      "loss": 2.5633,
+      "epoch": 0.15,
+      "grad_norm": 18.969608306884766,
+      "learning_rate": 1.898171777601138e-05,
+      "loss": 1.0799,
       "step": 1218
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 17.265287399291992,
-      "learning_rate": 1.7558384283852864e-05,
-      "loss": 2.3523,
+      "epoch": 0.15,
+      "grad_norm": 9.720231056213379,
+      "learning_rate": 1.8980881060954694e-05,
+      "loss": 2.2871,
       "step": 1219
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 18.783292770385742,
-      "learning_rate": 1.7556379673248474e-05,
-      "loss": 2.3425,
+      "epoch": 0.15,
+      "grad_norm": 13.65944766998291,
+      "learning_rate": 1.8980044345898005e-05,
+      "loss": 1.9212,
       "step": 1220
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 21.62091827392578,
-      "learning_rate": 1.755437506264408e-05,
-      "loss": 2.7889,
+      "epoch": 0.15,
+      "grad_norm": 10.61152172088623,
+      "learning_rate": 1.8979207630841318e-05,
+      "loss": 1.9664,
       "step": 1221
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 19.435802459716797,
-      "learning_rate": 1.7552370452039694e-05,
-      "loss": 3.4363,
+      "epoch": 0.15,
+      "grad_norm": 18.863014221191406,
+      "learning_rate": 1.8978370915784632e-05,
+      "loss": 3.9371,
       "step": 1222
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 13.962774276733398,
-      "learning_rate": 1.75503658414353e-05,
-      "loss": 1.6731,
+      "epoch": 0.15,
+      "grad_norm": 16.500232696533203,
+      "learning_rate": 1.8977534200727942e-05,
+      "loss": 4.1268,
       "step": 1223
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 10.637103080749512,
-      "learning_rate": 1.754836123083091e-05,
-      "loss": 1.9421,
+      "epoch": 0.15,
+      "grad_norm": 11.271100044250488,
+      "learning_rate": 1.8976697485671256e-05,
+      "loss": 2.2966,
       "step": 1224
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 12.985700607299805,
-      "learning_rate": 1.7546356620226525e-05,
-      "loss": 2.2144,
+      "epoch": 0.15,
+      "grad_norm": 13.12723445892334,
+      "learning_rate": 1.897586077061457e-05,
+      "loss": 2.3253,
       "step": 1225
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 16.512508392333984,
-      "learning_rate": 1.754435200962213e-05,
-      "loss": 2.4662,
+      "epoch": 0.15,
+      "grad_norm": 11.523720741271973,
+      "learning_rate": 1.8975024055557883e-05,
+      "loss": 2.9859,
       "step": 1226
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 20.497188568115234,
-      "learning_rate": 1.754234739901774e-05,
-      "loss": 2.0858,
+      "epoch": 0.15,
+      "grad_norm": 12.031673431396484,
+      "learning_rate": 1.8974187340501193e-05,
+      "loss": 2.6248,
       "step": 1227
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 32.291175842285156,
-      "learning_rate": 1.754034278841335e-05,
-      "loss": 3.3415,
+      "epoch": 0.15,
+      "grad_norm": 7.525265216827393,
+      "learning_rate": 1.8973350625444507e-05,
+      "loss": 1.3119,
       "step": 1228
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 16.312284469604492,
-      "learning_rate": 1.753833817780896e-05,
-      "loss": 2.2323,
+      "epoch": 0.15,
+      "grad_norm": 14.541187286376953,
+      "learning_rate": 1.897251391038782e-05,
+      "loss": 4.317,
       "step": 1229
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 23.16538429260254,
-      "learning_rate": 1.753633356720457e-05,
-      "loss": 3.7011,
+      "epoch": 0.15,
+      "grad_norm": 8.85118293762207,
+      "learning_rate": 1.897167719533113e-05,
+      "loss": 1.5031,
       "step": 1230
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 9.652095794677734,
-      "learning_rate": 1.7534328956600182e-05,
-      "loss": 2.5907,
+      "epoch": 0.15,
+      "grad_norm": 24.35873794555664,
+      "learning_rate": 1.8970840480274445e-05,
+      "loss": 2.8286,
       "step": 1231
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 16.361186981201172,
-      "learning_rate": 1.7532324345995792e-05,
-      "loss": 3.1489,
+      "epoch": 0.15,
+      "grad_norm": 18.828296661376953,
+      "learning_rate": 1.8970003765217758e-05,
+      "loss": 3.0941,
       "step": 1232
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 24.299665451049805,
-      "learning_rate": 1.7530319735391402e-05,
-      "loss": 2.751,
+      "epoch": 0.15,
+      "grad_norm": 89.44055938720703,
+      "learning_rate": 1.8969167050161072e-05,
+      "loss": 2.4806,
       "step": 1233
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 11.663421630859375,
-      "learning_rate": 1.7528315124787012e-05,
-      "loss": 2.2756,
+      "epoch": 0.15,
+      "grad_norm": 7.748295783996582,
+      "learning_rate": 1.8968330335104382e-05,
+      "loss": 0.6644,
       "step": 1234
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 16.520639419555664,
-      "learning_rate": 1.7526310514182622e-05,
-      "loss": 1.8753,
+      "epoch": 0.15,
+      "grad_norm": 7.684282302856445,
+      "learning_rate": 1.8967493620047696e-05,
+      "loss": 0.4885,
       "step": 1235
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 30.01593780517578,
-      "learning_rate": 1.7524305903578232e-05,
-      "loss": 1.1438,
+      "epoch": 0.16,
+      "grad_norm": 16.291973114013672,
+      "learning_rate": 1.8966656904991006e-05,
+      "loss": 3.0854,
       "step": 1236
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 28.729244232177734,
-      "learning_rate": 1.7522301292973842e-05,
-      "loss": 3.4131,
+      "epoch": 0.16,
+      "grad_norm": 11.370176315307617,
+      "learning_rate": 1.896582018993432e-05,
+      "loss": 4.0331,
       "step": 1237
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 18.10856819152832,
-      "learning_rate": 1.7520296682369452e-05,
-      "loss": 2.8804,
+      "epoch": 0.16,
+      "grad_norm": 19.60243034362793,
+      "learning_rate": 1.8964983474877633e-05,
+      "loss": 2.9217,
       "step": 1238
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 18.52988052368164,
-      "learning_rate": 1.7518292071765062e-05,
-      "loss": 2.57,
+      "epoch": 0.16,
+      "grad_norm": 10.804051399230957,
+      "learning_rate": 1.8964146759820944e-05,
+      "loss": 2.2526,
       "step": 1239
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 18.071670532226562,
-      "learning_rate": 1.751628746116067e-05,
-      "loss": 2.5888,
+      "epoch": 0.16,
+      "grad_norm": 13.438443183898926,
+      "learning_rate": 1.8963310044764257e-05,
+      "loss": 3.4662,
       "step": 1240
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 9.437912940979004,
-      "learning_rate": 1.7514282850556283e-05,
-      "loss": 2.1698,
+      "epoch": 0.16,
+      "grad_norm": 11.870586395263672,
+      "learning_rate": 1.8962473329707568e-05,
+      "loss": 2.2085,
       "step": 1241
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 11.90031623840332,
-      "learning_rate": 1.751227823995189e-05,
-      "loss": 2.5627,
+      "epoch": 0.16,
+      "grad_norm": 14.314711570739746,
+      "learning_rate": 1.896163661465088e-05,
+      "loss": 3.4219,
       "step": 1242
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 12.437149047851562,
-      "learning_rate": 1.75102736293475e-05,
-      "loss": 2.2992,
+      "epoch": 0.16,
+      "grad_norm": 14.737003326416016,
+      "learning_rate": 1.8960799899594195e-05,
+      "loss": 1.5509,
       "step": 1243
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 27.860820770263672,
-      "learning_rate": 1.7508269018743113e-05,
-      "loss": 3.5747,
+      "epoch": 0.16,
+      "grad_norm": 21.78973388671875,
+      "learning_rate": 1.8959963184537505e-05,
+      "loss": 1.1379,
       "step": 1244
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 26.30794334411621,
-      "learning_rate": 1.750626440813872e-05,
-      "loss": 3.8743,
+      "epoch": 0.16,
+      "grad_norm": 19.68343734741211,
+      "learning_rate": 1.895912646948082e-05,
+      "loss": 4.0445,
       "step": 1245
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 17.703901290893555,
-      "learning_rate": 1.750425979753433e-05,
-      "loss": 2.338,
+      "epoch": 0.16,
+      "grad_norm": 14.491902351379395,
+      "learning_rate": 1.8958289754424132e-05,
+      "loss": 3.4564,
       "step": 1246
     },
     {
-      "epoch": 0.37,
-      "grad_norm": 10.545022010803223,
-      "learning_rate": 1.750225518692994e-05,
-      "loss": 1.9899,
+      "epoch": 0.16,
+      "grad_norm": 10.514927864074707,
+      "learning_rate": 1.8957453039367446e-05,
+      "loss": 2.9613,
       "step": 1247
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 24.279394149780273,
-      "learning_rate": 1.750025057632555e-05,
-      "loss": 2.9463,
+      "epoch": 0.16,
+      "grad_norm": 17.163677215576172,
+      "learning_rate": 1.8956616324310756e-05,
+      "loss": 3.3785,
       "step": 1248
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 21.58678436279297,
-      "learning_rate": 1.749824596572116e-05,
-      "loss": 3.4896,
+      "epoch": 0.16,
+      "grad_norm": 7.404734134674072,
+      "learning_rate": 1.895577960925407e-05,
+      "loss": 2.0438,
       "step": 1249
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 13.904131889343262,
-      "learning_rate": 1.749624135511677e-05,
-      "loss": 3.1997,
+      "epoch": 0.16,
+      "grad_norm": 14.131943702697754,
+      "learning_rate": 1.8954942894197384e-05,
+      "loss": 1.5331,
       "step": 1250
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 10.544210433959961,
-      "learning_rate": 1.749423674451238e-05,
-      "loss": 1.9443,
+      "epoch": 0.16,
+      "grad_norm": 15.94304084777832,
+      "learning_rate": 1.8954106179140694e-05,
+      "loss": 2.2588,
       "step": 1251
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 16.909954071044922,
-      "learning_rate": 1.749223213390799e-05,
-      "loss": 3.1345,
+      "epoch": 0.16,
+      "grad_norm": 16.104053497314453,
+      "learning_rate": 1.8953269464084008e-05,
+      "loss": 2.8138,
       "step": 1252
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 18.661376953125,
-      "learning_rate": 1.74902275233036e-05,
-      "loss": 3.7376,
+      "epoch": 0.16,
+      "grad_norm": 12.07219123840332,
+      "learning_rate": 1.895243274902732e-05,
+      "loss": 3.3369,
       "step": 1253
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 16.603628158569336,
-      "learning_rate": 1.748822291269921e-05,
-      "loss": 2.108,
+      "epoch": 0.16,
+      "grad_norm": 21.96693229675293,
+      "learning_rate": 1.895159603397063e-05,
+      "loss": 3.9521,
       "step": 1254
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 12.670201301574707,
-      "learning_rate": 1.748621830209482e-05,
-      "loss": 3.4042,
+      "epoch": 0.16,
+      "grad_norm": 16.473649978637695,
+      "learning_rate": 1.8950759318913945e-05,
+      "loss": 3.1058,
       "step": 1255
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 12.029052734375,
-      "learning_rate": 1.748421369149043e-05,
-      "loss": 2.9556,
+      "epoch": 0.16,
+      "grad_norm": 34.58412551879883,
+      "learning_rate": 1.894992260385726e-05,
+      "loss": 2.8908,
       "step": 1256
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 28.61595916748047,
-      "learning_rate": 1.748220908088604e-05,
-      "loss": 2.0854,
+      "epoch": 0.16,
+      "grad_norm": 13.569806098937988,
+      "learning_rate": 1.8949085888800572e-05,
+      "loss": 1.7347,
       "step": 1257
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 22.586305618286133,
-      "learning_rate": 1.748020447028165e-05,
-      "loss": 2.461,
+      "epoch": 0.16,
+      "grad_norm": 17.082672119140625,
+      "learning_rate": 1.8948249173743883e-05,
+      "loss": 1.3953,
       "step": 1258
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 14.331897735595703,
-      "learning_rate": 1.7478199859677257e-05,
-      "loss": 3.5532,
+      "epoch": 0.16,
+      "grad_norm": 10.1463041305542,
+      "learning_rate": 1.8947412458687196e-05,
+      "loss": 1.7843,
       "step": 1259
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 18.208744049072266,
-      "learning_rate": 1.747619524907287e-05,
-      "loss": 2.59,
+      "epoch": 0.16,
+      "grad_norm": 45.11806869506836,
+      "learning_rate": 1.894657574363051e-05,
+      "loss": 3.1561,
       "step": 1260
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 26.50733757019043,
-      "learning_rate": 1.7474190638468477e-05,
-      "loss": 3.567,
+      "epoch": 0.16,
+      "grad_norm": 9.12941837310791,
+      "learning_rate": 1.894573902857382e-05,
+      "loss": 1.6232,
       "step": 1261
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 17.948020935058594,
-      "learning_rate": 1.7472186027864088e-05,
-      "loss": 2.5004,
+      "epoch": 0.16,
+      "grad_norm": 17.47677993774414,
+      "learning_rate": 1.8944902313517134e-05,
+      "loss": 4.1126,
       "step": 1262
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 11.927952766418457,
-      "learning_rate": 1.74701814172597e-05,
-      "loss": 2.1265,
+      "epoch": 0.16,
+      "grad_norm": 12.587291717529297,
+      "learning_rate": 1.8944065598460447e-05,
+      "loss": 2.5521,
       "step": 1263
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 26.811809539794922,
-      "learning_rate": 1.7468176806655308e-05,
-      "loss": 2.7967,
+      "epoch": 0.16,
+      "grad_norm": 18.3865966796875,
+      "learning_rate": 1.8943228883403758e-05,
+      "loss": 4.6634,
       "step": 1264
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 82.8468246459961,
-      "learning_rate": 1.7466172196050918e-05,
-      "loss": 2.652,
+      "epoch": 0.16,
+      "grad_norm": 12.10347843170166,
+      "learning_rate": 1.894239216834707e-05,
+      "loss": 3.8171,
       "step": 1265
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 13.454938888549805,
-      "learning_rate": 1.7464167585446528e-05,
-      "loss": 2.282,
+      "epoch": 0.16,
+      "grad_norm": 9.453478813171387,
+      "learning_rate": 1.8941555453290385e-05,
+      "loss": 1.9866,
       "step": 1266
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 20.55266571044922,
-      "learning_rate": 1.7462162974842138e-05,
-      "loss": 2.3296,
+      "epoch": 0.16,
+      "grad_norm": 13.832612037658691,
+      "learning_rate": 1.8940718738233695e-05,
+      "loss": 3.8677,
       "step": 1267
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 10.279245376586914,
-      "learning_rate": 1.7460158364237748e-05,
-      "loss": 1.3453,
+      "epoch": 0.16,
+      "grad_norm": 10.849949836730957,
+      "learning_rate": 1.893988202317701e-05,
+      "loss": 2.711,
       "step": 1268
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 11.651695251464844,
-      "learning_rate": 1.7458153753633358e-05,
-      "loss": 2.5631,
+      "epoch": 0.16,
+      "grad_norm": 11.43234634399414,
+      "learning_rate": 1.893904530812032e-05,
+      "loss": 2.7987,
       "step": 1269
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 12.197574615478516,
-      "learning_rate": 1.7456149143028968e-05,
-      "loss": 3.0425,
+      "epoch": 0.16,
+      "grad_norm": 10.973919868469238,
+      "learning_rate": 1.8938208593063633e-05,
+      "loss": 3.1193,
       "step": 1270
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 24.782493591308594,
-      "learning_rate": 1.7454144532424578e-05,
-      "loss": 3.1392,
+      "epoch": 0.16,
+      "grad_norm": 12.14253044128418,
+      "learning_rate": 1.8937371878006947e-05,
+      "loss": 3.347,
       "step": 1271
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 10.844775199890137,
-      "learning_rate": 1.745213992182019e-05,
-      "loss": 2.9512,
+      "epoch": 0.16,
+      "grad_norm": 14.993719100952148,
+      "learning_rate": 1.8936535162950257e-05,
+      "loss": 0.8552,
       "step": 1272
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 13.511686325073242,
-      "learning_rate": 1.7450135311215795e-05,
-      "loss": 2.8739,
+      "epoch": 0.16,
+      "grad_norm": 11.494132041931152,
+      "learning_rate": 1.893569844789357e-05,
+      "loss": 2.2367,
       "step": 1273
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 21.19719886779785,
-      "learning_rate": 1.744813070061141e-05,
-      "loss": 2.8494,
+      "epoch": 0.16,
+      "grad_norm": 28.249052047729492,
+      "learning_rate": 1.8934861732836884e-05,
+      "loss": 1.8822,
       "step": 1274
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 17.959545135498047,
-      "learning_rate": 1.744612609000702e-05,
-      "loss": 2.8389,
+      "epoch": 0.16,
+      "grad_norm": 15.573881149291992,
+      "learning_rate": 1.8934025017780194e-05,
+      "loss": 4.3407,
       "step": 1275
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 17.719860076904297,
-      "learning_rate": 1.7444121479402625e-05,
-      "loss": 2.7844,
+      "epoch": 0.16,
+      "grad_norm": 9.760501861572266,
+      "learning_rate": 1.8933188302723508e-05,
+      "loss": 0.9577,
       "step": 1276
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 15.047807693481445,
-      "learning_rate": 1.744211686879824e-05,
-      "loss": 2.5462,
+      "epoch": 0.16,
+      "grad_norm": 30.468326568603516,
+      "learning_rate": 1.893235158766682e-05,
+      "loss": 2.6084,
       "step": 1277
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 18.358489990234375,
-      "learning_rate": 1.7440112258193846e-05,
-      "loss": 2.8936,
+      "epoch": 0.16,
+      "grad_norm": 18.64023780822754,
+      "learning_rate": 1.8931514872610135e-05,
+      "loss": 4.2978,
       "step": 1278
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 20.895231246948242,
-      "learning_rate": 1.743810764758946e-05,
-      "loss": 3.1691,
+      "epoch": 0.16,
+      "grad_norm": 10.93867301940918,
+      "learning_rate": 1.8930678157553446e-05,
+      "loss": 3.2249,
       "step": 1279
     },
     {
-      "epoch": 0.38,
-      "grad_norm": 10.4979248046875,
-      "learning_rate": 1.743610303698507e-05,
-      "loss": 1.5841,
+      "epoch": 0.16,
+      "grad_norm": 51.95521545410156,
+      "learning_rate": 1.892984144249676e-05,
+      "loss": 2.4306,
       "step": 1280
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 14.42455768585205,
-      "learning_rate": 1.7434098426380676e-05,
-      "loss": 2.7656,
+      "epoch": 0.16,
+      "grad_norm": 19.859397888183594,
+      "learning_rate": 1.8929004727440073e-05,
+      "loss": 3.4901,
       "step": 1281
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 11.873161315917969,
-      "learning_rate": 1.743209381577629e-05,
-      "loss": 2.9719,
+      "epoch": 0.16,
+      "grad_norm": 19.423580169677734,
+      "learning_rate": 1.8928168012383383e-05,
+      "loss": 3.4095,
       "step": 1282
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 23.232919692993164,
-      "learning_rate": 1.7430089205171896e-05,
-      "loss": 2.5282,
+      "epoch": 0.16,
+      "grad_norm": 14.165048599243164,
+      "learning_rate": 1.8927331297326697e-05,
+      "loss": 3.5648,
       "step": 1283
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 15.954151153564453,
-      "learning_rate": 1.7428084594567506e-05,
-      "loss": 3.3602,
+      "epoch": 0.16,
+      "grad_norm": 10.209765434265137,
+      "learning_rate": 1.892649458227001e-05,
+      "loss": 3.1595,
       "step": 1284
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 15.332113265991211,
-      "learning_rate": 1.7426079983963116e-05,
-      "loss": 3.1367,
+      "epoch": 0.16,
+      "grad_norm": 23.574323654174805,
+      "learning_rate": 1.8925657867213324e-05,
+      "loss": 2.8012,
       "step": 1285
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 17.88947296142578,
-      "learning_rate": 1.7424075373358726e-05,
-      "loss": 2.8707,
+      "epoch": 0.16,
+      "grad_norm": 8.216002464294434,
+      "learning_rate": 1.8924821152156634e-05,
+      "loss": 1.5007,
       "step": 1286
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 14.455220222473145,
-      "learning_rate": 1.7422070762754336e-05,
-      "loss": 2.7743,
+      "epoch": 0.16,
+      "grad_norm": 20.161785125732422,
+      "learning_rate": 1.8923984437099948e-05,
+      "loss": 3.0664,
       "step": 1287
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 12.738212585449219,
-      "learning_rate": 1.7420066152149946e-05,
-      "loss": 2.4288,
+      "epoch": 0.16,
+      "grad_norm": 11.79356575012207,
+      "learning_rate": 1.892314772204326e-05,
+      "loss": 2.6138,
       "step": 1288
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 17.96286392211914,
-      "learning_rate": 1.7418061541545556e-05,
-      "loss": 2.1277,
+      "epoch": 0.16,
+      "grad_norm": 13.224976539611816,
+      "learning_rate": 1.8922311006986572e-05,
+      "loss": 4.1931,
       "step": 1289
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 10.881392478942871,
-      "learning_rate": 1.7416056930941167e-05,
-      "loss": 3.3985,
+      "epoch": 0.16,
+      "grad_norm": 11.43084716796875,
+      "learning_rate": 1.8921474291929886e-05,
+      "loss": 3.2066,
       "step": 1290
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 9.816652297973633,
-      "learning_rate": 1.7414052320336777e-05,
-      "loss": 1.9979,
+      "epoch": 0.16,
+      "grad_norm": 14.025525093078613,
+      "learning_rate": 1.89206375768732e-05,
+      "loss": 3.5726,
       "step": 1291
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 14.846339225769043,
-      "learning_rate": 1.7412047709732387e-05,
-      "loss": 3.2407,
+      "epoch": 0.16,
+      "grad_norm": 10.88570785522461,
+      "learning_rate": 1.891980086181651e-05,
+      "loss": 2.0154,
       "step": 1292
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 24.287704467773438,
-      "learning_rate": 1.7410043099127997e-05,
-      "loss": 3.1519,
+      "epoch": 0.16,
+      "grad_norm": 14.027239799499512,
+      "learning_rate": 1.8918964146759823e-05,
+      "loss": 2.4733,
       "step": 1293
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 18.814102172851562,
-      "learning_rate": 1.7408038488523607e-05,
-      "loss": 3.9634,
+      "epoch": 0.16,
+      "grad_norm": 9.714178085327148,
+      "learning_rate": 1.8918127431703133e-05,
+      "loss": 1.456,
       "step": 1294
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 21.42952537536621,
-      "learning_rate": 1.7406033877919214e-05,
-      "loss": 3.3032,
+      "epoch": 0.16,
+      "grad_norm": 19.824508666992188,
+      "learning_rate": 1.8917290716646447e-05,
+      "loss": 4.0455,
       "step": 1295
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 14.19246768951416,
-      "learning_rate": 1.7404029267314827e-05,
-      "loss": 2.3004,
+      "epoch": 0.16,
+      "grad_norm": 1605.49951171875,
+      "learning_rate": 1.891645400158976e-05,
+      "loss": 1.6104,
       "step": 1296
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 32.35289001464844,
-      "learning_rate": 1.7402024656710434e-05,
-      "loss": 4.9394,
+      "epoch": 0.16,
+      "grad_norm": 12.466790199279785,
+      "learning_rate": 1.891561728653307e-05,
+      "loss": 2.2651,
       "step": 1297
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 14.341626167297363,
-      "learning_rate": 1.7400020046106044e-05,
-      "loss": 2.9232,
+      "epoch": 0.16,
+      "grad_norm": 17.17475128173828,
+      "learning_rate": 1.8914780571476385e-05,
+      "loss": 3.2729,
       "step": 1298
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 13.265176773071289,
-      "learning_rate": 1.7398015435501657e-05,
-      "loss": 3.7512,
+      "epoch": 0.16,
+      "grad_norm": 16.722414016723633,
+      "learning_rate": 1.8913943856419698e-05,
+      "loss": 1.8439,
       "step": 1299
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 15.265084266662598,
-      "learning_rate": 1.7396010824897264e-05,
-      "loss": 3.0586,
+      "epoch": 0.16,
+      "grad_norm": 7.308617115020752,
+      "learning_rate": 1.891310714136301e-05,
+      "loss": 1.2779,
       "step": 1300
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 19.14535903930664,
-      "learning_rate": 1.7394006214292874e-05,
-      "loss": 3.9271,
+      "epoch": 0.16,
+      "grad_norm": 12.488905906677246,
+      "learning_rate": 1.8912270426306322e-05,
+      "loss": 3.6065,
       "step": 1301
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 19.937034606933594,
-      "learning_rate": 1.7392001603688484e-05,
-      "loss": 1.9253,
+      "epoch": 0.16,
+      "grad_norm": 11.767340660095215,
+      "learning_rate": 1.8911433711249636e-05,
+      "loss": 2.3138,
       "step": 1302
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 15.879293441772461,
-      "learning_rate": 1.7389996993084094e-05,
-      "loss": 2.5909,
+      "epoch": 0.16,
+      "grad_norm": 12.223610877990723,
+      "learning_rate": 1.8910596996192946e-05,
+      "loss": 4.5897,
       "step": 1303
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 22.646526336669922,
-      "learning_rate": 1.7387992382479704e-05,
-      "loss": 2.1147,
+      "epoch": 0.16,
+      "grad_norm": 8.791810989379883,
+      "learning_rate": 1.890976028113626e-05,
+      "loss": 0.9778,
       "step": 1304
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 19.122434616088867,
-      "learning_rate": 1.7385987771875314e-05,
-      "loss": 2.6576,
+      "epoch": 0.16,
+      "grad_norm": 11.121882438659668,
+      "learning_rate": 1.8908923566079573e-05,
+      "loss": 2.1943,
       "step": 1305
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 16.39354705810547,
-      "learning_rate": 1.7383983161270924e-05,
-      "loss": 2.3048,
+      "epoch": 0.16,
+      "grad_norm": 15.698457717895508,
+      "learning_rate": 1.8908086851022887e-05,
+      "loss": 2.093,
       "step": 1306
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 21.781461715698242,
-      "learning_rate": 1.7381978550666535e-05,
-      "loss": 2.5542,
+      "epoch": 0.16,
+      "grad_norm": 14.07865047454834,
+      "learning_rate": 1.8907250135966197e-05,
+      "loss": 3.0307,
       "step": 1307
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 13.316234588623047,
-      "learning_rate": 1.7379973940062145e-05,
-      "loss": 2.669,
+      "epoch": 0.16,
+      "grad_norm": 14.06962776184082,
+      "learning_rate": 1.890641342090951e-05,
+      "loss": 3.8779,
       "step": 1308
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 11.472354888916016,
-      "learning_rate": 1.7377969329457755e-05,
-      "loss": 2.7657,
+      "epoch": 0.16,
+      "grad_norm": 9.522274017333984,
+      "learning_rate": 1.8905576705852825e-05,
+      "loss": 2.6916,
       "step": 1309
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 25.965810775756836,
-      "learning_rate": 1.7375964718853365e-05,
-      "loss": 2.8167,
+      "epoch": 0.16,
+      "grad_norm": 9.4116849899292,
+      "learning_rate": 1.8904739990796135e-05,
+      "loss": 1.5084,
       "step": 1310
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 17.545780181884766,
-      "learning_rate": 1.7373960108248975e-05,
-      "loss": 3.5917,
+      "epoch": 0.16,
+      "grad_norm": 16.273611068725586,
+      "learning_rate": 1.890390327573945e-05,
+      "loss": 4.6093,
       "step": 1311
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 22.960935592651367,
-      "learning_rate": 1.7371955497644585e-05,
-      "loss": 3.4562,
+      "epoch": 0.16,
+      "grad_norm": 16.295133590698242,
+      "learning_rate": 1.8903066560682762e-05,
+      "loss": 2.3653,
       "step": 1312
     },
     {
-      "epoch": 0.39,
-      "grad_norm": 11.906625747680664,
-      "learning_rate": 1.7369950887040195e-05,
-      "loss": 3.293,
+      "epoch": 0.16,
+      "grad_norm": 24.667890548706055,
+      "learning_rate": 1.8902229845626076e-05,
+      "loss": 3.506,
       "step": 1313
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 14.520302772521973,
-      "learning_rate": 1.7367946276435802e-05,
-      "loss": 2.368,
+      "epoch": 0.16,
+      "grad_norm": 17.03211212158203,
+      "learning_rate": 1.8901393130569386e-05,
+      "loss": 2.8704,
       "step": 1314
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 9.339329719543457,
-      "learning_rate": 1.7365941665831415e-05,
-      "loss": 2.1992,
+      "epoch": 0.17,
+      "grad_norm": 13.481083869934082,
+      "learning_rate": 1.89005564155127e-05,
+      "loss": 1.1945,
       "step": 1315
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 19.66999626159668,
-      "learning_rate": 1.7363937055227022e-05,
-      "loss": 3.1317,
+      "epoch": 0.17,
+      "grad_norm": 13.902771949768066,
+      "learning_rate": 1.8899719700456013e-05,
+      "loss": 4.1782,
       "step": 1316
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 19.995849609375,
-      "learning_rate": 1.7361932444622632e-05,
-      "loss": 2.606,
+      "epoch": 0.17,
+      "grad_norm": 15.049631118774414,
+      "learning_rate": 1.8898882985399324e-05,
+      "loss": 2.4135,
       "step": 1317
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 16.427093505859375,
-      "learning_rate": 1.7359927834018245e-05,
-      "loss": 1.7637,
+      "epoch": 0.17,
+      "grad_norm": 13.669821739196777,
+      "learning_rate": 1.8898046270342637e-05,
+      "loss": 3.7981,
       "step": 1318
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 14.198533058166504,
-      "learning_rate": 1.7357923223413852e-05,
-      "loss": 2.904,
+      "epoch": 0.17,
+      "grad_norm": 9.362961769104004,
+      "learning_rate": 1.8897209555285947e-05,
+      "loss": 1.0743,
       "step": 1319
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 12.009779930114746,
-      "learning_rate": 1.7355918612809462e-05,
-      "loss": 2.2005,
-      "step": 1320
-    },
-    {
-      "epoch": 0.4,
-      "eval_loss": 0.39250195026397705,
-      "eval_runtime": 43.6691,
-      "eval_samples_per_second": 33.868,
-      "eval_steps_per_second": 33.868,
+      "epoch": 0.17,
+      "grad_norm": 14.232973098754883,
+      "learning_rate": 1.889637284022926e-05,
+      "loss": 2.9732,
       "step": 1320
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 11.49736499786377,
-      "learning_rate": 1.7353914002205072e-05,
-      "loss": 2.9098,
+      "epoch": 0.17,
+      "grad_norm": 14.265029907226562,
+      "learning_rate": 1.8895536125172575e-05,
+      "loss": 2.7579,
       "step": 1321
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 12.25181770324707,
-      "learning_rate": 1.7351909391600682e-05,
-      "loss": 2.83,
+      "epoch": 0.17,
+      "grad_norm": 12.623398780822754,
+      "learning_rate": 1.8894699410115885e-05,
+      "loss": 1.7141,
       "step": 1322
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 17.131059646606445,
-      "learning_rate": 1.7349904780996293e-05,
-      "loss": 2.6729,
+      "epoch": 0.17,
+      "grad_norm": 22.552587509155273,
+      "learning_rate": 1.88938626950592e-05,
+      "loss": 4.1298,
       "step": 1323
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 16.65376853942871,
-      "learning_rate": 1.7347900170391903e-05,
-      "loss": 2.9689,
+      "epoch": 0.17,
+      "grad_norm": 13.751575469970703,
+      "learning_rate": 1.8893025980002512e-05,
+      "loss": 2.101,
       "step": 1324
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 13.16373062133789,
-      "learning_rate": 1.7345895559787513e-05,
-      "loss": 2.5196,
+      "epoch": 0.17,
+      "grad_norm": 10.549819946289062,
+      "learning_rate": 1.8892189264945823e-05,
+      "loss": 2.7987,
       "step": 1325
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 15.137256622314453,
-      "learning_rate": 1.7343890949183123e-05,
-      "loss": 3.2273,
+      "epoch": 0.17,
+      "grad_norm": 10.506394386291504,
+      "learning_rate": 1.8891352549889136e-05,
+      "loss": 4.0793,
       "step": 1326
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 12.256842613220215,
-      "learning_rate": 1.7341886338578733e-05,
-      "loss": 3.0272,
+      "epoch": 0.17,
+      "grad_norm": 12.712037086486816,
+      "learning_rate": 1.889051583483245e-05,
+      "loss": 3.0174,
       "step": 1327
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 46.50150680541992,
-      "learning_rate": 1.7339881727974343e-05,
-      "loss": 3.4704,
+      "epoch": 0.17,
+      "grad_norm": 13.093432426452637,
+      "learning_rate": 1.888967911977576e-05,
+      "loss": 1.6325,
       "step": 1328
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 23.8785400390625,
-      "learning_rate": 1.7337877117369953e-05,
-      "loss": 3.1262,
+      "epoch": 0.17,
+      "grad_norm": 10.350591659545898,
+      "learning_rate": 1.8888842404719074e-05,
+      "loss": 0.7694,
       "step": 1329
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 17.71267318725586,
-      "learning_rate": 1.7335872506765563e-05,
-      "loss": 2.1541,
+      "epoch": 0.17,
+      "grad_norm": 10.909330368041992,
+      "learning_rate": 1.8888005689662387e-05,
+      "loss": 1.5713,
       "step": 1330
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 22.225250244140625,
-      "learning_rate": 1.7333867896161173e-05,
-      "loss": 1.5919,
+      "epoch": 0.17,
+      "grad_norm": 14.973225593566895,
+      "learning_rate": 1.8887168974605698e-05,
+      "loss": 3.2756,
       "step": 1331
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 16.465585708618164,
-      "learning_rate": 1.7331863285556783e-05,
-      "loss": 2.3168,
+      "epoch": 0.17,
+      "grad_norm": 9.871883392333984,
+      "learning_rate": 1.888633225954901e-05,
+      "loss": 1.5585,
       "step": 1332
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 24.607051849365234,
-      "learning_rate": 1.732985867495239e-05,
-      "loss": 3.1243,
+      "epoch": 0.17,
+      "grad_norm": 15.263958930969238,
+      "learning_rate": 1.8885495544492325e-05,
+      "loss": 1.3708,
       "step": 1333
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 19.95473861694336,
-      "learning_rate": 1.7327854064348003e-05,
-      "loss": 2.5079,
+      "epoch": 0.17,
+      "grad_norm": 15.60806655883789,
+      "learning_rate": 1.888465882943564e-05,
+      "loss": 2.9389,
       "step": 1334
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 19.784517288208008,
-      "learning_rate": 1.7325849453743614e-05,
-      "loss": 3.3156,
+      "epoch": 0.17,
+      "grad_norm": 13.39920425415039,
+      "learning_rate": 1.888382211437895e-05,
+      "loss": 3.3344,
       "step": 1335
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 13.695775032043457,
-      "learning_rate": 1.732384484313922e-05,
-      "loss": 3.1165,
+      "epoch": 0.17,
+      "grad_norm": 9.534605026245117,
+      "learning_rate": 1.8882985399322263e-05,
+      "loss": 2.4543,
       "step": 1336
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 11.951885223388672,
-      "learning_rate": 1.7321840232534834e-05,
-      "loss": 2.9321,
+      "epoch": 0.17,
+      "grad_norm": 11.92212200164795,
+      "learning_rate": 1.8882148684265576e-05,
+      "loss": 2.1972,
       "step": 1337
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 13.507436752319336,
-      "learning_rate": 1.731983562193044e-05,
-      "loss": 2.4538,
+      "epoch": 0.17,
+      "grad_norm": 12.673669815063477,
+      "learning_rate": 1.8881311969208886e-05,
+      "loss": 2.1967,
       "step": 1338
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 11.957839965820312,
-      "learning_rate": 1.731783101132605e-05,
-      "loss": 3.1245,
+      "epoch": 0.17,
+      "grad_norm": 15.573685646057129,
+      "learning_rate": 1.88804752541522e-05,
+      "loss": 3.0234,
       "step": 1339
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 14.829548835754395,
-      "learning_rate": 1.731582640072166e-05,
-      "loss": 2.3632,
+      "epoch": 0.17,
+      "grad_norm": 23.35232162475586,
+      "learning_rate": 1.8879638539095514e-05,
+      "loss": 2.9469,
       "step": 1340
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 14.587291717529297,
-      "learning_rate": 1.731382179011727e-05,
-      "loss": 2.8577,
+      "epoch": 0.17,
+      "grad_norm": 20.39870834350586,
+      "learning_rate": 1.8878801824038827e-05,
+      "loss": 4.6102,
       "step": 1341
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 13.556631088256836,
-      "learning_rate": 1.731181717951288e-05,
-      "loss": 2.0475,
+      "epoch": 0.17,
+      "grad_norm": 18.56085205078125,
+      "learning_rate": 1.8877965108982138e-05,
+      "loss": 3.432,
       "step": 1342
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 13.010260581970215,
-      "learning_rate": 1.730981256890849e-05,
-      "loss": 2.6603,
+      "epoch": 0.17,
+      "grad_norm": 15.179999351501465,
+      "learning_rate": 1.887712839392545e-05,
+      "loss": 4.0547,
       "step": 1343
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 11.09408187866211,
-      "learning_rate": 1.73078079583041e-05,
-      "loss": 2.5755,
+      "epoch": 0.17,
+      "grad_norm": 11.116223335266113,
+      "learning_rate": 1.8876291678868765e-05,
+      "loss": 1.5694,
       "step": 1344
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 13.849974632263184,
-      "learning_rate": 1.730580334769971e-05,
-      "loss": 2.8472,
+      "epoch": 0.17,
+      "grad_norm": 12.547582626342773,
+      "learning_rate": 1.8875454963812075e-05,
+      "loss": 4.2236,
       "step": 1345
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 21.211544036865234,
-      "learning_rate": 1.730379873709532e-05,
-      "loss": 2.7465,
+      "epoch": 0.17,
+      "grad_norm": 10.222655296325684,
+      "learning_rate": 1.887461824875539e-05,
+      "loss": 4.0636,
       "step": 1346
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 15.509568214416504,
-      "learning_rate": 1.730179412649093e-05,
-      "loss": 3.4552,
+      "epoch": 0.17,
+      "grad_norm": 8.905462265014648,
+      "learning_rate": 1.88737815336987e-05,
+      "loss": 2.6271,
       "step": 1347
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 13.904120445251465,
-      "learning_rate": 1.729978951588654e-05,
-      "loss": 2.9963,
+      "epoch": 0.17,
+      "grad_norm": 12.223560333251953,
+      "learning_rate": 1.8872944818642013e-05,
+      "loss": 2.2671,
       "step": 1348
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 17.5559024810791,
-      "learning_rate": 1.729778490528215e-05,
-      "loss": 2.4842,
+      "epoch": 0.17,
+      "grad_norm": 16.041154861450195,
+      "learning_rate": 1.8872108103585326e-05,
+      "loss": 3.0192,
       "step": 1349
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 15.476587295532227,
-      "learning_rate": 1.729578029467776e-05,
-      "loss": 2.5928,
+      "epoch": 0.17,
+      "grad_norm": 24.999361038208008,
+      "learning_rate": 1.8871271388528637e-05,
+      "loss": 2.1426,
       "step": 1350
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 15.32288932800293,
-      "learning_rate": 1.729377568407337e-05,
-      "loss": 3.4517,
+      "epoch": 0.17,
+      "grad_norm": 12.167609214782715,
+      "learning_rate": 1.887043467347195e-05,
+      "loss": 2.9919,
       "step": 1351
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 15.49992561340332,
-      "learning_rate": 1.7291771073468978e-05,
-      "loss": 3.042,
+      "epoch": 0.17,
+      "grad_norm": 11.232583045959473,
+      "learning_rate": 1.886959795841526e-05,
+      "loss": 3.7923,
       "step": 1352
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 13.955615043640137,
-      "learning_rate": 1.728976646286459e-05,
-      "loss": 2.5861,
+      "epoch": 0.17,
+      "grad_norm": 11.472769737243652,
+      "learning_rate": 1.8868761243358574e-05,
+      "loss": 2.5913,
       "step": 1353
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 12.550683975219727,
-      "learning_rate": 1.7287761852260202e-05,
-      "loss": 3.6951,
+      "epoch": 0.17,
+      "grad_norm": 24.40163230895996,
+      "learning_rate": 1.8867924528301888e-05,
+      "loss": 3.6598,
       "step": 1354
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 22.899044036865234,
-      "learning_rate": 1.728575724165581e-05,
-      "loss": 2.9966,
+      "epoch": 0.17,
+      "grad_norm": 12.834221839904785,
+      "learning_rate": 1.88670878132452e-05,
+      "loss": 2.6269,
       "step": 1355
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 12.468682289123535,
-      "learning_rate": 1.7283752631051422e-05,
-      "loss": 2.1763,
+      "epoch": 0.17,
+      "grad_norm": 14.640671730041504,
+      "learning_rate": 1.8866251098188512e-05,
+      "loss": 3.9581,
       "step": 1356
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 22.829370498657227,
-      "learning_rate": 1.728174802044703e-05,
-      "loss": 3.2563,
+      "epoch": 0.17,
+      "grad_norm": 14.906844139099121,
+      "learning_rate": 1.8865414383131825e-05,
+      "loss": 3.0435,
       "step": 1357
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 27.643447875976562,
-      "learning_rate": 1.727974340984264e-05,
-      "loss": 2.7474,
+      "epoch": 0.17,
+      "grad_norm": 11.10807991027832,
+      "learning_rate": 1.886457766807514e-05,
+      "loss": 3.1026,
       "step": 1358
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 17.19843101501465,
-      "learning_rate": 1.727773879923825e-05,
-      "loss": 2.9442,
+      "epoch": 0.17,
+      "grad_norm": 16.644826889038086,
+      "learning_rate": 1.886374095301845e-05,
+      "loss": 1.8461,
       "step": 1359
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 13.734451293945312,
-      "learning_rate": 1.727573418863386e-05,
-      "loss": 2.8518,
+      "epoch": 0.17,
+      "grad_norm": 11.831157684326172,
+      "learning_rate": 1.8862904237961763e-05,
+      "loss": 1.422,
       "step": 1360
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 11.024998664855957,
-      "learning_rate": 1.727372957802947e-05,
-      "loss": 2.2833,
+      "epoch": 0.17,
+      "grad_norm": 19.2175350189209,
+      "learning_rate": 1.8862067522905077e-05,
+      "loss": 2.5451,
       "step": 1361
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 19.57155990600586,
-      "learning_rate": 1.727172496742508e-05,
-      "loss": 2.9668,
+      "epoch": 0.17,
+      "grad_norm": 15.262014389038086,
+      "learning_rate": 1.886123080784839e-05,
+      "loss": 2.6348,
       "step": 1362
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 24.14830207824707,
-      "learning_rate": 1.726972035682069e-05,
-      "loss": 2.2259,
+      "epoch": 0.17,
+      "grad_norm": 10.91917896270752,
+      "learning_rate": 1.88603940927917e-05,
+      "loss": 2.6969,
       "step": 1363
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 13.887985229492188,
-      "learning_rate": 1.72677157462163e-05,
-      "loss": 2.4764,
+      "epoch": 0.17,
+      "grad_norm": 11.875226020812988,
+      "learning_rate": 1.8859557377735014e-05,
+      "loss": 2.2805,
       "step": 1364
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 11.10314655303955,
-      "learning_rate": 1.726571113561191e-05,
-      "loss": 2.1367,
+      "epoch": 0.17,
+      "grad_norm": 21.78172492980957,
+      "learning_rate": 1.8858720662678328e-05,
+      "loss": 2.5054,
       "step": 1365
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 11.575247764587402,
-      "learning_rate": 1.726370652500752e-05,
-      "loss": 2.396,
+      "epoch": 0.17,
+      "grad_norm": 14.415453910827637,
+      "learning_rate": 1.8857883947621638e-05,
+      "loss": 2.788,
       "step": 1366
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 17.140932083129883,
-      "learning_rate": 1.726170191440313e-05,
-      "loss": 3.0233,
+      "epoch": 0.17,
+      "grad_norm": 9.646689414978027,
+      "learning_rate": 1.8857047232564952e-05,
+      "loss": 1.9306,
       "step": 1367
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 18.503704071044922,
-      "learning_rate": 1.725969730379874e-05,
-      "loss": 3.2255,
+      "epoch": 0.17,
+      "grad_norm": 11.422354698181152,
+      "learning_rate": 1.8856210517508265e-05,
+      "loss": 1.9116,
       "step": 1368
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 14.695094108581543,
-      "learning_rate": 1.7257692693194346e-05,
-      "loss": 3.1332,
+      "epoch": 0.17,
+      "grad_norm": 19.322690963745117,
+      "learning_rate": 1.885537380245158e-05,
+      "loss": 4.9606,
       "step": 1369
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 12.079935073852539,
-      "learning_rate": 1.725568808258996e-05,
-      "loss": 2.4825,
+      "epoch": 0.17,
+      "grad_norm": 11.154664993286133,
+      "learning_rate": 1.885453708739489e-05,
+      "loss": 3.4789,
       "step": 1370
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 23.080753326416016,
-      "learning_rate": 1.7253683471985566e-05,
-      "loss": 2.7521,
+      "epoch": 0.17,
+      "grad_norm": 16.747013092041016,
+      "learning_rate": 1.8853700372338203e-05,
+      "loss": 3.9088,
       "step": 1371
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 13.837876319885254,
-      "learning_rate": 1.7251678861381176e-05,
-      "loss": 2.5238,
+      "epoch": 0.17,
+      "grad_norm": 13.053373336791992,
+      "learning_rate": 1.8852863657281513e-05,
+      "loss": 3.8792,
       "step": 1372
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 19.977251052856445,
-      "learning_rate": 1.724967425077679e-05,
-      "loss": 3.2955,
+      "epoch": 0.17,
+      "grad_norm": 10.83530044555664,
+      "learning_rate": 1.8852026942224827e-05,
+      "loss": 3.0796,
       "step": 1373
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 13.569607734680176,
-      "learning_rate": 1.7247669640172397e-05,
-      "loss": 2.7767,
+      "epoch": 0.17,
+      "grad_norm": 15.559985160827637,
+      "learning_rate": 1.885119022716814e-05,
+      "loss": 4.2939,
       "step": 1374
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 14.404850959777832,
-      "learning_rate": 1.7245665029568007e-05,
-      "loss": 2.6134,
+      "epoch": 0.17,
+      "grad_norm": 11.467248916625977,
+      "learning_rate": 1.885035351211145e-05,
+      "loss": 1.7113,
       "step": 1375
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 16.575212478637695,
-      "learning_rate": 1.7243660418963617e-05,
-      "loss": 1.7531,
+      "epoch": 0.17,
+      "grad_norm": 9.12670612335205,
+      "learning_rate": 1.8849516797054764e-05,
+      "loss": 2.0589,
       "step": 1376
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 21.60881233215332,
-      "learning_rate": 1.7241655808359227e-05,
-      "loss": 2.9619,
+      "epoch": 0.17,
+      "grad_norm": 9.2252197265625,
+      "learning_rate": 1.8848680081998078e-05,
+      "loss": 1.4876,
       "step": 1377
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 17.051956176757812,
-      "learning_rate": 1.7239651197754837e-05,
-      "loss": 3.0801,
+      "epoch": 0.17,
+      "grad_norm": 12.349713325500488,
+      "learning_rate": 1.884784336694139e-05,
+      "loss": 1.7325,
       "step": 1378
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 37.58320999145508,
-      "learning_rate": 1.7237646587150447e-05,
-      "loss": 2.3241,
+      "epoch": 0.17,
+      "grad_norm": 16.01396369934082,
+      "learning_rate": 1.8847006651884702e-05,
+      "loss": 2.7937,
       "step": 1379
     },
     {
-      "epoch": 0.41,
-      "grad_norm": 12.96102237701416,
-      "learning_rate": 1.7235641976546057e-05,
-      "loss": 2.2837,
+      "epoch": 0.17,
+      "grad_norm": 13.64192008972168,
+      "learning_rate": 1.8846169936828012e-05,
+      "loss": 2.941,
       "step": 1380
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 34.093780517578125,
-      "learning_rate": 1.7233637365941667e-05,
-      "loss": 2.8085,
+      "epoch": 0.17,
+      "grad_norm": 16.084604263305664,
+      "learning_rate": 1.8845333221771326e-05,
+      "loss": 2.1773,
       "step": 1381
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 17.705703735351562,
-      "learning_rate": 1.7231632755337277e-05,
-      "loss": 3.4047,
+      "epoch": 0.17,
+      "grad_norm": 20.62796401977539,
+      "learning_rate": 1.884449650671464e-05,
+      "loss": 1.8308,
       "step": 1382
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 16.184062957763672,
-      "learning_rate": 1.7229628144732887e-05,
-      "loss": 3.4847,
+      "epoch": 0.17,
+      "grad_norm": 18.79698371887207,
+      "learning_rate": 1.8843659791657953e-05,
+      "loss": 2.6251,
       "step": 1383
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 12.74220085144043,
-      "learning_rate": 1.7227623534128497e-05,
-      "loss": 1.8471,
+      "epoch": 0.17,
+      "grad_norm": 27.760475158691406,
+      "learning_rate": 1.8842823076601263e-05,
+      "loss": 5.3229,
       "step": 1384
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 26.621505737304688,
-      "learning_rate": 1.7225618923524108e-05,
-      "loss": 3.4508,
+      "epoch": 0.17,
+      "grad_norm": 11.442704200744629,
+      "learning_rate": 1.8841986361544577e-05,
+      "loss": 2.6164,
       "step": 1385
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 23.22762107849121,
-      "learning_rate": 1.7223614312919718e-05,
-      "loss": 2.8567,
+      "epoch": 0.17,
+      "grad_norm": 21.72187042236328,
+      "learning_rate": 1.884114964648789e-05,
+      "loss": 2.2087,
       "step": 1386
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 13.235843658447266,
-      "learning_rate": 1.7221609702315328e-05,
-      "loss": 2.368,
+      "epoch": 0.17,
+      "grad_norm": 18.298065185546875,
+      "learning_rate": 1.88403129314312e-05,
+      "loss": 2.0928,
       "step": 1387
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 10.793066024780273,
-      "learning_rate": 1.7219605091710934e-05,
-      "loss": 1.9442,
+      "epoch": 0.17,
+      "grad_norm": 14.109087944030762,
+      "learning_rate": 1.8839476216374515e-05,
+      "loss": 2.6915,
       "step": 1388
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 18.45147705078125,
-      "learning_rate": 1.7217600481106548e-05,
-      "loss": 3.3398,
+      "epoch": 0.17,
+      "grad_norm": 25.29339027404785,
+      "learning_rate": 1.883863950131783e-05,
+      "loss": 3.4751,
       "step": 1389
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 14.921577453613281,
-      "learning_rate": 1.7215595870502155e-05,
-      "loss": 2.1594,
+      "epoch": 0.17,
+      "grad_norm": 14.645580291748047,
+      "learning_rate": 1.8837802786261142e-05,
+      "loss": 2.9767,
       "step": 1390
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 14.961502075195312,
-      "learning_rate": 1.7213591259897765e-05,
-      "loss": 3.3974,
+      "epoch": 0.17,
+      "grad_norm": 9.96379566192627,
+      "learning_rate": 1.8836966071204452e-05,
+      "loss": 4.0804,
       "step": 1391
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 19.07111167907715,
-      "learning_rate": 1.7211586649293378e-05,
-      "loss": 3.1362,
+      "epoch": 0.17,
+      "grad_norm": 16.089723587036133,
+      "learning_rate": 1.8836129356147766e-05,
+      "loss": 2.7156,
       "step": 1392
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 10.949179649353027,
-      "learning_rate": 1.7209582038688985e-05,
-      "loss": 2.0993,
+      "epoch": 0.17,
+      "grad_norm": 42.31618118286133,
+      "learning_rate": 1.883529264109108e-05,
+      "loss": 2.8604,
       "step": 1393
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 12.562166213989258,
-      "learning_rate": 1.7207577428084595e-05,
-      "loss": 2.0554,
+      "epoch": 0.17,
+      "grad_norm": 12.02553653717041,
+      "learning_rate": 1.883445592603439e-05,
+      "loss": 3.5257,
       "step": 1394
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 11.35824966430664,
-      "learning_rate": 1.7205572817480205e-05,
-      "loss": 2.1912,
+      "epoch": 0.18,
+      "grad_norm": 24.402360916137695,
+      "learning_rate": 1.8833619210977703e-05,
+      "loss": 3.3099,
       "step": 1395
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 16.764873504638672,
-      "learning_rate": 1.7203568206875815e-05,
-      "loss": 2.5602,
+      "epoch": 0.18,
+      "grad_norm": 13.76604175567627,
+      "learning_rate": 1.8832782495921017e-05,
+      "loss": 2.6733,
       "step": 1396
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 14.467211723327637,
-      "learning_rate": 1.7201563596271425e-05,
-      "loss": 2.6329,
+      "epoch": 0.18,
+      "grad_norm": 15.612774848937988,
+      "learning_rate": 1.883194578086433e-05,
+      "loss": 2.981,
       "step": 1397
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 11.931891441345215,
-      "learning_rate": 1.7199558985667035e-05,
-      "loss": 2.2821,
+      "epoch": 0.18,
+      "grad_norm": 14.17855453491211,
+      "learning_rate": 1.883110906580764e-05,
+      "loss": 1.7441,
       "step": 1398
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 22.51726722717285,
-      "learning_rate": 1.7197554375062645e-05,
-      "loss": 2.7927,
+      "epoch": 0.18,
+      "grad_norm": 18.615697860717773,
+      "learning_rate": 1.8830272350750955e-05,
+      "loss": 2.8956,
       "step": 1399
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 10.18954086303711,
-      "learning_rate": 1.7195549764458255e-05,
-      "loss": 1.9007,
+      "epoch": 0.18,
+      "grad_norm": 33.42338562011719,
+      "learning_rate": 1.8829435635694265e-05,
+      "loss": 3.1567,
       "step": 1400
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 15.76696491241455,
-      "learning_rate": 1.7193545153853866e-05,
-      "loss": 2.0431,
+      "epoch": 0.18,
+      "grad_norm": 14.481989860534668,
+      "learning_rate": 1.882859892063758e-05,
+      "loss": 2.4805,
       "step": 1401
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 11.40339183807373,
-      "learning_rate": 1.7191540543249476e-05,
-      "loss": 2.5742,
+      "epoch": 0.18,
+      "grad_norm": 15.88669490814209,
+      "learning_rate": 1.8827762205580892e-05,
+      "loss": 1.9319,
       "step": 1402
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 12.021610260009766,
-      "learning_rate": 1.7189535932645086e-05,
-      "loss": 3.6196,
+      "epoch": 0.18,
+      "grad_norm": 21.62628173828125,
+      "learning_rate": 1.8826925490524202e-05,
+      "loss": 1.8743,
       "step": 1403
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 19.796781539916992,
-      "learning_rate": 1.7187531322040696e-05,
-      "loss": 2.8469,
+      "epoch": 0.18,
+      "grad_norm": 15.85845947265625,
+      "learning_rate": 1.8826088775467516e-05,
+      "loss": 3.7623,
       "step": 1404
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 14.170371055603027,
-      "learning_rate": 1.7185526711436306e-05,
-      "loss": 2.3962,
+      "epoch": 0.18,
+      "grad_norm": 12.276082038879395,
+      "learning_rate": 1.8825252060410826e-05,
+      "loss": 1.6724,
       "step": 1405
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 11.088910102844238,
-      "learning_rate": 1.7183522100831916e-05,
-      "loss": 2.6827,
+      "epoch": 0.18,
+      "grad_norm": 17.465612411499023,
+      "learning_rate": 1.882441534535414e-05,
+      "loss": 3.3751,
       "step": 1406
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 136.58908081054688,
-      "learning_rate": 1.7181517490227523e-05,
-      "loss": 2.3091,
+      "epoch": 0.18,
+      "grad_norm": 6.7463765144348145,
+      "learning_rate": 1.8823578630297454e-05,
+      "loss": 1.2683,
       "step": 1407
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 12.442429542541504,
-      "learning_rate": 1.7179512879623136e-05,
-      "loss": 2.4494,
+      "epoch": 0.18,
+      "grad_norm": 17.999980926513672,
+      "learning_rate": 1.8822741915240764e-05,
+      "loss": 1.6499,
       "step": 1408
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 14.868009567260742,
-      "learning_rate": 1.7177508269018746e-05,
-      "loss": 2.5188,
+      "epoch": 0.18,
+      "grad_norm": 8.493429183959961,
+      "learning_rate": 1.8821905200184078e-05,
+      "loss": 2.7235,
       "step": 1409
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 12.164691925048828,
-      "learning_rate": 1.7175503658414353e-05,
-      "loss": 2.378,
+      "epoch": 0.18,
+      "grad_norm": 68.3912124633789,
+      "learning_rate": 1.882106848512739e-05,
+      "loss": 4.2291,
       "step": 1410
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 15.50973129272461,
-      "learning_rate": 1.7173499047809966e-05,
-      "loss": 4.2229,
+      "epoch": 0.18,
+      "grad_norm": 6.770368576049805,
+      "learning_rate": 1.8820231770070705e-05,
+      "loss": 0.8155,
       "step": 1411
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 16.49388885498047,
-      "learning_rate": 1.7171494437205573e-05,
-      "loss": 2.5579,
+      "epoch": 0.18,
+      "grad_norm": 16.140487670898438,
+      "learning_rate": 1.8819395055014015e-05,
+      "loss": 2.7073,
       "step": 1412
     },
     {
-      "epoch": 0.42,
-      "grad_norm": 14.063419342041016,
-      "learning_rate": 1.7169489826601183e-05,
-      "loss": 3.0148,
+      "epoch": 0.18,
+      "grad_norm": 24.126846313476562,
+      "learning_rate": 1.881855833995733e-05,
+      "loss": 2.2317,
       "step": 1413
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 37.99775314331055,
-      "learning_rate": 1.7167485215996793e-05,
-      "loss": 2.9612,
+      "epoch": 0.18,
+      "grad_norm": 10.037046432495117,
+      "learning_rate": 1.8817721624900642e-05,
+      "loss": 2.8032,
       "step": 1414
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 18.0521183013916,
-      "learning_rate": 1.7165480605392403e-05,
-      "loss": 2.871,
+      "epoch": 0.18,
+      "grad_norm": 9.873966217041016,
+      "learning_rate": 1.8816884909843953e-05,
+      "loss": 2.6386,
       "step": 1415
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 17.04564094543457,
-      "learning_rate": 1.7163475994788013e-05,
-      "loss": 2.6688,
+      "epoch": 0.18,
+      "grad_norm": 16.003158569335938,
+      "learning_rate": 1.8816048194787266e-05,
+      "loss": 2.1387,
       "step": 1416
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 13.493517875671387,
-      "learning_rate": 1.7161471384183624e-05,
-      "loss": 2.1012,
+      "epoch": 0.18,
+      "grad_norm": 14.439873695373535,
+      "learning_rate": 1.881521147973058e-05,
+      "loss": 2.8796,
       "step": 1417
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 13.597625732421875,
-      "learning_rate": 1.7159466773579234e-05,
-      "loss": 3.3529,
+      "epoch": 0.18,
+      "grad_norm": 24.364248275756836,
+      "learning_rate": 1.8814374764673894e-05,
+      "loss": 2.6972,
       "step": 1418
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 17.114131927490234,
-      "learning_rate": 1.7157462162974844e-05,
-      "loss": 3.2373,
+      "epoch": 0.18,
+      "grad_norm": 19.170869827270508,
+      "learning_rate": 1.8813538049617204e-05,
+      "loss": 3.3193,
       "step": 1419
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 18.538379669189453,
-      "learning_rate": 1.7155457552370454e-05,
-      "loss": 2.4973,
+      "epoch": 0.18,
+      "grad_norm": 13.805689811706543,
+      "learning_rate": 1.8812701334560518e-05,
+      "loss": 4.1914,
       "step": 1420
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 23.284835815429688,
-      "learning_rate": 1.7153452941766064e-05,
-      "loss": 3.6339,
+      "epoch": 0.18,
+      "grad_norm": 17.165374755859375,
+      "learning_rate": 1.881186461950383e-05,
+      "loss": 2.8776,
       "step": 1421
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 19.382251739501953,
-      "learning_rate": 1.7151448331161674e-05,
-      "loss": 3.1989,
+      "epoch": 0.18,
+      "grad_norm": 11.152364730834961,
+      "learning_rate": 1.881102790444714e-05,
+      "loss": 2.1746,
       "step": 1422
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 17.688274383544922,
-      "learning_rate": 1.7149443720557284e-05,
-      "loss": 2.1218,
+      "epoch": 0.18,
+      "grad_norm": 19.310604095458984,
+      "learning_rate": 1.8810191189390455e-05,
+      "loss": 1.6934,
       "step": 1423
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 24.966703414916992,
-      "learning_rate": 1.7147439109952894e-05,
-      "loss": 2.9954,
+      "epoch": 0.18,
+      "grad_norm": 25.408475875854492,
+      "learning_rate": 1.880935447433377e-05,
+      "loss": 2.5396,
       "step": 1424
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 31.188905715942383,
-      "learning_rate": 1.7145434499348504e-05,
-      "loss": 2.9035,
+      "epoch": 0.18,
+      "grad_norm": 18.383811950683594,
+      "learning_rate": 1.880851775927708e-05,
+      "loss": 2.7541,
       "step": 1425
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 13.262977600097656,
-      "learning_rate": 1.714342988874411e-05,
-      "loss": 1.3303,
+      "epoch": 0.18,
+      "grad_norm": 11.278987884521484,
+      "learning_rate": 1.8807681044220393e-05,
+      "loss": 1.5525,
       "step": 1426
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 13.84818172454834,
-      "learning_rate": 1.7141425278139724e-05,
-      "loss": 2.3282,
+      "epoch": 0.18,
+      "grad_norm": 10.96129322052002,
+      "learning_rate": 1.8806844329163706e-05,
+      "loss": 1.641,
       "step": 1427
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 55.73940658569336,
-      "learning_rate": 1.7139420667535334e-05,
-      "loss": 3.5851,
+      "epoch": 0.18,
+      "grad_norm": 21.37160873413086,
+      "learning_rate": 1.8806007614107017e-05,
+      "loss": 2.2955,
       "step": 1428
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 11.50109577178955,
-      "learning_rate": 1.713741605693094e-05,
-      "loss": 2.6434,
+      "epoch": 0.18,
+      "grad_norm": 13.814507484436035,
+      "learning_rate": 1.880517089905033e-05,
+      "loss": 3.0471,
       "step": 1429
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 17.894203186035156,
-      "learning_rate": 1.7135411446326555e-05,
-      "loss": 2.5986,
+      "epoch": 0.18,
+      "grad_norm": 12.956681251525879,
+      "learning_rate": 1.880433418399364e-05,
+      "loss": 1.7278,
       "step": 1430
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 33.426055908203125,
-      "learning_rate": 1.713340683572216e-05,
-      "loss": 2.7803,
+      "epoch": 0.18,
+      "grad_norm": 12.984116554260254,
+      "learning_rate": 1.8803497468936954e-05,
+      "loss": 2.5082,
       "step": 1431
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 10.518821716308594,
-      "learning_rate": 1.713140222511777e-05,
-      "loss": 2.1774,
+      "epoch": 0.18,
+      "grad_norm": 14.6563138961792,
+      "learning_rate": 1.8802660753880268e-05,
+      "loss": 3.9653,
       "step": 1432
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 13.691694259643555,
-      "learning_rate": 1.712939761451338e-05,
-      "loss": 2.4883,
+      "epoch": 0.18,
+      "grad_norm": 6.648709297180176,
+      "learning_rate": 1.8801824038823578e-05,
+      "loss": 0.3524,
       "step": 1433
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 19.365102767944336,
-      "learning_rate": 1.712739300390899e-05,
-      "loss": 2.6138,
+      "epoch": 0.18,
+      "grad_norm": 17.727685928344727,
+      "learning_rate": 1.8800987323766892e-05,
+      "loss": 2.1727,
       "step": 1434
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 31.23158836364746,
-      "learning_rate": 1.71253883933046e-05,
-      "loss": 2.905,
+      "epoch": 0.18,
+      "grad_norm": 23.565282821655273,
+      "learning_rate": 1.8800150608710205e-05,
+      "loss": 2.1334,
       "step": 1435
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 23.67169761657715,
-      "learning_rate": 1.7123383782700212e-05,
-      "loss": 3.4515,
+      "epoch": 0.18,
+      "grad_norm": 13.650922775268555,
+      "learning_rate": 1.8799313893653516e-05,
+      "loss": 2.2225,
       "step": 1436
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 11.991448402404785,
-      "learning_rate": 1.7121379172095822e-05,
-      "loss": 2.562,
+      "epoch": 0.18,
+      "grad_norm": 10.054190635681152,
+      "learning_rate": 1.879847717859683e-05,
+      "loss": 3.5198,
       "step": 1437
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 34.79066467285156,
-      "learning_rate": 1.7119374561491432e-05,
-      "loss": 3.1665,
+      "epoch": 0.18,
+      "grad_norm": 11.719369888305664,
+      "learning_rate": 1.8797640463540143e-05,
+      "loss": 1.1041,
       "step": 1438
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 24.111236572265625,
-      "learning_rate": 1.7117369950887042e-05,
-      "loss": 3.2707,
+      "epoch": 0.18,
+      "grad_norm": 16.8707218170166,
+      "learning_rate": 1.8796803748483457e-05,
+      "loss": 2.3575,
       "step": 1439
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 36.61920928955078,
-      "learning_rate": 1.7115365340282652e-05,
-      "loss": 2.3113,
-      "step": 1440
-    },
-    {
-      "epoch": 0.43,
-      "eval_loss": 0.4355357885360718,
-      "eval_runtime": 43.4342,
-      "eval_samples_per_second": 34.051,
-      "eval_steps_per_second": 34.051,
+      "epoch": 0.18,
+      "grad_norm": 13.826940536499023,
+      "learning_rate": 1.8795967033426767e-05,
+      "loss": 2.2994,
       "step": 1440
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 25.45790672302246,
-      "learning_rate": 1.7113360729678262e-05,
-      "loss": 2.8492,
+      "epoch": 0.18,
+      "grad_norm": 13.495528221130371,
+      "learning_rate": 1.879513031837008e-05,
+      "loss": 1.6467,
       "step": 1441
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 8.575899124145508,
-      "learning_rate": 1.7111356119073872e-05,
-      "loss": 2.934,
+      "epoch": 0.18,
+      "grad_norm": 8.298262596130371,
+      "learning_rate": 1.8794293603313394e-05,
+      "loss": 2.4304,
       "step": 1442
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 14.703217506408691,
-      "learning_rate": 1.710935150846948e-05,
-      "loss": 2.8837,
+      "epoch": 0.18,
+      "grad_norm": 11.978004455566406,
+      "learning_rate": 1.8793456888256704e-05,
+      "loss": 1.8029,
       "step": 1443
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 45.12884521484375,
-      "learning_rate": 1.7107346897865092e-05,
-      "loss": 4.0134,
+      "epoch": 0.18,
+      "grad_norm": 17.661035537719727,
+      "learning_rate": 1.8792620173200018e-05,
+      "loss": 1.3806,
       "step": 1444
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 17.04209327697754,
-      "learning_rate": 1.71053422872607e-05,
-      "loss": 2.9208,
+      "epoch": 0.18,
+      "grad_norm": 16.51836585998535,
+      "learning_rate": 1.879178345814333e-05,
+      "loss": 2.9111,
       "step": 1445
     },
     {
-      "epoch": 0.43,
-      "grad_norm": 11.436484336853027,
-      "learning_rate": 1.710333767665631e-05,
-      "loss": 2.5288,
+      "epoch": 0.18,
+      "grad_norm": 40.740848541259766,
+      "learning_rate": 1.8790946743086645e-05,
+      "loss": 2.2159,
       "step": 1446
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 11.852707862854004,
-      "learning_rate": 1.7101333066051923e-05,
-      "loss": 2.4064,
+      "epoch": 0.18,
+      "grad_norm": 12.385976791381836,
+      "learning_rate": 1.8790110028029956e-05,
+      "loss": 2.6437,
       "step": 1447
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 18.86704444885254,
-      "learning_rate": 1.709932845544753e-05,
-      "loss": 2.924,
+      "epoch": 0.18,
+      "grad_norm": 19.24806022644043,
+      "learning_rate": 1.878927331297327e-05,
+      "loss": 2.9698,
       "step": 1448
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 26.129491806030273,
-      "learning_rate": 1.7097323844843143e-05,
-      "loss": 2.4179,
+      "epoch": 0.18,
+      "grad_norm": 31.226917266845703,
+      "learning_rate": 1.8788436597916583e-05,
+      "loss": 4.2987,
       "step": 1449
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 13.224143028259277,
-      "learning_rate": 1.709531923423875e-05,
-      "loss": 3.0095,
+      "epoch": 0.18,
+      "grad_norm": 13.335999488830566,
+      "learning_rate": 1.8787599882859893e-05,
+      "loss": 3.1112,
       "step": 1450
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 12.908326148986816,
-      "learning_rate": 1.709331462363436e-05,
-      "loss": 2.6045,
+      "epoch": 0.18,
+      "grad_norm": 13.00089168548584,
+      "learning_rate": 1.8786763167803207e-05,
+      "loss": 5.4781,
       "step": 1451
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 17.967126846313477,
-      "learning_rate": 1.7091310013029973e-05,
-      "loss": 1.9252,
+      "epoch": 0.18,
+      "grad_norm": 17.49352264404297,
+      "learning_rate": 1.878592645274652e-05,
+      "loss": 3.781,
       "step": 1452
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 15.611589431762695,
-      "learning_rate": 1.708930540242558e-05,
-      "loss": 3.118,
+      "epoch": 0.18,
+      "grad_norm": 12.117816925048828,
+      "learning_rate": 1.878508973768983e-05,
+      "loss": 2.2466,
       "step": 1453
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 19.359302520751953,
-      "learning_rate": 1.708730079182119e-05,
-      "loss": 2.8996,
+      "epoch": 0.18,
+      "grad_norm": 18.171384811401367,
+      "learning_rate": 1.8784253022633144e-05,
+      "loss": 2.1827,
       "step": 1454
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 15.188103675842285,
-      "learning_rate": 1.70852961812168e-05,
-      "loss": 3.3204,
+      "epoch": 0.18,
+      "grad_norm": 14.677664756774902,
+      "learning_rate": 1.8783416307576458e-05,
+      "loss": 4.503,
       "step": 1455
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 24.560422897338867,
-      "learning_rate": 1.708329157061241e-05,
-      "loss": 2.7835,
+      "epoch": 0.18,
+      "grad_norm": 7.858498573303223,
+      "learning_rate": 1.8782579592519768e-05,
+      "loss": 1.752,
       "step": 1456
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 13.218994140625,
-      "learning_rate": 1.708128696000802e-05,
-      "loss": 2.3175,
+      "epoch": 0.18,
+      "grad_norm": 16.817955017089844,
+      "learning_rate": 1.8781742877463082e-05,
+      "loss": 1.8967,
       "step": 1457
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 13.680419921875,
-      "learning_rate": 1.707928234940363e-05,
-      "loss": 2.2045,
+      "epoch": 0.18,
+      "grad_norm": 9.136285781860352,
+      "learning_rate": 1.8780906162406392e-05,
+      "loss": 2.2303,
       "step": 1458
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 15.42464828491211,
-      "learning_rate": 1.707727773879924e-05,
-      "loss": 2.7285,
+      "epoch": 0.18,
+      "grad_norm": 10.163810729980469,
+      "learning_rate": 1.8780069447349706e-05,
+      "loss": 4.6567,
       "step": 1459
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 25.709156036376953,
-      "learning_rate": 1.707527312819485e-05,
-      "loss": 3.6841,
+      "epoch": 0.18,
+      "grad_norm": 15.327980995178223,
+      "learning_rate": 1.877923273229302e-05,
+      "loss": 2.154,
       "step": 1460
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 15.961677551269531,
-      "learning_rate": 1.707326851759046e-05,
-      "loss": 3.4276,
+      "epoch": 0.18,
+      "grad_norm": 17.83934783935547,
+      "learning_rate": 1.877839601723633e-05,
+      "loss": 2.4298,
       "step": 1461
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 15.101017951965332,
-      "learning_rate": 1.7071263906986067e-05,
-      "loss": 2.935,
+      "epoch": 0.18,
+      "grad_norm": 11.144344329833984,
+      "learning_rate": 1.8777559302179643e-05,
+      "loss": 1.4427,
       "step": 1462
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 15.92037582397461,
-      "learning_rate": 1.706925929638168e-05,
-      "loss": 2.3067,
+      "epoch": 0.18,
+      "grad_norm": 13.788614273071289,
+      "learning_rate": 1.8776722587122957e-05,
+      "loss": 2.1013,
       "step": 1463
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 17.06378173828125,
-      "learning_rate": 1.7067254685777287e-05,
-      "loss": 2.2178,
+      "epoch": 0.18,
+      "grad_norm": 11.165741920471191,
+      "learning_rate": 1.8775885872066267e-05,
+      "loss": 3.3155,
       "step": 1464
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 15.061570167541504,
-      "learning_rate": 1.7065250075172897e-05,
-      "loss": 2.2647,
+      "epoch": 0.18,
+      "grad_norm": 16.798688888549805,
+      "learning_rate": 1.877504915700958e-05,
+      "loss": 3.3293,
       "step": 1465
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 11.610605239868164,
-      "learning_rate": 1.706324546456851e-05,
-      "loss": 2.2437,
+      "epoch": 0.18,
+      "grad_norm": 16.24199867248535,
+      "learning_rate": 1.8774212441952895e-05,
+      "loss": 3.1168,
       "step": 1466
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 9.191241264343262,
-      "learning_rate": 1.7061240853964118e-05,
-      "loss": 2.011,
+      "epoch": 0.18,
+      "grad_norm": 8.398725509643555,
+      "learning_rate": 1.8773375726896208e-05,
+      "loss": 1.4716,
       "step": 1467
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 29.908700942993164,
-      "learning_rate": 1.7059236243359728e-05,
-      "loss": 3.4803,
+      "epoch": 0.18,
+      "grad_norm": 12.215764045715332,
+      "learning_rate": 1.877253901183952e-05,
+      "loss": 3.7943,
       "step": 1468
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 31.88568115234375,
-      "learning_rate": 1.7057231632755338e-05,
-      "loss": 1.611,
+      "epoch": 0.18,
+      "grad_norm": 30.421791076660156,
+      "learning_rate": 1.8771702296782832e-05,
+      "loss": 4.2063,
       "step": 1469
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 18.888980865478516,
-      "learning_rate": 1.7055227022150948e-05,
-      "loss": 3.0609,
+      "epoch": 0.18,
+      "grad_norm": 8.843764305114746,
+      "learning_rate": 1.8770865581726146e-05,
+      "loss": 0.7433,
       "step": 1470
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 15.416746139526367,
-      "learning_rate": 1.7053222411546558e-05,
-      "loss": 2.4115,
+      "epoch": 0.18,
+      "grad_norm": 14.626254081726074,
+      "learning_rate": 1.8770028866669456e-05,
+      "loss": 2.1776,
       "step": 1471
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 16.257118225097656,
-      "learning_rate": 1.7051217800942168e-05,
-      "loss": 3.6617,
+      "epoch": 0.18,
+      "grad_norm": 12.44121265411377,
+      "learning_rate": 1.876919215161277e-05,
+      "loss": 3.0034,
       "step": 1472
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 14.70289134979248,
-      "learning_rate": 1.7049213190337778e-05,
-      "loss": 2.0395,
+      "epoch": 0.18,
+      "grad_norm": 31.2838077545166,
+      "learning_rate": 1.8768355436556083e-05,
+      "loss": 3.8373,
       "step": 1473
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 14.966999053955078,
-      "learning_rate": 1.7047208579733388e-05,
-      "loss": 2.7337,
+      "epoch": 0.18,
+      "grad_norm": 15.197158813476562,
+      "learning_rate": 1.8767518721499397e-05,
+      "loss": 3.3477,
       "step": 1474
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 19.118864059448242,
-      "learning_rate": 1.7045203969128998e-05,
-      "loss": 2.5427,
+      "epoch": 0.19,
+      "grad_norm": 13.24867057800293,
+      "learning_rate": 1.8766682006442707e-05,
+      "loss": 4.5144,
       "step": 1475
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 21.44363021850586,
-      "learning_rate": 1.704319935852461e-05,
-      "loss": 3.0899,
+      "epoch": 0.19,
+      "grad_norm": 8.047499656677246,
+      "learning_rate": 1.876584529138602e-05,
+      "loss": 0.7997,
       "step": 1476
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 16.222488403320312,
-      "learning_rate": 1.704119474792022e-05,
-      "loss": 2.3837,
+      "epoch": 0.19,
+      "grad_norm": 12.983343124389648,
+      "learning_rate": 1.8765008576329335e-05,
+      "loss": 1.9454,
       "step": 1477
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 13.13603401184082,
-      "learning_rate": 1.703919013731583e-05,
-      "loss": 2.4675,
+      "epoch": 0.19,
+      "grad_norm": 16.986835479736328,
+      "learning_rate": 1.8764171861272645e-05,
+      "loss": 2.1344,
       "step": 1478
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 20.3343563079834,
-      "learning_rate": 1.703718552671144e-05,
-      "loss": 2.9424,
+      "epoch": 0.19,
+      "grad_norm": 11.096803665161133,
+      "learning_rate": 1.876333514621596e-05,
+      "loss": 2.1649,
       "step": 1479
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 19.545503616333008,
-      "learning_rate": 1.703518091610705e-05,
-      "loss": 2.9953,
+      "epoch": 0.19,
+      "grad_norm": 16.026342391967773,
+      "learning_rate": 1.8762498431159272e-05,
+      "loss": 2.6944,
       "step": 1480
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 18.053640365600586,
-      "learning_rate": 1.7033176305502655e-05,
-      "loss": 2.11,
+      "epoch": 0.19,
+      "grad_norm": 17.424779891967773,
+      "learning_rate": 1.8761661716102582e-05,
+      "loss": 2.2067,
       "step": 1481
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 18.4859561920166,
-      "learning_rate": 1.703117169489827e-05,
-      "loss": 3.0607,
+      "epoch": 0.19,
+      "grad_norm": 22.919485092163086,
+      "learning_rate": 1.8760825001045896e-05,
+      "loss": 2.8243,
       "step": 1482
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 13.820666313171387,
-      "learning_rate": 1.702916708429388e-05,
-      "loss": 2.8445,
+      "epoch": 0.19,
+      "grad_norm": 11.948049545288086,
+      "learning_rate": 1.8759988285989206e-05,
+      "loss": 3.5818,
       "step": 1483
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 15.42257022857666,
-      "learning_rate": 1.7027162473689486e-05,
-      "loss": 2.8965,
+      "epoch": 0.19,
+      "grad_norm": 6.547436714172363,
+      "learning_rate": 1.875915157093252e-05,
+      "loss": 0.3148,
       "step": 1484
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 27.479591369628906,
-      "learning_rate": 1.70251578630851e-05,
-      "loss": 2.9632,
+      "epoch": 0.19,
+      "grad_norm": 10.880529403686523,
+      "learning_rate": 1.8758314855875834e-05,
+      "loss": 2.3092,
       "step": 1485
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 21.065073013305664,
-      "learning_rate": 1.7023153252480706e-05,
-      "loss": 2.7442,
+      "epoch": 0.19,
+      "grad_norm": 27.949281692504883,
+      "learning_rate": 1.8757478140819144e-05,
+      "loss": 3.5417,
       "step": 1486
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 7.838624477386475,
-      "learning_rate": 1.7021148641876316e-05,
-      "loss": 2.0983,
+      "epoch": 0.19,
+      "grad_norm": 11.91715145111084,
+      "learning_rate": 1.8756641425762457e-05,
+      "loss": 1.8293,
       "step": 1487
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 17.929758071899414,
-      "learning_rate": 1.7019144031271926e-05,
-      "loss": 2.3848,
+      "epoch": 0.19,
+      "grad_norm": 12.612098693847656,
+      "learning_rate": 1.875580471070577e-05,
+      "loss": 1.8214,
       "step": 1488
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 10.071520805358887,
-      "learning_rate": 1.7017139420667536e-05,
-      "loss": 1.5338,
+      "epoch": 0.19,
+      "grad_norm": 14.226095199584961,
+      "learning_rate": 1.875496799564908e-05,
+      "loss": 2.8455,
       "step": 1489
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 15.892718315124512,
-      "learning_rate": 1.7015134810063146e-05,
-      "loss": 2.5901,
+      "epoch": 0.19,
+      "grad_norm": 12.34013557434082,
+      "learning_rate": 1.8754131280592395e-05,
+      "loss": 2.5628,
       "step": 1490
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 13.212607383728027,
-      "learning_rate": 1.7013130199458756e-05,
-      "loss": 3.0064,
+      "epoch": 0.19,
+      "grad_norm": 14.521828651428223,
+      "learning_rate": 1.875329456553571e-05,
+      "loss": 2.5909,
       "step": 1491
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 21.533849716186523,
-      "learning_rate": 1.7011125588854366e-05,
-      "loss": 2.9632,
+      "epoch": 0.19,
+      "grad_norm": 10.721088409423828,
+      "learning_rate": 1.875245785047902e-05,
+      "loss": 3.2115,
       "step": 1492
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 18.14595603942871,
-      "learning_rate": 1.7009120978249976e-05,
-      "loss": 1.7862,
+      "epoch": 0.19,
+      "grad_norm": 28.61941909790039,
+      "learning_rate": 1.8751621135422333e-05,
+      "loss": 3.6909,
       "step": 1493
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 12.642379760742188,
-      "learning_rate": 1.7007116367645586e-05,
-      "loss": 2.7063,
+      "epoch": 0.19,
+      "grad_norm": 9.875435829162598,
+      "learning_rate": 1.8750784420365646e-05,
+      "loss": 1.6156,
       "step": 1494
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 18.708799362182617,
-      "learning_rate": 1.7005111757041197e-05,
-      "loss": 2.7772,
+      "epoch": 0.19,
+      "grad_norm": 13.883271217346191,
+      "learning_rate": 1.874994770530896e-05,
+      "loss": 0.3683,
       "step": 1495
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 29.240921020507812,
-      "learning_rate": 1.7003107146436807e-05,
-      "loss": 1.7695,
+      "epoch": 0.19,
+      "grad_norm": 46.34034729003906,
+      "learning_rate": 1.874911099025227e-05,
+      "loss": 3.1741,
       "step": 1496
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 18.173198699951172,
-      "learning_rate": 1.7001102535832417e-05,
-      "loss": 2.528,
+      "epoch": 0.19,
+      "grad_norm": 17.740985870361328,
+      "learning_rate": 1.8748274275195584e-05,
+      "loss": 3.2501,
       "step": 1497
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 10.611312866210938,
-      "learning_rate": 1.6999097925228027e-05,
-      "loss": 2.3301,
+      "epoch": 0.19,
+      "grad_norm": 11.90066146850586,
+      "learning_rate": 1.8747437560138897e-05,
+      "loss": 3.7449,
       "step": 1498
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 17.193771362304688,
-      "learning_rate": 1.6997093314623637e-05,
-      "loss": 1.8752,
+      "epoch": 0.19,
+      "grad_norm": 26.554466247558594,
+      "learning_rate": 1.8746600845082208e-05,
+      "loss": 2.4212,
       "step": 1499
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 17.25593376159668,
-      "learning_rate": 1.6995088704019244e-05,
-      "loss": 3.6294,
+      "epoch": 0.19,
+      "grad_norm": 19.78257179260254,
+      "learning_rate": 1.874576413002552e-05,
+      "loss": 4.9226,
       "step": 1500
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 18.063823699951172,
-      "learning_rate": 1.6993084093414857e-05,
-      "loss": 3.0548,
+      "epoch": 0.19,
+      "grad_norm": 18.166656494140625,
+      "learning_rate": 1.8744927414968835e-05,
+      "loss": 2.0438,
       "step": 1501
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 22.38393783569336,
-      "learning_rate": 1.6991079482810467e-05,
-      "loss": 2.6282,
+      "epoch": 0.19,
+      "grad_norm": 9.067051887512207,
+      "learning_rate": 1.874409069991215e-05,
+      "loss": 0.8948,
       "step": 1502
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 11.480286598205566,
-      "learning_rate": 1.6989074872206074e-05,
-      "loss": 1.8335,
+      "epoch": 0.19,
+      "grad_norm": 9.333136558532715,
+      "learning_rate": 1.874325398485546e-05,
+      "loss": 1.4343,
       "step": 1503
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 26.893537521362305,
-      "learning_rate": 1.6987070261601687e-05,
-      "loss": 2.7152,
+      "epoch": 0.19,
+      "grad_norm": 20.27508544921875,
+      "learning_rate": 1.8742417269798773e-05,
+      "loss": 2.5113,
       "step": 1504
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 11.621118545532227,
-      "learning_rate": 1.6985065650997294e-05,
-      "loss": 2.3662,
+      "epoch": 0.19,
+      "grad_norm": 18.887235641479492,
+      "learning_rate": 1.8741580554742086e-05,
+      "loss": 2.838,
       "step": 1505
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 15.238231658935547,
-      "learning_rate": 1.6983061040392904e-05,
-      "loss": 2.4623,
+      "epoch": 0.19,
+      "grad_norm": 19.680187225341797,
+      "learning_rate": 1.8740743839685396e-05,
+      "loss": 1.8419,
       "step": 1506
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 20.844465255737305,
-      "learning_rate": 1.6981056429788514e-05,
-      "loss": 2.5909,
+      "epoch": 0.19,
+      "grad_norm": 9.750263214111328,
+      "learning_rate": 1.873990712462871e-05,
+      "loss": 3.3147,
       "step": 1507
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 11.749519348144531,
-      "learning_rate": 1.6979051819184124e-05,
-      "loss": 1.8395,
+      "epoch": 0.19,
+      "grad_norm": 36.575958251953125,
+      "learning_rate": 1.8739070409572024e-05,
+      "loss": 2.0897,
       "step": 1508
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 15.4244384765625,
-      "learning_rate": 1.6977047208579734e-05,
-      "loss": 3.7336,
+      "epoch": 0.19,
+      "grad_norm": 33.13447952270508,
+      "learning_rate": 1.8738233694515334e-05,
+      "loss": 4.701,
       "step": 1509
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 11.57870864868164,
-      "learning_rate": 1.6975042597975344e-05,
-      "loss": 2.3881,
+      "epoch": 0.19,
+      "grad_norm": 10.510684967041016,
+      "learning_rate": 1.8737396979458648e-05,
+      "loss": 1.1409,
       "step": 1510
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 21.28474235534668,
-      "learning_rate": 1.6973037987370954e-05,
-      "loss": 2.8733,
+      "epoch": 0.19,
+      "grad_norm": 18.95481300354004,
+      "learning_rate": 1.8736560264401958e-05,
+      "loss": 1.5948,
       "step": 1511
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 13.479619026184082,
-      "learning_rate": 1.6971033376766565e-05,
-      "loss": 1.693,
+      "epoch": 0.19,
+      "grad_norm": 18.029176712036133,
+      "learning_rate": 1.873572354934527e-05,
+      "loss": 2.1726,
       "step": 1512
     },
     {
-      "epoch": 0.45,
-      "grad_norm": 17.223281860351562,
-      "learning_rate": 1.6969028766162175e-05,
-      "loss": 2.2971,
+      "epoch": 0.19,
+      "grad_norm": 13.098097801208496,
+      "learning_rate": 1.8734886834288585e-05,
+      "loss": 3.0375,
       "step": 1513
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 31.056154251098633,
-      "learning_rate": 1.6967024155557785e-05,
-      "loss": 2.5655,
+      "epoch": 0.19,
+      "grad_norm": 8.929461479187012,
+      "learning_rate": 1.8734050119231896e-05,
+      "loss": 2.3063,
       "step": 1514
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 60.43313980102539,
-      "learning_rate": 1.6965019544953395e-05,
-      "loss": 3.5182,
+      "epoch": 0.19,
+      "grad_norm": 16.9406795501709,
+      "learning_rate": 1.873321340417521e-05,
+      "loss": 2.5458,
       "step": 1515
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 15.120966911315918,
-      "learning_rate": 1.6963014934349005e-05,
-      "loss": 2.7201,
+      "epoch": 0.19,
+      "grad_norm": 15.653889656066895,
+      "learning_rate": 1.8732376689118523e-05,
+      "loss": 3.3106,
       "step": 1516
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 19.221208572387695,
-      "learning_rate": 1.6961010323744615e-05,
-      "loss": 3.3837,
+      "epoch": 0.19,
+      "grad_norm": 11.295975685119629,
+      "learning_rate": 1.8731539974061833e-05,
+      "loss": 0.8551,
       "step": 1517
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 13.70907974243164,
-      "learning_rate": 1.6959005713140225e-05,
-      "loss": 1.8185,
+      "epoch": 0.19,
+      "grad_norm": 10.627366065979004,
+      "learning_rate": 1.8730703259005147e-05,
+      "loss": 3.8078,
       "step": 1518
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 25.266483306884766,
-      "learning_rate": 1.6957001102535832e-05,
-      "loss": 3.0797,
+      "epoch": 0.19,
+      "grad_norm": 17.634016036987305,
+      "learning_rate": 1.872986654394846e-05,
+      "loss": 3.6376,
       "step": 1519
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 16.34473419189453,
-      "learning_rate": 1.6954996491931445e-05,
-      "loss": 2.689,
+      "epoch": 0.19,
+      "grad_norm": 8.201330184936523,
+      "learning_rate": 1.872902982889177e-05,
+      "loss": 1.2747,
       "step": 1520
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 21.672119140625,
-      "learning_rate": 1.6952991881327055e-05,
-      "loss": 2.4117,
+      "epoch": 0.19,
+      "grad_norm": 30.411375045776367,
+      "learning_rate": 1.8728193113835084e-05,
+      "loss": 3.7582,
       "step": 1521
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 12.567007064819336,
-      "learning_rate": 1.6950987270722662e-05,
-      "loss": 2.5169,
+      "epoch": 0.19,
+      "grad_norm": 9.317015647888184,
+      "learning_rate": 1.8727356398778398e-05,
+      "loss": 2.8275,
       "step": 1522
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 54.59273147583008,
-      "learning_rate": 1.6948982660118275e-05,
-      "loss": 3.4834,
+      "epoch": 0.19,
+      "grad_norm": 14.676838874816895,
+      "learning_rate": 1.872651968372171e-05,
+      "loss": 3.5349,
       "step": 1523
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 18.03861427307129,
-      "learning_rate": 1.6946978049513882e-05,
-      "loss": 2.824,
+      "epoch": 0.19,
+      "grad_norm": 17.38626480102539,
+      "learning_rate": 1.8725682968665022e-05,
+      "loss": 3.3976,
       "step": 1524
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 40.20027542114258,
-      "learning_rate": 1.6944973438909492e-05,
-      "loss": 2.5567,
+      "epoch": 0.19,
+      "grad_norm": 20.100765228271484,
+      "learning_rate": 1.8724846253608335e-05,
+      "loss": 3.4542,
       "step": 1525
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 12.565523147583008,
-      "learning_rate": 1.6942968828305106e-05,
-      "loss": 3.1967,
+      "epoch": 0.19,
+      "grad_norm": 86.41800689697266,
+      "learning_rate": 1.872400953855165e-05,
+      "loss": 3.4961,
       "step": 1526
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 12.132319450378418,
-      "learning_rate": 1.6940964217700712e-05,
-      "loss": 2.5713,
+      "epoch": 0.19,
+      "grad_norm": 11.995071411132812,
+      "learning_rate": 1.872317282349496e-05,
+      "loss": 3.2023,
       "step": 1527
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 28.399211883544922,
-      "learning_rate": 1.6938959607096323e-05,
-      "loss": 2.7361,
+      "epoch": 0.19,
+      "grad_norm": 43.101261138916016,
+      "learning_rate": 1.8722336108438273e-05,
+      "loss": 1.917,
       "step": 1528
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 20.132675170898438,
-      "learning_rate": 1.6936954996491933e-05,
-      "loss": 2.6361,
+      "epoch": 0.19,
+      "grad_norm": 7.935921669006348,
+      "learning_rate": 1.8721499393381587e-05,
+      "loss": 1.2631,
       "step": 1529
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 19.281675338745117,
-      "learning_rate": 1.6934950385887543e-05,
-      "loss": 2.8946,
+      "epoch": 0.19,
+      "grad_norm": 14.268749237060547,
+      "learning_rate": 1.87206626783249e-05,
+      "loss": 4.7786,
       "step": 1530
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 11.40804672241211,
-      "learning_rate": 1.6932945775283153e-05,
-      "loss": 2.4466,
+      "epoch": 0.19,
+      "grad_norm": 13.571290016174316,
+      "learning_rate": 1.871982596326821e-05,
+      "loss": 1.4386,
       "step": 1531
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 19.21536636352539,
-      "learning_rate": 1.6930941164678763e-05,
-      "loss": 2.9093,
+      "epoch": 0.19,
+      "grad_norm": 28.266124725341797,
+      "learning_rate": 1.8718989248211524e-05,
+      "loss": 3.2817,
       "step": 1532
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 13.203646659851074,
-      "learning_rate": 1.6928936554074373e-05,
-      "loss": 2.3911,
+      "epoch": 0.19,
+      "grad_norm": 10.189460754394531,
+      "learning_rate": 1.8718152533154838e-05,
+      "loss": 2.2712,
       "step": 1533
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 11.949692726135254,
-      "learning_rate": 1.6926931943469983e-05,
-      "loss": 2.8228,
+      "epoch": 0.19,
+      "grad_norm": 8.047140121459961,
+      "learning_rate": 1.8717315818098148e-05,
+      "loss": 2.6007,
       "step": 1534
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 14.0568208694458,
-      "learning_rate": 1.6924927332865593e-05,
-      "loss": 2.7099,
+      "epoch": 0.19,
+      "grad_norm": 12.582521438598633,
+      "learning_rate": 1.8716479103041462e-05,
+      "loss": 2.3067,
       "step": 1535
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 55.63922882080078,
-      "learning_rate": 1.69229227222612e-05,
-      "loss": 2.8945,
+      "epoch": 0.19,
+      "grad_norm": 21.55182647705078,
+      "learning_rate": 1.8715642387984772e-05,
+      "loss": 3.2436,
       "step": 1536
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 9.444130897521973,
-      "learning_rate": 1.6920918111656813e-05,
-      "loss": 1.714,
+      "epoch": 0.19,
+      "grad_norm": 9.425176620483398,
+      "learning_rate": 1.8714805672928086e-05,
+      "loss": 3.6856,
       "step": 1537
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 12.79077434539795,
-      "learning_rate": 1.6918913501052423e-05,
-      "loss": 3.4578,
+      "epoch": 0.19,
+      "grad_norm": 15.825239181518555,
+      "learning_rate": 1.87139689578714e-05,
+      "loss": 0.7055,
       "step": 1538
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 12.505064010620117,
-      "learning_rate": 1.691690889044803e-05,
-      "loss": 1.7876,
+      "epoch": 0.19,
+      "grad_norm": 11.436142921447754,
+      "learning_rate": 1.871313224281471e-05,
+      "loss": 1.6424,
       "step": 1539
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 19.741012573242188,
-      "learning_rate": 1.6914904279843644e-05,
-      "loss": 2.2345,
+      "epoch": 0.19,
+      "grad_norm": 25.982975006103516,
+      "learning_rate": 1.8712295527758023e-05,
+      "loss": 2.9231,
       "step": 1540
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 26.9569149017334,
-      "learning_rate": 1.691289966923925e-05,
-      "loss": 2.3699,
+      "epoch": 0.19,
+      "grad_norm": 21.5576171875,
+      "learning_rate": 1.8711458812701337e-05,
+      "loss": 2.6003,
       "step": 1541
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 21.60605812072754,
-      "learning_rate": 1.691089505863486e-05,
-      "loss": 3.1069,
+      "epoch": 0.19,
+      "grad_norm": 12.502641677856445,
+      "learning_rate": 1.8710622097644647e-05,
+      "loss": 2.9228,
       "step": 1542
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 14.085561752319336,
-      "learning_rate": 1.690889044803047e-05,
-      "loss": 2.5667,
+      "epoch": 0.19,
+      "grad_norm": 9.891007423400879,
+      "learning_rate": 1.870978538258796e-05,
+      "loss": 3.1422,
       "step": 1543
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 19.00534439086914,
-      "learning_rate": 1.690688583742608e-05,
-      "loss": 2.816,
+      "epoch": 0.19,
+      "grad_norm": 16.10243034362793,
+      "learning_rate": 1.8708948667531274e-05,
+      "loss": 2.2476,
       "step": 1544
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 14.58003044128418,
-      "learning_rate": 1.690488122682169e-05,
-      "loss": 3.1388,
+      "epoch": 0.19,
+      "grad_norm": 12.385344505310059,
+      "learning_rate": 1.8708111952474585e-05,
+      "loss": 3.9671,
       "step": 1545
     },
     {
-      "epoch": 0.46,
-      "grad_norm": 23.24366569519043,
-      "learning_rate": 1.69028766162173e-05,
-      "loss": 1.9925,
+      "epoch": 0.19,
+      "grad_norm": 22.93207359313965,
+      "learning_rate": 1.87072752374179e-05,
+      "loss": 3.9058,
       "step": 1546
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 18.700740814208984,
-      "learning_rate": 1.690087200561291e-05,
-      "loss": 1.9935,
+      "epoch": 0.19,
+      "grad_norm": 11.850655555725098,
+      "learning_rate": 1.8706438522361212e-05,
+      "loss": 3.4488,
       "step": 1547
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 20.79172134399414,
-      "learning_rate": 1.689886739500852e-05,
-      "loss": 3.1285,
+      "epoch": 0.19,
+      "grad_norm": 27.660612106323242,
+      "learning_rate": 1.8705601807304522e-05,
+      "loss": 3.3126,
       "step": 1548
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 22.56122589111328,
-      "learning_rate": 1.689686278440413e-05,
-      "loss": 3.1273,
+      "epoch": 0.19,
+      "grad_norm": 13.83594036102295,
+      "learning_rate": 1.8704765092247836e-05,
+      "loss": 3.2726,
       "step": 1549
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 12.774810791015625,
-      "learning_rate": 1.689485817379974e-05,
-      "loss": 2.8262,
+      "epoch": 0.19,
+      "grad_norm": 6.115479946136475,
+      "learning_rate": 1.870392837719115e-05,
+      "loss": 2.2482,
       "step": 1550
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 16.529850006103516,
-      "learning_rate": 1.689285356319535e-05,
-      "loss": 2.3473,
+      "epoch": 0.19,
+      "grad_norm": 16.649349212646484,
+      "learning_rate": 1.8703091662134463e-05,
+      "loss": 3.2705,
       "step": 1551
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 20.23556137084961,
-      "learning_rate": 1.689084895259096e-05,
-      "loss": 1.8937,
+      "epoch": 0.19,
+      "grad_norm": 26.15545654296875,
+      "learning_rate": 1.8702254947077774e-05,
+      "loss": 4.2139,
       "step": 1552
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 23.965484619140625,
-      "learning_rate": 1.688884434198657e-05,
-      "loss": 2.5148,
+      "epoch": 0.19,
+      "grad_norm": 9.073019981384277,
+      "learning_rate": 1.8701418232021087e-05,
+      "loss": 1.6893,
       "step": 1553
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 11.394997596740723,
-      "learning_rate": 1.688683973138218e-05,
-      "loss": 2.7078,
+      "epoch": 0.2,
+      "grad_norm": 13.878929138183594,
+      "learning_rate": 1.87005815169644e-05,
+      "loss": 2.7366,
       "step": 1554
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 14.20688247680664,
-      "learning_rate": 1.6884835120777788e-05,
-      "loss": 1.9273,
+      "epoch": 0.2,
+      "grad_norm": 15.587919235229492,
+      "learning_rate": 1.869974480190771e-05,
+      "loss": 2.722,
       "step": 1555
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 8.693448066711426,
-      "learning_rate": 1.68828305101734e-05,
-      "loss": 1.7773,
+      "epoch": 0.2,
+      "grad_norm": 14.054646492004395,
+      "learning_rate": 1.8698908086851025e-05,
+      "loss": 3.5987,
       "step": 1556
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 14.1410493850708,
-      "learning_rate": 1.688082589956901e-05,
-      "loss": 2.5604,
+      "epoch": 0.2,
+      "grad_norm": 21.463497161865234,
+      "learning_rate": 1.869807137179434e-05,
+      "loss": 2.3904,
       "step": 1557
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 29.748794555664062,
-      "learning_rate": 1.6878821288964618e-05,
-      "loss": 2.5162,
+      "epoch": 0.2,
+      "grad_norm": 13.603147506713867,
+      "learning_rate": 1.8697234656737652e-05,
+      "loss": 3.0553,
       "step": 1558
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 10.367297172546387,
-      "learning_rate": 1.6876816678360232e-05,
-      "loss": 3.0466,
+      "epoch": 0.2,
+      "grad_norm": 6.39517068862915,
+      "learning_rate": 1.8696397941680962e-05,
+      "loss": 2.7938,
       "step": 1559
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 16.686874389648438,
-      "learning_rate": 1.687481206775584e-05,
-      "loss": 3.304,
-      "step": 1560
-    },
-    {
-      "epoch": 0.47,
-      "eval_loss": 0.3973664343357086,
-      "eval_runtime": 43.4098,
-      "eval_samples_per_second": 34.071,
-      "eval_steps_per_second": 34.071,
+      "epoch": 0.2,
+      "grad_norm": 17.777820587158203,
+      "learning_rate": 1.8695561226624276e-05,
+      "loss": 3.9284,
       "step": 1560
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 14.685885429382324,
-      "learning_rate": 1.687280745715145e-05,
-      "loss": 2.3338,
+      "epoch": 0.2,
+      "grad_norm": 6.8158111572265625,
+      "learning_rate": 1.869472451156759e-05,
+      "loss": 0.7039,
       "step": 1561
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 20.013853073120117,
-      "learning_rate": 1.687080284654706e-05,
-      "loss": 2.3406,
+      "epoch": 0.2,
+      "grad_norm": 16.261743545532227,
+      "learning_rate": 1.86938877965109e-05,
+      "loss": 3.0032,
       "step": 1562
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 14.868475914001465,
-      "learning_rate": 1.686879823594267e-05,
-      "loss": 1.8712,
+      "epoch": 0.2,
+      "grad_norm": 11.026827812194824,
+      "learning_rate": 1.8693051081454213e-05,
+      "loss": 2.5343,
       "step": 1563
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 20.89446449279785,
-      "learning_rate": 1.686679362533828e-05,
-      "loss": 2.7493,
+      "epoch": 0.2,
+      "grad_norm": 17.33503532409668,
+      "learning_rate": 1.8692214366397524e-05,
+      "loss": 1.871,
       "step": 1564
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 34.56494903564453,
-      "learning_rate": 1.686478901473389e-05,
-      "loss": 2.8977,
+      "epoch": 0.2,
+      "grad_norm": 11.103533744812012,
+      "learning_rate": 1.8691377651340837e-05,
+      "loss": 3.0134,
       "step": 1565
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 14.068147659301758,
-      "learning_rate": 1.68627844041295e-05,
-      "loss": 2.5127,
+      "epoch": 0.2,
+      "grad_norm": 8.425973892211914,
+      "learning_rate": 1.869054093628415e-05,
+      "loss": 0.9428,
       "step": 1566
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 18.625877380371094,
-      "learning_rate": 1.686077979352511e-05,
-      "loss": 2.8985,
+      "epoch": 0.2,
+      "grad_norm": 18.52934455871582,
+      "learning_rate": 1.868970422122746e-05,
+      "loss": 2.4734,
       "step": 1567
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 11.287114143371582,
-      "learning_rate": 1.685877518292072e-05,
-      "loss": 1.6863,
+      "epoch": 0.2,
+      "grad_norm": 17.02704429626465,
+      "learning_rate": 1.8688867506170775e-05,
+      "loss": 3.5376,
       "step": 1568
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 19.42438316345215,
-      "learning_rate": 1.685677057231633e-05,
-      "loss": 2.5056,
+      "epoch": 0.2,
+      "grad_norm": 14.429878234863281,
+      "learning_rate": 1.8688030791114085e-05,
+      "loss": 3.239,
       "step": 1569
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 13.997237205505371,
-      "learning_rate": 1.685476596171194e-05,
-      "loss": 2.6668,
+      "epoch": 0.2,
+      "grad_norm": 11.238717079162598,
+      "learning_rate": 1.86871940760574e-05,
+      "loss": 1.6583,
       "step": 1570
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 10.12904167175293,
-      "learning_rate": 1.685276135110755e-05,
-      "loss": 2.0919,
+      "epoch": 0.2,
+      "grad_norm": 11.936284065246582,
+      "learning_rate": 1.8686357361000713e-05,
+      "loss": 4.6975,
       "step": 1571
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 23.484573364257812,
-      "learning_rate": 1.685075674050316e-05,
-      "loss": 2.9543,
+      "epoch": 0.2,
+      "grad_norm": 19.51447105407715,
+      "learning_rate": 1.8685520645944026e-05,
+      "loss": 1.3431,
       "step": 1572
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 21.227386474609375,
-      "learning_rate": 1.684875212989877e-05,
-      "loss": 2.0336,
+      "epoch": 0.2,
+      "grad_norm": 17.191984176635742,
+      "learning_rate": 1.8684683930887336e-05,
+      "loss": 1.7212,
       "step": 1573
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 30.29522132873535,
-      "learning_rate": 1.6846747519294376e-05,
-      "loss": 2.6453,
+      "epoch": 0.2,
+      "grad_norm": 12.201562881469727,
+      "learning_rate": 1.868384721583065e-05,
+      "loss": 3.2341,
       "step": 1574
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 25.412912368774414,
-      "learning_rate": 1.684474290868999e-05,
-      "loss": 3.1794,
+      "epoch": 0.2,
+      "grad_norm": 18.343507766723633,
+      "learning_rate": 1.8683010500773964e-05,
+      "loss": 3.3134,
       "step": 1575
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 26.188493728637695,
-      "learning_rate": 1.68427382980856e-05,
-      "loss": 2.9667,
+      "epoch": 0.2,
+      "grad_norm": 13.997854232788086,
+      "learning_rate": 1.8682173785717274e-05,
+      "loss": 2.8495,
       "step": 1576
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 15.816750526428223,
-      "learning_rate": 1.6840733687481206e-05,
-      "loss": 2.0286,
+      "epoch": 0.2,
+      "grad_norm": 9.156922340393066,
+      "learning_rate": 1.8681337070660588e-05,
+      "loss": 0.344,
       "step": 1577
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 17.47620964050293,
-      "learning_rate": 1.683872907687682e-05,
-      "loss": 3.0095,
+      "epoch": 0.2,
+      "grad_norm": 17.772275924682617,
+      "learning_rate": 1.86805003556039e-05,
+      "loss": 3.108,
       "step": 1578
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 14.732656478881836,
-      "learning_rate": 1.6836724466272427e-05,
-      "loss": 3.2241,
+      "epoch": 0.2,
+      "grad_norm": 15.730620384216309,
+      "learning_rate": 1.8679663640547215e-05,
+      "loss": 3.0182,
       "step": 1579
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 15.800749778747559,
-      "learning_rate": 1.6834719855668037e-05,
-      "loss": 2.4581,
+      "epoch": 0.2,
+      "grad_norm": 145.39260864257812,
+      "learning_rate": 1.8678826925490525e-05,
+      "loss": 2.7213,
       "step": 1580
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 14.257530212402344,
-      "learning_rate": 1.683271524506365e-05,
-      "loss": 2.6594,
+      "epoch": 0.2,
+      "grad_norm": 15.215822219848633,
+      "learning_rate": 1.867799021043384e-05,
+      "loss": 2.2526,
       "step": 1581
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 18.350427627563477,
-      "learning_rate": 1.6830710634459257e-05,
-      "loss": 2.8644,
+      "epoch": 0.2,
+      "grad_norm": 52.26152038574219,
+      "learning_rate": 1.8677153495377152e-05,
+      "loss": 2.0332,
       "step": 1582
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 17.27747344970703,
-      "learning_rate": 1.6828706023854867e-05,
-      "loss": 2.8629,
+      "epoch": 0.2,
+      "grad_norm": 9.461394309997559,
+      "learning_rate": 1.8676316780320463e-05,
+      "loss": 0.6025,
       "step": 1583
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 13.865232467651367,
-      "learning_rate": 1.6826701413250477e-05,
-      "loss": 2.115,
+      "epoch": 0.2,
+      "grad_norm": 14.551637649536133,
+      "learning_rate": 1.8675480065263776e-05,
+      "loss": 1.4883,
       "step": 1584
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 26.224536895751953,
-      "learning_rate": 1.6824696802646087e-05,
-      "loss": 3.4931,
+      "epoch": 0.2,
+      "grad_norm": 29.688762664794922,
+      "learning_rate": 1.867464335020709e-05,
+      "loss": 2.9508,
       "step": 1585
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 10.468141555786133,
-      "learning_rate": 1.6822692192041697e-05,
-      "loss": 1.9319,
+      "epoch": 0.2,
+      "grad_norm": 11.327323913574219,
+      "learning_rate": 1.8673806635150404e-05,
+      "loss": 3.3841,
       "step": 1586
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 10.993379592895508,
-      "learning_rate": 1.6820687581437307e-05,
-      "loss": 2.8885,
+      "epoch": 0.2,
+      "grad_norm": 18.282663345336914,
+      "learning_rate": 1.8672969920093714e-05,
+      "loss": 3.8014,
       "step": 1587
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 18.79608726501465,
-      "learning_rate": 1.6818682970832917e-05,
-      "loss": 3.4616,
+      "epoch": 0.2,
+      "grad_norm": 10.740460395812988,
+      "learning_rate": 1.8672133205037028e-05,
+      "loss": 3.5532,
       "step": 1588
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 23.80633544921875,
-      "learning_rate": 1.6816678360228528e-05,
-      "loss": 3.5015,
+      "epoch": 0.2,
+      "grad_norm": 23.3198184967041,
+      "learning_rate": 1.8671296489980338e-05,
+      "loss": 3.5516,
       "step": 1589
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 13.126861572265625,
-      "learning_rate": 1.6814673749624138e-05,
-      "loss": 2.2426,
+      "epoch": 0.2,
+      "grad_norm": 15.05379581451416,
+      "learning_rate": 1.867045977492365e-05,
+      "loss": 3.7878,
       "step": 1590
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 18.61846923828125,
-      "learning_rate": 1.6812669139019748e-05,
-      "loss": 2.8252,
+      "epoch": 0.2,
+      "grad_norm": 19.40207290649414,
+      "learning_rate": 1.8669623059866965e-05,
+      "loss": 4.2777,
       "step": 1591
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 9.052312850952148,
-      "learning_rate": 1.6810664528415358e-05,
-      "loss": 2.1735,
+      "epoch": 0.2,
+      "grad_norm": 13.312829971313477,
+      "learning_rate": 1.8668786344810275e-05,
+      "loss": 3.4725,
       "step": 1592
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 9.982820510864258,
-      "learning_rate": 1.6808659917810964e-05,
-      "loss": 1.8825,
+      "epoch": 0.2,
+      "grad_norm": 10.412389755249023,
+      "learning_rate": 1.866794962975359e-05,
+      "loss": 1.704,
       "step": 1593
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 18.984390258789062,
-      "learning_rate": 1.6806655307206578e-05,
-      "loss": 3.0772,
+      "epoch": 0.2,
+      "grad_norm": 13.629739761352539,
+      "learning_rate": 1.86671129146969e-05,
+      "loss": 1.4708,
       "step": 1594
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 19.849212646484375,
-      "learning_rate": 1.6804650696602188e-05,
-      "loss": 2.8981,
+      "epoch": 0.2,
+      "grad_norm": 10.407073974609375,
+      "learning_rate": 1.8666276199640213e-05,
+      "loss": 1.7156,
       "step": 1595
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 15.359475135803223,
-      "learning_rate": 1.6802646085997795e-05,
-      "loss": 2.7745,
+      "epoch": 0.2,
+      "grad_norm": 13.543862342834473,
+      "learning_rate": 1.8665439484583527e-05,
+      "loss": 4.2424,
       "step": 1596
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 13.44256591796875,
-      "learning_rate": 1.6800641475393408e-05,
-      "loss": 2.4322,
+      "epoch": 0.2,
+      "grad_norm": 11.773777961730957,
+      "learning_rate": 1.8664602769526837e-05,
+      "loss": 1.0708,
       "step": 1597
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 19.221418380737305,
-      "learning_rate": 1.6798636864789015e-05,
-      "loss": 1.8762,
+      "epoch": 0.2,
+      "grad_norm": 13.88772201538086,
+      "learning_rate": 1.866376605447015e-05,
+      "loss": 1.8847,
       "step": 1598
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 20.303268432617188,
-      "learning_rate": 1.6796632254184625e-05,
-      "loss": 1.7788,
+      "epoch": 0.2,
+      "grad_norm": 32.566158294677734,
+      "learning_rate": 1.8662929339413464e-05,
+      "loss": 4.1873,
       "step": 1599
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 19.06624412536621,
-      "learning_rate": 1.679462764358024e-05,
-      "loss": 3.7968,
+      "epoch": 0.2,
+      "grad_norm": 23.395357131958008,
+      "learning_rate": 1.8662092624356778e-05,
+      "loss": 3.2244,
       "step": 1600
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 10.663858413696289,
-      "learning_rate": 1.6792623032975845e-05,
-      "loss": 1.1838,
+      "epoch": 0.2,
+      "eval_loss": 0.21643704175949097,
+      "eval_runtime": 95.2668,
+      "eval_samples_per_second": 37.18,
+      "eval_steps_per_second": 37.18,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.963095664978027,
+      "learning_rate": 1.8661255909300088e-05,
+      "loss": 1.789,
       "step": 1601
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 16.027814865112305,
-      "learning_rate": 1.6790618422371455e-05,
-      "loss": 2.7883,
+      "epoch": 0.2,
+      "grad_norm": 10.993432998657227,
+      "learning_rate": 1.8660419194243402e-05,
+      "loss": 2.4976,
       "step": 1602
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 15.941658020019531,
-      "learning_rate": 1.6788613811767065e-05,
-      "loss": 2.9455,
+      "epoch": 0.2,
+      "grad_norm": 13.061546325683594,
+      "learning_rate": 1.8659582479186715e-05,
+      "loss": 2.7292,
       "step": 1603
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 16.636554718017578,
-      "learning_rate": 1.6786609201162675e-05,
-      "loss": 2.8941,
+      "epoch": 0.2,
+      "grad_norm": 17.390151977539062,
+      "learning_rate": 1.8658745764130026e-05,
+      "loss": 3.6218,
       "step": 1604
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 11.56485366821289,
-      "learning_rate": 1.6784604590558285e-05,
-      "loss": 2.961,
+      "epoch": 0.2,
+      "grad_norm": 15.30650520324707,
+      "learning_rate": 1.865790904907334e-05,
+      "loss": 5.6267,
       "step": 1605
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 13.037918090820312,
-      "learning_rate": 1.6782599979953896e-05,
-      "loss": 2.292,
+      "epoch": 0.2,
+      "grad_norm": 19.716659545898438,
+      "learning_rate": 1.8657072334016653e-05,
+      "loss": 3.9792,
       "step": 1606
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 16.472829818725586,
-      "learning_rate": 1.6780595369349506e-05,
-      "loss": 3.5172,
+      "epoch": 0.2,
+      "grad_norm": 18.685670852661133,
+      "learning_rate": 1.8656235618959963e-05,
+      "loss": 2.7784,
       "step": 1607
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 21.085201263427734,
-      "learning_rate": 1.6778590758745116e-05,
-      "loss": 2.8983,
+      "epoch": 0.2,
+      "grad_norm": 12.650303840637207,
+      "learning_rate": 1.8655398903903277e-05,
+      "loss": 2.6199,
       "step": 1608
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 43.585323333740234,
-      "learning_rate": 1.6776586148140726e-05,
-      "loss": 2.7387,
+      "epoch": 0.2,
+      "grad_norm": 9.990423202514648,
+      "learning_rate": 1.865456218884659e-05,
+      "loss": 1.2948,
       "step": 1609
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 12.80986213684082,
-      "learning_rate": 1.6774581537536332e-05,
-      "loss": 2.1192,
+      "epoch": 0.2,
+      "grad_norm": 12.272200584411621,
+      "learning_rate": 1.8653725473789904e-05,
+      "loss": 1.7085,
       "step": 1610
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 24.946006774902344,
-      "learning_rate": 1.6772576926931946e-05,
-      "loss": 3.2683,
+      "epoch": 0.2,
+      "grad_norm": 9.659276008605957,
+      "learning_rate": 1.8652888758733214e-05,
+      "loss": 3.2644,
       "step": 1611
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 10.049246788024902,
-      "learning_rate": 1.6770572316327556e-05,
-      "loss": 2.5447,
+      "epoch": 0.2,
+      "grad_norm": 8.444182395935059,
+      "learning_rate": 1.8652052043676528e-05,
+      "loss": 1.3415,
       "step": 1612
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 13.763422012329102,
-      "learning_rate": 1.6768567705723163e-05,
-      "loss": 2.1387,
+      "epoch": 0.2,
+      "grad_norm": 10.08949089050293,
+      "learning_rate": 1.8651215328619842e-05,
+      "loss": 1.0016,
       "step": 1613
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 18.731260299682617,
-      "learning_rate": 1.6766563095118776e-05,
-      "loss": 1.9977,
+      "epoch": 0.2,
+      "grad_norm": 11.300054550170898,
+      "learning_rate": 1.8650378613563152e-05,
+      "loss": 2.5986,
       "step": 1614
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 11.16000747680664,
-      "learning_rate": 1.6764558484514383e-05,
-      "loss": 1.9304,
+      "epoch": 0.2,
+      "grad_norm": 13.937424659729004,
+      "learning_rate": 1.8649541898506466e-05,
+      "loss": 1.9747,
       "step": 1615
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 15.650193214416504,
-      "learning_rate": 1.6762553873909993e-05,
-      "loss": 3.0744,
+      "epoch": 0.2,
+      "grad_norm": 12.886364936828613,
+      "learning_rate": 1.864870518344978e-05,
+      "loss": 1.5003,
       "step": 1616
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 19.635541915893555,
-      "learning_rate": 1.6760549263305603e-05,
-      "loss": 2.1045,
+      "epoch": 0.2,
+      "grad_norm": 14.680170059204102,
+      "learning_rate": 1.864786846839309e-05,
+      "loss": 3.1874,
       "step": 1617
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 17.894126892089844,
-      "learning_rate": 1.6758544652701213e-05,
-      "loss": 2.6789,
+      "epoch": 0.2,
+      "grad_norm": 16.69063377380371,
+      "learning_rate": 1.8647031753336403e-05,
+      "loss": 3.7762,
       "step": 1618
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 20.0896053314209,
-      "learning_rate": 1.6756540042096823e-05,
-      "loss": 2.6357,
+      "epoch": 0.2,
+      "grad_norm": 19.271764755249023,
+      "learning_rate": 1.8646195038279717e-05,
+      "loss": 2.1444,
       "step": 1619
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 18.415935516357422,
-      "learning_rate": 1.6754535431492433e-05,
-      "loss": 2.9713,
+      "epoch": 0.2,
+      "grad_norm": 11.025174140930176,
+      "learning_rate": 1.8645358323223027e-05,
+      "loss": 2.8204,
       "step": 1620
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 22.75372314453125,
-      "learning_rate": 1.6752530820888043e-05,
-      "loss": 3.3434,
+      "epoch": 0.2,
+      "grad_norm": 12.710426330566406,
+      "learning_rate": 1.864452160816634e-05,
+      "loss": 1.4955,
       "step": 1621
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 20.71872901916504,
-      "learning_rate": 1.6750526210283654e-05,
-      "loss": 2.466,
+      "epoch": 0.2,
+      "grad_norm": 47.94896697998047,
+      "learning_rate": 1.864368489310965e-05,
+      "loss": 1.9602,
       "step": 1622
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 15.37236499786377,
-      "learning_rate": 1.6748521599679264e-05,
-      "loss": 3.1419,
+      "epoch": 0.2,
+      "grad_norm": 13.462227821350098,
+      "learning_rate": 1.8642848178052965e-05,
+      "loss": 2.2038,
       "step": 1623
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 16.39763641357422,
-      "learning_rate": 1.6746516989074874e-05,
-      "loss": 2.6307,
+      "epoch": 0.2,
+      "grad_norm": 15.98450756072998,
+      "learning_rate": 1.8642011462996278e-05,
+      "loss": 4.13,
       "step": 1624
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 12.221854209899902,
-      "learning_rate": 1.6744512378470484e-05,
-      "loss": 3.121,
+      "epoch": 0.2,
+      "grad_norm": 9.845216751098633,
+      "learning_rate": 1.864117474793959e-05,
+      "loss": 0.5645,
       "step": 1625
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 12.343154907226562,
-      "learning_rate": 1.6742507767866094e-05,
-      "loss": 2.0893,
+      "epoch": 0.2,
+      "grad_norm": 16.83765411376953,
+      "learning_rate": 1.8640338032882902e-05,
+      "loss": 1.9015,
       "step": 1626
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 15.398106575012207,
-      "learning_rate": 1.6740503157261704e-05,
-      "loss": 2.126,
+      "epoch": 0.2,
+      "grad_norm": 21.475820541381836,
+      "learning_rate": 1.8639501317826216e-05,
+      "loss": 3.4947,
       "step": 1627
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 16.43120574951172,
-      "learning_rate": 1.6738498546657314e-05,
-      "loss": 3.3734,
+      "epoch": 0.2,
+      "grad_norm": 20.425395965576172,
+      "learning_rate": 1.8638664602769526e-05,
+      "loss": 2.7783,
       "step": 1628
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 11.764989852905273,
-      "learning_rate": 1.673649393605292e-05,
-      "loss": 1.5352,
+      "epoch": 0.2,
+      "grad_norm": 9.8743257522583,
+      "learning_rate": 1.863782788771284e-05,
+      "loss": 2.3828,
       "step": 1629
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 24.872587203979492,
-      "learning_rate": 1.6734489325448534e-05,
-      "loss": 2.9974,
+      "epoch": 0.2,
+      "grad_norm": 9.016057014465332,
+      "learning_rate": 1.8636991172656153e-05,
+      "loss": 1.0653,
       "step": 1630
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 16.981983184814453,
-      "learning_rate": 1.6732484714844144e-05,
-      "loss": 2.1175,
+      "epoch": 0.2,
+      "grad_norm": 23.619979858398438,
+      "learning_rate": 1.8636154457599467e-05,
+      "loss": 2.5996,
       "step": 1631
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 20.261165618896484,
-      "learning_rate": 1.673048010423975e-05,
-      "loss": 3.5279,
+      "epoch": 0.2,
+      "grad_norm": 19.225311279296875,
+      "learning_rate": 1.8635317742542777e-05,
+      "loss": 3.6662,
       "step": 1632
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 26.34723663330078,
-      "learning_rate": 1.6728475493635364e-05,
-      "loss": 3.3491,
+      "epoch": 0.2,
+      "grad_norm": 26.433652877807617,
+      "learning_rate": 1.863448102748609e-05,
+      "loss": 2.8623,
       "step": 1633
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 18.659168243408203,
-      "learning_rate": 1.672647088303097e-05,
-      "loss": 2.9663,
+      "epoch": 0.21,
+      "grad_norm": 228.86788940429688,
+      "learning_rate": 1.8633644312429405e-05,
+      "loss": 3.3586,
       "step": 1634
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 11.186785697937012,
-      "learning_rate": 1.672446627242658e-05,
-      "loss": 2.1397,
+      "epoch": 0.21,
+      "grad_norm": 9.119266510009766,
+      "learning_rate": 1.8632807597372715e-05,
+      "loss": 1.3587,
       "step": 1635
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 9.327524185180664,
-      "learning_rate": 1.672246166182219e-05,
-      "loss": 1.744,
+      "epoch": 0.21,
+      "grad_norm": 29.663616180419922,
+      "learning_rate": 1.863197088231603e-05,
+      "loss": 4.7198,
       "step": 1636
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 17.69675064086914,
-      "learning_rate": 1.67204570512178e-05,
-      "loss": 2.9865,
+      "epoch": 0.21,
+      "grad_norm": 20.09743309020996,
+      "learning_rate": 1.8631134167259342e-05,
+      "loss": 2.0186,
       "step": 1637
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 21.239151000976562,
-      "learning_rate": 1.671845244061341e-05,
-      "loss": 2.8212,
+      "epoch": 0.21,
+      "grad_norm": 18.866987228393555,
+      "learning_rate": 1.8630297452202656e-05,
+      "loss": 1.6415,
       "step": 1638
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 41.4120979309082,
-      "learning_rate": 1.671644783000902e-05,
-      "loss": 2.7785,
+      "epoch": 0.21,
+      "grad_norm": 9.930262565612793,
+      "learning_rate": 1.8629460737145966e-05,
+      "loss": 1.4622,
       "step": 1639
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 11.790454864501953,
-      "learning_rate": 1.671444321940463e-05,
-      "loss": 2.2797,
+      "epoch": 0.21,
+      "grad_norm": 7.938800811767578,
+      "learning_rate": 1.862862402208928e-05,
+      "loss": 2.2411,
       "step": 1640
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 17.779918670654297,
-      "learning_rate": 1.6712438608800242e-05,
-      "loss": 3.0112,
+      "epoch": 0.21,
+      "grad_norm": 59.838722229003906,
+      "learning_rate": 1.8627787307032593e-05,
+      "loss": 1.5835,
       "step": 1641
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 16.839773178100586,
-      "learning_rate": 1.6710433998195852e-05,
-      "loss": 2.4351,
+      "epoch": 0.21,
+      "grad_norm": 14.412603378295898,
+      "learning_rate": 1.8626950591975904e-05,
+      "loss": 2.4392,
       "step": 1642
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 15.281671524047852,
-      "learning_rate": 1.6708429387591462e-05,
-      "loss": 1.9823,
+      "epoch": 0.21,
+      "grad_norm": 12.059946060180664,
+      "learning_rate": 1.8626113876919217e-05,
+      "loss": 1.2654,
       "step": 1643
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 19.457735061645508,
-      "learning_rate": 1.6706424776987072e-05,
-      "loss": 3.6049,
+      "epoch": 0.21,
+      "grad_norm": 9.609929084777832,
+      "learning_rate": 1.862527716186253e-05,
+      "loss": 2.3594,
       "step": 1644
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 16.624021530151367,
-      "learning_rate": 1.6704420166382682e-05,
-      "loss": 3.3101,
+      "epoch": 0.21,
+      "grad_norm": 16.069400787353516,
+      "learning_rate": 1.862444044680584e-05,
+      "loss": 2.0094,
       "step": 1645
     },
     {
-      "epoch": 0.49,
-      "grad_norm": 14.08040714263916,
-      "learning_rate": 1.6702415555778292e-05,
-      "loss": 2.4127,
+      "epoch": 0.21,
+      "grad_norm": 10.690958976745605,
+      "learning_rate": 1.8623603731749155e-05,
+      "loss": 1.2511,
       "step": 1646
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 23.419265747070312,
-      "learning_rate": 1.6700410945173902e-05,
-      "loss": 3.5622,
+      "epoch": 0.21,
+      "grad_norm": 25.660438537597656,
+      "learning_rate": 1.8622767016692465e-05,
+      "loss": 2.8155,
       "step": 1647
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 12.655960083007812,
-      "learning_rate": 1.669840633456951e-05,
-      "loss": 1.7092,
+      "epoch": 0.21,
+      "grad_norm": 14.544229507446289,
+      "learning_rate": 1.862193030163578e-05,
+      "loss": 2.4869,
       "step": 1648
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 10.199254989624023,
-      "learning_rate": 1.6696401723965122e-05,
-      "loss": 1.9251,
+      "epoch": 0.21,
+      "grad_norm": 17.626832962036133,
+      "learning_rate": 1.8621093586579092e-05,
+      "loss": 2.6804,
       "step": 1649
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 30.36039161682129,
-      "learning_rate": 1.6694397113360732e-05,
-      "loss": 1.9794,
+      "epoch": 0.21,
+      "grad_norm": 19.331235885620117,
+      "learning_rate": 1.8620256871522403e-05,
+      "loss": 1.5767,
       "step": 1650
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 20.610151290893555,
-      "learning_rate": 1.669239250275634e-05,
-      "loss": 2.0625,
+      "epoch": 0.21,
+      "grad_norm": 18.1611270904541,
+      "learning_rate": 1.8619420156465716e-05,
+      "loss": 1.382,
       "step": 1651
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 11.706425666809082,
-      "learning_rate": 1.6690387892151953e-05,
-      "loss": 1.8749,
+      "epoch": 0.21,
+      "grad_norm": 16.357955932617188,
+      "learning_rate": 1.861858344140903e-05,
+      "loss": 3.0114,
       "step": 1652
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 9.867399215698242,
-      "learning_rate": 1.668838328154756e-05,
-      "loss": 2.4526,
+      "epoch": 0.21,
+      "grad_norm": 12.59953784942627,
+      "learning_rate": 1.861774672635234e-05,
+      "loss": 2.191,
       "step": 1653
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 21.617414474487305,
-      "learning_rate": 1.668637867094317e-05,
-      "loss": 4.5465,
+      "epoch": 0.21,
+      "grad_norm": 14.794685363769531,
+      "learning_rate": 1.8616910011295654e-05,
+      "loss": 3.3433,
       "step": 1654
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 21.315481185913086,
-      "learning_rate": 1.6684374060338783e-05,
-      "loss": 2.3748,
+      "epoch": 0.21,
+      "grad_norm": 15.138568878173828,
+      "learning_rate": 1.8616073296238968e-05,
+      "loss": 0.7934,
       "step": 1655
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 19.910005569458008,
-      "learning_rate": 1.668236944973439e-05,
-      "loss": 2.4007,
+      "epoch": 0.21,
+      "grad_norm": 16.43613052368164,
+      "learning_rate": 1.8615236581182278e-05,
+      "loss": 3.3455,
       "step": 1656
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 11.719648361206055,
-      "learning_rate": 1.668036483913e-05,
-      "loss": 2.5453,
+      "epoch": 0.21,
+      "grad_norm": 24.419034957885742,
+      "learning_rate": 1.861439986612559e-05,
+      "loss": 2.7773,
       "step": 1657
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 14.366843223571777,
-      "learning_rate": 1.667836022852561e-05,
-      "loss": 2.7823,
+      "epoch": 0.21,
+      "grad_norm": 16.706079483032227,
+      "learning_rate": 1.8613563151068905e-05,
+      "loss": 3.6568,
       "step": 1658
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 20.43113136291504,
-      "learning_rate": 1.667635561792122e-05,
-      "loss": 3.0117,
+      "epoch": 0.21,
+      "grad_norm": 13.282275199890137,
+      "learning_rate": 1.861272643601222e-05,
+      "loss": 2.1749,
       "step": 1659
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 25.725727081298828,
-      "learning_rate": 1.667435100731683e-05,
-      "loss": 1.9852,
+      "epoch": 0.21,
+      "grad_norm": 8.940580368041992,
+      "learning_rate": 1.861188972095553e-05,
+      "loss": 2.3572,
       "step": 1660
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 12.890719413757324,
-      "learning_rate": 1.667234639671244e-05,
-      "loss": 2.8881,
+      "epoch": 0.21,
+      "grad_norm": 11.337531089782715,
+      "learning_rate": 1.8611053005898843e-05,
+      "loss": 3.3588,
       "step": 1661
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 15.555781364440918,
-      "learning_rate": 1.667034178610805e-05,
-      "loss": 3.1708,
+      "epoch": 0.21,
+      "grad_norm": 18.37208366394043,
+      "learning_rate": 1.8610216290842156e-05,
+      "loss": 2.1123,
       "step": 1662
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 13.397451400756836,
-      "learning_rate": 1.666833717550366e-05,
-      "loss": 2.0334,
+      "epoch": 0.21,
+      "grad_norm": 14.14362907409668,
+      "learning_rate": 1.8609379575785467e-05,
+      "loss": 4.3833,
       "step": 1663
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 12.852067947387695,
-      "learning_rate": 1.666633256489927e-05,
-      "loss": 3.193,
+      "epoch": 0.21,
+      "grad_norm": 15.549084663391113,
+      "learning_rate": 1.860854286072878e-05,
+      "loss": 3.4541,
       "step": 1664
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 17.795087814331055,
-      "learning_rate": 1.666432795429488e-05,
-      "loss": 2.1889,
+      "epoch": 0.21,
+      "grad_norm": 17.050514221191406,
+      "learning_rate": 1.8607706145672094e-05,
+      "loss": 2.1293,
       "step": 1665
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 10.84477710723877,
-      "learning_rate": 1.666232334369049e-05,
-      "loss": 2.1249,
+      "epoch": 0.21,
+      "grad_norm": 106.90633392333984,
+      "learning_rate": 1.8606869430615408e-05,
+      "loss": 2.8493,
       "step": 1666
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 13.947750091552734,
-      "learning_rate": 1.66603187330861e-05,
-      "loss": 2.1903,
+      "epoch": 0.21,
+      "grad_norm": 13.605634689331055,
+      "learning_rate": 1.8606032715558718e-05,
+      "loss": 1.3126,
       "step": 1667
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 15.641617774963379,
-      "learning_rate": 1.665831412248171e-05,
-      "loss": 1.4892,
+      "epoch": 0.21,
+      "grad_norm": 18.58948516845703,
+      "learning_rate": 1.860519600050203e-05,
+      "loss": 3.1841,
       "step": 1668
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 17.841720581054688,
-      "learning_rate": 1.665630951187732e-05,
-      "loss": 2.2156,
+      "epoch": 0.21,
+      "grad_norm": 12.607535362243652,
+      "learning_rate": 1.8604359285445345e-05,
+      "loss": 3.6446,
       "step": 1669
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 21.81488800048828,
-      "learning_rate": 1.6654304901272927e-05,
-      "loss": 2.7059,
+      "epoch": 0.21,
+      "grad_norm": 5.925419330596924,
+      "learning_rate": 1.8603522570388655e-05,
+      "loss": 0.8371,
       "step": 1670
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 10.093061447143555,
-      "learning_rate": 1.665230029066854e-05,
-      "loss": 2.087,
+      "epoch": 0.21,
+      "grad_norm": 13.583077430725098,
+      "learning_rate": 1.860268585533197e-05,
+      "loss": 1.5817,
       "step": 1671
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 12.38084602355957,
-      "learning_rate": 1.6650295680064148e-05,
-      "loss": 2.6622,
+      "epoch": 0.21,
+      "grad_norm": 17.19635772705078,
+      "learning_rate": 1.8601849140275283e-05,
+      "loss": 4.8983,
       "step": 1672
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 13.34916877746582,
-      "learning_rate": 1.6648291069459758e-05,
-      "loss": 3.5172,
+      "epoch": 0.21,
+      "grad_norm": 14.493489265441895,
+      "learning_rate": 1.8601012425218593e-05,
+      "loss": 3.0678,
       "step": 1673
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 14.111736297607422,
-      "learning_rate": 1.664628645885537e-05,
-      "loss": 2.5334,
+      "epoch": 0.21,
+      "grad_norm": 10.430756568908691,
+      "learning_rate": 1.8600175710161907e-05,
+      "loss": 5.0035,
       "step": 1674
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 19.57823944091797,
-      "learning_rate": 1.6644281848250978e-05,
-      "loss": 2.8923,
+      "epoch": 0.21,
+      "grad_norm": 41.852569580078125,
+      "learning_rate": 1.8599338995105217e-05,
+      "loss": 2.7717,
       "step": 1675
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 20.907072067260742,
-      "learning_rate": 1.6642277237646588e-05,
-      "loss": 2.7302,
+      "epoch": 0.21,
+      "grad_norm": 26.517459869384766,
+      "learning_rate": 1.859850228004853e-05,
+      "loss": 4.5046,
       "step": 1676
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 12.42391300201416,
-      "learning_rate": 1.6640272627042198e-05,
-      "loss": 2.4441,
+      "epoch": 0.21,
+      "grad_norm": 18.16859245300293,
+      "learning_rate": 1.8597665564991844e-05,
+      "loss": 4.4326,
       "step": 1677
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 39.446815490722656,
-      "learning_rate": 1.6638268016437808e-05,
-      "loss": 3.3596,
+      "epoch": 0.21,
+      "grad_norm": 12.79591178894043,
+      "learning_rate": 1.8596828849935154e-05,
+      "loss": 2.8772,
       "step": 1678
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 10.514243125915527,
-      "learning_rate": 1.6636263405833418e-05,
-      "loss": 1.9339,
+      "epoch": 0.21,
+      "grad_norm": 32.08021926879883,
+      "learning_rate": 1.8595992134878468e-05,
+      "loss": 1.3034,
       "step": 1679
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 22.982282638549805,
-      "learning_rate": 1.6634258795229028e-05,
-      "loss": 2.6203,
-      "step": 1680
-    },
-    {
-      "epoch": 0.51,
-      "eval_loss": 0.421953409910202,
-      "eval_runtime": 43.6652,
-      "eval_samples_per_second": 33.871,
-      "eval_steps_per_second": 33.871,
+      "epoch": 0.21,
+      "grad_norm": 23.306074142456055,
+      "learning_rate": 1.859515541982178e-05,
+      "loss": 3.1075,
       "step": 1680
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 31.43738555908203,
-      "learning_rate": 1.663225418462464e-05,
-      "loss": 2.5882,
+      "epoch": 0.21,
+      "grad_norm": 9.065752983093262,
+      "learning_rate": 1.8594318704765092e-05,
+      "loss": 2.7193,
       "step": 1681
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 10.706387519836426,
-      "learning_rate": 1.663024957402025e-05,
-      "loss": 2.1867,
+      "epoch": 0.21,
+      "grad_norm": 77.59383392333984,
+      "learning_rate": 1.8593481989708406e-05,
+      "loss": 1.0788,
       "step": 1682
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 15.15072250366211,
-      "learning_rate": 1.662824496341586e-05,
-      "loss": 2.9213,
+      "epoch": 0.21,
+      "grad_norm": 29.036340713500977,
+      "learning_rate": 1.859264527465172e-05,
+      "loss": 3.8931,
       "step": 1683
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 9.624457359313965,
-      "learning_rate": 1.6626240352811465e-05,
-      "loss": 1.8701,
+      "epoch": 0.21,
+      "grad_norm": 20.310911178588867,
+      "learning_rate": 1.859180855959503e-05,
+      "loss": 2.2261,
       "step": 1684
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 9.49343490600586,
-      "learning_rate": 1.662423574220708e-05,
-      "loss": 1.612,
+      "epoch": 0.21,
+      "grad_norm": 14.476409912109375,
+      "learning_rate": 1.8590971844538343e-05,
+      "loss": 2.332,
       "step": 1685
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 28.716997146606445,
-      "learning_rate": 1.662223113160269e-05,
-      "loss": 3.3154,
+      "epoch": 0.21,
+      "grad_norm": 7.494418621063232,
+      "learning_rate": 1.8590135129481657e-05,
+      "loss": 0.7061,
       "step": 1686
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 9.385660171508789,
-      "learning_rate": 1.66202265209983e-05,
-      "loss": 2.5589,
+      "epoch": 0.21,
+      "grad_norm": 8.152714729309082,
+      "learning_rate": 1.858929841442497e-05,
+      "loss": 2.0858,
       "step": 1687
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 14.978555679321289,
-      "learning_rate": 1.661822191039391e-05,
-      "loss": 2.1595,
+      "epoch": 0.21,
+      "grad_norm": 9.176353454589844,
+      "learning_rate": 1.858846169936828e-05,
+      "loss": 1.212,
       "step": 1688
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 18.335002899169922,
-      "learning_rate": 1.6616217299789516e-05,
-      "loss": 2.4189,
+      "epoch": 0.21,
+      "grad_norm": 23.76990509033203,
+      "learning_rate": 1.8587624984311594e-05,
+      "loss": 3.4625,
       "step": 1689
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 16.582286834716797,
-      "learning_rate": 1.661421268918513e-05,
-      "loss": 3.3232,
+      "epoch": 0.21,
+      "grad_norm": 19.800321578979492,
+      "learning_rate": 1.8586788269254908e-05,
+      "loss": 1.423,
       "step": 1690
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 13.659370422363281,
-      "learning_rate": 1.6612208078580736e-05,
-      "loss": 2.2705,
+      "epoch": 0.21,
+      "grad_norm": 12.460290908813477,
+      "learning_rate": 1.8585951554198218e-05,
+      "loss": 1.7099,
       "step": 1691
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 13.425541877746582,
-      "learning_rate": 1.6610203467976346e-05,
-      "loss": 2.4391,
+      "epoch": 0.21,
+      "grad_norm": 11.062760353088379,
+      "learning_rate": 1.8585114839141532e-05,
+      "loss": 2.7525,
       "step": 1692
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 22.200546264648438,
-      "learning_rate": 1.660819885737196e-05,
-      "loss": 2.4271,
+      "epoch": 0.21,
+      "grad_norm": 9.447924613952637,
+      "learning_rate": 1.8584278124084846e-05,
+      "loss": 2.3637,
       "step": 1693
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 14.445297241210938,
-      "learning_rate": 1.6606194246767566e-05,
-      "loss": 2.3692,
+      "epoch": 0.21,
+      "grad_norm": 14.916996955871582,
+      "learning_rate": 1.858344140902816e-05,
+      "loss": 2.6405,
       "step": 1694
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 14.760397911071777,
-      "learning_rate": 1.6604189636163176e-05,
-      "loss": 2.3689,
+      "epoch": 0.21,
+      "grad_norm": 10.311155319213867,
+      "learning_rate": 1.858260469397147e-05,
+      "loss": 2.2143,
       "step": 1695
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 10.618377685546875,
-      "learning_rate": 1.6602185025558786e-05,
-      "loss": 2.1075,
+      "epoch": 0.21,
+      "grad_norm": 12.385705947875977,
+      "learning_rate": 1.8581767978914783e-05,
+      "loss": 3.1643,
       "step": 1696
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 17.689512252807617,
-      "learning_rate": 1.6600180414954396e-05,
-      "loss": 3.0902,
+      "epoch": 0.21,
+      "grad_norm": 15.891322135925293,
+      "learning_rate": 1.8580931263858097e-05,
+      "loss": 2.2877,
       "step": 1697
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 8.896835327148438,
-      "learning_rate": 1.6598175804350006e-05,
-      "loss": 2.3301,
+      "epoch": 0.21,
+      "grad_norm": 20.978282928466797,
+      "learning_rate": 1.8580094548801407e-05,
+      "loss": 1.7149,
       "step": 1698
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 9.717402458190918,
-      "learning_rate": 1.6596171193745616e-05,
-      "loss": 1.9009,
+      "epoch": 0.21,
+      "grad_norm": 52.45249557495117,
+      "learning_rate": 1.857925783374472e-05,
+      "loss": 2.7727,
       "step": 1699
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 16.815967559814453,
-      "learning_rate": 1.6594166583141227e-05,
-      "loss": 3.0338,
+      "epoch": 0.21,
+      "grad_norm": 16.271333694458008,
+      "learning_rate": 1.857842111868803e-05,
+      "loss": 3.4275,
       "step": 1700
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 33.1158447265625,
-      "learning_rate": 1.6592161972536837e-05,
-      "loss": 3.0146,
+      "epoch": 0.21,
+      "grad_norm": 10.972915649414062,
+      "learning_rate": 1.8577584403631345e-05,
+      "loss": 3.8511,
       "step": 1701
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 12.088573455810547,
-      "learning_rate": 1.6590157361932447e-05,
-      "loss": 3.0398,
+      "epoch": 0.21,
+      "grad_norm": 19.936786651611328,
+      "learning_rate": 1.8576747688574658e-05,
+      "loss": 2.3473,
       "step": 1702
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 24.571584701538086,
-      "learning_rate": 1.6588152751328053e-05,
-      "loss": 3.3241,
+      "epoch": 0.21,
+      "grad_norm": 14.996315956115723,
+      "learning_rate": 1.857591097351797e-05,
+      "loss": 3.5185,
       "step": 1703
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 23.808401107788086,
-      "learning_rate": 1.6586148140723667e-05,
-      "loss": 3.3574,
+      "epoch": 0.21,
+      "grad_norm": 13.889389038085938,
+      "learning_rate": 1.8575074258461282e-05,
+      "loss": 4.0015,
       "step": 1704
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 20.748384475708008,
-      "learning_rate": 1.6584143530119277e-05,
-      "loss": 3.0009,
+      "epoch": 0.21,
+      "grad_norm": 15.474664688110352,
+      "learning_rate": 1.8574237543404592e-05,
+      "loss": 3.308,
       "step": 1705
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 24.722551345825195,
-      "learning_rate": 1.6582138919514884e-05,
-      "loss": 2.8073,
+      "epoch": 0.21,
+      "grad_norm": 9.600517272949219,
+      "learning_rate": 1.8573400828347906e-05,
+      "loss": 4.5351,
       "step": 1706
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 16.103742599487305,
-      "learning_rate": 1.6580134308910497e-05,
-      "loss": 3.2797,
+      "epoch": 0.21,
+      "grad_norm": 11.001420974731445,
+      "learning_rate": 1.857256411329122e-05,
+      "loss": 2.574,
       "step": 1707
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 17.112245559692383,
-      "learning_rate": 1.6578129698306104e-05,
-      "loss": 2.8844,
+      "epoch": 0.21,
+      "grad_norm": 16.208568572998047,
+      "learning_rate": 1.8571727398234533e-05,
+      "loss": 2.8805,
       "step": 1708
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 12.628554344177246,
-      "learning_rate": 1.6576125087701714e-05,
-      "loss": 2.5837,
+      "epoch": 0.21,
+      "grad_norm": 9.639992713928223,
+      "learning_rate": 1.8570890683177844e-05,
+      "loss": 2.6542,
       "step": 1709
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 15.136996269226074,
-      "learning_rate": 1.6574120477097327e-05,
-      "loss": 2.8925,
+      "epoch": 0.21,
+      "grad_norm": 14.849407196044922,
+      "learning_rate": 1.8570053968121157e-05,
+      "loss": 1.8041,
       "step": 1710
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 15.211194038391113,
-      "learning_rate": 1.6572115866492934e-05,
-      "loss": 2.2617,
+      "epoch": 0.21,
+      "grad_norm": 13.08426570892334,
+      "learning_rate": 1.856921725306447e-05,
+      "loss": 2.8605,
       "step": 1711
     },
     {
-      "epoch": 0.51,
-      "grad_norm": 13.138132095336914,
-      "learning_rate": 1.6570111255888544e-05,
-      "loss": 2.1384,
+      "epoch": 0.21,
+      "grad_norm": 15.084927558898926,
+      "learning_rate": 1.856838053800778e-05,
+      "loss": 2.7685,
       "step": 1712
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 15.303442001342773,
-      "learning_rate": 1.6568106645284154e-05,
-      "loss": 3.7071,
+      "epoch": 0.21,
+      "grad_norm": 12.851953506469727,
+      "learning_rate": 1.8567543822951095e-05,
+      "loss": 3.5303,
       "step": 1713
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 16.974143981933594,
-      "learning_rate": 1.6566102034679764e-05,
-      "loss": 2.0731,
+      "epoch": 0.22,
+      "grad_norm": 11.996140480041504,
+      "learning_rate": 1.856670710789441e-05,
+      "loss": 3.8478,
       "step": 1714
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 22.324352264404297,
-      "learning_rate": 1.6564097424075374e-05,
-      "loss": 2.5875,
+      "epoch": 0.22,
+      "grad_norm": 8.756845474243164,
+      "learning_rate": 1.8565870392837722e-05,
+      "loss": 0.718,
       "step": 1715
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 17.300701141357422,
-      "learning_rate": 1.6562092813470984e-05,
-      "loss": 2.9271,
+      "epoch": 0.22,
+      "grad_norm": 16.061248779296875,
+      "learning_rate": 1.8565033677781032e-05,
+      "loss": 3.183,
       "step": 1716
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 20.104694366455078,
-      "learning_rate": 1.6560088202866595e-05,
-      "loss": 3.3379,
+      "epoch": 0.22,
+      "grad_norm": 12.82740592956543,
+      "learning_rate": 1.8564196962724346e-05,
+      "loss": 2.9077,
       "step": 1717
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 11.9177827835083,
-      "learning_rate": 1.6558083592262205e-05,
-      "loss": 2.2438,
+      "epoch": 0.22,
+      "grad_norm": 11.336594581604004,
+      "learning_rate": 1.856336024766766e-05,
+      "loss": 3.04,
       "step": 1718
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 12.473027229309082,
-      "learning_rate": 1.6556078981657815e-05,
-      "loss": 2.8623,
+      "epoch": 0.22,
+      "grad_norm": 9.98063850402832,
+      "learning_rate": 1.856252353261097e-05,
+      "loss": 3.3082,
       "step": 1719
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 16.957218170166016,
-      "learning_rate": 1.6554074371053425e-05,
-      "loss": 2.7224,
+      "epoch": 0.22,
+      "grad_norm": 14.953742027282715,
+      "learning_rate": 1.8561686817554284e-05,
+      "loss": 4.0954,
       "step": 1720
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 9.168384552001953,
-      "learning_rate": 1.6552069760449035e-05,
-      "loss": 1.4801,
+      "epoch": 0.22,
+      "grad_norm": 11.000082969665527,
+      "learning_rate": 1.8560850102497597e-05,
+      "loss": 2.8741,
       "step": 1721
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 14.333914756774902,
-      "learning_rate": 1.655006514984464e-05,
-      "loss": 1.834,
+      "epoch": 0.22,
+      "grad_norm": 31.708345413208008,
+      "learning_rate": 1.856001338744091e-05,
+      "loss": 3.8666,
       "step": 1722
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 8.387497901916504,
-      "learning_rate": 1.6548060539240255e-05,
-      "loss": 1.4313,
+      "epoch": 0.22,
+      "grad_norm": 21.968839645385742,
+      "learning_rate": 1.855917667238422e-05,
+      "loss": 1.3914,
       "step": 1723
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 10.79047679901123,
-      "learning_rate": 1.6546055928635865e-05,
-      "loss": 3.2971,
+      "epoch": 0.22,
+      "grad_norm": 13.351322174072266,
+      "learning_rate": 1.8558339957327535e-05,
+      "loss": 2.0592,
       "step": 1724
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 10.988299369812012,
-      "learning_rate": 1.6544051318031472e-05,
-      "loss": 2.5413,
+      "epoch": 0.22,
+      "grad_norm": 12.113936424255371,
+      "learning_rate": 1.855750324227085e-05,
+      "loss": 2.0291,
       "step": 1725
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 16.327035903930664,
-      "learning_rate": 1.6542046707427085e-05,
-      "loss": 2.1365,
+      "epoch": 0.22,
+      "grad_norm": 10.533174514770508,
+      "learning_rate": 1.855666652721416e-05,
+      "loss": 2.5006,
       "step": 1726
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 10.375070571899414,
-      "learning_rate": 1.6540042096822692e-05,
-      "loss": 2.7831,
+      "epoch": 0.22,
+      "grad_norm": 8.11629581451416,
+      "learning_rate": 1.8555829812157472e-05,
+      "loss": 2.5132,
       "step": 1727
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 11.78280258178711,
-      "learning_rate": 1.6538037486218302e-05,
-      "loss": 2.6977,
+      "epoch": 0.22,
+      "grad_norm": 14.704581260681152,
+      "learning_rate": 1.8554993097100783e-05,
+      "loss": 3.4344,
       "step": 1728
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 25.220386505126953,
-      "learning_rate": 1.6536032875613916e-05,
-      "loss": 2.6352,
+      "epoch": 0.22,
+      "grad_norm": 12.885751724243164,
+      "learning_rate": 1.8554156382044096e-05,
+      "loss": 1.8441,
       "step": 1729
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 38.015220642089844,
-      "learning_rate": 1.6534028265009522e-05,
-      "loss": 2.9953,
+      "epoch": 0.22,
+      "grad_norm": 16.6527042388916,
+      "learning_rate": 1.855331966698741e-05,
+      "loss": 2.329,
       "step": 1730
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 16.3960018157959,
-      "learning_rate": 1.6532023654405132e-05,
-      "loss": 2.1179,
+      "epoch": 0.22,
+      "grad_norm": 15.419515609741211,
+      "learning_rate": 1.855248295193072e-05,
+      "loss": 2.7254,
       "step": 1731
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 17.71184730529785,
-      "learning_rate": 1.6530019043800742e-05,
-      "loss": 2.3667,
+      "epoch": 0.22,
+      "grad_norm": 16.190248489379883,
+      "learning_rate": 1.8551646236874034e-05,
+      "loss": 2.6196,
       "step": 1732
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 11.10054874420166,
-      "learning_rate": 1.6528014433196353e-05,
-      "loss": 1.6067,
+      "epoch": 0.22,
+      "grad_norm": 18.080875396728516,
+      "learning_rate": 1.8550809521817344e-05,
+      "loss": 0.8927,
       "step": 1733
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 19.382043838500977,
-      "learning_rate": 1.6526009822591963e-05,
-      "loss": 3.0277,
+      "epoch": 0.22,
+      "grad_norm": 20.11777687072754,
+      "learning_rate": 1.8549972806760658e-05,
+      "loss": 3.1772,
       "step": 1734
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 64.88998413085938,
-      "learning_rate": 1.6524005211987573e-05,
-      "loss": 1.9251,
+      "epoch": 0.22,
+      "grad_norm": 9.067489624023438,
+      "learning_rate": 1.854913609170397e-05,
+      "loss": 2.5178,
       "step": 1735
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 16.99798583984375,
-      "learning_rate": 1.6522000601383183e-05,
-      "loss": 2.2818,
+      "epoch": 0.22,
+      "grad_norm": 10.937751770019531,
+      "learning_rate": 1.8548299376647285e-05,
+      "loss": 3.259,
       "step": 1736
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 11.27112102508545,
-      "learning_rate": 1.6519995990778793e-05,
-      "loss": 2.4658,
+      "epoch": 0.22,
+      "grad_norm": 10.825780868530273,
+      "learning_rate": 1.8547462661590595e-05,
+      "loss": 0.5879,
       "step": 1737
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 16.193252563476562,
-      "learning_rate": 1.6517991380174403e-05,
-      "loss": 2.023,
+      "epoch": 0.22,
+      "grad_norm": 30.348739624023438,
+      "learning_rate": 1.854662594653391e-05,
+      "loss": 3.1827,
       "step": 1738
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 11.178709983825684,
-      "learning_rate": 1.6515986769570013e-05,
-      "loss": 2.0991,
+      "epoch": 0.22,
+      "grad_norm": 19.03845977783203,
+      "learning_rate": 1.8545789231477223e-05,
+      "loss": 2.0507,
       "step": 1739
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 12.193035125732422,
-      "learning_rate": 1.6513982158965623e-05,
-      "loss": 2.4835,
+      "epoch": 0.22,
+      "grad_norm": 19.138124465942383,
+      "learning_rate": 1.8544952516420533e-05,
+      "loss": 2.5662,
       "step": 1740
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 11.212671279907227,
-      "learning_rate": 1.6511977548361233e-05,
-      "loss": 1.8206,
+      "epoch": 0.22,
+      "grad_norm": 17.365880966186523,
+      "learning_rate": 1.8544115801363846e-05,
+      "loss": 4.8537,
       "step": 1741
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 7.858899116516113,
-      "learning_rate": 1.6509972937756843e-05,
-      "loss": 2.2607,
+      "epoch": 0.22,
+      "grad_norm": 14.248538970947266,
+      "learning_rate": 1.854327908630716e-05,
+      "loss": 2.9112,
       "step": 1742
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 34.4583740234375,
-      "learning_rate": 1.6507968327152453e-05,
-      "loss": 2.8326,
+      "epoch": 0.22,
+      "grad_norm": 11.511013984680176,
+      "learning_rate": 1.8542442371250474e-05,
+      "loss": 1.631,
       "step": 1743
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 15.727729797363281,
-      "learning_rate": 1.650596371654806e-05,
-      "loss": 3.1574,
+      "epoch": 0.22,
+      "grad_norm": 17.367900848388672,
+      "learning_rate": 1.8541605656193784e-05,
+      "loss": 3.3912,
       "step": 1744
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 25.254718780517578,
-      "learning_rate": 1.6503959105943674e-05,
-      "loss": 3.1741,
+      "epoch": 0.22,
+      "grad_norm": 10.672541618347168,
+      "learning_rate": 1.8540768941137098e-05,
+      "loss": 2.0005,
       "step": 1745
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 17.325634002685547,
-      "learning_rate": 1.650195449533928e-05,
-      "loss": 2.9572,
+      "epoch": 0.22,
+      "grad_norm": 27.745975494384766,
+      "learning_rate": 1.853993222608041e-05,
+      "loss": 3.8916,
       "step": 1746
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 23.073179244995117,
-      "learning_rate": 1.649994988473489e-05,
-      "loss": 1.7773,
+      "epoch": 0.22,
+      "grad_norm": 23.153770446777344,
+      "learning_rate": 1.853909551102372e-05,
+      "loss": 2.1451,
       "step": 1747
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 16.520126342773438,
-      "learning_rate": 1.6497945274130504e-05,
-      "loss": 3.4431,
+      "epoch": 0.22,
+      "grad_norm": 11.951934814453125,
+      "learning_rate": 1.8538258795967035e-05,
+      "loss": 1.9602,
       "step": 1748
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 18.78542709350586,
-      "learning_rate": 1.649594066352611e-05,
-      "loss": 2.6952,
+      "epoch": 0.22,
+      "grad_norm": 12.270580291748047,
+      "learning_rate": 1.853742208091035e-05,
+      "loss": 1.9256,
       "step": 1749
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 19.45966911315918,
-      "learning_rate": 1.649393605292172e-05,
-      "loss": 2.005,
+      "epoch": 0.22,
+      "grad_norm": 19.807544708251953,
+      "learning_rate": 1.8536585365853663e-05,
+      "loss": 3.9123,
       "step": 1750
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 11.350322723388672,
-      "learning_rate": 1.649193144231733e-05,
-      "loss": 1.6326,
+      "epoch": 0.22,
+      "grad_norm": 13.968215942382812,
+      "learning_rate": 1.8535748650796973e-05,
+      "loss": 1.5282,
       "step": 1751
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 18.4759521484375,
-      "learning_rate": 1.648992683171294e-05,
-      "loss": 2.4786,
+      "epoch": 0.22,
+      "grad_norm": 9.959256172180176,
+      "learning_rate": 1.8534911935740286e-05,
+      "loss": 2.4537,
       "step": 1752
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 15.755661010742188,
-      "learning_rate": 1.648792222110855e-05,
-      "loss": 1.7824,
+      "epoch": 0.22,
+      "grad_norm": 10.980196952819824,
+      "learning_rate": 1.8534075220683597e-05,
+      "loss": 1.7011,
       "step": 1753
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 46.813907623291016,
-      "learning_rate": 1.648591761050416e-05,
-      "loss": 3.5372,
+      "epoch": 0.22,
+      "grad_norm": 13.64299201965332,
+      "learning_rate": 1.853323850562691e-05,
+      "loss": 2.0559,
       "step": 1754
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 11.6588134765625,
-      "learning_rate": 1.648391299989977e-05,
-      "loss": 1.7987,
+      "epoch": 0.22,
+      "grad_norm": 16.080829620361328,
+      "learning_rate": 1.8532401790570224e-05,
+      "loss": 3.1677,
       "step": 1755
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 15.711779594421387,
-      "learning_rate": 1.648190838929538e-05,
-      "loss": 2.7252,
+      "epoch": 0.22,
+      "grad_norm": 16.70003890991211,
+      "learning_rate": 1.8531565075513534e-05,
+      "loss": 2.5473,
       "step": 1756
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 9.573955535888672,
-      "learning_rate": 1.647990377869099e-05,
-      "loss": 2.4631,
+      "epoch": 0.22,
+      "grad_norm": 12.727276802062988,
+      "learning_rate": 1.8530728360456848e-05,
+      "loss": 3.6629,
       "step": 1757
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 9.92371940612793,
-      "learning_rate": 1.64778991680866e-05,
-      "loss": 2.0796,
+      "epoch": 0.22,
+      "grad_norm": 15.42647933959961,
+      "learning_rate": 1.8529891645400158e-05,
+      "loss": 3.0492,
       "step": 1758
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 26.95187759399414,
-      "learning_rate": 1.647589455748221e-05,
-      "loss": 2.8893,
+      "epoch": 0.22,
+      "grad_norm": 76.11067962646484,
+      "learning_rate": 1.8529054930343472e-05,
+      "loss": 0.5017,
       "step": 1759
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 9.273636817932129,
-      "learning_rate": 1.647388994687782e-05,
-      "loss": 2.0131,
+      "epoch": 0.22,
+      "grad_norm": 10.16500186920166,
+      "learning_rate": 1.8528218215286785e-05,
+      "loss": 3.383,
       "step": 1760
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 19.220373153686523,
-      "learning_rate": 1.647188533627343e-05,
-      "loss": 3.017,
+      "epoch": 0.22,
+      "grad_norm": 12.610782623291016,
+      "learning_rate": 1.8527381500230096e-05,
+      "loss": 2.814,
       "step": 1761
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 34.35186767578125,
-      "learning_rate": 1.646988072566904e-05,
-      "loss": 3.3928,
+      "epoch": 0.22,
+      "grad_norm": 11.346853256225586,
+      "learning_rate": 1.852654478517341e-05,
+      "loss": 1.7856,
       "step": 1762
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 36.41823959350586,
-      "learning_rate": 1.6467876115064648e-05,
-      "loss": 3.6049,
+      "epoch": 0.22,
+      "grad_norm": 7.411301612854004,
+      "learning_rate": 1.8525708070116723e-05,
+      "loss": 1.8628,
       "step": 1763
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 14.849743843078613,
-      "learning_rate": 1.6465871504460262e-05,
-      "loss": 1.9525,
+      "epoch": 0.22,
+      "grad_norm": 14.970037460327148,
+      "learning_rate": 1.8524871355060037e-05,
+      "loss": 1.8636,
       "step": 1764
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 20.26209259033203,
-      "learning_rate": 1.646386689385587e-05,
-      "loss": 2.2319,
+      "epoch": 0.22,
+      "grad_norm": 24.586057662963867,
+      "learning_rate": 1.8524034640003347e-05,
+      "loss": 3.4602,
       "step": 1765
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 24.784208297729492,
-      "learning_rate": 1.646186228325148e-05,
-      "loss": 2.5943,
+      "epoch": 0.22,
+      "grad_norm": 6.546807289123535,
+      "learning_rate": 1.852319792494666e-05,
+      "loss": 0.668,
       "step": 1766
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 8.97102165222168,
-      "learning_rate": 1.6459857672647092e-05,
-      "loss": 2.0995,
+      "epoch": 0.22,
+      "grad_norm": 11.386066436767578,
+      "learning_rate": 1.8522361209889974e-05,
+      "loss": 1.7019,
       "step": 1767
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 12.025717735290527,
-      "learning_rate": 1.64578530620427e-05,
-      "loss": 1.3348,
+      "epoch": 0.22,
+      "grad_norm": 15.573068618774414,
+      "learning_rate": 1.8521524494833284e-05,
+      "loss": 3.1102,
       "step": 1768
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 22.216552734375,
-      "learning_rate": 1.645584845143831e-05,
-      "loss": 2.9476,
+      "epoch": 0.22,
+      "grad_norm": 18.69664192199707,
+      "learning_rate": 1.8520687779776598e-05,
+      "loss": 1.7954,
       "step": 1769
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 16.27435874938965,
-      "learning_rate": 1.645384384083392e-05,
-      "loss": 2.9931,
+      "epoch": 0.22,
+      "grad_norm": 9.728567123413086,
+      "learning_rate": 1.8519851064719912e-05,
+      "loss": 1.1618,
       "step": 1770
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 23.192441940307617,
-      "learning_rate": 1.645183923022953e-05,
-      "loss": 3.2747,
+      "epoch": 0.22,
+      "grad_norm": 19.483200073242188,
+      "learning_rate": 1.8519014349663225e-05,
+      "loss": 3.4606,
       "step": 1771
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 26.984561920166016,
-      "learning_rate": 1.644983461962514e-05,
-      "loss": 3.2546,
+      "epoch": 0.22,
+      "grad_norm": 14.560922622680664,
+      "learning_rate": 1.8518177634606536e-05,
+      "loss": 3.6956,
       "step": 1772
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 16.29502296447754,
-      "learning_rate": 1.644783000902075e-05,
-      "loss": 2.9974,
+      "epoch": 0.22,
+      "grad_norm": 24.239315032958984,
+      "learning_rate": 1.851734091954985e-05,
+      "loss": 1.1153,
       "step": 1773
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 15.848845481872559,
-      "learning_rate": 1.644582539841636e-05,
-      "loss": 3.1271,
+      "epoch": 0.22,
+      "grad_norm": 15.974748611450195,
+      "learning_rate": 1.8516504204493163e-05,
+      "loss": 3.7795,
       "step": 1774
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 34.28373336791992,
-      "learning_rate": 1.644382078781197e-05,
-      "loss": 3.1274,
+      "epoch": 0.22,
+      "grad_norm": 12.274819374084473,
+      "learning_rate": 1.8515667489436473e-05,
+      "loss": 2.016,
       "step": 1775
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 18.812297821044922,
-      "learning_rate": 1.644181617720758e-05,
-      "loss": 2.0213,
+      "epoch": 0.22,
+      "grad_norm": 8.790360450744629,
+      "learning_rate": 1.8514830774379787e-05,
+      "loss": 2.2788,
       "step": 1776
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 23.698867797851562,
-      "learning_rate": 1.6439811566603186e-05,
-      "loss": 3.1921,
+      "epoch": 0.22,
+      "grad_norm": 12.69150447845459,
+      "learning_rate": 1.85139940593231e-05,
+      "loss": 1.2259,
       "step": 1777
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 8.073270797729492,
-      "learning_rate": 1.64378069559988e-05,
-      "loss": 2.1989,
+      "epoch": 0.22,
+      "grad_norm": 12.652657508850098,
+      "learning_rate": 1.8513157344266414e-05,
+      "loss": 2.6977,
       "step": 1778
     },
     {
-      "epoch": 0.53,
-      "grad_norm": 18.90115737915039,
-      "learning_rate": 1.643580234539441e-05,
-      "loss": 3.2046,
+      "epoch": 0.22,
+      "grad_norm": 182.5238037109375,
+      "learning_rate": 1.8512320629209724e-05,
+      "loss": 2.9445,
       "step": 1779
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 12.170257568359375,
-      "learning_rate": 1.6433797734790016e-05,
-      "loss": 3.0221,
+      "epoch": 0.22,
+      "grad_norm": 14.875434875488281,
+      "learning_rate": 1.8511483914153038e-05,
+      "loss": 1.0557,
       "step": 1780
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 19.89399528503418,
-      "learning_rate": 1.643179312418563e-05,
-      "loss": 3.4413,
+      "epoch": 0.22,
+      "grad_norm": 15.446181297302246,
+      "learning_rate": 1.851064719909635e-05,
+      "loss": 2.3366,
       "step": 1781
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 14.685297012329102,
-      "learning_rate": 1.6429788513581237e-05,
-      "loss": 2.2774,
+      "epoch": 0.22,
+      "grad_norm": 10.390913963317871,
+      "learning_rate": 1.8509810484039662e-05,
+      "loss": 1.2231,
       "step": 1782
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 29.041440963745117,
-      "learning_rate": 1.6427783902976847e-05,
-      "loss": 2.4153,
+      "epoch": 0.22,
+      "grad_norm": 12.480498313903809,
+      "learning_rate": 1.8508973768982976e-05,
+      "loss": 1.8289,
       "step": 1783
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 13.65660285949707,
-      "learning_rate": 1.642577929237246e-05,
-      "loss": 2.6716,
+      "epoch": 0.22,
+      "grad_norm": 8.062909126281738,
+      "learning_rate": 1.8508137053926286e-05,
+      "loss": 2.944,
       "step": 1784
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 11.705273628234863,
-      "learning_rate": 1.6423774681768067e-05,
-      "loss": 1.8001,
+      "epoch": 0.22,
+      "grad_norm": 27.470788955688477,
+      "learning_rate": 1.85073003388696e-05,
+      "loss": 3.3835,
       "step": 1785
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 25.235740661621094,
-      "learning_rate": 1.6421770071163677e-05,
-      "loss": 3.1693,
+      "epoch": 0.22,
+      "grad_norm": 15.176369667053223,
+      "learning_rate": 1.850646362381291e-05,
+      "loss": 2.6695,
       "step": 1786
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 15.884099960327148,
-      "learning_rate": 1.6419765460559287e-05,
-      "loss": 2.4285,
+      "epoch": 0.22,
+      "grad_norm": 16.90056037902832,
+      "learning_rate": 1.8505626908756223e-05,
+      "loss": 2.991,
       "step": 1787
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 52.04803466796875,
-      "learning_rate": 1.6417760849954897e-05,
-      "loss": 4.0988,
+      "epoch": 0.22,
+      "grad_norm": 13.010893821716309,
+      "learning_rate": 1.8504790193699537e-05,
+      "loss": 3.5261,
       "step": 1788
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 42.220130920410156,
-      "learning_rate": 1.6415756239350507e-05,
-      "loss": 2.3157,
+      "epoch": 0.22,
+      "grad_norm": 31.022720336914062,
+      "learning_rate": 1.8503953478642847e-05,
+      "loss": 3.7701,
       "step": 1789
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 19.528057098388672,
-      "learning_rate": 1.6413751628746117e-05,
-      "loss": 2.153,
+      "epoch": 0.22,
+      "grad_norm": 8.040053367614746,
+      "learning_rate": 1.850311676358616e-05,
+      "loss": 0.5569,
       "step": 1790
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 10.022256851196289,
-      "learning_rate": 1.6411747018141727e-05,
-      "loss": 2.6832,
+      "epoch": 0.22,
+      "grad_norm": 5.776892185211182,
+      "learning_rate": 1.8502280048529475e-05,
+      "loss": 0.5414,
       "step": 1791
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 14.05746841430664,
-      "learning_rate": 1.6409742407537337e-05,
-      "loss": 2.6188,
+      "epoch": 0.22,
+      "grad_norm": 10.185456275939941,
+      "learning_rate": 1.850144333347279e-05,
+      "loss": 2.0107,
       "step": 1792
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 17.109386444091797,
-      "learning_rate": 1.6407737796932947e-05,
-      "loss": 2.4536,
+      "epoch": 0.23,
+      "grad_norm": 12.830659866333008,
+      "learning_rate": 1.85006066184161e-05,
+      "loss": 2.1403,
       "step": 1793
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 14.443997383117676,
-      "learning_rate": 1.6405733186328558e-05,
-      "loss": 2.4826,
+      "epoch": 0.23,
+      "grad_norm": 26.65049934387207,
+      "learning_rate": 1.8499769903359412e-05,
+      "loss": 2.5158,
       "step": 1794
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 29.712779998779297,
-      "learning_rate": 1.6403728575724168e-05,
-      "loss": 3.0167,
+      "epoch": 0.23,
+      "grad_norm": 8.251920700073242,
+      "learning_rate": 1.8498933188302726e-05,
+      "loss": 2.5884,
       "step": 1795
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 19.298267364501953,
-      "learning_rate": 1.6401723965119778e-05,
-      "loss": 2.9807,
+      "epoch": 0.23,
+      "grad_norm": 14.79460620880127,
+      "learning_rate": 1.8498096473246036e-05,
+      "loss": 2.9623,
       "step": 1796
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 15.033485412597656,
-      "learning_rate": 1.6399719354515388e-05,
-      "loss": 2.6656,
+      "epoch": 0.23,
+      "grad_norm": 11.604616165161133,
+      "learning_rate": 1.849725975818935e-05,
+      "loss": 1.7841,
       "step": 1797
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 16.12212371826172,
-      "learning_rate": 1.6397714743910998e-05,
-      "loss": 2.3691,
+      "epoch": 0.23,
+      "grad_norm": 13.62449836730957,
+      "learning_rate": 1.8496423043132663e-05,
+      "loss": 2.9062,
       "step": 1798
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 26.953134536743164,
-      "learning_rate": 1.6395710133306605e-05,
-      "loss": 3.2797,
+      "epoch": 0.23,
+      "grad_norm": 21.44681167602539,
+      "learning_rate": 1.8495586328075977e-05,
+      "loss": 3.5239,
       "step": 1799
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 9.630372047424316,
-      "learning_rate": 1.6393705522702218e-05,
-      "loss": 1.7721,
-      "step": 1800
-    },
-    {
-      "epoch": 0.54,
-      "eval_loss": 0.3416885435581207,
-      "eval_runtime": 43.4913,
-      "eval_samples_per_second": 34.007,
-      "eval_steps_per_second": 34.007,
+      "epoch": 0.23,
+      "grad_norm": 10.320944786071777,
+      "learning_rate": 1.8494749613019287e-05,
+      "loss": 2.2619,
       "step": 1800
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 18.797924041748047,
-      "learning_rate": 1.6391700912097825e-05,
-      "loss": 1.5968,
+      "epoch": 0.23,
+      "grad_norm": 9.892986297607422,
+      "learning_rate": 1.84939128979626e-05,
+      "loss": 1.4009,
       "step": 1801
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 31.31778907775879,
-      "learning_rate": 1.6389696301493435e-05,
-      "loss": 2.9072,
+      "epoch": 0.23,
+      "grad_norm": 7.621852397918701,
+      "learning_rate": 1.8493076182905915e-05,
+      "loss": 0.2668,
       "step": 1802
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 14.117547035217285,
-      "learning_rate": 1.6387691690889048e-05,
-      "loss": 2.4949,
+      "epoch": 0.23,
+      "grad_norm": 10.474602699279785,
+      "learning_rate": 1.8492239467849225e-05,
+      "loss": 3.0311,
       "step": 1803
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 12.579538345336914,
-      "learning_rate": 1.6385687080284655e-05,
-      "loss": 1.8951,
+      "epoch": 0.23,
+      "grad_norm": 13.508172988891602,
+      "learning_rate": 1.849140275279254e-05,
+      "loss": 3.6226,
       "step": 1804
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 23.75373077392578,
-      "learning_rate": 1.6383682469680265e-05,
-      "loss": 1.9489,
+      "epoch": 0.23,
+      "grad_norm": 19.923870086669922,
+      "learning_rate": 1.8490566037735852e-05,
+      "loss": 2.9188,
       "step": 1805
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 13.660398483276367,
-      "learning_rate": 1.6381677859075875e-05,
-      "loss": 1.6119,
+      "epoch": 0.23,
+      "grad_norm": 7.190248489379883,
+      "learning_rate": 1.8489729322679162e-05,
+      "loss": 1.7364,
       "step": 1806
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 60.78506851196289,
-      "learning_rate": 1.6379673248471485e-05,
-      "loss": 2.7651,
+      "epoch": 0.23,
+      "grad_norm": 13.450784683227539,
+      "learning_rate": 1.8488892607622476e-05,
+      "loss": 1.7814,
       "step": 1807
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 14.931727409362793,
-      "learning_rate": 1.6377668637867095e-05,
-      "loss": 2.2028,
+      "epoch": 0.23,
+      "grad_norm": 12.926718711853027,
+      "learning_rate": 1.848805589256579e-05,
+      "loss": 2.7234,
       "step": 1808
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 10.574440956115723,
-      "learning_rate": 1.6375664027262705e-05,
-      "loss": 2.6472,
+      "epoch": 0.23,
+      "grad_norm": 6.936464309692383,
+      "learning_rate": 1.84872191775091e-05,
+      "loss": 0.7802,
       "step": 1809
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 14.71934986114502,
-      "learning_rate": 1.6373659416658315e-05,
-      "loss": 2.3538,
+      "epoch": 0.23,
+      "grad_norm": 27.081253051757812,
+      "learning_rate": 1.8486382462452414e-05,
+      "loss": 3.1689,
       "step": 1810
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 16.00834846496582,
-      "learning_rate": 1.6371654806053926e-05,
-      "loss": 2.8576,
+      "epoch": 0.23,
+      "grad_norm": 14.529179573059082,
+      "learning_rate": 1.8485545747395724e-05,
+      "loss": 3.096,
       "step": 1811
     },
     {
-      "epoch": 0.54,
-      "grad_norm": 15.810260772705078,
-      "learning_rate": 1.6369650195449536e-05,
-      "loss": 2.9982,
+      "epoch": 0.23,
+      "grad_norm": 11.196646690368652,
+      "learning_rate": 1.8484709032339038e-05,
+      "loss": 2.3944,
       "step": 1812
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 29.83978843688965,
-      "learning_rate": 1.6367645584845146e-05,
-      "loss": 3.0221,
+      "epoch": 0.23,
+      "grad_norm": 20.353017807006836,
+      "learning_rate": 1.848387231728235e-05,
+      "loss": 2.2399,
       "step": 1813
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 16.944915771484375,
-      "learning_rate": 1.6365640974240756e-05,
-      "loss": 2.8306,
+      "epoch": 0.23,
+      "grad_norm": 11.980401039123535,
+      "learning_rate": 1.848303560222566e-05,
+      "loss": 1.1062,
       "step": 1814
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 22.454910278320312,
-      "learning_rate": 1.6363636363636366e-05,
-      "loss": 2.3444,
+      "epoch": 0.23,
+      "grad_norm": 8.730816841125488,
+      "learning_rate": 1.8482198887168975e-05,
+      "loss": 1.8447,
       "step": 1815
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 16.41376495361328,
-      "learning_rate": 1.6361631753031976e-05,
-      "loss": 1.8047,
+      "epoch": 0.23,
+      "grad_norm": 23.536231994628906,
+      "learning_rate": 1.848136217211229e-05,
+      "loss": 4.1924,
       "step": 1816
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 4576.21875,
-      "learning_rate": 1.6359627142427586e-05,
-      "loss": 2.2836,
+      "epoch": 0.23,
+      "grad_norm": 22.22958755493164,
+      "learning_rate": 1.84805254570556e-05,
+      "loss": 1.1769,
       "step": 1817
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 15.997746467590332,
-      "learning_rate": 1.6357622531823193e-05,
-      "loss": 2.5703,
+      "epoch": 0.23,
+      "grad_norm": 12.718234062194824,
+      "learning_rate": 1.8479688741998913e-05,
+      "loss": 2.6205,
       "step": 1818
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 15.037249565124512,
-      "learning_rate": 1.6355617921218806e-05,
-      "loss": 3.2597,
+      "epoch": 0.23,
+      "grad_norm": 8.55314826965332,
+      "learning_rate": 1.8478852026942226e-05,
+      "loss": 0.3194,
       "step": 1819
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 13.758872032165527,
-      "learning_rate": 1.6353613310614413e-05,
-      "loss": 3.4616,
+      "epoch": 0.23,
+      "grad_norm": 20.39598846435547,
+      "learning_rate": 1.847801531188554e-05,
+      "loss": 1.7185,
       "step": 1820
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 10.099397659301758,
-      "learning_rate": 1.6351608700010023e-05,
-      "loss": 1.96,
+      "epoch": 0.23,
+      "grad_norm": 13.92740535736084,
+      "learning_rate": 1.847717859682885e-05,
+      "loss": 1.4368,
       "step": 1821
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 32.01512908935547,
-      "learning_rate": 1.6349604089405636e-05,
-      "loss": 4.0462,
+      "epoch": 0.23,
+      "grad_norm": 10.905973434448242,
+      "learning_rate": 1.8476341881772164e-05,
+      "loss": 1.3227,
       "step": 1822
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 17.568777084350586,
-      "learning_rate": 1.6347599478801243e-05,
-      "loss": 2.5459,
+      "epoch": 0.23,
+      "grad_norm": 17.807950973510742,
+      "learning_rate": 1.8475505166715478e-05,
+      "loss": 2.9048,
       "step": 1823
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 20.525028228759766,
-      "learning_rate": 1.6345594868196853e-05,
-      "loss": 2.6286,
+      "epoch": 0.23,
+      "grad_norm": 18.465328216552734,
+      "learning_rate": 1.8474668451658788e-05,
+      "loss": 1.9646,
       "step": 1824
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 18.882827758789062,
-      "learning_rate": 1.6343590257592463e-05,
-      "loss": 2.5098,
+      "epoch": 0.23,
+      "grad_norm": 13.115753173828125,
+      "learning_rate": 1.84738317366021e-05,
+      "loss": 3.1957,
       "step": 1825
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 10.823426246643066,
-      "learning_rate": 1.6341585646988073e-05,
-      "loss": 1.6721,
+      "epoch": 0.23,
+      "grad_norm": 11.76266860961914,
+      "learning_rate": 1.8472995021545415e-05,
+      "loss": 1.2549,
       "step": 1826
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 14.602689743041992,
-      "learning_rate": 1.6339581036383684e-05,
-      "loss": 2.6786,
+      "epoch": 0.23,
+      "grad_norm": 19.048133850097656,
+      "learning_rate": 1.847215830648873e-05,
+      "loss": 2.3595,
       "step": 1827
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 17.44394302368164,
-      "learning_rate": 1.6337576425779294e-05,
-      "loss": 2.5762,
+      "epoch": 0.23,
+      "grad_norm": 11.120207786560059,
+      "learning_rate": 1.847132159143204e-05,
+      "loss": 2.949,
       "step": 1828
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 19.712026596069336,
-      "learning_rate": 1.6335571815174904e-05,
-      "loss": 2.6143,
+      "epoch": 0.23,
+      "grad_norm": 7.369383335113525,
+      "learning_rate": 1.8470484876375353e-05,
+      "loss": 2.196,
       "step": 1829
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 25.916528701782227,
-      "learning_rate": 1.6333567204570514e-05,
-      "loss": 2.5159,
+      "epoch": 0.23,
+      "grad_norm": 9.80823040008545,
+      "learning_rate": 1.8469648161318666e-05,
+      "loss": 3.9954,
       "step": 1830
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 17.429548263549805,
-      "learning_rate": 1.6331562593966124e-05,
-      "loss": 3.5722,
+      "epoch": 0.23,
+      "grad_norm": 9.584776878356934,
+      "learning_rate": 1.8468811446261977e-05,
+      "loss": 1.239,
       "step": 1831
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 14.735082626342773,
-      "learning_rate": 1.6329557983361734e-05,
-      "loss": 2.0687,
+      "epoch": 0.23,
+      "grad_norm": 13.136709213256836,
+      "learning_rate": 1.846797473120529e-05,
+      "loss": 2.052,
       "step": 1832
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 19.132709503173828,
-      "learning_rate": 1.6327553372757344e-05,
-      "loss": 2.9178,
+      "epoch": 0.23,
+      "grad_norm": 11.339838981628418,
+      "learning_rate": 1.8467138016148604e-05,
+      "loss": 1.2994,
       "step": 1833
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 20.842416763305664,
-      "learning_rate": 1.6325548762152954e-05,
-      "loss": 3.3049,
+      "epoch": 0.23,
+      "grad_norm": 27.841861724853516,
+      "learning_rate": 1.8466301301091914e-05,
+      "loss": 3.0134,
       "step": 1834
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 14.198040008544922,
-      "learning_rate": 1.6323544151548564e-05,
-      "loss": 2.8631,
+      "epoch": 0.23,
+      "grad_norm": 13.050997734069824,
+      "learning_rate": 1.8465464586035228e-05,
+      "loss": 3.0181,
       "step": 1835
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 12.003619194030762,
-      "learning_rate": 1.6321539540944174e-05,
-      "loss": 2.0466,
+      "epoch": 0.23,
+      "grad_norm": 75.58671569824219,
+      "learning_rate": 1.846462787097854e-05,
+      "loss": 3.5269,
       "step": 1836
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 14.629396438598633,
-      "learning_rate": 1.631953493033978e-05,
-      "loss": 2.967,
+      "epoch": 0.23,
+      "grad_norm": 10.211413383483887,
+      "learning_rate": 1.8463791155921852e-05,
+      "loss": 2.2705,
       "step": 1837
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 15.829573631286621,
-      "learning_rate": 1.6317530319735394e-05,
-      "loss": 2.1286,
+      "epoch": 0.23,
+      "grad_norm": 14.928301811218262,
+      "learning_rate": 1.8462954440865165e-05,
+      "loss": 2.5123,
       "step": 1838
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 14.251692771911621,
-      "learning_rate": 1.6315525709131e-05,
-      "loss": 2.4073,
+      "epoch": 0.23,
+      "grad_norm": 33.32582473754883,
+      "learning_rate": 1.8462117725808476e-05,
+      "loss": 3.597,
       "step": 1839
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 25.430139541625977,
-      "learning_rate": 1.631352109852661e-05,
-      "loss": 2.9292,
+      "epoch": 0.23,
+      "grad_norm": 33.28379440307617,
+      "learning_rate": 1.846128101075179e-05,
+      "loss": 4.2034,
       "step": 1840
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 27.490047454833984,
-      "learning_rate": 1.6311516487922225e-05,
-      "loss": 3.4921,
+      "epoch": 0.23,
+      "grad_norm": 8.760516166687012,
+      "learning_rate": 1.8460444295695103e-05,
+      "loss": 2.4341,
       "step": 1841
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 9.101457595825195,
-      "learning_rate": 1.630951187731783e-05,
-      "loss": 2.1141,
+      "epoch": 0.23,
+      "grad_norm": 11.161774635314941,
+      "learning_rate": 1.8459607580638413e-05,
+      "loss": 2.317,
       "step": 1842
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 12.929884910583496,
-      "learning_rate": 1.630750726671344e-05,
-      "loss": 1.934,
+      "epoch": 0.23,
+      "grad_norm": 9.371776580810547,
+      "learning_rate": 1.8458770865581727e-05,
+      "loss": 2.9491,
       "step": 1843
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 11.764657020568848,
-      "learning_rate": 1.630550265610905e-05,
-      "loss": 2.0135,
+      "epoch": 0.23,
+      "grad_norm": 23.16387176513672,
+      "learning_rate": 1.845793415052504e-05,
+      "loss": 2.3188,
       "step": 1844
     },
     {
-      "epoch": 0.55,
-      "grad_norm": 15.838483810424805,
-      "learning_rate": 1.630349804550466e-05,
-      "loss": 2.7645,
+      "epoch": 0.23,
+      "grad_norm": 11.673296928405762,
+      "learning_rate": 1.845709743546835e-05,
+      "loss": 2.4927,
       "step": 1845
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 19.013673782348633,
-      "learning_rate": 1.6301493434900272e-05,
-      "loss": 2.2205,
+      "epoch": 0.23,
+      "grad_norm": 14.898637771606445,
+      "learning_rate": 1.8456260720411664e-05,
+      "loss": 2.679,
       "step": 1846
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 12.440937042236328,
-      "learning_rate": 1.6299488824295882e-05,
-      "loss": 3.1805,
+      "epoch": 0.23,
+      "grad_norm": 10.247369766235352,
+      "learning_rate": 1.8455424005354978e-05,
+      "loss": 2.9585,
       "step": 1847
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 18.47015380859375,
-      "learning_rate": 1.6297484213691492e-05,
-      "loss": 2.9228,
+      "epoch": 0.23,
+      "grad_norm": 14.360986709594727,
+      "learning_rate": 1.8454587290298292e-05,
+      "loss": 2.2375,
       "step": 1848
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 26.46478843688965,
-      "learning_rate": 1.6295479603087102e-05,
-      "loss": 3.1306,
+      "epoch": 0.23,
+      "grad_norm": 12.506513595581055,
+      "learning_rate": 1.8453750575241602e-05,
+      "loss": 2.4649,
       "step": 1849
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 20.930150985717773,
-      "learning_rate": 1.6293474992482712e-05,
-      "loss": 2.2428,
+      "epoch": 0.23,
+      "grad_norm": 15.462091445922852,
+      "learning_rate": 1.8452913860184916e-05,
+      "loss": 2.2894,
       "step": 1850
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 13.267045021057129,
-      "learning_rate": 1.629147038187832e-05,
-      "loss": 2.6697,
+      "epoch": 0.23,
+      "grad_norm": 13.013998031616211,
+      "learning_rate": 1.845207714512823e-05,
+      "loss": 3.056,
       "step": 1851
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 11.723921775817871,
-      "learning_rate": 1.6289465771273932e-05,
-      "loss": 2.8602,
+      "epoch": 0.23,
+      "grad_norm": 10.314661979675293,
+      "learning_rate": 1.845124043007154e-05,
+      "loss": 2.3853,
       "step": 1852
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 11.42064094543457,
-      "learning_rate": 1.6287461160669542e-05,
-      "loss": 2.9313,
+      "epoch": 0.23,
+      "grad_norm": 8.996312141418457,
+      "learning_rate": 1.8450403715014853e-05,
+      "loss": 1.611,
       "step": 1853
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 31.033870697021484,
-      "learning_rate": 1.628545655006515e-05,
-      "loss": 3.2394,
+      "epoch": 0.23,
+      "grad_norm": 19.560791015625,
+      "learning_rate": 1.8449566999958167e-05,
+      "loss": 5.2749,
       "step": 1854
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 17.858312606811523,
-      "learning_rate": 1.6283451939460762e-05,
-      "loss": 2.1223,
+      "epoch": 0.23,
+      "grad_norm": 14.763134956359863,
+      "learning_rate": 1.844873028490148e-05,
+      "loss": 1.3857,
       "step": 1855
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 13.080681800842285,
-      "learning_rate": 1.628144732885637e-05,
-      "loss": 2.4824,
+      "epoch": 0.23,
+      "grad_norm": 7.593055725097656,
+      "learning_rate": 1.844789356984479e-05,
+      "loss": 2.0708,
       "step": 1856
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 9.161316871643066,
-      "learning_rate": 1.6279442718251983e-05,
-      "loss": 1.5993,
+      "epoch": 0.23,
+      "grad_norm": 14.49431037902832,
+      "learning_rate": 1.8447056854788104e-05,
+      "loss": 2.6208,
       "step": 1857
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 15.82925796508789,
-      "learning_rate": 1.6277438107647593e-05,
-      "loss": 2.2007,
+      "epoch": 0.23,
+      "grad_norm": 6.890941619873047,
+      "learning_rate": 1.8446220139731418e-05,
+      "loss": 0.7461,
       "step": 1858
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 29.120468139648438,
-      "learning_rate": 1.62754334970432e-05,
-      "loss": 3.2814,
+      "epoch": 0.23,
+      "grad_norm": 13.095731735229492,
+      "learning_rate": 1.8445383424674728e-05,
+      "loss": 3.8139,
       "step": 1859
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 20.672487258911133,
-      "learning_rate": 1.6273428886438813e-05,
-      "loss": 2.6652,
+      "epoch": 0.23,
+      "grad_norm": 8.560629844665527,
+      "learning_rate": 1.8444546709618042e-05,
+      "loss": 3.0648,
       "step": 1860
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 18.959318161010742,
-      "learning_rate": 1.627142427583442e-05,
-      "loss": 2.5797,
+      "epoch": 0.23,
+      "grad_norm": 27.346872329711914,
+      "learning_rate": 1.8443709994561356e-05,
+      "loss": 1.476,
       "step": 1861
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 10.144336700439453,
-      "learning_rate": 1.626941966523003e-05,
-      "loss": 2.1145,
+      "epoch": 0.23,
+      "grad_norm": 16.805004119873047,
+      "learning_rate": 1.8442873279504666e-05,
+      "loss": 1.9073,
       "step": 1862
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 19.466264724731445,
-      "learning_rate": 1.626741505462564e-05,
-      "loss": 2.4139,
+      "epoch": 0.23,
+      "grad_norm": 10.577855110168457,
+      "learning_rate": 1.844203656444798e-05,
+      "loss": 2.9577,
       "step": 1863
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 13.246243476867676,
-      "learning_rate": 1.626541044402125e-05,
-      "loss": 2.2984,
+      "epoch": 0.23,
+      "grad_norm": 17.957332611083984,
+      "learning_rate": 1.844119984939129e-05,
+      "loss": 2.2535,
       "step": 1864
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 15.933266639709473,
-      "learning_rate": 1.626340583341686e-05,
-      "loss": 2.6167,
+      "epoch": 0.23,
+      "grad_norm": 23.276569366455078,
+      "learning_rate": 1.8440363134334603e-05,
+      "loss": 3.8149,
       "step": 1865
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 10.605395317077637,
-      "learning_rate": 1.626140122281247e-05,
-      "loss": 2.094,
+      "epoch": 0.23,
+      "grad_norm": 8.072980880737305,
+      "learning_rate": 1.8439526419277917e-05,
+      "loss": 1.0882,
       "step": 1866
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 14.293509483337402,
-      "learning_rate": 1.625939661220808e-05,
-      "loss": 2.4193,
+      "epoch": 0.23,
+      "grad_norm": 15.412882804870605,
+      "learning_rate": 1.8438689704221227e-05,
+      "loss": 4.7679,
       "step": 1867
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 14.355191230773926,
-      "learning_rate": 1.625739200160369e-05,
-      "loss": 3.3217,
+      "epoch": 0.23,
+      "grad_norm": 24.126787185668945,
+      "learning_rate": 1.843785298916454e-05,
+      "loss": 2.9979,
       "step": 1868
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 11.297273635864258,
-      "learning_rate": 1.62553873909993e-05,
-      "loss": 1.9198,
+      "epoch": 0.23,
+      "grad_norm": 11.626553535461426,
+      "learning_rate": 1.8437016274107855e-05,
+      "loss": 2.255,
       "step": 1869
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 19.611162185668945,
-      "learning_rate": 1.625338278039491e-05,
-      "loss": 2.3439,
+      "epoch": 0.23,
+      "grad_norm": 12.555859565734863,
+      "learning_rate": 1.8436179559051165e-05,
+      "loss": 3.1869,
       "step": 1870
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 9.530548095703125,
-      "learning_rate": 1.625137816979052e-05,
-      "loss": 1.1765,
+      "epoch": 0.23,
+      "grad_norm": 13.366609573364258,
+      "learning_rate": 1.843534284399448e-05,
+      "loss": 2.9125,
       "step": 1871
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 20.621294021606445,
-      "learning_rate": 1.624937355918613e-05,
-      "loss": 2.2362,
+      "epoch": 0.23,
+      "grad_norm": 16.570537567138672,
+      "learning_rate": 1.8434506128937792e-05,
+      "loss": 3.2158,
       "step": 1872
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 20.514188766479492,
-      "learning_rate": 1.6247368948581737e-05,
-      "loss": 3.2915,
+      "epoch": 0.24,
+      "grad_norm": 20.994041442871094,
+      "learning_rate": 1.8433669413881102e-05,
+      "loss": 1.722,
       "step": 1873
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 11.529969215393066,
-      "learning_rate": 1.624536433797735e-05,
-      "loss": 1.7275,
+      "epoch": 0.24,
+      "grad_norm": 11.375537872314453,
+      "learning_rate": 1.8432832698824416e-05,
+      "loss": 1.4647,
       "step": 1874
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 15.720181465148926,
-      "learning_rate": 1.6243359727372957e-05,
-      "loss": 2.0002,
+      "epoch": 0.24,
+      "grad_norm": 27.82518196105957,
+      "learning_rate": 1.843199598376773e-05,
+      "loss": 3.7956,
       "step": 1875
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 20.808303833007812,
-      "learning_rate": 1.6241355116768567e-05,
-      "loss": 3.8721,
+      "epoch": 0.24,
+      "grad_norm": 15.789790153503418,
+      "learning_rate": 1.8431159268711043e-05,
+      "loss": 0.1957,
       "step": 1876
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 11.557236671447754,
-      "learning_rate": 1.623935050616418e-05,
-      "loss": 1.6353,
+      "epoch": 0.24,
+      "grad_norm": 12.570982933044434,
+      "learning_rate": 1.8430322553654354e-05,
+      "loss": 2.4985,
       "step": 1877
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 13.909945487976074,
-      "learning_rate": 1.6237345895559788e-05,
-      "loss": 2.2762,
+      "epoch": 0.24,
+      "grad_norm": 13.817831993103027,
+      "learning_rate": 1.8429485838597667e-05,
+      "loss": 1.6574,
       "step": 1878
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 36.823299407958984,
-      "learning_rate": 1.6235341284955398e-05,
-      "loss": 2.6792,
+      "epoch": 0.24,
+      "grad_norm": 18.990236282348633,
+      "learning_rate": 1.842864912354098e-05,
+      "loss": 4.569,
       "step": 1879
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 73.27782440185547,
-      "learning_rate": 1.6233336674351008e-05,
-      "loss": 3.6506,
+      "epoch": 0.24,
+      "grad_norm": 9.704483032226562,
+      "learning_rate": 1.842781240848429e-05,
+      "loss": 1.0726,
       "step": 1880
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 15.239794731140137,
-      "learning_rate": 1.6231332063746618e-05,
-      "loss": 2.5237,
+      "epoch": 0.24,
+      "grad_norm": 16.10625648498535,
+      "learning_rate": 1.8426975693427605e-05,
+      "loss": 5.0485,
       "step": 1881
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 28.48020362854004,
-      "learning_rate": 1.6229327453142228e-05,
-      "loss": 2.8108,
+      "epoch": 0.24,
+      "grad_norm": 10.295207977294922,
+      "learning_rate": 1.842613897837092e-05,
+      "loss": 2.6966,
       "step": 1882
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 9.497574806213379,
-      "learning_rate": 1.6227322842537838e-05,
-      "loss": 1.4183,
+      "epoch": 0.24,
+      "grad_norm": 11.28648567199707,
+      "learning_rate": 1.8425302263314232e-05,
+      "loss": 3.5715,
       "step": 1883
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 12.64692211151123,
-      "learning_rate": 1.6225318231933448e-05,
-      "loss": 2.5578,
+      "epoch": 0.24,
+      "grad_norm": 6.4751482009887695,
+      "learning_rate": 1.8424465548257542e-05,
+      "loss": 1.2767,
       "step": 1884
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 23.391645431518555,
-      "learning_rate": 1.6223313621329058e-05,
-      "loss": 2.03,
+      "epoch": 0.24,
+      "grad_norm": 11.318130493164062,
+      "learning_rate": 1.8423628833200856e-05,
+      "loss": 2.1673,
       "step": 1885
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 25.81462860107422,
-      "learning_rate": 1.622130901072467e-05,
-      "loss": 2.3658,
+      "epoch": 0.24,
+      "grad_norm": 18.204483032226562,
+      "learning_rate": 1.842279211814417e-05,
+      "loss": 1.1198,
       "step": 1886
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 22.7410831451416,
-      "learning_rate": 1.621930440012028e-05,
-      "loss": 2.291,
+      "epoch": 0.24,
+      "grad_norm": 4.5781073570251465,
+      "learning_rate": 1.842195540308748e-05,
+      "loss": 0.1754,
       "step": 1887
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 14.532958030700684,
-      "learning_rate": 1.621729978951589e-05,
-      "loss": 2.3839,
+      "epoch": 0.24,
+      "grad_norm": 9.175607681274414,
+      "learning_rate": 1.8421118688030794e-05,
+      "loss": 0.4124,
       "step": 1888
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 9.742128372192383,
-      "learning_rate": 1.62152951789115e-05,
-      "loss": 1.9635,
+      "epoch": 0.24,
+      "grad_norm": 22.146230697631836,
+      "learning_rate": 1.8420281972974107e-05,
+      "loss": 2.7644,
       "step": 1889
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 11.445870399475098,
-      "learning_rate": 1.621329056830711e-05,
-      "loss": 2.448,
+      "epoch": 0.24,
+      "grad_norm": 9.474859237670898,
+      "learning_rate": 1.8419445257917418e-05,
+      "loss": 3.5436,
       "step": 1890
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 17.39885902404785,
-      "learning_rate": 1.621128595770272e-05,
-      "loss": 1.9545,
+      "epoch": 0.24,
+      "grad_norm": 18.500133514404297,
+      "learning_rate": 1.841860854286073e-05,
+      "loss": 1.9568,
       "step": 1891
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 11.735391616821289,
-      "learning_rate": 1.6209281347098325e-05,
-      "loss": 2.49,
+      "epoch": 0.24,
+      "grad_norm": 15.487665176391602,
+      "learning_rate": 1.841777182780404e-05,
+      "loss": 2.2543,
       "step": 1892
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 31.197790145874023,
-      "learning_rate": 1.620727673649394e-05,
-      "loss": 2.4989,
+      "epoch": 0.24,
+      "grad_norm": 35.86439514160156,
+      "learning_rate": 1.8416935112747355e-05,
+      "loss": 3.3658,
       "step": 1893
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 28.4355525970459,
-      "learning_rate": 1.6205272125889546e-05,
-      "loss": 2.9879,
+      "epoch": 0.24,
+      "grad_norm": 11.434447288513184,
+      "learning_rate": 1.841609839769067e-05,
+      "loss": 1.0026,
       "step": 1894
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 21.622291564941406,
-      "learning_rate": 1.6203267515285156e-05,
-      "loss": 3.1001,
+      "epoch": 0.24,
+      "grad_norm": 19.575517654418945,
+      "learning_rate": 1.841526168263398e-05,
+      "loss": 1.4819,
       "step": 1895
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 15.58934497833252,
-      "learning_rate": 1.620126290468077e-05,
-      "loss": 2.0499,
+      "epoch": 0.24,
+      "grad_norm": 17.874048233032227,
+      "learning_rate": 1.8414424967577293e-05,
+      "loss": 1.6523,
       "step": 1896
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 16.792016983032227,
-      "learning_rate": 1.6199258294076376e-05,
-      "loss": 2.5507,
+      "epoch": 0.24,
+      "grad_norm": 14.109743118286133,
+      "learning_rate": 1.8413588252520606e-05,
+      "loss": 3.1316,
       "step": 1897
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 15.837790489196777,
-      "learning_rate": 1.6197253683471986e-05,
-      "loss": 2.5776,
+      "epoch": 0.24,
+      "grad_norm": 14.72187614440918,
+      "learning_rate": 1.8412751537463917e-05,
+      "loss": 3.0954,
       "step": 1898
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 13.746726989746094,
-      "learning_rate": 1.6195249072867596e-05,
-      "loss": 1.7229,
+      "epoch": 0.24,
+      "grad_norm": 49.131553649902344,
+      "learning_rate": 1.841191482240723e-05,
+      "loss": 2.0844,
       "step": 1899
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 20.217098236083984,
-      "learning_rate": 1.6193244462263206e-05,
-      "loss": 2.7257,
+      "epoch": 0.24,
+      "grad_norm": 13.758391380310059,
+      "learning_rate": 1.8411078107350544e-05,
+      "loss": 4.0891,
       "step": 1900
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 14.193788528442383,
-      "learning_rate": 1.6191239851658816e-05,
-      "loss": 2.4532,
+      "epoch": 0.24,
+      "grad_norm": 11.789108276367188,
+      "learning_rate": 1.8410241392293854e-05,
+      "loss": 1.2636,
       "step": 1901
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 12.840964317321777,
-      "learning_rate": 1.6189235241054426e-05,
-      "loss": 1.3137,
+      "epoch": 0.24,
+      "grad_norm": 25.986669540405273,
+      "learning_rate": 1.8409404677237168e-05,
+      "loss": 2.2081,
       "step": 1902
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 10.857292175292969,
-      "learning_rate": 1.6187230630450036e-05,
-      "loss": 2.3681,
+      "epoch": 0.24,
+      "grad_norm": 15.064836502075195,
+      "learning_rate": 1.840856796218048e-05,
+      "loss": 2.0525,
       "step": 1903
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 11.725610733032227,
-      "learning_rate": 1.6185226019845646e-05,
-      "loss": 2.0677,
+      "epoch": 0.24,
+      "grad_norm": 7.967344760894775,
+      "learning_rate": 1.8407731247123795e-05,
+      "loss": 1.1078,
       "step": 1904
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 16.389385223388672,
-      "learning_rate": 1.6183221409241257e-05,
-      "loss": 2.2652,
+      "epoch": 0.24,
+      "grad_norm": 30.750852584838867,
+      "learning_rate": 1.8406894532067105e-05,
+      "loss": 1.8963,
       "step": 1905
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 20.219581604003906,
-      "learning_rate": 1.6181216798636867e-05,
-      "loss": 2.3222,
+      "epoch": 0.24,
+      "grad_norm": 7.841359615325928,
+      "learning_rate": 1.840605781701042e-05,
+      "loss": 2.8736,
       "step": 1906
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 15.775784492492676,
-      "learning_rate": 1.6179212188032477e-05,
-      "loss": 2.0435,
+      "epoch": 0.24,
+      "grad_norm": 14.268943786621094,
+      "learning_rate": 1.8405221101953733e-05,
+      "loss": 2.2186,
       "step": 1907
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 11.360732078552246,
-      "learning_rate": 1.6177207577428087e-05,
-      "loss": 2.0024,
+      "epoch": 0.24,
+      "grad_norm": 22.337039947509766,
+      "learning_rate": 1.8404384386897043e-05,
+      "loss": 2.3814,
       "step": 1908
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 26.330806732177734,
-      "learning_rate": 1.6175202966823697e-05,
-      "loss": 3.1264,
+      "epoch": 0.24,
+      "grad_norm": 15.489692687988281,
+      "learning_rate": 1.8403547671840357e-05,
+      "loss": 2.0205,
       "step": 1909
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 16.440622329711914,
-      "learning_rate": 1.6173198356219307e-05,
-      "loss": 2.6761,
+      "epoch": 0.24,
+      "grad_norm": 14.391831398010254,
+      "learning_rate": 1.840271095678367e-05,
+      "loss": 3.1901,
       "step": 1910
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 14.55276107788086,
-      "learning_rate": 1.6171193745614914e-05,
-      "loss": 2.2345,
+      "epoch": 0.24,
+      "grad_norm": 11.057231903076172,
+      "learning_rate": 1.8401874241726984e-05,
+      "loss": 3.2508,
       "step": 1911
     },
     {
-      "epoch": 0.57,
-      "grad_norm": 15.58680534362793,
-      "learning_rate": 1.6169189135010527e-05,
-      "loss": 2.8794,
+      "epoch": 0.24,
+      "grad_norm": 18.586143493652344,
+      "learning_rate": 1.8401037526670294e-05,
+      "loss": 2.3013,
       "step": 1912
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 27.979393005371094,
-      "learning_rate": 1.6167184524406137e-05,
-      "loss": 2.5893,
+      "epoch": 0.24,
+      "grad_norm": 8.885278701782227,
+      "learning_rate": 1.8400200811613608e-05,
+      "loss": 1.8043,
       "step": 1913
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 14.484015464782715,
-      "learning_rate": 1.6165179913801744e-05,
-      "loss": 2.4981,
+      "epoch": 0.24,
+      "grad_norm": 21.466678619384766,
+      "learning_rate": 1.839936409655692e-05,
+      "loss": 2.3803,
       "step": 1914
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 9.881906509399414,
-      "learning_rate": 1.6163175303197357e-05,
-      "loss": 2.1874,
+      "epoch": 0.24,
+      "grad_norm": 26.071767807006836,
+      "learning_rate": 1.839852738150023e-05,
+      "loss": 3.6033,
       "step": 1915
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 17.09141731262207,
-      "learning_rate": 1.6161170692592964e-05,
-      "loss": 2.9159,
+      "epoch": 0.24,
+      "grad_norm": 19.87377166748047,
+      "learning_rate": 1.8397690666443545e-05,
+      "loss": 2.1179,
       "step": 1916
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 18.972490310668945,
-      "learning_rate": 1.6159166081988574e-05,
-      "loss": 1.6856,
+      "epoch": 0.24,
+      "grad_norm": 15.754097938537598,
+      "learning_rate": 1.8396853951386856e-05,
+      "loss": 4.6128,
       "step": 1917
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 10.02529525756836,
-      "learning_rate": 1.6157161471384184e-05,
-      "loss": 1.7176,
+      "epoch": 0.24,
+      "grad_norm": 17.967514038085938,
+      "learning_rate": 1.839601723633017e-05,
+      "loss": 2.6943,
       "step": 1918
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 21.166505813598633,
-      "learning_rate": 1.6155156860779794e-05,
-      "loss": 2.5582,
+      "epoch": 0.24,
+      "grad_norm": 14.497222900390625,
+      "learning_rate": 1.8395180521273483e-05,
+      "loss": 3.1712,
       "step": 1919
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 34.15064239501953,
-      "learning_rate": 1.6153152250175404e-05,
-      "loss": 4.2157,
-      "step": 1920
-    },
-    {
-      "epoch": 0.58,
-      "eval_loss": 0.339510053396225,
-      "eval_runtime": 43.4353,
-      "eval_samples_per_second": 34.051,
-      "eval_steps_per_second": 34.051,
+      "epoch": 0.24,
+      "grad_norm": 17.02864646911621,
+      "learning_rate": 1.8394343806216793e-05,
+      "loss": 2.9962,
       "step": 1920
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 18.826351165771484,
-      "learning_rate": 1.6151147639571015e-05,
-      "loss": 2.8172,
+      "epoch": 0.24,
+      "grad_norm": 20.203519821166992,
+      "learning_rate": 1.8393507091160107e-05,
+      "loss": 2.0085,
       "step": 1921
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 28.641447067260742,
-      "learning_rate": 1.6149143028966625e-05,
-      "loss": 3.8235,
+      "epoch": 0.24,
+      "grad_norm": 11.49713134765625,
+      "learning_rate": 1.8392670376103417e-05,
+      "loss": 3.9497,
       "step": 1922
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 18.224559783935547,
-      "learning_rate": 1.6147138418362235e-05,
-      "loss": 1.6789,
+      "epoch": 0.24,
+      "grad_norm": 32.90269088745117,
+      "learning_rate": 1.839183366104673e-05,
+      "loss": 2.3974,
       "step": 1923
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 20.033323287963867,
-      "learning_rate": 1.6145133807757845e-05,
-      "loss": 2.2715,
+      "epoch": 0.24,
+      "grad_norm": 18.34015464782715,
+      "learning_rate": 1.8390996945990044e-05,
+      "loss": 3.679,
       "step": 1924
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 18.451032638549805,
-      "learning_rate": 1.6143129197153455e-05,
-      "loss": 3.6021,
+      "epoch": 0.24,
+      "grad_norm": 13.834124565124512,
+      "learning_rate": 1.8390160230933358e-05,
+      "loss": 1.8375,
       "step": 1925
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 13.090971946716309,
-      "learning_rate": 1.6141124586549065e-05,
-      "loss": 2.8423,
+      "epoch": 0.24,
+      "grad_norm": 10.36095905303955,
+      "learning_rate": 1.8389323515876668e-05,
+      "loss": 1.666,
       "step": 1926
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 31.57698631286621,
-      "learning_rate": 1.6139119975944675e-05,
-      "loss": 2.1646,
+      "epoch": 0.24,
+      "grad_norm": 11.952203750610352,
+      "learning_rate": 1.8388486800819982e-05,
+      "loss": 3.0777,
       "step": 1927
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 19.8615779876709,
-      "learning_rate": 1.6137115365340285e-05,
-      "loss": 2.4651,
+      "epoch": 0.24,
+      "grad_norm": 14.914665222167969,
+      "learning_rate": 1.8387650085763296e-05,
+      "loss": 3.1809,
       "step": 1928
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 17.440027236938477,
-      "learning_rate": 1.6135110754735895e-05,
-      "loss": 2.2323,
+      "epoch": 0.24,
+      "grad_norm": 26.21071434020996,
+      "learning_rate": 1.8386813370706606e-05,
+      "loss": 3.7649,
       "step": 1929
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 12.359676361083984,
-      "learning_rate": 1.6133106144131502e-05,
-      "loss": 3.0975,
+      "epoch": 0.24,
+      "grad_norm": 12.51376724243164,
+      "learning_rate": 1.838597665564992e-05,
+      "loss": 1.9043,
       "step": 1930
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 7.871943473815918,
-      "learning_rate": 1.6131101533527115e-05,
-      "loss": 1.8966,
+      "epoch": 0.24,
+      "grad_norm": 11.830827713012695,
+      "learning_rate": 1.8385139940593233e-05,
+      "loss": 1.3338,
       "step": 1931
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 13.588000297546387,
-      "learning_rate": 1.6129096922922725e-05,
-      "loss": 2.7859,
+      "epoch": 0.24,
+      "grad_norm": 24.52362632751465,
+      "learning_rate": 1.8384303225536547e-05,
+      "loss": 1.8047,
       "step": 1932
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 16.730295181274414,
-      "learning_rate": 1.6127092312318332e-05,
-      "loss": 2.4625,
+      "epoch": 0.24,
+      "grad_norm": 12.227885246276855,
+      "learning_rate": 1.8383466510479857e-05,
+      "loss": 3.0994,
       "step": 1933
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 13.648134231567383,
-      "learning_rate": 1.6125087701713946e-05,
-      "loss": 2.5004,
+      "epoch": 0.24,
+      "grad_norm": 12.330459594726562,
+      "learning_rate": 1.838262979542317e-05,
+      "loss": 2.7654,
       "step": 1934
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 19.520750045776367,
-      "learning_rate": 1.6123083091109552e-05,
-      "loss": 3.5212,
+      "epoch": 0.24,
+      "grad_norm": 18.657487869262695,
+      "learning_rate": 1.8381793080366484e-05,
+      "loss": 3.6686,
       "step": 1935
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 19.73345184326172,
-      "learning_rate": 1.6121078480505162e-05,
-      "loss": 2.2163,
+      "epoch": 0.24,
+      "grad_norm": 14.701178550720215,
+      "learning_rate": 1.8380956365309795e-05,
+      "loss": 2.7045,
       "step": 1936
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 11.382865905761719,
-      "learning_rate": 1.6119073869900772e-05,
-      "loss": 1.5394,
+      "epoch": 0.24,
+      "grad_norm": 21.401046752929688,
+      "learning_rate": 1.8380119650253108e-05,
+      "loss": 2.1688,
       "step": 1937
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 11.32697868347168,
-      "learning_rate": 1.6117069259296383e-05,
-      "loss": 1.7365,
+      "epoch": 0.24,
+      "grad_norm": 11.859993934631348,
+      "learning_rate": 1.8379282935196422e-05,
+      "loss": 2.4173,
       "step": 1938
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 17.160417556762695,
-      "learning_rate": 1.6115064648691993e-05,
-      "loss": 3.0884,
+      "epoch": 0.24,
+      "grad_norm": 12.437385559082031,
+      "learning_rate": 1.8378446220139735e-05,
+      "loss": 3.579,
       "step": 1939
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 10.711620330810547,
-      "learning_rate": 1.6113060038087603e-05,
-      "loss": 2.0598,
+      "epoch": 0.24,
+      "grad_norm": 22.76668357849121,
+      "learning_rate": 1.8377609505083046e-05,
+      "loss": 4.4727,
       "step": 1940
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 15.657376289367676,
-      "learning_rate": 1.6111055427483213e-05,
-      "loss": 2.5557,
+      "epoch": 0.24,
+      "grad_norm": 23.67467498779297,
+      "learning_rate": 1.837677279002636e-05,
+      "loss": 3.4649,
       "step": 1941
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 17.509824752807617,
-      "learning_rate": 1.6109050816878823e-05,
-      "loss": 2.2766,
+      "epoch": 0.24,
+      "grad_norm": 15.801477432250977,
+      "learning_rate": 1.8375936074969673e-05,
+      "loss": 4.4846,
       "step": 1942
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 10.339923858642578,
-      "learning_rate": 1.6107046206274433e-05,
-      "loss": 1.8058,
+      "epoch": 0.24,
+      "grad_norm": 15.054116249084473,
+      "learning_rate": 1.8375099359912983e-05,
+      "loss": 2.2226,
       "step": 1943
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 16.32088851928711,
-      "learning_rate": 1.6105041595670043e-05,
-      "loss": 2.2214,
+      "epoch": 0.24,
+      "grad_norm": 11.429304122924805,
+      "learning_rate": 1.8374262644856297e-05,
+      "loss": 1.7485,
       "step": 1944
     },
     {
-      "epoch": 0.58,
-      "grad_norm": 34.30913543701172,
-      "learning_rate": 1.6103036985065653e-05,
-      "loss": 3.2443,
+      "epoch": 0.24,
+      "grad_norm": 18.58489227294922,
+      "learning_rate": 1.8373425929799607e-05,
+      "loss": 4.9844,
       "step": 1945
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 13.773527145385742,
-      "learning_rate": 1.6101032374461263e-05,
-      "loss": 3.0382,
+      "epoch": 0.24,
+      "grad_norm": 12.300481796264648,
+      "learning_rate": 1.837258921474292e-05,
+      "loss": 2.1001,
       "step": 1946
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 23.005949020385742,
-      "learning_rate": 1.609902776385687e-05,
-      "loss": 2.0551,
+      "epoch": 0.24,
+      "grad_norm": 12.821094512939453,
+      "learning_rate": 1.8371752499686235e-05,
+      "loss": 2.51,
       "step": 1947
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 32.260467529296875,
-      "learning_rate": 1.6097023153252483e-05,
-      "loss": 2.4883,
+      "epoch": 0.24,
+      "grad_norm": 10.40767765045166,
+      "learning_rate": 1.8370915784629545e-05,
+      "loss": 3.1537,
       "step": 1948
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 13.194154739379883,
-      "learning_rate": 1.609501854264809e-05,
-      "loss": 2.7221,
+      "epoch": 0.24,
+      "grad_norm": 35.25419616699219,
+      "learning_rate": 1.837007906957286e-05,
+      "loss": 3.2483,
       "step": 1949
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 29.038406372070312,
-      "learning_rate": 1.60930139320437e-05,
-      "loss": 2.7329,
+      "epoch": 0.24,
+      "grad_norm": 24.408248901367188,
+      "learning_rate": 1.836924235451617e-05,
+      "loss": 2.5151,
       "step": 1950
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 15.281466484069824,
-      "learning_rate": 1.6091009321439314e-05,
-      "loss": 2.5336,
+      "epoch": 0.24,
+      "grad_norm": 15.579033851623535,
+      "learning_rate": 1.8368405639459482e-05,
+      "loss": 3.2173,
       "step": 1951
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 14.665143013000488,
-      "learning_rate": 1.608900471083492e-05,
-      "loss": 1.9784,
+      "epoch": 0.24,
+      "grad_norm": 9.459357261657715,
+      "learning_rate": 1.8367568924402796e-05,
+      "loss": 3.3478,
       "step": 1952
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 17.575735092163086,
-      "learning_rate": 1.608700010023053e-05,
-      "loss": 1.5022,
+      "epoch": 0.25,
+      "grad_norm": 13.950507164001465,
+      "learning_rate": 1.836673220934611e-05,
+      "loss": 2.4268,
       "step": 1953
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 16.851743698120117,
-      "learning_rate": 1.608499548962614e-05,
-      "loss": 2.1047,
+      "epoch": 0.25,
+      "grad_norm": 16.881439208984375,
+      "learning_rate": 1.836589549428942e-05,
+      "loss": 1.0458,
       "step": 1954
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 17.620439529418945,
-      "learning_rate": 1.608299087902175e-05,
-      "loss": 2.4975,
+      "epoch": 0.25,
+      "grad_norm": 23.890779495239258,
+      "learning_rate": 1.8365058779232734e-05,
+      "loss": 3.3657,
       "step": 1955
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 18.211750030517578,
-      "learning_rate": 1.608098626841736e-05,
-      "loss": 2.441,
+      "epoch": 0.25,
+      "grad_norm": 11.523681640625,
+      "learning_rate": 1.8364222064176047e-05,
+      "loss": 1.7303,
       "step": 1956
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 38.04513168334961,
-      "learning_rate": 1.607898165781297e-05,
-      "loss": 3.2777,
+      "epoch": 0.25,
+      "grad_norm": 16.36073875427246,
+      "learning_rate": 1.8363385349119357e-05,
+      "loss": 1.4092,
       "step": 1957
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 18.044525146484375,
-      "learning_rate": 1.607697704720858e-05,
-      "loss": 3.4203,
+      "epoch": 0.25,
+      "grad_norm": 14.88504409790039,
+      "learning_rate": 1.836254863406267e-05,
+      "loss": 1.4235,
       "step": 1958
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 14.710494995117188,
-      "learning_rate": 1.607497243660419e-05,
-      "loss": 2.0862,
+      "epoch": 0.25,
+      "grad_norm": 30.882617950439453,
+      "learning_rate": 1.8361711919005985e-05,
+      "loss": 1.8152,
       "step": 1959
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 10.158746719360352,
-      "learning_rate": 1.60729678259998e-05,
-      "loss": 1.8249,
+      "epoch": 0.25,
+      "grad_norm": 8.716890335083008,
+      "learning_rate": 1.83608752039493e-05,
+      "loss": 1.9457,
       "step": 1960
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 14.311506271362305,
-      "learning_rate": 1.607096321539541e-05,
-      "loss": 2.0706,
+      "epoch": 0.25,
+      "grad_norm": 37.496131896972656,
+      "learning_rate": 1.836003848889261e-05,
+      "loss": 4.3405,
       "step": 1961
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 9.432246208190918,
-      "learning_rate": 1.606895860479102e-05,
-      "loss": 2.0623,
+      "epoch": 0.25,
+      "grad_norm": 7.394501686096191,
+      "learning_rate": 1.8359201773835922e-05,
+      "loss": 1.7728,
       "step": 1962
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 17.04642105102539,
-      "learning_rate": 1.606695399418663e-05,
-      "loss": 2.918,
+      "epoch": 0.25,
+      "grad_norm": 15.787137031555176,
+      "learning_rate": 1.8358365058779236e-05,
+      "loss": 3.8866,
       "step": 1963
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 11.562948226928711,
-      "learning_rate": 1.606494938358224e-05,
-      "loss": 2.5815,
+      "epoch": 0.25,
+      "grad_norm": 13.577347755432129,
+      "learning_rate": 1.8357528343722546e-05,
+      "loss": 2.8812,
       "step": 1964
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 18.638172149658203,
-      "learning_rate": 1.606294477297785e-05,
-      "loss": 2.995,
+      "epoch": 0.25,
+      "grad_norm": 14.8187837600708,
+      "learning_rate": 1.835669162866586e-05,
+      "loss": 3.5142,
       "step": 1965
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 14.144832611083984,
-      "learning_rate": 1.6060940162373458e-05,
-      "loss": 2.1113,
+      "epoch": 0.25,
+      "grad_norm": 9.960477828979492,
+      "learning_rate": 1.8355854913609174e-05,
+      "loss": 3.0019,
       "step": 1966
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 16.268152236938477,
-      "learning_rate": 1.605893555176907e-05,
-      "loss": 2.2524,
+      "epoch": 0.25,
+      "grad_norm": 10.168802261352539,
+      "learning_rate": 1.8355018198552484e-05,
+      "loss": 2.6793,
       "step": 1967
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 16.423564910888672,
-      "learning_rate": 1.6056930941164678e-05,
-      "loss": 2.2185,
+      "epoch": 0.25,
+      "grad_norm": 12.786600112915039,
+      "learning_rate": 1.8354181483495797e-05,
+      "loss": 0.4634,
       "step": 1968
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 28.028411865234375,
-      "learning_rate": 1.605492633056029e-05,
-      "loss": 2.556,
+      "epoch": 0.25,
+      "grad_norm": 10.668661117553711,
+      "learning_rate": 1.835334476843911e-05,
+      "loss": 2.2684,
       "step": 1969
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 21.24933624267578,
-      "learning_rate": 1.6052921719955902e-05,
-      "loss": 2.3257,
+      "epoch": 0.25,
+      "grad_norm": 20.51132583618164,
+      "learning_rate": 1.835250805338242e-05,
+      "loss": 2.2883,
       "step": 1970
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 15.96957778930664,
-      "learning_rate": 1.605091710935151e-05,
-      "loss": 2.617,
+      "epoch": 0.25,
+      "grad_norm": 13.205888748168945,
+      "learning_rate": 1.8351671338325735e-05,
+      "loss": 2.5053,
       "step": 1971
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 10.580945014953613,
-      "learning_rate": 1.604891249874712e-05,
-      "loss": 1.4834,
+      "epoch": 0.25,
+      "grad_norm": 17.55050277709961,
+      "learning_rate": 1.835083462326905e-05,
+      "loss": 2.3348,
       "step": 1972
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 12.444055557250977,
-      "learning_rate": 1.604690788814273e-05,
-      "loss": 2.222,
+      "epoch": 0.25,
+      "grad_norm": 16.795305252075195,
+      "learning_rate": 1.834999790821236e-05,
+      "loss": 2.2641,
       "step": 1973
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 12.718807220458984,
-      "learning_rate": 1.604490327753834e-05,
-      "loss": 1.986,
+      "epoch": 0.25,
+      "grad_norm": 9.307861328125,
+      "learning_rate": 1.8349161193155673e-05,
+      "loss": 1.1856,
       "step": 1974
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 11.073088645935059,
-      "learning_rate": 1.604289866693395e-05,
-      "loss": 2.3385,
+      "epoch": 0.25,
+      "grad_norm": 18.906524658203125,
+      "learning_rate": 1.8348324478098983e-05,
+      "loss": 2.9946,
       "step": 1975
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 22.884178161621094,
-      "learning_rate": 1.604089405632956e-05,
-      "loss": 3.075,
+      "epoch": 0.25,
+      "grad_norm": 12.760396957397461,
+      "learning_rate": 1.8347487763042296e-05,
+      "loss": 0.9203,
       "step": 1976
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 13.408631324768066,
-      "learning_rate": 1.603888944572517e-05,
-      "loss": 1.867,
+      "epoch": 0.25,
+      "grad_norm": 8.600692749023438,
+      "learning_rate": 1.834665104798561e-05,
+      "loss": 1.4238,
       "step": 1977
     },
     {
-      "epoch": 0.59,
-      "grad_norm": 20.047677993774414,
-      "learning_rate": 1.603688483512078e-05,
-      "loss": 2.3888,
+      "epoch": 0.25,
+      "grad_norm": 13.111052513122559,
+      "learning_rate": 1.834581433292892e-05,
+      "loss": 1.2993,
       "step": 1978
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 56.602203369140625,
-      "learning_rate": 1.603488022451639e-05,
-      "loss": 2.9184,
+      "epoch": 0.25,
+      "grad_norm": 10.117164611816406,
+      "learning_rate": 1.8344977617872234e-05,
+      "loss": 2.9947,
       "step": 1979
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 23.749616622924805,
-      "learning_rate": 1.6032875613912e-05,
-      "loss": 2.7749,
+      "epoch": 0.25,
+      "grad_norm": 18.097362518310547,
+      "learning_rate": 1.8344140902815548e-05,
+      "loss": 2.3721,
       "step": 1980
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 26.525043487548828,
-      "learning_rate": 1.603087100330761e-05,
-      "loss": 2.1839,
+      "epoch": 0.25,
+      "grad_norm": 26.985071182250977,
+      "learning_rate": 1.8343304187758858e-05,
+      "loss": 2.5174,
       "step": 1981
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 32.17291259765625,
-      "learning_rate": 1.602886639270322e-05,
-      "loss": 2.7472,
+      "epoch": 0.25,
+      "grad_norm": 13.80362606048584,
+      "learning_rate": 1.834246747270217e-05,
+      "loss": 2.1506,
       "step": 1982
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 13.829222679138184,
-      "learning_rate": 1.602686178209883e-05,
-      "loss": 2.3262,
+      "epoch": 0.25,
+      "grad_norm": 17.8221492767334,
+      "learning_rate": 1.8341630757645485e-05,
+      "loss": 2.271,
       "step": 1983
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 10.74743366241455,
-      "learning_rate": 1.602485717149444e-05,
-      "loss": 3.0373,
+      "epoch": 0.25,
+      "grad_norm": 13.265314102172852,
+      "learning_rate": 1.83407940425888e-05,
+      "loss": 2.7834,
       "step": 1984
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 19.987829208374023,
-      "learning_rate": 1.6022852560890046e-05,
-      "loss": 2.2646,
+      "epoch": 0.25,
+      "grad_norm": 24.445533752441406,
+      "learning_rate": 1.833995732753211e-05,
+      "loss": 1.5695,
       "step": 1985
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 15.228018760681152,
-      "learning_rate": 1.602084795028566e-05,
-      "loss": 2.8864,
+      "epoch": 0.25,
+      "grad_norm": 9.981922149658203,
+      "learning_rate": 1.8339120612475423e-05,
+      "loss": 1.6212,
       "step": 1986
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 39.494258880615234,
-      "learning_rate": 1.601884333968127e-05,
-      "loss": 3.7293,
+      "epoch": 0.25,
+      "grad_norm": 12.802506446838379,
+      "learning_rate": 1.8338283897418736e-05,
+      "loss": 1.9037,
       "step": 1987
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 21.58100700378418,
-      "learning_rate": 1.6016838729076877e-05,
-      "loss": 3.238,
+      "epoch": 0.25,
+      "grad_norm": 15.808159828186035,
+      "learning_rate": 1.8337447182362047e-05,
+      "loss": 1.8119,
       "step": 1988
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 26.791006088256836,
-      "learning_rate": 1.601483411847249e-05,
-      "loss": 2.7061,
+      "epoch": 0.25,
+      "grad_norm": 15.071918487548828,
+      "learning_rate": 1.833661046730536e-05,
+      "loss": 1.9275,
       "step": 1989
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 26.071443557739258,
-      "learning_rate": 1.6012829507868097e-05,
-      "loss": 2.4725,
+      "epoch": 0.25,
+      "grad_norm": 14.01952838897705,
+      "learning_rate": 1.8335773752248674e-05,
+      "loss": 2.6755,
       "step": 1990
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 14.587154388427734,
-      "learning_rate": 1.6010824897263707e-05,
-      "loss": 2.7513,
+      "epoch": 0.25,
+      "grad_norm": 10.879013061523438,
+      "learning_rate": 1.8334937037191988e-05,
+      "loss": 2.146,
       "step": 1991
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 19.951908111572266,
-      "learning_rate": 1.6008820286659317e-05,
-      "loss": 2.3058,
+      "epoch": 0.25,
+      "grad_norm": 14.858782768249512,
+      "learning_rate": 1.8334100322135298e-05,
+      "loss": 1.8593,
       "step": 1992
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 18.910688400268555,
-      "learning_rate": 1.6006815676054927e-05,
-      "loss": 2.4444,
+      "epoch": 0.25,
+      "grad_norm": 9.486717224121094,
+      "learning_rate": 1.833326360707861e-05,
+      "loss": 1.4072,
       "step": 1993
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 12.591381072998047,
-      "learning_rate": 1.6004811065450537e-05,
-      "loss": 3.0018,
+      "epoch": 0.25,
+      "grad_norm": 15.695525169372559,
+      "learning_rate": 1.8332426892021925e-05,
+      "loss": 1.6865,
       "step": 1994
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 25.672998428344727,
-      "learning_rate": 1.6002806454846147e-05,
-      "loss": 2.6811,
+      "epoch": 0.25,
+      "grad_norm": 10.566913604736328,
+      "learning_rate": 1.8331590176965235e-05,
+      "loss": 4.4794,
       "step": 1995
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 24.982601165771484,
-      "learning_rate": 1.6000801844241757e-05,
-      "loss": 3.1955,
+      "epoch": 0.25,
+      "grad_norm": 11.069722175598145,
+      "learning_rate": 1.833075346190855e-05,
+      "loss": 2.5513,
       "step": 1996
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 12.664204597473145,
-      "learning_rate": 1.5998797233637367e-05,
-      "loss": 2.2998,
+      "epoch": 0.25,
+      "grad_norm": 7.837258338928223,
+      "learning_rate": 1.8329916746851863e-05,
+      "loss": 2.3469,
       "step": 1997
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 10.164942741394043,
-      "learning_rate": 1.5996792623032977e-05,
-      "loss": 3.4713,
+      "epoch": 0.25,
+      "grad_norm": 11.02144718170166,
+      "learning_rate": 1.8329080031795173e-05,
+      "loss": 2.1457,
       "step": 1998
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 11.533797264099121,
-      "learning_rate": 1.5994788012428588e-05,
-      "loss": 2.7492,
+      "epoch": 0.25,
+      "grad_norm": 12.016775131225586,
+      "learning_rate": 1.8328243316738487e-05,
+      "loss": 2.4249,
       "step": 1999
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 31.470319747924805,
-      "learning_rate": 1.5992783401824198e-05,
-      "loss": 3.0072,
+      "epoch": 0.25,
+      "grad_norm": 13.773508071899414,
+      "learning_rate": 1.83274066016818e-05,
+      "loss": 3.2491,
       "step": 2000
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 17.791658401489258,
-      "learning_rate": 1.5990778791219808e-05,
-      "loss": 1.9317,
+      "epoch": 0.25,
+      "eval_loss": 0.17568634450435638,
+      "eval_runtime": 93.8646,
+      "eval_samples_per_second": 37.735,
+      "eval_steps_per_second": 37.735,
+      "step": 2000
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 28.11022186279297,
+      "learning_rate": 1.832656988662511e-05,
+      "loss": 3.2889,
       "step": 2001
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 31.150108337402344,
-      "learning_rate": 1.5988774180615418e-05,
-      "loss": 2.8161,
+      "epoch": 0.25,
+      "grad_norm": 11.31784725189209,
+      "learning_rate": 1.8325733171568424e-05,
+      "loss": 2.3865,
       "step": 2002
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 13.658658027648926,
-      "learning_rate": 1.5986769570011028e-05,
-      "loss": 1.9486,
+      "epoch": 0.25,
+      "grad_norm": 5.304311752319336,
+      "learning_rate": 1.8324896456511734e-05,
+      "loss": 0.5511,
       "step": 2003
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 13.074492454528809,
-      "learning_rate": 1.5984764959406635e-05,
-      "loss": 1.8306,
+      "epoch": 0.25,
+      "grad_norm": 19.37568473815918,
+      "learning_rate": 1.8324059741455048e-05,
+      "loss": 2.0012,
       "step": 2004
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 13.934167861938477,
-      "learning_rate": 1.5982760348802248e-05,
-      "loss": 2.0469,
+      "epoch": 0.25,
+      "grad_norm": 8.72088623046875,
+      "learning_rate": 1.8323223026398362e-05,
+      "loss": 1.6433,
       "step": 2005
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 19.25830078125,
-      "learning_rate": 1.5980755738197858e-05,
-      "loss": 3.0445,
+      "epoch": 0.25,
+      "grad_norm": 21.67353630065918,
+      "learning_rate": 1.8322386311341672e-05,
+      "loss": 2.4568,
       "step": 2006
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 12.957606315612793,
-      "learning_rate": 1.5978751127593465e-05,
-      "loss": 2.2492,
+      "epoch": 0.25,
+      "grad_norm": 16.720203399658203,
+      "learning_rate": 1.8321549596284986e-05,
+      "loss": 2.1404,
       "step": 2007
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 18.43289566040039,
-      "learning_rate": 1.5976746516989078e-05,
-      "loss": 3.2304,
+      "epoch": 0.25,
+      "grad_norm": 13.917689323425293,
+      "learning_rate": 1.83207128812283e-05,
+      "loss": 3.6468,
       "step": 2008
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 15.042205810546875,
-      "learning_rate": 1.5974741906384685e-05,
-      "loss": 2.5782,
+      "epoch": 0.25,
+      "grad_norm": 13.364958763122559,
+      "learning_rate": 1.831987616617161e-05,
+      "loss": 2.2146,
       "step": 2009
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 32.639183044433594,
-      "learning_rate": 1.5972737295780295e-05,
-      "loss": 2.2884,
+      "epoch": 0.25,
+      "grad_norm": 16.78025245666504,
+      "learning_rate": 1.8319039451114923e-05,
+      "loss": 3.4028,
       "step": 2010
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 14.541574478149414,
-      "learning_rate": 1.5970732685175905e-05,
-      "loss": 2.1057,
+      "epoch": 0.25,
+      "grad_norm": 14.87304401397705,
+      "learning_rate": 1.8318202736058237e-05,
+      "loss": 2.1005,
       "step": 2011
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 33.099979400634766,
-      "learning_rate": 1.5968728074571515e-05,
-      "loss": 2.8465,
+      "epoch": 0.25,
+      "grad_norm": 21.702777862548828,
+      "learning_rate": 1.831736602100155e-05,
+      "loss": 2.0948,
       "step": 2012
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 29.527706146240234,
-      "learning_rate": 1.5966723463967125e-05,
-      "loss": 1.8221,
+      "epoch": 0.25,
+      "grad_norm": 17.03411293029785,
+      "learning_rate": 1.831652930594486e-05,
+      "loss": 3.4917,
       "step": 2013
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 12.222841262817383,
-      "learning_rate": 1.5964718853362735e-05,
-      "loss": 1.9562,
+      "epoch": 0.25,
+      "grad_norm": 122.54450225830078,
+      "learning_rate": 1.8315692590888174e-05,
+      "loss": 1.5136,
       "step": 2014
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 11.430624961853027,
-      "learning_rate": 1.5962714242758345e-05,
-      "loss": 1.8148,
+      "epoch": 0.25,
+      "grad_norm": 8.881357192993164,
+      "learning_rate": 1.8314855875831488e-05,
+      "loss": 1.9177,
       "step": 2015
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 12.992792129516602,
-      "learning_rate": 1.5960709632153956e-05,
-      "loss": 2.0858,
+      "epoch": 0.25,
+      "grad_norm": 12.461234092712402,
+      "learning_rate": 1.83140191607748e-05,
+      "loss": 2.3942,
       "step": 2016
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 25.652822494506836,
-      "learning_rate": 1.5958705021549566e-05,
-      "loss": 3.0605,
+      "epoch": 0.25,
+      "grad_norm": 12.239439964294434,
+      "learning_rate": 1.8313182445718112e-05,
+      "loss": 3.2475,
       "step": 2017
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 9.427448272705078,
-      "learning_rate": 1.5956700410945176e-05,
-      "loss": 2.0717,
+      "epoch": 0.25,
+      "grad_norm": 15.35310173034668,
+      "learning_rate": 1.8312345730661426e-05,
+      "loss": 2.0139,
       "step": 2018
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 18.489501953125,
-      "learning_rate": 1.5954695800340786e-05,
-      "loss": 2.3319,
+      "epoch": 0.25,
+      "grad_norm": 11.228468894958496,
+      "learning_rate": 1.831150901560474e-05,
+      "loss": 2.1745,
       "step": 2019
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 14.960678100585938,
-      "learning_rate": 1.5952691189736396e-05,
-      "loss": 2.3096,
+      "epoch": 0.25,
+      "grad_norm": 9.156081199645996,
+      "learning_rate": 1.831067230054805e-05,
+      "loss": 1.7031,
       "step": 2020
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 17.850452423095703,
-      "learning_rate": 1.5950686579132003e-05,
-      "loss": 2.5948,
+      "epoch": 0.25,
+      "grad_norm": 16.42210578918457,
+      "learning_rate": 1.8309835585491363e-05,
+      "loss": 3.2483,
       "step": 2021
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 13.320516586303711,
-      "learning_rate": 1.5948681968527616e-05,
-      "loss": 2.0685,
+      "epoch": 0.25,
+      "grad_norm": 12.818464279174805,
+      "learning_rate": 1.8308998870434677e-05,
+      "loss": 1.0547,
       "step": 2022
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 12.267958641052246,
-      "learning_rate": 1.5946677357923223e-05,
-      "loss": 2.1863,
+      "epoch": 0.25,
+      "grad_norm": 9.632564544677734,
+      "learning_rate": 1.8308162155377987e-05,
+      "loss": 1.0357,
       "step": 2023
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 17.197010040283203,
-      "learning_rate": 1.5944672747318833e-05,
-      "loss": 2.5286,
+      "epoch": 0.25,
+      "grad_norm": 14.38925838470459,
+      "learning_rate": 1.83073254403213e-05,
+      "loss": 3.4575,
       "step": 2024
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 16.6329288482666,
-      "learning_rate": 1.5942668136714446e-05,
-      "loss": 2.0742,
+      "epoch": 0.25,
+      "grad_norm": 12.298242568969727,
+      "learning_rate": 1.8306488725264614e-05,
+      "loss": 1.9957,
       "step": 2025
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 36.75046157836914,
-      "learning_rate": 1.5940663526110053e-05,
-      "loss": 2.7239,
+      "epoch": 0.25,
+      "grad_norm": 15.588695526123047,
+      "learning_rate": 1.8305652010207925e-05,
+      "loss": 2.809,
       "step": 2026
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 14.18947982788086,
-      "learning_rate": 1.5938658915505663e-05,
-      "loss": 2.6615,
+      "epoch": 0.25,
+      "grad_norm": 12.689305305480957,
+      "learning_rate": 1.830481529515124e-05,
+      "loss": 1.6972,
       "step": 2027
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 13.752023696899414,
-      "learning_rate": 1.5936654304901273e-05,
-      "loss": 1.8863,
+      "epoch": 0.25,
+      "grad_norm": 16.787702560424805,
+      "learning_rate": 1.830397858009455e-05,
+      "loss": 3.0474,
       "step": 2028
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 17.094932556152344,
-      "learning_rate": 1.5934649694296883e-05,
-      "loss": 3.2093,
+      "epoch": 0.25,
+      "grad_norm": 10.33752155303955,
+      "learning_rate": 1.8303141865037862e-05,
+      "loss": 2.0853,
       "step": 2029
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 16.756057739257812,
-      "learning_rate": 1.5932645083692497e-05,
-      "loss": 2.6765,
+      "epoch": 0.25,
+      "grad_norm": 12.861720085144043,
+      "learning_rate": 1.8302305149981176e-05,
+      "loss": 2.6356,
       "step": 2030
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 21.48285484313965,
-      "learning_rate": 1.5930640473088103e-05,
-      "loss": 2.0404,
+      "epoch": 0.25,
+      "grad_norm": 6.6646809577941895,
+      "learning_rate": 1.8301468434924486e-05,
+      "loss": 3.4687,
       "step": 2031
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 31.83589744567871,
-      "learning_rate": 1.5928635862483714e-05,
-      "loss": 2.708,
+      "epoch": 0.26,
+      "grad_norm": 14.956864356994629,
+      "learning_rate": 1.83006317198678e-05,
+      "loss": 2.253,
       "step": 2032
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 50.02532958984375,
-      "learning_rate": 1.5926631251879324e-05,
-      "loss": 2.5092,
+      "epoch": 0.26,
+      "grad_norm": 14.440831184387207,
+      "learning_rate": 1.8299795004811113e-05,
+      "loss": 2.7597,
       "step": 2033
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 11.736242294311523,
-      "learning_rate": 1.5924626641274934e-05,
-      "loss": 1.6545,
+      "epoch": 0.26,
+      "grad_norm": 23.195627212524414,
+      "learning_rate": 1.8298958289754424e-05,
+      "loss": 3.7566,
       "step": 2034
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 22.306257247924805,
-      "learning_rate": 1.5922622030670544e-05,
-      "loss": 2.438,
+      "epoch": 0.26,
+      "grad_norm": 18.944862365722656,
+      "learning_rate": 1.8298121574697737e-05,
+      "loss": 1.6732,
       "step": 2035
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 19.088542938232422,
-      "learning_rate": 1.5920617420066154e-05,
-      "loss": 2.5573,
+      "epoch": 0.26,
+      "grad_norm": 14.188807487487793,
+      "learning_rate": 1.829728485964105e-05,
+      "loss": 0.6534,
       "step": 2036
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 43.53116989135742,
-      "learning_rate": 1.5918612809461764e-05,
-      "loss": 2.4721,
+      "epoch": 0.26,
+      "grad_norm": 14.124086380004883,
+      "learning_rate": 1.829644814458436e-05,
+      "loss": 4.062,
       "step": 2037
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 16.74660873413086,
-      "learning_rate": 1.5916608198857374e-05,
-      "loss": 2.1669,
+      "epoch": 0.26,
+      "grad_norm": 11.936144828796387,
+      "learning_rate": 1.8295611429527675e-05,
+      "loss": 2.4579,
       "step": 2038
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 18.83763885498047,
-      "learning_rate": 1.5914603588252984e-05,
-      "loss": 2.0639,
+      "epoch": 0.26,
+      "grad_norm": 17.59867286682129,
+      "learning_rate": 1.829477471447099e-05,
+      "loss": 4.4768,
       "step": 2039
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 12.473159790039062,
-      "learning_rate": 1.591259897764859e-05,
-      "loss": 2.0599,
-      "step": 2040
-    },
-    {
-      "epoch": 0.61,
-      "eval_loss": 0.35515907406806946,
-      "eval_runtime": 43.2723,
-      "eval_samples_per_second": 34.179,
-      "eval_steps_per_second": 34.179,
+      "epoch": 0.26,
+      "grad_norm": 14.712457656860352,
+      "learning_rate": 1.8293937999414302e-05,
+      "loss": 3.6839,
       "step": 2040
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 34.54806900024414,
-      "learning_rate": 1.5910594367044204e-05,
-      "loss": 2.8225,
+      "epoch": 0.26,
+      "grad_norm": 12.759182929992676,
+      "learning_rate": 1.8293101284357612e-05,
+      "loss": 1.928,
       "step": 2041
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 11.377670288085938,
-      "learning_rate": 1.5908589756439814e-05,
-      "loss": 1.3387,
+      "epoch": 0.26,
+      "grad_norm": 10.04848861694336,
+      "learning_rate": 1.8292264569300926e-05,
+      "loss": 0.9844,
       "step": 2042
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 21.465269088745117,
-      "learning_rate": 1.590658514583542e-05,
-      "loss": 3.1448,
+      "epoch": 0.26,
+      "grad_norm": 11.40523910522461,
+      "learning_rate": 1.829142785424424e-05,
+      "loss": 1.5692,
       "step": 2043
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 13.400733947753906,
-      "learning_rate": 1.5904580535231035e-05,
-      "loss": 1.9627,
+      "epoch": 0.26,
+      "grad_norm": 13.77685832977295,
+      "learning_rate": 1.829059113918755e-05,
+      "loss": 2.7652,
       "step": 2044
     },
     {
-      "epoch": 0.61,
-      "grad_norm": 8.886704444885254,
-      "learning_rate": 1.590257592462664e-05,
-      "loss": 1.8101,
+      "epoch": 0.26,
+      "grad_norm": 7.332242012023926,
+      "learning_rate": 1.8289754424130864e-05,
+      "loss": 1.2132,
       "step": 2045
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 17.08548355102539,
-      "learning_rate": 1.590057131402225e-05,
-      "loss": 2.1456,
+      "epoch": 0.26,
+      "grad_norm": 8.43189525604248,
+      "learning_rate": 1.8288917709074177e-05,
+      "loss": 0.9907,
       "step": 2046
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 16.1516170501709,
-      "learning_rate": 1.589856670341786e-05,
-      "loss": 2.642,
+      "epoch": 0.26,
+      "grad_norm": 17.869178771972656,
+      "learning_rate": 1.828808099401749e-05,
+      "loss": 2.8893,
       "step": 2047
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 16.168611526489258,
-      "learning_rate": 1.589656209281347e-05,
-      "loss": 2.2961,
+      "epoch": 0.26,
+      "grad_norm": 12.180258750915527,
+      "learning_rate": 1.82872442789608e-05,
+      "loss": 3.3625,
       "step": 2048
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 17.722412109375,
-      "learning_rate": 1.589455748220908e-05,
-      "loss": 2.2299,
+      "epoch": 0.26,
+      "grad_norm": 16.094064712524414,
+      "learning_rate": 1.8286407563904115e-05,
+      "loss": 1.7809,
       "step": 2049
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 17.987937927246094,
-      "learning_rate": 1.589255287160469e-05,
-      "loss": 2.1003,
+      "epoch": 0.26,
+      "grad_norm": 13.385570526123047,
+      "learning_rate": 1.828557084884743e-05,
+      "loss": 3.2022,
       "step": 2050
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 13.64321517944336,
-      "learning_rate": 1.5890548261000302e-05,
-      "loss": 2.5651,
+      "epoch": 0.26,
+      "grad_norm": 11.876225471496582,
+      "learning_rate": 1.828473413379074e-05,
+      "loss": 1.3704,
       "step": 2051
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 10.08339786529541,
-      "learning_rate": 1.5888543650395912e-05,
-      "loss": 2.679,
+      "epoch": 0.26,
+      "grad_norm": 15.90719985961914,
+      "learning_rate": 1.8283897418734052e-05,
+      "loss": 1.4698,
       "step": 2052
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 19.540874481201172,
-      "learning_rate": 1.5886539039791522e-05,
-      "loss": 1.9754,
+      "epoch": 0.26,
+      "grad_norm": 22.79240608215332,
+      "learning_rate": 1.8283060703677366e-05,
+      "loss": 2.6591,
       "step": 2053
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 11.823172569274902,
-      "learning_rate": 1.5884534429187132e-05,
-      "loss": 1.5079,
+      "epoch": 0.26,
+      "grad_norm": 13.034547805786133,
+      "learning_rate": 1.8282223988620676e-05,
+      "loss": 1.9659,
       "step": 2054
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 14.114450454711914,
-      "learning_rate": 1.5882529818582742e-05,
-      "loss": 2.3742,
+      "epoch": 0.26,
+      "grad_norm": 17.19205093383789,
+      "learning_rate": 1.828138727356399e-05,
+      "loss": 3.2907,
       "step": 2055
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 9.858566284179688,
-      "learning_rate": 1.5880525207978352e-05,
-      "loss": 3.4989,
+      "epoch": 0.26,
+      "grad_norm": 13.555797576904297,
+      "learning_rate": 1.82805505585073e-05,
+      "loss": 2.1088,
       "step": 2056
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 14.25406551361084,
-      "learning_rate": 1.5878520597373962e-05,
-      "loss": 3.3511,
+      "epoch": 0.26,
+      "grad_norm": 13.456886291503906,
+      "learning_rate": 1.8279713843450614e-05,
+      "loss": 2.3089,
       "step": 2057
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 16.247241973876953,
-      "learning_rate": 1.5876515986769572e-05,
-      "loss": 2.1215,
+      "epoch": 0.26,
+      "grad_norm": 15.142951965332031,
+      "learning_rate": 1.8278877128393928e-05,
+      "loss": 2.9177,
       "step": 2058
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 13.218579292297363,
-      "learning_rate": 1.587451137616518e-05,
-      "loss": 2.4852,
+      "epoch": 0.26,
+      "grad_norm": 12.207382202148438,
+      "learning_rate": 1.8278040413337238e-05,
+      "loss": 3.182,
       "step": 2059
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 13.075812339782715,
-      "learning_rate": 1.5872506765560793e-05,
-      "loss": 2.3496,
+      "epoch": 0.26,
+      "grad_norm": 5.149965286254883,
+      "learning_rate": 1.827720369828055e-05,
+      "loss": 0.2904,
       "step": 2060
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 26.715604782104492,
-      "learning_rate": 1.5870502154956403e-05,
-      "loss": 2.1378,
+      "epoch": 0.26,
+      "grad_norm": 19.729907989501953,
+      "learning_rate": 1.8276366983223865e-05,
+      "loss": 3.0673,
       "step": 2061
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 20.732059478759766,
-      "learning_rate": 1.586849754435201e-05,
-      "loss": 2.008,
+      "epoch": 0.26,
+      "grad_norm": 10.116591453552246,
+      "learning_rate": 1.8275530268167175e-05,
+      "loss": 1.1507,
       "step": 2062
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 18.93284797668457,
-      "learning_rate": 1.5866492933747623e-05,
-      "loss": 3.4783,
+      "epoch": 0.26,
+      "grad_norm": 10.737142562866211,
+      "learning_rate": 1.827469355311049e-05,
+      "loss": 2.4385,
       "step": 2063
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 9.962442398071289,
-      "learning_rate": 1.586448832314323e-05,
-      "loss": 2.3671,
+      "epoch": 0.26,
+      "grad_norm": 9.6014986038208,
+      "learning_rate": 1.8273856838053803e-05,
+      "loss": 2.2556,
       "step": 2064
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 9.932574272155762,
-      "learning_rate": 1.586248371253884e-05,
-      "loss": 1.7524,
+      "epoch": 0.26,
+      "grad_norm": 16.13591957092285,
+      "learning_rate": 1.8273020122997113e-05,
+      "loss": 2.2409,
       "step": 2065
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 16.17888069152832,
-      "learning_rate": 1.586047910193445e-05,
-      "loss": 1.6364,
+      "epoch": 0.26,
+      "grad_norm": 17.769683837890625,
+      "learning_rate": 1.8272183407940427e-05,
+      "loss": 3.6554,
       "step": 2066
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 24.62932014465332,
-      "learning_rate": 1.585847449133006e-05,
-      "loss": 1.8643,
+      "epoch": 0.26,
+      "grad_norm": 21.505428314208984,
+      "learning_rate": 1.827134669288374e-05,
+      "loss": 3.0136,
       "step": 2067
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 17.033721923828125,
-      "learning_rate": 1.585646988072567e-05,
-      "loss": 2.2953,
+      "epoch": 0.26,
+      "grad_norm": 10.269015312194824,
+      "learning_rate": 1.8270509977827054e-05,
+      "loss": 1.3763,
       "step": 2068
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 24.557415008544922,
-      "learning_rate": 1.585446527012128e-05,
-      "loss": 3.0784,
+      "epoch": 0.26,
+      "grad_norm": 8.491415977478027,
+      "learning_rate": 1.8269673262770364e-05,
+      "loss": 1.7759,
       "step": 2069
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 27.415164947509766,
-      "learning_rate": 1.585246065951689e-05,
-      "loss": 2.3325,
+      "epoch": 0.26,
+      "grad_norm": 21.678611755371094,
+      "learning_rate": 1.8268836547713678e-05,
+      "loss": 3.0247,
       "step": 2070
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 18.90687370300293,
-      "learning_rate": 1.58504560489125e-05,
-      "loss": 2.4899,
+      "epoch": 0.26,
+      "grad_norm": 9.085183143615723,
+      "learning_rate": 1.826799983265699e-05,
+      "loss": 1.5463,
       "step": 2071
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 12.064643859863281,
-      "learning_rate": 1.584845143830811e-05,
-      "loss": 2.6493,
+      "epoch": 0.26,
+      "grad_norm": 5.598382472991943,
+      "learning_rate": 1.8267163117600302e-05,
+      "loss": 0.2033,
       "step": 2072
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 15.442275047302246,
-      "learning_rate": 1.584644682770372e-05,
-      "loss": 1.9658,
+      "epoch": 0.26,
+      "grad_norm": 12.32597827911377,
+      "learning_rate": 1.8266326402543615e-05,
+      "loss": 2.0717,
       "step": 2073
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 14.710504531860352,
-      "learning_rate": 1.584444221709933e-05,
-      "loss": 2.2267,
+      "epoch": 0.26,
+      "grad_norm": 64.5411148071289,
+      "learning_rate": 1.826548968748693e-05,
+      "loss": 2.2075,
       "step": 2074
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 9.406432151794434,
-      "learning_rate": 1.584243760649494e-05,
-      "loss": 1.4927,
+      "epoch": 0.26,
+      "grad_norm": 15.841133117675781,
+      "learning_rate": 1.8264652972430243e-05,
+      "loss": 1.7682,
       "step": 2075
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 14.804086685180664,
-      "learning_rate": 1.584043299589055e-05,
-      "loss": 1.8073,
+      "epoch": 0.26,
+      "grad_norm": 13.668527603149414,
+      "learning_rate": 1.8263816257373553e-05,
+      "loss": 1.8377,
       "step": 2076
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 15.176630973815918,
-      "learning_rate": 1.583842838528616e-05,
-      "loss": 1.8248,
+      "epoch": 0.26,
+      "grad_norm": 22.06974220275879,
+      "learning_rate": 1.8262979542316867e-05,
+      "loss": 3.6966,
       "step": 2077
     },
     {
-      "epoch": 0.62,
-      "grad_norm": 20.446643829345703,
-      "learning_rate": 1.5836423774681767e-05,
-      "loss": 2.3386,
+      "epoch": 0.26,
+      "grad_norm": 11.24099063873291,
+      "learning_rate": 1.826214282726018e-05,
+      "loss": 3.6092,
       "step": 2078
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 20.13260269165039,
-      "learning_rate": 1.583441916407738e-05,
-      "loss": 2.7811,
+      "epoch": 0.26,
+      "grad_norm": 25.48215675354004,
+      "learning_rate": 1.826130611220349e-05,
+      "loss": 3.2326,
       "step": 2079
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 18.64963150024414,
-      "learning_rate": 1.583241455347299e-05,
-      "loss": 2.9017,
+      "epoch": 0.26,
+      "grad_norm": 16.542388916015625,
+      "learning_rate": 1.8260469397146804e-05,
+      "loss": 1.4559,
       "step": 2080
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 17.931795120239258,
-      "learning_rate": 1.5830409942868597e-05,
-      "loss": 2.3632,
+      "epoch": 0.26,
+      "grad_norm": 10.460676193237305,
+      "learning_rate": 1.8259632682090114e-05,
+      "loss": 2.46,
       "step": 2081
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 19.928274154663086,
-      "learning_rate": 1.582840533226421e-05,
-      "loss": 1.7388,
+      "epoch": 0.26,
+      "grad_norm": 51.06991958618164,
+      "learning_rate": 1.8258795967033428e-05,
+      "loss": 3.0062,
       "step": 2082
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 11.930707931518555,
-      "learning_rate": 1.5826400721659818e-05,
-      "loss": 1.5894,
+      "epoch": 0.26,
+      "grad_norm": 14.114103317260742,
+      "learning_rate": 1.825795925197674e-05,
+      "loss": 1.4743,
       "step": 2083
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 24.95637321472168,
-      "learning_rate": 1.5824396111055428e-05,
-      "loss": 3.1926,
+      "epoch": 0.26,
+      "grad_norm": 31.35147476196289,
+      "learning_rate": 1.8257122536920052e-05,
+      "loss": 4.4442,
       "step": 2084
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 14.697010040283203,
-      "learning_rate": 1.582239150045104e-05,
-      "loss": 3.0097,
+      "epoch": 0.26,
+      "grad_norm": 9.869153022766113,
+      "learning_rate": 1.8256285821863366e-05,
+      "loss": 2.6016,
       "step": 2085
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 9.35490894317627,
-      "learning_rate": 1.5820386889846648e-05,
-      "loss": 1.5664,
+      "epoch": 0.26,
+      "grad_norm": 17.398767471313477,
+      "learning_rate": 1.8255449106806676e-05,
+      "loss": 1.9688,
       "step": 2086
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 21.47760581970215,
-      "learning_rate": 1.5818382279242258e-05,
-      "loss": 2.8386,
+      "epoch": 0.26,
+      "grad_norm": 8.782936096191406,
+      "learning_rate": 1.825461239174999e-05,
+      "loss": 0.8068,
       "step": 2087
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 16.634397506713867,
-      "learning_rate": 1.5816377668637868e-05,
-      "loss": 2.8702,
+      "epoch": 0.26,
+      "grad_norm": 10.660070419311523,
+      "learning_rate": 1.8253775676693303e-05,
+      "loss": 0.635,
       "step": 2088
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 12.756913185119629,
-      "learning_rate": 1.5814373058033478e-05,
-      "loss": 2.456,
+      "epoch": 0.26,
+      "grad_norm": 16.110692977905273,
+      "learning_rate": 1.8252938961636617e-05,
+      "loss": 2.1616,
       "step": 2089
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 13.137290000915527,
-      "learning_rate": 1.5812368447429088e-05,
-      "loss": 1.8993,
+      "epoch": 0.26,
+      "grad_norm": 9.807194709777832,
+      "learning_rate": 1.8252102246579927e-05,
+      "loss": 1.3718,
       "step": 2090
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 23.007200241088867,
-      "learning_rate": 1.58103638368247e-05,
-      "loss": 2.7142,
+      "epoch": 0.26,
+      "grad_norm": 11.816697120666504,
+      "learning_rate": 1.825126553152324e-05,
+      "loss": 3.6432,
       "step": 2091
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 10.685704231262207,
-      "learning_rate": 1.580835922622031e-05,
-      "loss": 1.872,
+      "epoch": 0.26,
+      "grad_norm": 14.993602752685547,
+      "learning_rate": 1.8250428816466554e-05,
+      "loss": 3.2196,
       "step": 2092
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 34.11006164550781,
-      "learning_rate": 1.580635461561592e-05,
-      "loss": 2.8411,
+      "epoch": 0.26,
+      "grad_norm": 19.83965492248535,
+      "learning_rate": 1.8249592101409865e-05,
+      "loss": 1.3436,
       "step": 2093
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 15.17668628692627,
-      "learning_rate": 1.580435000501153e-05,
-      "loss": 2.318,
+      "epoch": 0.26,
+      "grad_norm": 14.342920303344727,
+      "learning_rate": 1.8248755386353178e-05,
+      "loss": 1.0609,
       "step": 2094
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 10.943110466003418,
-      "learning_rate": 1.580234539440714e-05,
-      "loss": 3.1694,
+      "epoch": 0.26,
+      "grad_norm": 43.393165588378906,
+      "learning_rate": 1.8247918671296492e-05,
+      "loss": 1.8596,
       "step": 2095
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 17.957090377807617,
-      "learning_rate": 1.580034078380275e-05,
-      "loss": 2.8144,
+      "epoch": 0.26,
+      "grad_norm": 16.4021053314209,
+      "learning_rate": 1.8247081956239806e-05,
+      "loss": 4.3693,
       "step": 2096
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 21.06376075744629,
-      "learning_rate": 1.5798336173198355e-05,
-      "loss": 2.4499,
+      "epoch": 0.26,
+      "grad_norm": 10.059845924377441,
+      "learning_rate": 1.8246245241183116e-05,
+      "loss": 1.4066,
       "step": 2097
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 17.400442123413086,
-      "learning_rate": 1.579633156259397e-05,
-      "loss": 1.1497,
+      "epoch": 0.26,
+      "grad_norm": 16.443490982055664,
+      "learning_rate": 1.824540852612643e-05,
+      "loss": 1.9437,
       "step": 2098
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 14.588909149169922,
-      "learning_rate": 1.579432695198958e-05,
-      "loss": 3.2946,
+      "epoch": 0.26,
+      "grad_norm": 10.47969913482666,
+      "learning_rate": 1.8244571811069743e-05,
+      "loss": 2.004,
       "step": 2099
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 13.294282913208008,
-      "learning_rate": 1.5792322341385186e-05,
-      "loss": 1.6623,
+      "epoch": 0.26,
+      "grad_norm": 11.944048881530762,
+      "learning_rate": 1.8243735096013053e-05,
+      "loss": 3.0902,
       "step": 2100
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 23.731582641601562,
-      "learning_rate": 1.57903177307808e-05,
-      "loss": 3.3506,
+      "epoch": 0.26,
+      "grad_norm": 12.932525634765625,
+      "learning_rate": 1.8242898380956367e-05,
+      "loss": 2.3718,
       "step": 2101
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 15.64577579498291,
-      "learning_rate": 1.5788313120176406e-05,
-      "loss": 2.6303,
+      "epoch": 0.26,
+      "grad_norm": 11.039018630981445,
+      "learning_rate": 1.824206166589968e-05,
+      "loss": 4.8056,
       "step": 2102
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 19.747045516967773,
-      "learning_rate": 1.5786308509572016e-05,
-      "loss": 2.5166,
+      "epoch": 0.26,
+      "grad_norm": 24.806896209716797,
+      "learning_rate": 1.8241224950842994e-05,
+      "loss": 2.4179,
       "step": 2103
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 42.6046142578125,
-      "learning_rate": 1.578430389896763e-05,
-      "loss": 2.8453,
+      "epoch": 0.26,
+      "grad_norm": 13.5587158203125,
+      "learning_rate": 1.8240388235786305e-05,
+      "loss": 3.4312,
       "step": 2104
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 8.715738296508789,
-      "learning_rate": 1.5782299288363236e-05,
-      "loss": 1.9525,
+      "epoch": 0.26,
+      "grad_norm": 9.747724533081055,
+      "learning_rate": 1.8239551520729618e-05,
+      "loss": 1.2014,
       "step": 2105
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 31.00543975830078,
-      "learning_rate": 1.5780294677758846e-05,
-      "loss": 3.0852,
+      "epoch": 0.26,
+      "grad_norm": 8.556553840637207,
+      "learning_rate": 1.823871480567293e-05,
+      "loss": 2.2537,
       "step": 2106
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 10.329021453857422,
-      "learning_rate": 1.5778290067154456e-05,
-      "loss": 2.1254,
+      "epoch": 0.26,
+      "grad_norm": 13.39060115814209,
+      "learning_rate": 1.8237878090616242e-05,
+      "loss": 2.0256,
       "step": 2107
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 12.255505561828613,
-      "learning_rate": 1.5776285456550066e-05,
-      "loss": 2.1254,
+      "epoch": 0.26,
+      "grad_norm": 17.98282241821289,
+      "learning_rate": 1.8237041375559556e-05,
+      "loss": 3.5958,
       "step": 2108
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 13.093355178833008,
-      "learning_rate": 1.5774280845945676e-05,
-      "loss": 2.1318,
+      "epoch": 0.26,
+      "grad_norm": 15.861956596374512,
+      "learning_rate": 1.8236204660502866e-05,
+      "loss": 3.77,
       "step": 2109
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 34.728302001953125,
-      "learning_rate": 1.5772276235341287e-05,
-      "loss": 3.1983,
+      "epoch": 0.26,
+      "grad_norm": 19.806100845336914,
+      "learning_rate": 1.823536794544618e-05,
+      "loss": 1.8967,
       "step": 2110
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 13.47143840789795,
-      "learning_rate": 1.5770271624736897e-05,
-      "loss": 2.0851,
+      "epoch": 0.26,
+      "grad_norm": 13.889492988586426,
+      "learning_rate": 1.8234531230389493e-05,
+      "loss": 1.456,
       "step": 2111
     },
     {
-      "epoch": 0.63,
-      "grad_norm": 13.41025161743164,
-      "learning_rate": 1.5768267014132507e-05,
-      "loss": 2.8133,
+      "epoch": 0.27,
+      "grad_norm": 19.10417366027832,
+      "learning_rate": 1.8233694515332804e-05,
+      "loss": 3.4118,
       "step": 2112
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 15.229413986206055,
-      "learning_rate": 1.5766262403528117e-05,
-      "loss": 2.8769,
+      "epoch": 0.27,
+      "grad_norm": 9.383771896362305,
+      "learning_rate": 1.8232857800276117e-05,
+      "loss": 1.2884,
       "step": 2113
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 29.792980194091797,
-      "learning_rate": 1.5764257792923724e-05,
-      "loss": 2.6786,
+      "epoch": 0.27,
+      "grad_norm": 12.032513618469238,
+      "learning_rate": 1.8232021085219428e-05,
+      "loss": 2.0497,
       "step": 2114
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 12.121004104614258,
-      "learning_rate": 1.5762253182319337e-05,
-      "loss": 2.3366,
+      "epoch": 0.27,
+      "grad_norm": 15.174636840820312,
+      "learning_rate": 1.823118437016274e-05,
+      "loss": 3.8727,
       "step": 2115
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 17.383033752441406,
-      "learning_rate": 1.5760248571714947e-05,
-      "loss": 2.2895,
+      "epoch": 0.27,
+      "grad_norm": 47.917633056640625,
+      "learning_rate": 1.8230347655106055e-05,
+      "loss": 2.8027,
       "step": 2116
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 28.382801055908203,
-      "learning_rate": 1.5758243961110554e-05,
-      "loss": 2.5891,
+      "epoch": 0.27,
+      "grad_norm": 6.924861431121826,
+      "learning_rate": 1.822951094004937e-05,
+      "loss": 1.0783,
       "step": 2117
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 14.136621475219727,
-      "learning_rate": 1.5756239350506167e-05,
-      "loss": 2.4559,
+      "epoch": 0.27,
+      "grad_norm": 11.44529914855957,
+      "learning_rate": 1.822867422499268e-05,
+      "loss": 2.6785,
       "step": 2118
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 12.245827674865723,
-      "learning_rate": 1.5754234739901774e-05,
-      "loss": 1.3721,
+      "epoch": 0.27,
+      "grad_norm": 26.966922760009766,
+      "learning_rate": 1.8227837509935992e-05,
+      "loss": 3.2264,
       "step": 2119
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 10.706925392150879,
-      "learning_rate": 1.5752230129297384e-05,
-      "loss": 2.2164,
+      "epoch": 0.27,
+      "grad_norm": 16.654945373535156,
+      "learning_rate": 1.8227000794879306e-05,
+      "loss": 3.6384,
       "step": 2120
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 16.694616317749023,
-      "learning_rate": 1.5750225518692994e-05,
-      "loss": 3.428,
+      "epoch": 0.27,
+      "grad_norm": 12.091151237487793,
+      "learning_rate": 1.8226164079822616e-05,
+      "loss": 0.3314,
       "step": 2121
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 33.90262985229492,
-      "learning_rate": 1.5748220908088604e-05,
-      "loss": 3.3125,
+      "epoch": 0.27,
+      "grad_norm": 22.83612632751465,
+      "learning_rate": 1.822532736476593e-05,
+      "loss": 2.3924,
       "step": 2122
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 12.597078323364258,
-      "learning_rate": 1.5746216297484214e-05,
-      "loss": 1.9381,
+      "epoch": 0.27,
+      "grad_norm": 10.339336395263672,
+      "learning_rate": 1.8224490649709244e-05,
+      "loss": 1.4609,
       "step": 2123
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 13.725189208984375,
-      "learning_rate": 1.5744211686879824e-05,
-      "loss": 1.9921,
+      "epoch": 0.27,
+      "grad_norm": 12.055213928222656,
+      "learning_rate": 1.8223653934652557e-05,
+      "loss": 2.463,
       "step": 2124
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 13.38775634765625,
-      "learning_rate": 1.5742207076275434e-05,
-      "loss": 2.3561,
+      "epoch": 0.27,
+      "grad_norm": 16.07337188720703,
+      "learning_rate": 1.8222817219595867e-05,
+      "loss": 1.9958,
       "step": 2125
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 15.606164932250977,
-      "learning_rate": 1.5740202465671045e-05,
-      "loss": 3.1076,
+      "epoch": 0.27,
+      "grad_norm": 8.979070663452148,
+      "learning_rate": 1.822198050453918e-05,
+      "loss": 2.3272,
       "step": 2126
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 17.07231330871582,
-      "learning_rate": 1.5738197855066655e-05,
-      "loss": 2.0324,
+      "epoch": 0.27,
+      "grad_norm": 8.645263671875,
+      "learning_rate": 1.8221143789482495e-05,
+      "loss": 2.1392,
       "step": 2127
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 18.28791618347168,
-      "learning_rate": 1.5736193244462265e-05,
-      "loss": 2.8287,
+      "epoch": 0.27,
+      "grad_norm": 16.874176025390625,
+      "learning_rate": 1.8220307074425805e-05,
+      "loss": 2.6234,
       "step": 2128
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 20.407081604003906,
-      "learning_rate": 1.5734188633857875e-05,
-      "loss": 2.3472,
+      "epoch": 0.27,
+      "grad_norm": 12.335731506347656,
+      "learning_rate": 1.821947035936912e-05,
+      "loss": 2.9914,
       "step": 2129
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 12.980356216430664,
-      "learning_rate": 1.5732184023253485e-05,
-      "loss": 3.1184,
+      "epoch": 0.27,
+      "grad_norm": 21.420886993408203,
+      "learning_rate": 1.8218633644312432e-05,
+      "loss": 2.7485,
       "step": 2130
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 19.251155853271484,
-      "learning_rate": 1.5730179412649095e-05,
-      "loss": 2.2474,
+      "epoch": 0.27,
+      "grad_norm": 13.961335182189941,
+      "learning_rate": 1.8217796929255746e-05,
+      "loss": 1.9668,
       "step": 2131
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 8.661418914794922,
-      "learning_rate": 1.5728174802044705e-05,
-      "loss": 1.4024,
+      "epoch": 0.27,
+      "grad_norm": 13.38197135925293,
+      "learning_rate": 1.8216960214199056e-05,
+      "loss": 2.7635,
       "step": 2132
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 24.52813148498535,
-      "learning_rate": 1.5726170191440312e-05,
-      "loss": 2.4654,
+      "epoch": 0.27,
+      "grad_norm": 9.04694652557373,
+      "learning_rate": 1.821612349914237e-05,
+      "loss": 1.0293,
       "step": 2133
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 12.824371337890625,
-      "learning_rate": 1.5724165580835925e-05,
-      "loss": 1.5831,
+      "epoch": 0.27,
+      "grad_norm": 25.329591751098633,
+      "learning_rate": 1.821528678408568e-05,
+      "loss": 2.931,
       "step": 2134
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 11.373884201049805,
-      "learning_rate": 1.5722160970231535e-05,
-      "loss": 2.043,
+      "epoch": 0.27,
+      "grad_norm": 9.510565757751465,
+      "learning_rate": 1.8214450069028994e-05,
+      "loss": 3.2181,
       "step": 2135
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 24.369014739990234,
-      "learning_rate": 1.5720156359627142e-05,
-      "loss": 2.4182,
+      "epoch": 0.27,
+      "grad_norm": 17.192440032958984,
+      "learning_rate": 1.8213613353972307e-05,
+      "loss": 1.9588,
       "step": 2136
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 12.161373138427734,
-      "learning_rate": 1.5718151749022755e-05,
-      "loss": 1.8717,
+      "epoch": 0.27,
+      "grad_norm": 8.89498519897461,
+      "learning_rate": 1.8212776638915618e-05,
+      "loss": 0.9073,
       "step": 2137
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 22.5483341217041,
-      "learning_rate": 1.5716147138418362e-05,
-      "loss": 1.8705,
+      "epoch": 0.27,
+      "grad_norm": 8.90566349029541,
+      "learning_rate": 1.821193992385893e-05,
+      "loss": 2.5312,
       "step": 2138
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 16.376750946044922,
-      "learning_rate": 1.5714142527813972e-05,
-      "loss": 2.359,
+      "epoch": 0.27,
+      "grad_norm": 7.776777267456055,
+      "learning_rate": 1.821110320880224e-05,
+      "loss": 1.8593,
       "step": 2139
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 12.507425308227539,
-      "learning_rate": 1.5712137917209582e-05,
-      "loss": 1.8627,
+      "epoch": 0.27,
+      "grad_norm": 10.225785255432129,
+      "learning_rate": 1.8210266493745555e-05,
+      "loss": 3.0292,
       "step": 2140
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 12.094954490661621,
-      "learning_rate": 1.5710133306605192e-05,
-      "loss": 1.5144,
+      "epoch": 0.27,
+      "grad_norm": 8.556358337402344,
+      "learning_rate": 1.820942977868887e-05,
+      "loss": 2.0255,
       "step": 2141
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 12.721661567687988,
-      "learning_rate": 1.5708128696000802e-05,
-      "loss": 2.4135,
+      "epoch": 0.27,
+      "grad_norm": 11.723958969116211,
+      "learning_rate": 1.820859306363218e-05,
+      "loss": 3.7294,
       "step": 2142
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 16.270231246948242,
-      "learning_rate": 1.5706124085396413e-05,
-      "loss": 1.5671,
+      "epoch": 0.27,
+      "grad_norm": 11.870851516723633,
+      "learning_rate": 1.8207756348575493e-05,
+      "loss": 1.7791,
       "step": 2143
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 22.51713752746582,
-      "learning_rate": 1.5704119474792023e-05,
-      "loss": 2.8953,
+      "epoch": 0.27,
+      "grad_norm": 9.81495189666748,
+      "learning_rate": 1.8206919633518806e-05,
+      "loss": 2.0573,
       "step": 2144
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 14.330571174621582,
-      "learning_rate": 1.5702114864187633e-05,
-      "loss": 1.948,
+      "epoch": 0.27,
+      "grad_norm": 10.432618141174316,
+      "learning_rate": 1.820608291846212e-05,
+      "loss": 3.0315,
       "step": 2145
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 16.00999641418457,
-      "learning_rate": 1.5700110253583243e-05,
-      "loss": 2.0595,
+      "epoch": 0.27,
+      "grad_norm": 20.963991165161133,
+      "learning_rate": 1.820524620340543e-05,
+      "loss": 1.2123,
       "step": 2146
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 23.264543533325195,
-      "learning_rate": 1.5698105642978853e-05,
-      "loss": 2.7985,
+      "epoch": 0.27,
+      "grad_norm": 10.1554594039917,
+      "learning_rate": 1.8204409488348744e-05,
+      "loss": 1.1561,
       "step": 2147
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 12.527949333190918,
-      "learning_rate": 1.5696101032374463e-05,
-      "loss": 2.2662,
+      "epoch": 0.27,
+      "grad_norm": 28.186546325683594,
+      "learning_rate": 1.8203572773292058e-05,
+      "loss": 2.2849,
       "step": 2148
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 23.13153076171875,
-      "learning_rate": 1.5694096421770073e-05,
-      "loss": 3.2819,
+      "epoch": 0.27,
+      "grad_norm": 14.456684112548828,
+      "learning_rate": 1.8202736058235368e-05,
+      "loss": 1.4981,
       "step": 2149
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 11.828460693359375,
-      "learning_rate": 1.5692091811165683e-05,
-      "loss": 1.9632,
+      "epoch": 0.27,
+      "grad_norm": 14.477010726928711,
+      "learning_rate": 1.820189934317868e-05,
+      "loss": 3.8286,
       "step": 2150
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 30.631380081176758,
-      "learning_rate": 1.5690087200561293e-05,
-      "loss": 2.5282,
+      "epoch": 0.27,
+      "grad_norm": 12.221043586730957,
+      "learning_rate": 1.8201062628121995e-05,
+      "loss": 1.7472,
       "step": 2151
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 16.613862991333008,
-      "learning_rate": 1.56880825899569e-05,
-      "loss": 2.6759,
+      "epoch": 0.27,
+      "grad_norm": 11.797484397888184,
+      "learning_rate": 1.820022591306531e-05,
+      "loss": 2.4295,
       "step": 2152
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 36.72877502441406,
-      "learning_rate": 1.5686077979352513e-05,
-      "loss": 2.0691,
+      "epoch": 0.27,
+      "grad_norm": 12.780076026916504,
+      "learning_rate": 1.819938919800862e-05,
+      "loss": 1.7441,
       "step": 2153
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 11.451091766357422,
-      "learning_rate": 1.5684073368748123e-05,
-      "loss": 1.2804,
+      "epoch": 0.27,
+      "grad_norm": 35.99589157104492,
+      "learning_rate": 1.8198552482951933e-05,
+      "loss": 2.7663,
       "step": 2154
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 13.62600326538086,
-      "learning_rate": 1.568206875814373e-05,
-      "loss": 1.9118,
+      "epoch": 0.27,
+      "grad_norm": 18.572803497314453,
+      "learning_rate": 1.8197715767895246e-05,
+      "loss": 2.4358,
       "step": 2155
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 21.245529174804688,
-      "learning_rate": 1.5680064147539344e-05,
-      "loss": 2.7814,
+      "epoch": 0.27,
+      "grad_norm": 21.449501037597656,
+      "learning_rate": 1.8196879052838557e-05,
+      "loss": 2.8693,
       "step": 2156
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 22.127309799194336,
-      "learning_rate": 1.567805953693495e-05,
-      "loss": 1.8782,
+      "epoch": 0.27,
+      "grad_norm": 11.441527366638184,
+      "learning_rate": 1.819604233778187e-05,
+      "loss": 0.954,
       "step": 2157
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 11.301019668579102,
-      "learning_rate": 1.567605492633056e-05,
-      "loss": 3.1518,
+      "epoch": 0.27,
+      "grad_norm": 12.363370895385742,
+      "learning_rate": 1.8195205622725184e-05,
+      "loss": 1.6574,
       "step": 2158
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 11.62056827545166,
-      "learning_rate": 1.5674050315726174e-05,
-      "loss": 2.3903,
+      "epoch": 0.27,
+      "grad_norm": 6.905948162078857,
+      "learning_rate": 1.8194368907668494e-05,
+      "loss": 0.8388,
       "step": 2159
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 15.686822891235352,
-      "learning_rate": 1.567204570512178e-05,
-      "loss": 2.4537,
-      "step": 2160
-    },
-    {
-      "epoch": 0.65,
-      "eval_loss": 0.3099038302898407,
-      "eval_runtime": 43.3953,
-      "eval_samples_per_second": 34.082,
-      "eval_steps_per_second": 34.082,
+      "epoch": 0.27,
+      "grad_norm": 13.503583908081055,
+      "learning_rate": 1.8193532192611808e-05,
+      "loss": 3.3977,
       "step": 2160
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 14.51310920715332,
-      "learning_rate": 1.567004109451739e-05,
-      "loss": 2.5808,
+      "epoch": 0.27,
+      "grad_norm": 9.192841529846191,
+      "learning_rate": 1.819269547755512e-05,
+      "loss": 1.6161,
       "step": 2161
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 15.057507514953613,
-      "learning_rate": 1.5668036483913e-05,
-      "loss": 2.8596,
+      "epoch": 0.27,
+      "grad_norm": 24.90398406982422,
+      "learning_rate": 1.8191858762498432e-05,
+      "loss": 2.0326,
       "step": 2162
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 12.982969284057617,
-      "learning_rate": 1.566603187330861e-05,
-      "loss": 2.5042,
+      "epoch": 0.27,
+      "grad_norm": 10.266580581665039,
+      "learning_rate": 1.8191022047441745e-05,
+      "loss": 2.0516,
       "step": 2163
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 11.829024314880371,
-      "learning_rate": 1.566402726270422e-05,
-      "loss": 1.9732,
+      "epoch": 0.27,
+      "grad_norm": 16.922138214111328,
+      "learning_rate": 1.819018533238506e-05,
+      "loss": 2.1421,
       "step": 2164
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 14.070128440856934,
-      "learning_rate": 1.566202265209983e-05,
-      "loss": 2.6257,
+      "epoch": 0.27,
+      "grad_norm": 20.214248657226562,
+      "learning_rate": 1.818934861732837e-05,
+      "loss": 2.6699,
       "step": 2165
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 10.464433670043945,
-      "learning_rate": 1.566001804149544e-05,
-      "loss": 2.178,
+      "epoch": 0.27,
+      "grad_norm": 10.546091079711914,
+      "learning_rate": 1.8188511902271683e-05,
+      "loss": 3.4302,
       "step": 2166
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 11.11734390258789,
-      "learning_rate": 1.565801343089105e-05,
-      "loss": 1.5285,
+      "epoch": 0.27,
+      "grad_norm": 14.921259880065918,
+      "learning_rate": 1.8187675187214993e-05,
+      "loss": 2.9676,
       "step": 2167
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 14.957688331604004,
-      "learning_rate": 1.565600882028666e-05,
-      "loss": 2.1816,
+      "epoch": 0.27,
+      "grad_norm": 18.569194793701172,
+      "learning_rate": 1.8186838472158307e-05,
+      "loss": 2.3316,
       "step": 2168
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 16.353477478027344,
-      "learning_rate": 1.565400420968227e-05,
-      "loss": 2.2476,
+      "epoch": 0.27,
+      "grad_norm": 21.49285888671875,
+      "learning_rate": 1.818600175710162e-05,
+      "loss": 1.8356,
       "step": 2169
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 9.818117141723633,
-      "learning_rate": 1.565199959907788e-05,
-      "loss": 2.1336,
+      "epoch": 0.27,
+      "grad_norm": 10.457573890686035,
+      "learning_rate": 1.818516504204493e-05,
+      "loss": 1.3546,
       "step": 2170
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 20.093345642089844,
-      "learning_rate": 1.564999498847349e-05,
-      "loss": 2.6549,
+      "epoch": 0.27,
+      "grad_norm": 25.030502319335938,
+      "learning_rate": 1.8184328326988245e-05,
+      "loss": 2.2491,
       "step": 2171
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 18.352497100830078,
-      "learning_rate": 1.56479903778691e-05,
-      "loss": 2.6063,
+      "epoch": 0.27,
+      "grad_norm": 12.664494514465332,
+      "learning_rate": 1.8183491611931558e-05,
+      "loss": 2.7344,
       "step": 2172
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 26.143571853637695,
-      "learning_rate": 1.5645985767264712e-05,
-      "loss": 2.8258,
+      "epoch": 0.27,
+      "grad_norm": 11.065912246704102,
+      "learning_rate": 1.8182654896874872e-05,
+      "loss": 1.6745,
       "step": 2173
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 18.577131271362305,
-      "learning_rate": 1.564398115666032e-05,
-      "loss": 1.8137,
+      "epoch": 0.27,
+      "grad_norm": 30.694799423217773,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 3.3537,
       "step": 2174
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 11.283987045288086,
-      "learning_rate": 1.5641976546055932e-05,
-      "loss": 1.9124,
+      "epoch": 0.27,
+      "grad_norm": 11.940272331237793,
+      "learning_rate": 1.8180981466761496e-05,
+      "loss": 2.2207,
       "step": 2175
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 19.12086296081543,
-      "learning_rate": 1.563997193545154e-05,
-      "loss": 2.3072,
+      "epoch": 0.27,
+      "grad_norm": 20.852827072143555,
+      "learning_rate": 1.818014475170481e-05,
+      "loss": 4.1757,
       "step": 2176
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 11.322100639343262,
-      "learning_rate": 1.563796732484715e-05,
-      "loss": 1.8458,
+      "epoch": 0.27,
+      "grad_norm": 15.129714012145996,
+      "learning_rate": 1.817930803664812e-05,
+      "loss": 2.8836,
       "step": 2177
     },
     {
-      "epoch": 0.65,
-      "grad_norm": 16.309368133544922,
-      "learning_rate": 1.5635962714242762e-05,
-      "loss": 2.3407,
+      "epoch": 0.27,
+      "grad_norm": 20.83080291748047,
+      "learning_rate": 1.8178471321591433e-05,
+      "loss": 2.491,
       "step": 2178
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 42.83634567260742,
-      "learning_rate": 1.563395810363837e-05,
-      "loss": 2.9269,
+      "epoch": 0.27,
+      "grad_norm": 10.54280948638916,
+      "learning_rate": 1.8177634606534747e-05,
+      "loss": 2.227,
       "step": 2179
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 16.00682830810547,
-      "learning_rate": 1.563195349303398e-05,
-      "loss": 2.2228,
+      "epoch": 0.27,
+      "grad_norm": 13.582934379577637,
+      "learning_rate": 1.817679789147806e-05,
+      "loss": 4.0549,
       "step": 2180
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 10.3877592086792,
-      "learning_rate": 1.562994888242959e-05,
-      "loss": 2.0959,
+      "epoch": 0.27,
+      "grad_norm": 11.53133773803711,
+      "learning_rate": 1.817596117642137e-05,
+      "loss": 3.4779,
       "step": 2181
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 23.170366287231445,
-      "learning_rate": 1.56279442718252e-05,
-      "loss": 2.8749,
+      "epoch": 0.27,
+      "grad_norm": 14.917023658752441,
+      "learning_rate": 1.8175124461364684e-05,
+      "loss": 1.7925,
       "step": 2182
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 10.100824356079102,
-      "learning_rate": 1.562593966122081e-05,
-      "loss": 2.0848,
+      "epoch": 0.27,
+      "grad_norm": 23.791954040527344,
+      "learning_rate": 1.8174287746307998e-05,
+      "loss": 2.4726,
       "step": 2183
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 17.805927276611328,
-      "learning_rate": 1.562393505061642e-05,
-      "loss": 2.6061,
+      "epoch": 0.27,
+      "grad_norm": 6.664028644561768,
+      "learning_rate": 1.817345103125131e-05,
+      "loss": 0.2051,
       "step": 2184
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 18.987218856811523,
-      "learning_rate": 1.562193044001203e-05,
-      "loss": 2.7442,
+      "epoch": 0.27,
+      "grad_norm": 14.408364295959473,
+      "learning_rate": 1.8172614316194622e-05,
+      "loss": 3.4575,
       "step": 2185
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 44.760196685791016,
-      "learning_rate": 1.561992582940764e-05,
-      "loss": 2.3028,
+      "epoch": 0.27,
+      "grad_norm": 13.059243202209473,
+      "learning_rate": 1.8171777601137936e-05,
+      "loss": 3.5806,
       "step": 2186
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 17.44770622253418,
-      "learning_rate": 1.561792121880325e-05,
-      "loss": 3.0411,
+      "epoch": 0.27,
+      "grad_norm": 12.887029647827148,
+      "learning_rate": 1.8170940886081246e-05,
+      "loss": 1.003,
       "step": 2187
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 15.366558074951172,
-      "learning_rate": 1.5615916608198856e-05,
-      "loss": 2.4066,
+      "epoch": 0.27,
+      "grad_norm": 14.370407104492188,
+      "learning_rate": 1.817010417102456e-05,
+      "loss": 4.1245,
       "step": 2188
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 14.17463493347168,
-      "learning_rate": 1.561391199759447e-05,
-      "loss": 1.7436,
+      "epoch": 0.27,
+      "grad_norm": 21.14535903930664,
+      "learning_rate": 1.8169267455967873e-05,
+      "loss": 1.0258,
       "step": 2189
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 11.214127540588379,
-      "learning_rate": 1.561190738699008e-05,
-      "loss": 1.9083,
+      "epoch": 0.27,
+      "grad_norm": 16.043947219848633,
+      "learning_rate": 1.8168430740911184e-05,
+      "loss": 2.7536,
       "step": 2190
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 14.602449417114258,
-      "learning_rate": 1.5609902776385686e-05,
-      "loss": 2.8559,
+      "epoch": 0.27,
+      "grad_norm": 21.75202178955078,
+      "learning_rate": 1.8167594025854497e-05,
+      "loss": 2.3325,
       "step": 2191
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 16.023332595825195,
-      "learning_rate": 1.56078981657813e-05,
-      "loss": 1.8949,
+      "epoch": 0.28,
+      "grad_norm": 18.115833282470703,
+      "learning_rate": 1.8166757310797807e-05,
+      "loss": 2.7141,
       "step": 2192
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 20.160932540893555,
-      "learning_rate": 1.5605893555176907e-05,
-      "loss": 2.0295,
+      "epoch": 0.28,
+      "grad_norm": 12.496354103088379,
+      "learning_rate": 1.816592059574112e-05,
+      "loss": 1.4812,
       "step": 2193
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 12.389837265014648,
-      "learning_rate": 1.5603888944572517e-05,
-      "loss": 2.2384,
+      "epoch": 0.28,
+      "grad_norm": 17.79195785522461,
+      "learning_rate": 1.8165083880684435e-05,
+      "loss": 2.2022,
       "step": 2194
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 18.265491485595703,
-      "learning_rate": 1.5601884333968127e-05,
-      "loss": 2.4571,
+      "epoch": 0.28,
+      "grad_norm": 16.170005798339844,
+      "learning_rate": 1.8164247165627745e-05,
+      "loss": 3.9978,
       "step": 2195
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 14.70784854888916,
-      "learning_rate": 1.5599879723363737e-05,
-      "loss": 2.1679,
+      "epoch": 0.28,
+      "grad_norm": 7.060325622558594,
+      "learning_rate": 1.816341045057106e-05,
+      "loss": 0.5294,
       "step": 2196
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 16.466394424438477,
-      "learning_rate": 1.5597875112759347e-05,
-      "loss": 1.9488,
+      "epoch": 0.28,
+      "grad_norm": 25.24850082397461,
+      "learning_rate": 1.8162573735514372e-05,
+      "loss": 2.0798,
       "step": 2197
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 21.658117294311523,
-      "learning_rate": 1.5595870502154957e-05,
-      "loss": 2.7079,
+      "epoch": 0.28,
+      "grad_norm": 10.852115631103516,
+      "learning_rate": 1.8161737020457683e-05,
+      "loss": 2.0862,
       "step": 2198
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 17.353038787841797,
-      "learning_rate": 1.5593865891550567e-05,
-      "loss": 2.9999,
+      "epoch": 0.28,
+      "grad_norm": 17.65023422241211,
+      "learning_rate": 1.8160900305400996e-05,
+      "loss": 2.637,
       "step": 2199
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 15.819283485412598,
-      "learning_rate": 1.5591861280946177e-05,
-      "loss": 2.1287,
+      "epoch": 0.28,
+      "grad_norm": 10.618857383728027,
+      "learning_rate": 1.816006359034431e-05,
+      "loss": 1.9751,
       "step": 2200
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 13.62539291381836,
-      "learning_rate": 1.5589856670341787e-05,
-      "loss": 1.9569,
+      "epoch": 0.28,
+      "grad_norm": 13.295710563659668,
+      "learning_rate": 1.8159226875287623e-05,
+      "loss": 2.2849,
       "step": 2201
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 18.662357330322266,
-      "learning_rate": 1.5587852059737397e-05,
-      "loss": 2.51,
+      "epoch": 0.28,
+      "grad_norm": 29.24386215209961,
+      "learning_rate": 1.8158390160230934e-05,
+      "loss": 4.5359,
       "step": 2202
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 26.726213455200195,
-      "learning_rate": 1.5585847449133007e-05,
-      "loss": 1.9874,
+      "epoch": 0.28,
+      "grad_norm": 18.27402114868164,
+      "learning_rate": 1.8157553445174247e-05,
+      "loss": 3.1177,
       "step": 2203
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 16.964014053344727,
-      "learning_rate": 1.5583842838528618e-05,
-      "loss": 3.0422,
+      "epoch": 0.28,
+      "grad_norm": 21.276901245117188,
+      "learning_rate": 1.815671673011756e-05,
+      "loss": 1.0634,
       "step": 2204
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 24.55567169189453,
-      "learning_rate": 1.5581838227924228e-05,
-      "loss": 2.4687,
+      "epoch": 0.28,
+      "grad_norm": 12.932526588439941,
+      "learning_rate": 1.815588001506087e-05,
+      "loss": 1.5843,
       "step": 2205
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 12.898763656616211,
-      "learning_rate": 1.5579833617319838e-05,
-      "loss": 1.8264,
+      "epoch": 0.28,
+      "grad_norm": 13.635990142822266,
+      "learning_rate": 1.8155043300004185e-05,
+      "loss": 2.964,
       "step": 2206
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 14.676507949829102,
-      "learning_rate": 1.5577829006715444e-05,
-      "loss": 1.942,
+      "epoch": 0.28,
+      "grad_norm": 8.697279930114746,
+      "learning_rate": 1.81542065849475e-05,
+      "loss": 3.8756,
       "step": 2207
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 21.591907501220703,
-      "learning_rate": 1.5575824396111058e-05,
-      "loss": 2.193,
+      "epoch": 0.28,
+      "grad_norm": 12.8943452835083,
+      "learning_rate": 1.8153369869890812e-05,
+      "loss": 2.171,
       "step": 2208
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 22.116395950317383,
-      "learning_rate": 1.5573819785506668e-05,
-      "loss": 1.8098,
+      "epoch": 0.28,
+      "grad_norm": 14.581306457519531,
+      "learning_rate": 1.8152533154834123e-05,
+      "loss": 2.451,
       "step": 2209
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 17.048828125,
-      "learning_rate": 1.5571815174902275e-05,
-      "loss": 2.4309,
+      "epoch": 0.28,
+      "grad_norm": 14.167716979980469,
+      "learning_rate": 1.8151696439777436e-05,
+      "loss": 4.8333,
       "step": 2210
     },
     {
-      "epoch": 0.66,
-      "grad_norm": 17.50075912475586,
-      "learning_rate": 1.5569810564297888e-05,
-      "loss": 3.5984,
+      "epoch": 0.28,
+      "grad_norm": 40.78544998168945,
+      "learning_rate": 1.815085972472075e-05,
+      "loss": 3.447,
       "step": 2211
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 42.22193145751953,
-      "learning_rate": 1.5567805953693495e-05,
-      "loss": 2.6408,
+      "epoch": 0.28,
+      "grad_norm": 13.381462097167969,
+      "learning_rate": 1.815002300966406e-05,
+      "loss": 1.2005,
       "step": 2212
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 16.618879318237305,
-      "learning_rate": 1.5565801343089105e-05,
-      "loss": 1.9168,
+      "epoch": 0.28,
+      "grad_norm": 12.654776573181152,
+      "learning_rate": 1.8149186294607374e-05,
+      "loss": 3.1519,
       "step": 2213
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 20.115299224853516,
-      "learning_rate": 1.5563796732484715e-05,
-      "loss": 2.5345,
+      "epoch": 0.28,
+      "grad_norm": 15.005374908447266,
+      "learning_rate": 1.8148349579550687e-05,
+      "loss": 3.3541,
       "step": 2214
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 9.575844764709473,
-      "learning_rate": 1.5561792121880325e-05,
-      "loss": 2.5817,
+      "epoch": 0.28,
+      "grad_norm": 14.365894317626953,
+      "learning_rate": 1.8147512864493998e-05,
+      "loss": 2.2676,
       "step": 2215
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 16.084333419799805,
-      "learning_rate": 1.5559787511275935e-05,
-      "loss": 2.8486,
+      "epoch": 0.28,
+      "grad_norm": 28.20368766784668,
+      "learning_rate": 1.814667614943731e-05,
+      "loss": 2.9791,
       "step": 2216
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 9.678840637207031,
-      "learning_rate": 1.5557782900671545e-05,
-      "loss": 2.0502,
+      "epoch": 0.28,
+      "grad_norm": 9.189824104309082,
+      "learning_rate": 1.814583943438062e-05,
+      "loss": 0.9645,
       "step": 2217
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 25.424049377441406,
-      "learning_rate": 1.5555778290067155e-05,
-      "loss": 2.3241,
+      "epoch": 0.28,
+      "grad_norm": 10.372204780578613,
+      "learning_rate": 1.8145002719323935e-05,
+      "loss": 1.2545,
       "step": 2218
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 14.452130317687988,
-      "learning_rate": 1.5553773679462765e-05,
-      "loss": 2.2107,
+      "epoch": 0.28,
+      "grad_norm": 11.494043350219727,
+      "learning_rate": 1.814416600426725e-05,
+      "loss": 1.083,
       "step": 2219
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 34.95116424560547,
-      "learning_rate": 1.5551769068858375e-05,
-      "loss": 2.9797,
+      "epoch": 0.28,
+      "grad_norm": 14.215128898620605,
+      "learning_rate": 1.814332928921056e-05,
+      "loss": 1.2213,
       "step": 2220
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 10.273086547851562,
-      "learning_rate": 1.5549764458253986e-05,
-      "loss": 1.676,
+      "epoch": 0.28,
+      "grad_norm": 13.489676475524902,
+      "learning_rate": 1.8142492574153873e-05,
+      "loss": 3.0496,
       "step": 2221
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 16.252601623535156,
-      "learning_rate": 1.5547759847649596e-05,
-      "loss": 2.6206,
+      "epoch": 0.28,
+      "grad_norm": 15.215855598449707,
+      "learning_rate": 1.8141655859097186e-05,
+      "loss": 2.7327,
       "step": 2222
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 10.310708045959473,
-      "learning_rate": 1.5545755237045206e-05,
-      "loss": 2.1657,
+      "epoch": 0.28,
+      "grad_norm": 9.644189834594727,
+      "learning_rate": 1.8140819144040497e-05,
+      "loss": 0.9311,
       "step": 2223
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 11.262001037597656,
-      "learning_rate": 1.5543750626440816e-05,
-      "loss": 1.9435,
+      "epoch": 0.28,
+      "grad_norm": 7.358780860900879,
+      "learning_rate": 1.813998242898381e-05,
+      "loss": 1.696,
       "step": 2224
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 20.996740341186523,
-      "learning_rate": 1.5541746015836426e-05,
-      "loss": 2.7842,
+      "epoch": 0.28,
+      "grad_norm": 22.613245010375977,
+      "learning_rate": 1.8139145713927124e-05,
+      "loss": 2.0394,
       "step": 2225
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 25.680212020874023,
-      "learning_rate": 1.5539741405232033e-05,
-      "loss": 2.4817,
+      "epoch": 0.28,
+      "grad_norm": 11.782620429992676,
+      "learning_rate": 1.8138308998870434e-05,
+      "loss": 1.9239,
       "step": 2226
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 11.317098617553711,
-      "learning_rate": 1.5537736794627646e-05,
-      "loss": 1.2717,
+      "epoch": 0.28,
+      "grad_norm": 25.47810935974121,
+      "learning_rate": 1.8137472283813748e-05,
+      "loss": 3.1678,
       "step": 2227
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 14.966314315795898,
-      "learning_rate": 1.5535732184023256e-05,
-      "loss": 2.5357,
+      "epoch": 0.28,
+      "grad_norm": 12.637117385864258,
+      "learning_rate": 1.813663556875706e-05,
+      "loss": 2.666,
       "step": 2228
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 22.922651290893555,
-      "learning_rate": 1.5533727573418863e-05,
-      "loss": 3.0417,
+      "epoch": 0.28,
+      "grad_norm": 18.55261993408203,
+      "learning_rate": 1.8135798853700375e-05,
+      "loss": 2.0364,
       "step": 2229
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 21.57832908630371,
-      "learning_rate": 1.5531722962814476e-05,
-      "loss": 2.1712,
+      "epoch": 0.28,
+      "grad_norm": 16.093608856201172,
+      "learning_rate": 1.8134962138643685e-05,
+      "loss": 4.8497,
       "step": 2230
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 22.31909942626953,
-      "learning_rate": 1.5529718352210083e-05,
-      "loss": 2.7337,
+      "epoch": 0.28,
+      "grad_norm": 14.051105499267578,
+      "learning_rate": 1.8134125423587e-05,
+      "loss": 2.0579,
       "step": 2231
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 10.842183113098145,
-      "learning_rate": 1.5527713741605693e-05,
-      "loss": 1.886,
+      "epoch": 0.28,
+      "grad_norm": 17.554685592651367,
+      "learning_rate": 1.8133288708530313e-05,
+      "loss": 2.6494,
       "step": 2232
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 17.99361228942871,
-      "learning_rate": 1.5525709131001307e-05,
-      "loss": 2.3479,
+      "epoch": 0.28,
+      "grad_norm": 22.024080276489258,
+      "learning_rate": 1.8132451993473623e-05,
+      "loss": 3.6693,
       "step": 2233
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 36.20960998535156,
-      "learning_rate": 1.5523704520396913e-05,
-      "loss": 2.8728,
+      "epoch": 0.28,
+      "grad_norm": 19.70965003967285,
+      "learning_rate": 1.8131615278416937e-05,
+      "loss": 2.4351,
       "step": 2234
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 14.748534202575684,
-      "learning_rate": 1.5521699909792523e-05,
-      "loss": 2.0896,
+      "epoch": 0.28,
+      "grad_norm": 12.084982872009277,
+      "learning_rate": 1.813077856336025e-05,
+      "loss": 1.5988,
       "step": 2235
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 10.2194242477417,
-      "learning_rate": 1.5519695299188133e-05,
-      "loss": 2.0784,
+      "epoch": 0.28,
+      "grad_norm": 9.847620010375977,
+      "learning_rate": 1.8129941848303564e-05,
+      "loss": 1.1611,
       "step": 2236
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 19.986635208129883,
-      "learning_rate": 1.5517690688583744e-05,
-      "loss": 2.4334,
+      "epoch": 0.28,
+      "grad_norm": 18.409879684448242,
+      "learning_rate": 1.8129105133246874e-05,
+      "loss": 2.0993,
       "step": 2237
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 19.66067886352539,
-      "learning_rate": 1.5515686077979354e-05,
-      "loss": 2.684,
+      "epoch": 0.28,
+      "grad_norm": 8.946453094482422,
+      "learning_rate": 1.8128268418190188e-05,
+      "loss": 1.5504,
       "step": 2238
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 70.83124542236328,
-      "learning_rate": 1.5513681467374964e-05,
-      "loss": 2.5154,
+      "epoch": 0.28,
+      "grad_norm": 18.910524368286133,
+      "learning_rate": 1.81274317031335e-05,
+      "loss": 1.4382,
       "step": 2239
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 23.612783432006836,
-      "learning_rate": 1.5511676856770574e-05,
-      "loss": 2.886,
+      "epoch": 0.28,
+      "grad_norm": 10.438148498535156,
+      "learning_rate": 1.8126594988076812e-05,
+      "loss": 1.585,
       "step": 2240
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 15.822755813598633,
-      "learning_rate": 1.5509672246166184e-05,
-      "loss": 2.0681,
+      "epoch": 0.28,
+      "grad_norm": 22.281301498413086,
+      "learning_rate": 1.8125758273020125e-05,
+      "loss": 1.9504,
       "step": 2241
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 25.98380470275879,
-      "learning_rate": 1.5507667635561794e-05,
-      "loss": 2.8038,
+      "epoch": 0.28,
+      "grad_norm": 9.4960355758667,
+      "learning_rate": 1.812492155796344e-05,
+      "loss": 1.915,
       "step": 2242
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 27.250049591064453,
-      "learning_rate": 1.5505663024957404e-05,
-      "loss": 2.5481,
+      "epoch": 0.28,
+      "grad_norm": 18.760650634765625,
+      "learning_rate": 1.812408484290675e-05,
+      "loss": 1.9631,
       "step": 2243
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 36.15570831298828,
-      "learning_rate": 1.5503658414353014e-05,
-      "loss": 2.1858,
+      "epoch": 0.28,
+      "grad_norm": 50.82057189941406,
+      "learning_rate": 1.8123248127850063e-05,
+      "loss": 2.4465,
       "step": 2244
     },
     {
-      "epoch": 0.67,
-      "grad_norm": 10.935135841369629,
-      "learning_rate": 1.5501653803748624e-05,
-      "loss": 1.1376,
+      "epoch": 0.28,
+      "grad_norm": 14.840272903442383,
+      "learning_rate": 1.8122411412793373e-05,
+      "loss": 4.0736,
       "step": 2245
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 21.463272094726562,
-      "learning_rate": 1.5499649193144234e-05,
-      "loss": 2.5708,
+      "epoch": 0.28,
+      "grad_norm": 27.440397262573242,
+      "learning_rate": 1.8121574697736687e-05,
+      "loss": 3.074,
       "step": 2246
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 12.168373107910156,
-      "learning_rate": 1.5497644582539844e-05,
-      "loss": 2.3351,
+      "epoch": 0.28,
+      "grad_norm": 11.00921630859375,
+      "learning_rate": 1.812073798268e-05,
+      "loss": 3.1562,
       "step": 2247
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 20.284778594970703,
-      "learning_rate": 1.549563997193545e-05,
-      "loss": 2.2741,
+      "epoch": 0.28,
+      "grad_norm": 9.403813362121582,
+      "learning_rate": 1.811990126762331e-05,
+      "loss": 2.7584,
       "step": 2248
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 21.697620391845703,
-      "learning_rate": 1.5493635361331065e-05,
-      "loss": 3.3955,
+      "epoch": 0.28,
+      "grad_norm": 9.529999732971191,
+      "learning_rate": 1.8119064552566624e-05,
+      "loss": 2.7479,
       "step": 2249
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 15.422367095947266,
-      "learning_rate": 1.549163075072667e-05,
-      "loss": 2.879,
+      "epoch": 0.28,
+      "grad_norm": 9.866724967956543,
+      "learning_rate": 1.8118227837509938e-05,
+      "loss": 2.7984,
       "step": 2250
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 17.34280776977539,
-      "learning_rate": 1.548962614012228e-05,
-      "loss": 2.5866,
+      "epoch": 0.28,
+      "grad_norm": 23.09242057800293,
+      "learning_rate": 1.811739112245325e-05,
+      "loss": 3.4462,
       "step": 2251
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 17.975322723388672,
-      "learning_rate": 1.5487621529517895e-05,
-      "loss": 2.3053,
+      "epoch": 0.28,
+      "grad_norm": 10.374611854553223,
+      "learning_rate": 1.8116554407396562e-05,
+      "loss": 1.1058,
       "step": 2252
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 30.330408096313477,
-      "learning_rate": 1.54856169189135e-05,
-      "loss": 3.002,
+      "epoch": 0.28,
+      "grad_norm": 22.321758270263672,
+      "learning_rate": 1.8115717692339876e-05,
+      "loss": 1.7572,
       "step": 2253
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 17.363624572753906,
-      "learning_rate": 1.548361230830911e-05,
-      "loss": 2.323,
+      "epoch": 0.28,
+      "grad_norm": 21.010087966918945,
+      "learning_rate": 1.8114880977283186e-05,
+      "loss": 3.2659,
       "step": 2254
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 14.340827941894531,
-      "learning_rate": 1.548160769770472e-05,
-      "loss": 1.8183,
+      "epoch": 0.28,
+      "grad_norm": 11.888099670410156,
+      "learning_rate": 1.81140442622265e-05,
+      "loss": 1.7821,
       "step": 2255
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 11.480280876159668,
-      "learning_rate": 1.5479603087100332e-05,
-      "loss": 1.9501,
+      "epoch": 0.28,
+      "grad_norm": 14.751751899719238,
+      "learning_rate": 1.8113207547169813e-05,
+      "loss": 3.6896,
       "step": 2256
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 10.279202461242676,
-      "learning_rate": 1.5477598476495942e-05,
-      "loss": 2.1804,
+      "epoch": 0.28,
+      "grad_norm": 19.622156143188477,
+      "learning_rate": 1.8112370832113127e-05,
+      "loss": 1.9646,
       "step": 2257
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 7.849579334259033,
-      "learning_rate": 1.5475593865891552e-05,
-      "loss": 2.1096,
+      "epoch": 0.28,
+      "grad_norm": 24.655155181884766,
+      "learning_rate": 1.8111534117056437e-05,
+      "loss": 3.0499,
       "step": 2258
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 15.246369361877441,
-      "learning_rate": 1.5473589255287162e-05,
-      "loss": 2.5827,
+      "epoch": 0.28,
+      "grad_norm": 22.66373062133789,
+      "learning_rate": 1.811069740199975e-05,
+      "loss": 2.765,
       "step": 2259
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 11.893373489379883,
-      "learning_rate": 1.5471584644682772e-05,
-      "loss": 1.4826,
+      "epoch": 0.28,
+      "grad_norm": 14.57512378692627,
+      "learning_rate": 1.8109860686943064e-05,
+      "loss": 3.0029,
       "step": 2260
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 22.876691818237305,
-      "learning_rate": 1.5469580034078382e-05,
-      "loss": 2.7584,
+      "epoch": 0.28,
+      "grad_norm": 7.522922039031982,
+      "learning_rate": 1.8109023971886375e-05,
+      "loss": 0.8426,
       "step": 2261
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 22.17548179626465,
-      "learning_rate": 1.546757542347399e-05,
-      "loss": 2.8086,
+      "epoch": 0.28,
+      "grad_norm": 19.920307159423828,
+      "learning_rate": 1.8108187256829688e-05,
+      "loss": 3.0463,
       "step": 2262
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 19.29637908935547,
-      "learning_rate": 1.5465570812869602e-05,
-      "loss": 2.4008,
+      "epoch": 0.28,
+      "grad_norm": 7.890814304351807,
+      "learning_rate": 1.8107350541773002e-05,
+      "loss": 3.4674,
       "step": 2263
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 18.765453338623047,
-      "learning_rate": 1.5463566202265212e-05,
-      "loss": 2.7496,
+      "epoch": 0.28,
+      "grad_norm": 16.865583419799805,
+      "learning_rate": 1.8106513826716316e-05,
+      "loss": 1.4047,
       "step": 2264
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 15.087578773498535,
-      "learning_rate": 1.5461561591660823e-05,
-      "loss": 2.6039,
+      "epoch": 0.28,
+      "grad_norm": 24.833620071411133,
+      "learning_rate": 1.8105677111659626e-05,
+      "loss": 3.3807,
       "step": 2265
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 9.94806957244873,
-      "learning_rate": 1.5459556981056433e-05,
-      "loss": 1.0356,
+      "epoch": 0.28,
+      "grad_norm": 24.19931411743164,
+      "learning_rate": 1.810484039660294e-05,
+      "loss": 1.5681,
       "step": 2266
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 20.82723617553711,
-      "learning_rate": 1.545755237045204e-05,
-      "loss": 2.3768,
+      "epoch": 0.28,
+      "grad_norm": 13.846046447753906,
+      "learning_rate": 1.8104003681546253e-05,
+      "loss": 3.3777,
       "step": 2267
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 11.183643341064453,
-      "learning_rate": 1.5455547759847653e-05,
-      "loss": 2.0118,
+      "epoch": 0.28,
+      "grad_norm": 9.359046936035156,
+      "learning_rate": 1.8103166966489563e-05,
+      "loss": 1.6452,
       "step": 2268
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 11.922686576843262,
-      "learning_rate": 1.545354314924326e-05,
-      "loss": 2.3468,
+      "epoch": 0.28,
+      "grad_norm": 11.376022338867188,
+      "learning_rate": 1.8102330251432877e-05,
+      "loss": 1.0759,
       "step": 2269
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 10.599013328552246,
-      "learning_rate": 1.545153853863887e-05,
-      "loss": 1.6547,
+      "epoch": 0.28,
+      "grad_norm": 21.50180435180664,
+      "learning_rate": 1.8101493536376187e-05,
+      "loss": 1.9246,
       "step": 2270
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 9.613250732421875,
-      "learning_rate": 1.5449533928034483e-05,
-      "loss": 2.3148,
+      "epoch": 0.29,
+      "grad_norm": 12.495444297790527,
+      "learning_rate": 1.81006568213195e-05,
+      "loss": 2.3863,
       "step": 2271
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 20.166654586791992,
-      "learning_rate": 1.544752931743009e-05,
-      "loss": 2.3727,
+      "epoch": 0.29,
+      "grad_norm": 8.674577713012695,
+      "learning_rate": 1.8099820106262815e-05,
+      "loss": 0.7363,
       "step": 2272
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 14.838998794555664,
-      "learning_rate": 1.54455247068257e-05,
-      "loss": 1.8119,
+      "epoch": 0.29,
+      "grad_norm": 10.121570587158203,
+      "learning_rate": 1.8098983391206125e-05,
+      "loss": 1.8153,
       "step": 2273
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 15.088861465454102,
-      "learning_rate": 1.544352009622131e-05,
-      "loss": 1.2957,
+      "epoch": 0.29,
+      "grad_norm": 18.356172561645508,
+      "learning_rate": 1.809814667614944e-05,
+      "loss": 1.9094,
       "step": 2274
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 16.902069091796875,
-      "learning_rate": 1.544151548561692e-05,
-      "loss": 2.637,
+      "epoch": 0.29,
+      "grad_norm": 10.023639678955078,
+      "learning_rate": 1.8097309961092752e-05,
+      "loss": 0.9665,
       "step": 2275
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 39.02754592895508,
-      "learning_rate": 1.543951087501253e-05,
-      "loss": 2.1298,
+      "epoch": 0.29,
+      "grad_norm": 27.681039810180664,
+      "learning_rate": 1.8096473246036062e-05,
+      "loss": 1.8041,
       "step": 2276
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 13.487964630126953,
-      "learning_rate": 1.543750626440814e-05,
-      "loss": 2.4923,
+      "epoch": 0.29,
+      "grad_norm": 17.335979461669922,
+      "learning_rate": 1.8095636530979376e-05,
+      "loss": 1.7052,
       "step": 2277
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 28.00465965270996,
-      "learning_rate": 1.543550165380375e-05,
-      "loss": 2.9849,
+      "epoch": 0.29,
+      "grad_norm": 8.04034423828125,
+      "learning_rate": 1.809479981592269e-05,
+      "loss": 2.1678,
       "step": 2278
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 12.890748977661133,
-      "learning_rate": 1.543349704319936e-05,
-      "loss": 2.6115,
+      "epoch": 0.29,
+      "grad_norm": 10.94822883605957,
+      "learning_rate": 1.8093963100866e-05,
+      "loss": 2.656,
       "step": 2279
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 14.795134544372559,
-      "learning_rate": 1.543149243259497e-05,
-      "loss": 1.9501,
-      "step": 2280
-    },
-    {
-      "epoch": 0.69,
-      "eval_loss": 0.27775686979293823,
-      "eval_runtime": 43.6573,
-      "eval_samples_per_second": 33.878,
-      "eval_steps_per_second": 33.878,
+      "epoch": 0.29,
+      "grad_norm": 12.786932945251465,
+      "learning_rate": 1.8093126385809314e-05,
+      "loss": 1.507,
       "step": 2280
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 17.335163116455078,
-      "learning_rate": 1.5429487821990577e-05,
-      "loss": 2.4545,
+      "epoch": 0.29,
+      "grad_norm": 14.343307495117188,
+      "learning_rate": 1.8092289670752627e-05,
+      "loss": 2.3808,
       "step": 2281
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 19.485933303833008,
-      "learning_rate": 1.542748321138619e-05,
-      "loss": 1.9536,
+      "epoch": 0.29,
+      "grad_norm": 18.072303771972656,
+      "learning_rate": 1.8091452955695938e-05,
+      "loss": 1.1889,
       "step": 2282
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 17.225101470947266,
-      "learning_rate": 1.54254786007818e-05,
-      "loss": 1.7893,
+      "epoch": 0.29,
+      "grad_norm": 20.963382720947266,
+      "learning_rate": 1.809061624063925e-05,
+      "loss": 3.0112,
       "step": 2283
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 12.618383407592773,
-      "learning_rate": 1.5423473990177407e-05,
-      "loss": 2.0978,
+      "epoch": 0.29,
+      "grad_norm": 13.710307121276855,
+      "learning_rate": 1.8089779525582565e-05,
+      "loss": 2.0044,
       "step": 2284
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 21.85410499572754,
-      "learning_rate": 1.542146937957302e-05,
-      "loss": 2.1725,
+      "epoch": 0.29,
+      "grad_norm": 10.80042839050293,
+      "learning_rate": 1.808894281052588e-05,
+      "loss": 1.8148,
       "step": 2285
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 13.22624397277832,
-      "learning_rate": 1.5419464768968628e-05,
-      "loss": 1.9183,
+      "epoch": 0.29,
+      "grad_norm": 11.106392860412598,
+      "learning_rate": 1.808810609546919e-05,
+      "loss": 3.2806,
       "step": 2286
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 14.312644004821777,
-      "learning_rate": 1.5417460158364238e-05,
-      "loss": 1.764,
+      "epoch": 0.29,
+      "grad_norm": 12.298428535461426,
+      "learning_rate": 1.8087269380412502e-05,
+      "loss": 3.5851,
       "step": 2287
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 17.018287658691406,
-      "learning_rate": 1.541545554775985e-05,
-      "loss": 1.6649,
+      "epoch": 0.29,
+      "grad_norm": 16.31171989440918,
+      "learning_rate": 1.8086432665355816e-05,
+      "loss": 2.3223,
       "step": 2288
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 19.239450454711914,
-      "learning_rate": 1.5413450937155458e-05,
-      "loss": 2.6163,
+      "epoch": 0.29,
+      "grad_norm": 29.622272491455078,
+      "learning_rate": 1.8085595950299126e-05,
+      "loss": 1.8154,
       "step": 2289
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 8.873007774353027,
-      "learning_rate": 1.5411446326551068e-05,
-      "loss": 2.4475,
+      "epoch": 0.29,
+      "grad_norm": 8.188494682312012,
+      "learning_rate": 1.808475923524244e-05,
+      "loss": 0.9744,
       "step": 2290
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 34.43484878540039,
-      "learning_rate": 1.5409441715946678e-05,
-      "loss": 3.171,
+      "epoch": 0.29,
+      "grad_norm": 16.53975486755371,
+      "learning_rate": 1.8083922520185754e-05,
+      "loss": 3.5978,
       "step": 2291
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 13.385025024414062,
-      "learning_rate": 1.5407437105342288e-05,
-      "loss": 2.4151,
+      "epoch": 0.29,
+      "grad_norm": 13.368452072143555,
+      "learning_rate": 1.8083085805129067e-05,
+      "loss": 3.1355,
       "step": 2292
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 23.839799880981445,
-      "learning_rate": 1.5405432494737898e-05,
-      "loss": 3.4928,
+      "epoch": 0.29,
+      "grad_norm": 15.011205673217773,
+      "learning_rate": 1.8082249090072378e-05,
+      "loss": 3.6453,
       "step": 2293
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 13.791393280029297,
-      "learning_rate": 1.5403427884133508e-05,
-      "loss": 1.5945,
+      "epoch": 0.29,
+      "grad_norm": 7.424562931060791,
+      "learning_rate": 1.808141237501569e-05,
+      "loss": 0.7775,
       "step": 2294
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 15.341072082519531,
-      "learning_rate": 1.5401423273529118e-05,
-      "loss": 1.6761,
+      "epoch": 0.29,
+      "grad_norm": 16.634830474853516,
+      "learning_rate": 1.8080575659959005e-05,
+      "loss": 1.8949,
       "step": 2295
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 16.246986389160156,
-      "learning_rate": 1.539941866292473e-05,
-      "loss": 1.9111,
+      "epoch": 0.29,
+      "grad_norm": 17.14111328125,
+      "learning_rate": 1.8079738944902315e-05,
+      "loss": 3.6253,
       "step": 2296
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 21.339174270629883,
-      "learning_rate": 1.539741405232034e-05,
-      "loss": 2.3112,
+      "epoch": 0.29,
+      "grad_norm": 16.829402923583984,
+      "learning_rate": 1.807890222984563e-05,
+      "loss": 2.47,
       "step": 2297
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 19.16057777404785,
-      "learning_rate": 1.539540944171595e-05,
-      "loss": 1.8671,
+      "epoch": 0.29,
+      "grad_norm": 21.543121337890625,
+      "learning_rate": 1.807806551478894e-05,
+      "loss": 2.6683,
       "step": 2298
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 10.452445983886719,
-      "learning_rate": 1.539340483111156e-05,
-      "loss": 1.1408,
+      "epoch": 0.29,
+      "grad_norm": 19.138608932495117,
+      "learning_rate": 1.8077228799732253e-05,
+      "loss": 1.9961,
       "step": 2299
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 12.354015350341797,
-      "learning_rate": 1.5391400220507165e-05,
-      "loss": 2.4143,
+      "epoch": 0.29,
+      "grad_norm": 10.469704627990723,
+      "learning_rate": 1.8076392084675566e-05,
+      "loss": 1.7902,
       "step": 2300
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 12.557357788085938,
-      "learning_rate": 1.538939560990278e-05,
-      "loss": 1.8995,
+      "epoch": 0.29,
+      "grad_norm": 17.763158798217773,
+      "learning_rate": 1.8075555369618877e-05,
+      "loss": 2.2397,
       "step": 2301
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 16.86603546142578,
-      "learning_rate": 1.538739099929839e-05,
-      "loss": 2.0175,
+      "epoch": 0.29,
+      "grad_norm": 22.89780616760254,
+      "learning_rate": 1.807471865456219e-05,
+      "loss": 3.7139,
       "step": 2302
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 22.319530487060547,
-      "learning_rate": 1.5385386388693996e-05,
-      "loss": 3.6267,
+      "epoch": 0.29,
+      "grad_norm": 10.7036714553833,
+      "learning_rate": 1.80738819395055e-05,
+      "loss": 2.5254,
       "step": 2303
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 21.675506591796875,
-      "learning_rate": 1.538338177808961e-05,
-      "loss": 2.8276,
+      "epoch": 0.29,
+      "grad_norm": 13.151268005371094,
+      "learning_rate": 1.8073045224448814e-05,
+      "loss": 2.0414,
       "step": 2304
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 25.780988693237305,
-      "learning_rate": 1.5381377167485216e-05,
-      "loss": 2.1939,
+      "epoch": 0.29,
+      "grad_norm": 13.832414627075195,
+      "learning_rate": 1.8072208509392128e-05,
+      "loss": 1.5838,
       "step": 2305
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 21.725067138671875,
-      "learning_rate": 1.5379372556880826e-05,
-      "loss": 2.4273,
+      "epoch": 0.29,
+      "grad_norm": 15.708240509033203,
+      "learning_rate": 1.807137179433544e-05,
+      "loss": 1.7294,
       "step": 2306
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 13.079160690307617,
-      "learning_rate": 1.537736794627644e-05,
-      "loss": 2.748,
+      "epoch": 0.29,
+      "grad_norm": 13.90760326385498,
+      "learning_rate": 1.807053507927875e-05,
+      "loss": 2.8137,
       "step": 2307
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 70.98606872558594,
-      "learning_rate": 1.5375363335672046e-05,
-      "loss": 3.8798,
+      "epoch": 0.29,
+      "grad_norm": 9.946630477905273,
+      "learning_rate": 1.8069698364222065e-05,
+      "loss": 2.2482,
       "step": 2308
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 30.048757553100586,
-      "learning_rate": 1.5373358725067656e-05,
-      "loss": 1.797,
+      "epoch": 0.29,
+      "grad_norm": 10.19784927368164,
+      "learning_rate": 1.806886164916538e-05,
+      "loss": 1.0116,
       "step": 2309
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 19.938982009887695,
-      "learning_rate": 1.5371354114463266e-05,
-      "loss": 2.9236,
+      "epoch": 0.29,
+      "grad_norm": 8.03913688659668,
+      "learning_rate": 1.806802493410869e-05,
+      "loss": 1.6907,
       "step": 2310
     },
     {
-      "epoch": 0.69,
-      "grad_norm": 19.096269607543945,
-      "learning_rate": 1.5369349503858876e-05,
-      "loss": 3.0036,
+      "epoch": 0.29,
+      "grad_norm": 10.337860107421875,
+      "learning_rate": 1.8067188219052003e-05,
+      "loss": 2.5385,
       "step": 2311
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 16.5783634185791,
-      "learning_rate": 1.5367344893254486e-05,
-      "loss": 2.3479,
+      "epoch": 0.29,
+      "grad_norm": 18.989763259887695,
+      "learning_rate": 1.8066351503995317e-05,
+      "loss": 3.5781,
       "step": 2312
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 11.655875205993652,
-      "learning_rate": 1.5365340282650096e-05,
-      "loss": 2.0129,
+      "epoch": 0.29,
+      "grad_norm": 12.300849914550781,
+      "learning_rate": 1.806551478893863e-05,
+      "loss": 3.7757,
       "step": 2313
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 12.814486503601074,
-      "learning_rate": 1.5363335672045706e-05,
-      "loss": 2.283,
+      "epoch": 0.29,
+      "grad_norm": 83.98790740966797,
+      "learning_rate": 1.806467807388194e-05,
+      "loss": 1.8347,
       "step": 2314
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 12.23183822631836,
-      "learning_rate": 1.5361331061441317e-05,
-      "loss": 2.0969,
+      "epoch": 0.29,
+      "grad_norm": 12.340867042541504,
+      "learning_rate": 1.8063841358825254e-05,
+      "loss": 2.302,
       "step": 2315
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 23.845361709594727,
-      "learning_rate": 1.5359326450836927e-05,
-      "loss": 2.6363,
+      "epoch": 0.29,
+      "grad_norm": 15.20331859588623,
+      "learning_rate": 1.8063004643768568e-05,
+      "loss": 1.1444,
       "step": 2316
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 22.525117874145508,
-      "learning_rate": 1.5357321840232537e-05,
-      "loss": 2.7907,
+      "epoch": 0.29,
+      "grad_norm": 8.978998184204102,
+      "learning_rate": 1.8062167928711878e-05,
+      "loss": 1.4229,
       "step": 2317
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 35.314666748046875,
-      "learning_rate": 1.5355317229628147e-05,
-      "loss": 2.434,
+      "epoch": 0.29,
+      "grad_norm": 13.404335021972656,
+      "learning_rate": 1.806133121365519e-05,
+      "loss": 2.6539,
       "step": 2318
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 10.943061828613281,
-      "learning_rate": 1.5353312619023757e-05,
-      "loss": 1.8863,
+      "epoch": 0.29,
+      "grad_norm": 13.901432991027832,
+      "learning_rate": 1.8060494498598505e-05,
+      "loss": 1.7427,
       "step": 2319
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 16.291006088256836,
-      "learning_rate": 1.5351308008419367e-05,
-      "loss": 1.8541,
+      "epoch": 0.29,
+      "grad_norm": 17.40442657470703,
+      "learning_rate": 1.8059657783541816e-05,
+      "loss": 3.412,
       "step": 2320
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 12.258851051330566,
-      "learning_rate": 1.5349303397814977e-05,
-      "loss": 2.3217,
+      "epoch": 0.29,
+      "grad_norm": 13.368928909301758,
+      "learning_rate": 1.805882106848513e-05,
+      "loss": 1.7645,
       "step": 2321
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 12.439240455627441,
-      "learning_rate": 1.5347298787210584e-05,
-      "loss": 2.1214,
+      "epoch": 0.29,
+      "grad_norm": 31.82679557800293,
+      "learning_rate": 1.8057984353428443e-05,
+      "loss": 2.8897,
       "step": 2322
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 24.879018783569336,
-      "learning_rate": 1.5345294176606197e-05,
-      "loss": 2.4778,
+      "epoch": 0.29,
+      "grad_norm": 18.478469848632812,
+      "learning_rate": 1.8057147638371753e-05,
+      "loss": 2.7191,
       "step": 2323
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 14.103094100952148,
-      "learning_rate": 1.5343289566001804e-05,
-      "loss": 3.2186,
+      "epoch": 0.29,
+      "grad_norm": 19.266084671020508,
+      "learning_rate": 1.8056310923315067e-05,
+      "loss": 1.6193,
       "step": 2324
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 16.0571346282959,
-      "learning_rate": 1.5341284955397414e-05,
-      "loss": 1.7382,
+      "epoch": 0.29,
+      "grad_norm": 8.752995491027832,
+      "learning_rate": 1.805547420825838e-05,
+      "loss": 0.6984,
       "step": 2325
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 14.569293975830078,
-      "learning_rate": 1.5339280344793027e-05,
-      "loss": 2.6728,
+      "epoch": 0.29,
+      "grad_norm": 6.644970893859863,
+      "learning_rate": 1.805463749320169e-05,
+      "loss": 1.3478,
       "step": 2326
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 19.245534896850586,
-      "learning_rate": 1.5337275734188634e-05,
-      "loss": 2.1571,
+      "epoch": 0.29,
+      "grad_norm": 9.999451637268066,
+      "learning_rate": 1.8053800778145004e-05,
+      "loss": 1.3756,
       "step": 2327
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 35.83250427246094,
-      "learning_rate": 1.5335271123584244e-05,
-      "loss": 3.574,
+      "epoch": 0.29,
+      "grad_norm": 9.796958923339844,
+      "learning_rate": 1.8052964063088318e-05,
+      "loss": 1.3429,
       "step": 2328
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 22.268566131591797,
-      "learning_rate": 1.5333266512979854e-05,
-      "loss": 2.8414,
+      "epoch": 0.29,
+      "grad_norm": 13.106298446655273,
+      "learning_rate": 1.8052127348031628e-05,
+      "loss": 1.7764,
       "step": 2329
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 13.335097312927246,
-      "learning_rate": 1.5331261902375464e-05,
-      "loss": 2.2446,
+      "epoch": 0.29,
+      "grad_norm": 14.986753463745117,
+      "learning_rate": 1.8051290632974942e-05,
+      "loss": 0.9655,
       "step": 2330
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 21.128978729248047,
-      "learning_rate": 1.5329257291771075e-05,
-      "loss": 2.1243,
+      "epoch": 0.29,
+      "grad_norm": 15.127579689025879,
+      "learning_rate": 1.8050453917918252e-05,
+      "loss": 2.3554,
       "step": 2331
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 15.031562805175781,
-      "learning_rate": 1.5327252681166685e-05,
-      "loss": 2.1856,
+      "epoch": 0.29,
+      "grad_norm": 11.969268798828125,
+      "learning_rate": 1.8049617202861566e-05,
+      "loss": 2.6626,
       "step": 2332
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 24.040231704711914,
-      "learning_rate": 1.5325248070562295e-05,
-      "loss": 2.7496,
+      "epoch": 0.29,
+      "grad_norm": 10.662618637084961,
+      "learning_rate": 1.804878048780488e-05,
+      "loss": 2.7009,
       "step": 2333
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 21.210121154785156,
-      "learning_rate": 1.5323243459957905e-05,
-      "loss": 2.1727,
+      "epoch": 0.29,
+      "grad_norm": 10.531999588012695,
+      "learning_rate": 1.804794377274819e-05,
+      "loss": 1.5971,
       "step": 2334
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 14.999757766723633,
-      "learning_rate": 1.5321238849353515e-05,
-      "loss": 1.8486,
+      "epoch": 0.29,
+      "grad_norm": 11.093401908874512,
+      "learning_rate": 1.8047107057691503e-05,
+      "loss": 2.4525,
       "step": 2335
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 21.83721160888672,
-      "learning_rate": 1.5319234238749125e-05,
-      "loss": 2.4315,
+      "epoch": 0.29,
+      "grad_norm": 15.833708763122559,
+      "learning_rate": 1.8046270342634817e-05,
+      "loss": 3.6285,
       "step": 2336
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 10.136129379272461,
-      "learning_rate": 1.5317229628144735e-05,
-      "loss": 1.8972,
+      "epoch": 0.29,
+      "grad_norm": 12.989253044128418,
+      "learning_rate": 1.804543362757813e-05,
+      "loss": 1.9026,
       "step": 2337
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 14.037278175354004,
-      "learning_rate": 1.5315225017540345e-05,
-      "loss": 1.9107,
+      "epoch": 0.29,
+      "grad_norm": 15.443399429321289,
+      "learning_rate": 1.804459691252144e-05,
+      "loss": 2.1652,
       "step": 2338
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 12.793769836425781,
-      "learning_rate": 1.5313220406935955e-05,
-      "loss": 1.8938,
+      "epoch": 0.29,
+      "grad_norm": 18.758634567260742,
+      "learning_rate": 1.8043760197464755e-05,
+      "loss": 3.9224,
       "step": 2339
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 46.66535568237305,
-      "learning_rate": 1.5311215796331565e-05,
-      "loss": 3.1435,
+      "epoch": 0.29,
+      "grad_norm": 14.861361503601074,
+      "learning_rate": 1.8042923482408068e-05,
+      "loss": 3.4067,
       "step": 2340
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 24.93342399597168,
-      "learning_rate": 1.5309211185727172e-05,
-      "loss": 2.3244,
+      "epoch": 0.29,
+      "grad_norm": 50.64338684082031,
+      "learning_rate": 1.804208676735138e-05,
+      "loss": 3.8945,
       "step": 2341
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 17.550573348999023,
-      "learning_rate": 1.5307206575122785e-05,
-      "loss": 1.4479,
+      "epoch": 0.29,
+      "grad_norm": 10.846446990966797,
+      "learning_rate": 1.8041250052294692e-05,
+      "loss": 3.1811,
       "step": 2342
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 23.929763793945312,
-      "learning_rate": 1.5305201964518392e-05,
-      "loss": 2.6339,
+      "epoch": 0.29,
+      "grad_norm": 8.02611255645752,
+      "learning_rate": 1.8040413337238006e-05,
+      "loss": 0.9488,
       "step": 2343
     },
     {
-      "epoch": 0.7,
-      "grad_norm": 18.37746238708496,
-      "learning_rate": 1.5303197353914002e-05,
-      "loss": 1.916,
+      "epoch": 0.29,
+      "grad_norm": 23.65349578857422,
+      "learning_rate": 1.803957662218132e-05,
+      "loss": 1.5523,
       "step": 2344
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 32.94269943237305,
-      "learning_rate": 1.5301192743309616e-05,
-      "loss": 2.7435,
+      "epoch": 0.29,
+      "grad_norm": 25.175683975219727,
+      "learning_rate": 1.803873990712463e-05,
+      "loss": 3.753,
       "step": 2345
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 14.292465209960938,
-      "learning_rate": 1.5299188132705222e-05,
-      "loss": 1.8105,
+      "epoch": 0.29,
+      "grad_norm": 14.84681510925293,
+      "learning_rate": 1.8037903192067943e-05,
+      "loss": 1.4025,
       "step": 2346
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 12.711530685424805,
-      "learning_rate": 1.5297183522100832e-05,
-      "loss": 1.9106,
+      "epoch": 0.29,
+      "grad_norm": 8.496464729309082,
+      "learning_rate": 1.8037066477011257e-05,
+      "loss": 1.9409,
       "step": 2347
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 16.853271484375,
-      "learning_rate": 1.5295178911496443e-05,
-      "loss": 1.9825,
+      "epoch": 0.29,
+      "grad_norm": 9.734086036682129,
+      "learning_rate": 1.8036229761954567e-05,
+      "loss": 1.5214,
       "step": 2348
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 17.740068435668945,
-      "learning_rate": 1.5293174300892053e-05,
-      "loss": 1.5591,
+      "epoch": 0.29,
+      "grad_norm": 10.568015098571777,
+      "learning_rate": 1.803539304689788e-05,
+      "loss": 1.6711,
       "step": 2349
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 19.151620864868164,
-      "learning_rate": 1.5291169690287663e-05,
-      "loss": 3.9344,
+      "epoch": 0.29,
+      "grad_norm": 18.141674041748047,
+      "learning_rate": 1.8034556331841195e-05,
+      "loss": 2.8501,
       "step": 2350
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 14.547835350036621,
-      "learning_rate": 1.5289165079683273e-05,
-      "loss": 2.0179,
+      "epoch": 0.3,
+      "grad_norm": 10.642548561096191,
+      "learning_rate": 1.8033719616784505e-05,
+      "loss": 0.7835,
       "step": 2351
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 14.701192855834961,
-      "learning_rate": 1.5287160469078883e-05,
-      "loss": 2.1844,
+      "epoch": 0.3,
+      "grad_norm": 11.74319839477539,
+      "learning_rate": 1.803288290172782e-05,
+      "loss": 1.8802,
       "step": 2352
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 20.547576904296875,
-      "learning_rate": 1.5285155858474493e-05,
-      "loss": 2.5008,
+      "epoch": 0.3,
+      "grad_norm": 19.375816345214844,
+      "learning_rate": 1.8032046186671132e-05,
+      "loss": 1.8404,
       "step": 2353
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 13.004501342773438,
-      "learning_rate": 1.5283151247870103e-05,
-      "loss": 1.8492,
+      "epoch": 0.3,
+      "grad_norm": 8.290349006652832,
+      "learning_rate": 1.8031209471614442e-05,
+      "loss": 0.4438,
       "step": 2354
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 18.574153900146484,
-      "learning_rate": 1.528114663726571e-05,
-      "loss": 1.7889,
+      "epoch": 0.3,
+      "grad_norm": 10.130553245544434,
+      "learning_rate": 1.8030372756557756e-05,
+      "loss": 1.3496,
       "step": 2355
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 21.489116668701172,
-      "learning_rate": 1.5279142026661323e-05,
-      "loss": 2.2442,
+      "epoch": 0.3,
+      "grad_norm": 13.821660995483398,
+      "learning_rate": 1.8029536041501066e-05,
+      "loss": 1.9425,
       "step": 2356
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 11.905816078186035,
-      "learning_rate": 1.5277137416056933e-05,
-      "loss": 1.6658,
+      "epoch": 0.3,
+      "grad_norm": 15.139850616455078,
+      "learning_rate": 1.802869932644438e-05,
+      "loss": 4.0812,
       "step": 2357
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 16.373111724853516,
-      "learning_rate": 1.527513280545254e-05,
-      "loss": 1.7949,
+      "epoch": 0.3,
+      "grad_norm": 22.017791748046875,
+      "learning_rate": 1.8027862611387694e-05,
+      "loss": 3.3954,
       "step": 2358
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 14.20676326751709,
-      "learning_rate": 1.5273128194848153e-05,
-      "loss": 2.0334,
+      "epoch": 0.3,
+      "grad_norm": 5.533475875854492,
+      "learning_rate": 1.8027025896331004e-05,
+      "loss": 0.5178,
       "step": 2359
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 14.752305030822754,
-      "learning_rate": 1.527112358424376e-05,
-      "loss": 2.1005,
+      "epoch": 0.3,
+      "grad_norm": 17.990325927734375,
+      "learning_rate": 1.8026189181274317e-05,
+      "loss": 2.4957,
       "step": 2360
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 13.963676452636719,
-      "learning_rate": 1.526911897363937e-05,
-      "loss": 2.9856,
+      "epoch": 0.3,
+      "grad_norm": 13.587029457092285,
+      "learning_rate": 1.802535246621763e-05,
+      "loss": 1.8408,
       "step": 2361
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 11.170400619506836,
-      "learning_rate": 1.5267114363034984e-05,
-      "loss": 1.8121,
+      "epoch": 0.3,
+      "grad_norm": 6.42975378036499,
+      "learning_rate": 1.802451575116094e-05,
+      "loss": 1.965,
       "step": 2362
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 17.13726043701172,
-      "learning_rate": 1.526510975243059e-05,
-      "loss": 2.1476,
+      "epoch": 0.3,
+      "grad_norm": 24.11678123474121,
+      "learning_rate": 1.8023679036104255e-05,
+      "loss": 2.8355,
       "step": 2363
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 79.9980697631836,
-      "learning_rate": 1.52631051418262e-05,
-      "loss": 3.3246,
+      "epoch": 0.3,
+      "grad_norm": 7.636974334716797,
+      "learning_rate": 1.802284232104757e-05,
+      "loss": 0.7662,
       "step": 2364
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 21.73980712890625,
-      "learning_rate": 1.526110053122181e-05,
-      "loss": 2.9358,
+      "epoch": 0.3,
+      "grad_norm": 9.104048728942871,
+      "learning_rate": 1.8022005605990882e-05,
+      "loss": 1.3221,
       "step": 2365
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 13.559164047241211,
-      "learning_rate": 1.525909592061742e-05,
-      "loss": 2.0114,
+      "epoch": 0.3,
+      "grad_norm": 20.817089080810547,
+      "learning_rate": 1.8021168890934193e-05,
+      "loss": 2.1917,
       "step": 2366
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 14.406414031982422,
-      "learning_rate": 1.525709131001303e-05,
-      "loss": 1.1554,
+      "epoch": 0.3,
+      "grad_norm": 12.99535846710205,
+      "learning_rate": 1.8020332175877506e-05,
+      "loss": 1.9057,
       "step": 2367
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 11.396732330322266,
-      "learning_rate": 1.5255086699408641e-05,
-      "loss": 2.1299,
+      "epoch": 0.3,
+      "grad_norm": 17.81003761291504,
+      "learning_rate": 1.801949546082082e-05,
+      "loss": 1.3834,
       "step": 2368
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 22.816299438476562,
-      "learning_rate": 1.5253082088804253e-05,
-      "loss": 2.8568,
+      "epoch": 0.3,
+      "grad_norm": 23.86208152770996,
+      "learning_rate": 1.801865874576413e-05,
+      "loss": 3.3293,
       "step": 2369
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 19.31545066833496,
-      "learning_rate": 1.5251077478199861e-05,
-      "loss": 2.9191,
+      "epoch": 0.3,
+      "grad_norm": 12.152969360351562,
+      "learning_rate": 1.8017822030707444e-05,
+      "loss": 1.6765,
       "step": 2370
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 20.314498901367188,
-      "learning_rate": 1.5249072867595471e-05,
-      "loss": 1.6643,
+      "epoch": 0.3,
+      "grad_norm": 38.21599197387695,
+      "learning_rate": 1.8016985315650757e-05,
+      "loss": 1.3122,
       "step": 2371
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 21.570932388305664,
-      "learning_rate": 1.524706825699108e-05,
-      "loss": 2.4345,
+      "epoch": 0.3,
+      "grad_norm": 19.01609230041504,
+      "learning_rate": 1.801614860059407e-05,
+      "loss": 3.2396,
       "step": 2372
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 12.214997291564941,
-      "learning_rate": 1.5245063646386691e-05,
-      "loss": 1.8357,
+      "epoch": 0.3,
+      "grad_norm": 30.038503646850586,
+      "learning_rate": 1.801531188553738e-05,
+      "loss": 2.8291,
       "step": 2373
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 15.286229133605957,
-      "learning_rate": 1.5243059035782301e-05,
-      "loss": 1.6836,
+      "epoch": 0.3,
+      "grad_norm": 16.662433624267578,
+      "learning_rate": 1.8014475170480695e-05,
+      "loss": 2.5137,
       "step": 2374
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 18.07011604309082,
-      "learning_rate": 1.524105442517791e-05,
-      "loss": 2.6134,
+      "epoch": 0.3,
+      "grad_norm": 13.939742088317871,
+      "learning_rate": 1.801363845542401e-05,
+      "loss": 1.557,
       "step": 2375
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 13.150678634643555,
-      "learning_rate": 1.5239049814573522e-05,
-      "loss": 2.336,
+      "epoch": 0.3,
+      "grad_norm": 8.272865295410156,
+      "learning_rate": 1.801280174036732e-05,
+      "loss": 2.2698,
       "step": 2376
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 14.786459922790527,
-      "learning_rate": 1.523704520396913e-05,
-      "loss": 1.8392,
+      "epoch": 0.3,
+      "grad_norm": 22.545406341552734,
+      "learning_rate": 1.8011965025310633e-05,
+      "loss": 4.3373,
       "step": 2377
     },
     {
-      "epoch": 0.71,
-      "grad_norm": 12.723509788513184,
-      "learning_rate": 1.523504059336474e-05,
-      "loss": 2.4568,
+      "epoch": 0.3,
+      "grad_norm": 18.655996322631836,
+      "learning_rate": 1.8011128310253946e-05,
+      "loss": 3.117,
       "step": 2378
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 17.796947479248047,
-      "learning_rate": 1.5233035982760348e-05,
-      "loss": 2.8035,
+      "epoch": 0.3,
+      "grad_norm": 9.008397102355957,
+      "learning_rate": 1.8010291595197256e-05,
+      "loss": 2.2974,
       "step": 2379
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 7.9330854415893555,
-      "learning_rate": 1.523103137215596e-05,
-      "loss": 1.0556,
+      "epoch": 0.3,
+      "grad_norm": 11.4547758102417,
+      "learning_rate": 1.800945488014057e-05,
+      "loss": 2.4967,
       "step": 2380
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 20.23641586303711,
-      "learning_rate": 1.522902676155157e-05,
-      "loss": 2.4805,
+      "epoch": 0.3,
+      "grad_norm": 15.299418449401855,
+      "learning_rate": 1.800861816508388e-05,
+      "loss": 3.1593,
       "step": 2381
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 14.428129196166992,
-      "learning_rate": 1.5227022150947179e-05,
-      "loss": 2.3162,
+      "epoch": 0.3,
+      "grad_norm": 9.479822158813477,
+      "learning_rate": 1.8007781450027194e-05,
+      "loss": 1.0239,
       "step": 2382
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 19.485916137695312,
-      "learning_rate": 1.522501754034279e-05,
-      "loss": 2.8538,
+      "epoch": 0.3,
+      "grad_norm": 17.806331634521484,
+      "learning_rate": 1.8006944734970508e-05,
+      "loss": 3.0007,
       "step": 2383
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 23.651782989501953,
-      "learning_rate": 1.5223012929738399e-05,
-      "loss": 2.2296,
+      "epoch": 0.3,
+      "grad_norm": 8.056856155395508,
+      "learning_rate": 1.8006108019913818e-05,
+      "loss": 1.9608,
       "step": 2384
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 20.420745849609375,
-      "learning_rate": 1.5221008319134009e-05,
-      "loss": 2.8712,
+      "epoch": 0.3,
+      "grad_norm": 24.79678726196289,
+      "learning_rate": 1.800527130485713e-05,
+      "loss": 4.0215,
       "step": 2385
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 13.436408042907715,
-      "learning_rate": 1.5219003708529619e-05,
-      "loss": 2.4493,
+      "epoch": 0.3,
+      "grad_norm": 7.317874908447266,
+      "learning_rate": 1.8004434589800445e-05,
+      "loss": 2.7529,
       "step": 2386
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 9.011188507080078,
-      "learning_rate": 1.5216999097925229e-05,
-      "loss": 1.8621,
+      "epoch": 0.3,
+      "grad_norm": 11.379854202270508,
+      "learning_rate": 1.8003597874743755e-05,
+      "loss": 3.1902,
       "step": 2387
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 24.090200424194336,
-      "learning_rate": 1.521499448732084e-05,
-      "loss": 2.7823,
+      "epoch": 0.3,
+      "grad_norm": 7.822103023529053,
+      "learning_rate": 1.800276115968707e-05,
+      "loss": 0.639,
       "step": 2388
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 22.469762802124023,
-      "learning_rate": 1.521298987671645e-05,
-      "loss": 2.1563,
+      "epoch": 0.3,
+      "grad_norm": 10.701630592346191,
+      "learning_rate": 1.8001924444630383e-05,
+      "loss": 1.1973,
       "step": 2389
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 20.96700668334961,
-      "learning_rate": 1.521098526611206e-05,
-      "loss": 2.3123,
+      "epoch": 0.3,
+      "grad_norm": 4.938593864440918,
+      "learning_rate": 1.8001087729573693e-05,
+      "loss": 0.2573,
       "step": 2390
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 13.538233757019043,
-      "learning_rate": 1.5208980655507668e-05,
-      "loss": 1.8336,
+      "epoch": 0.3,
+      "grad_norm": 14.3922119140625,
+      "learning_rate": 1.8000251014517007e-05,
+      "loss": 3.0215,
       "step": 2391
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 19.34774398803711,
-      "learning_rate": 1.520697604490328e-05,
-      "loss": 2.2524,
+      "epoch": 0.3,
+      "grad_norm": 32.6680908203125,
+      "learning_rate": 1.799941429946032e-05,
+      "loss": 3.7568,
       "step": 2392
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 23.267269134521484,
-      "learning_rate": 1.520497143429889e-05,
-      "loss": 3.195,
+      "epoch": 0.3,
+      "grad_norm": 61.6491584777832,
+      "learning_rate": 1.7998577584403634e-05,
+      "loss": 3.0698,
       "step": 2393
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 43.13511657714844,
-      "learning_rate": 1.5202966823694498e-05,
-      "loss": 2.9556,
+      "epoch": 0.3,
+      "grad_norm": 9.55639362335205,
+      "learning_rate": 1.7997740869346944e-05,
+      "loss": 1.6788,
       "step": 2394
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 12.41313362121582,
-      "learning_rate": 1.520096221309011e-05,
-      "loss": 2.4461,
+      "epoch": 0.3,
+      "grad_norm": 16.137413024902344,
+      "learning_rate": 1.7996904154290258e-05,
+      "loss": 1.9237,
       "step": 2395
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 16.363807678222656,
-      "learning_rate": 1.5198957602485718e-05,
-      "loss": 2.2686,
+      "epoch": 0.3,
+      "grad_norm": 22.509485244750977,
+      "learning_rate": 1.799606743923357e-05,
+      "loss": 3.7312,
       "step": 2396
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 18.332998275756836,
-      "learning_rate": 1.5196952991881328e-05,
-      "loss": 2.3713,
+      "epoch": 0.3,
+      "grad_norm": 15.136279106140137,
+      "learning_rate": 1.7995230724176882e-05,
+      "loss": 2.7475,
       "step": 2397
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 15.461286544799805,
-      "learning_rate": 1.5194948381276937e-05,
-      "loss": 2.0574,
+      "epoch": 0.3,
+      "grad_norm": 8.282188415527344,
+      "learning_rate": 1.7994394009120195e-05,
+      "loss": 2.5397,
       "step": 2398
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 32.771728515625,
-      "learning_rate": 1.5192943770672548e-05,
-      "loss": 2.7732,
+      "epoch": 0.3,
+      "grad_norm": 21.439027786254883,
+      "learning_rate": 1.799355729406351e-05,
+      "loss": 2.3746,
       "step": 2399
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 15.005850791931152,
-      "learning_rate": 1.5190939160068158e-05,
-      "loss": 2.0059,
+      "epoch": 0.3,
+      "grad_norm": 15.517571449279785,
+      "learning_rate": 1.7992720579006823e-05,
+      "loss": 1.6803,
       "step": 2400
     },
     {
-      "epoch": 0.72,
-      "eval_loss": 0.2799599766731262,
-      "eval_runtime": 43.2863,
-      "eval_samples_per_second": 34.168,
-      "eval_steps_per_second": 34.168,
+      "epoch": 0.3,
+      "eval_loss": 0.14937622845172882,
+      "eval_runtime": 93.6313,
+      "eval_samples_per_second": 37.829,
+      "eval_steps_per_second": 37.829,
       "step": 2400
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 18.537982940673828,
-      "learning_rate": 1.5188934549463767e-05,
-      "loss": 2.3335,
+      "epoch": 0.3,
+      "grad_norm": 6.837953090667725,
+      "learning_rate": 1.7991883863950133e-05,
+      "loss": 1.6215,
       "step": 2401
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 28.75159454345703,
-      "learning_rate": 1.5186929938859379e-05,
-      "loss": 2.5293,
+      "epoch": 0.3,
+      "grad_norm": 9.428163528442383,
+      "learning_rate": 1.7991047148893447e-05,
+      "loss": 1.1436,
       "step": 2402
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 19.689199447631836,
-      "learning_rate": 1.5184925328254987e-05,
-      "loss": 3.4202,
+      "epoch": 0.3,
+      "grad_norm": 12.235206604003906,
+      "learning_rate": 1.799021043383676e-05,
+      "loss": 2.8743,
       "step": 2403
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 16.85619354248047,
-      "learning_rate": 1.5182920717650597e-05,
-      "loss": 2.1249,
+      "epoch": 0.3,
+      "grad_norm": 16.27605628967285,
+      "learning_rate": 1.798937371878007e-05,
+      "loss": 4.4169,
       "step": 2404
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 15.209796905517578,
-      "learning_rate": 1.5180916107046209e-05,
-      "loss": 2.1942,
+      "epoch": 0.3,
+      "grad_norm": 23.758119583129883,
+      "learning_rate": 1.7988537003723384e-05,
+      "loss": 5.8766,
       "step": 2405
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 19.145883560180664,
-      "learning_rate": 1.5178911496441817e-05,
-      "loss": 2.3506,
+      "epoch": 0.3,
+      "grad_norm": 7.420042514801025,
+      "learning_rate": 1.7987700288666698e-05,
+      "loss": 3.2284,
       "step": 2406
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 13.587536811828613,
-      "learning_rate": 1.5176906885837427e-05,
-      "loss": 2.483,
+      "epoch": 0.3,
+      "grad_norm": 10.003791809082031,
+      "learning_rate": 1.7986863573610008e-05,
+      "loss": 0.8627,
       "step": 2407
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 14.813488960266113,
-      "learning_rate": 1.5174902275233036e-05,
-      "loss": 1.6431,
+      "epoch": 0.3,
+      "grad_norm": 9.849807739257812,
+      "learning_rate": 1.7986026858553322e-05,
+      "loss": 3.7947,
       "step": 2408
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 7.390352249145508,
-      "learning_rate": 1.5172897664628648e-05,
-      "loss": 1.329,
+      "epoch": 0.3,
+      "grad_norm": 13.273680686950684,
+      "learning_rate": 1.7985190143496632e-05,
+      "loss": 3.0434,
       "step": 2409
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 124.1446304321289,
-      "learning_rate": 1.5170893054024256e-05,
-      "loss": 2.8408,
+      "epoch": 0.3,
+      "grad_norm": 23.513259887695312,
+      "learning_rate": 1.7984353428439946e-05,
+      "loss": 1.9259,
       "step": 2410
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 31.081998825073242,
-      "learning_rate": 1.5168888443419868e-05,
-      "loss": 2.8232,
+      "epoch": 0.3,
+      "grad_norm": 5.763058662414551,
+      "learning_rate": 1.798351671338326e-05,
+      "loss": 0.232,
       "step": 2411
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 18.641647338867188,
-      "learning_rate": 1.5166883832815478e-05,
-      "loss": 1.8497,
+      "epoch": 0.3,
+      "grad_norm": 27.04803466796875,
+      "learning_rate": 1.798267999832657e-05,
+      "loss": 3.0844,
       "step": 2412
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 17.504941940307617,
-      "learning_rate": 1.5164879222211086e-05,
-      "loss": 2.4579,
+      "epoch": 0.3,
+      "grad_norm": 10.611763954162598,
+      "learning_rate": 1.7981843283269883e-05,
+      "loss": 3.6497,
       "step": 2413
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 11.096992492675781,
-      "learning_rate": 1.5162874611606698e-05,
-      "loss": 2.1797,
+      "epoch": 0.3,
+      "grad_norm": 8.736261367797852,
+      "learning_rate": 1.7981006568213197e-05,
+      "loss": 2.5193,
       "step": 2414
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 14.132731437683105,
-      "learning_rate": 1.5160870001002306e-05,
-      "loss": 1.5483,
+      "epoch": 0.3,
+      "grad_norm": 27.361827850341797,
+      "learning_rate": 1.7980169853156507e-05,
+      "loss": 1.9007,
       "step": 2415
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 49.675132751464844,
-      "learning_rate": 1.5158865390397916e-05,
-      "loss": 2.4514,
+      "epoch": 0.3,
+      "grad_norm": 6.731814861297607,
+      "learning_rate": 1.797933313809982e-05,
+      "loss": 1.5845,
       "step": 2416
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 12.76675796508789,
-      "learning_rate": 1.5156860779793528e-05,
-      "loss": 1.2279,
+      "epoch": 0.3,
+      "grad_norm": 12.201180458068848,
+      "learning_rate": 1.7978496423043134e-05,
+      "loss": 2.0134,
       "step": 2417
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 10.700181007385254,
-      "learning_rate": 1.5154856169189137e-05,
-      "loss": 2.0745,
+      "epoch": 0.3,
+      "grad_norm": 16.464221954345703,
+      "learning_rate": 1.7977659707986445e-05,
+      "loss": 2.1593,
       "step": 2418
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 17.42853546142578,
-      "learning_rate": 1.5152851558584747e-05,
-      "loss": 2.7441,
+      "epoch": 0.3,
+      "grad_norm": 10.752031326293945,
+      "learning_rate": 1.797682299292976e-05,
+      "loss": 1.6055,
       "step": 2419
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 8.670104026794434,
-      "learning_rate": 1.5150846947980355e-05,
-      "loss": 1.8953,
+      "epoch": 0.3,
+      "grad_norm": 15.68919849395752,
+      "learning_rate": 1.7975986277873072e-05,
+      "loss": 1.6983,
       "step": 2420
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 24.383087158203125,
-      "learning_rate": 1.5148842337375967e-05,
-      "loss": 2.6507,
+      "epoch": 0.3,
+      "grad_norm": 15.785876274108887,
+      "learning_rate": 1.7975149562816386e-05,
+      "loss": 1.7882,
       "step": 2421
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 12.565937042236328,
-      "learning_rate": 1.5146837726771575e-05,
-      "loss": 1.5887,
+      "epoch": 0.3,
+      "grad_norm": 7.236278533935547,
+      "learning_rate": 1.7974312847759696e-05,
+      "loss": 1.3744,
       "step": 2422
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 22.554595947265625,
-      "learning_rate": 1.5144833116167185e-05,
-      "loss": 2.4942,
+      "epoch": 0.3,
+      "grad_norm": 17.342849731445312,
+      "learning_rate": 1.797347613270301e-05,
+      "loss": 4.0136,
       "step": 2423
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 8.884187698364258,
-      "learning_rate": 1.5142828505562797e-05,
-      "loss": 1.2549,
+      "epoch": 0.3,
+      "grad_norm": 12.334209442138672,
+      "learning_rate": 1.7972639417646323e-05,
+      "loss": 2.6935,
       "step": 2424
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 15.001145362854004,
-      "learning_rate": 1.5140823894958406e-05,
-      "loss": 2.1248,
+      "epoch": 0.3,
+      "grad_norm": 20.487749099731445,
+      "learning_rate": 1.7971802702589633e-05,
+      "loss": 1.966,
       "step": 2425
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 13.459707260131836,
-      "learning_rate": 1.5138819284354016e-05,
-      "loss": 2.0756,
+      "epoch": 0.3,
+      "grad_norm": 36.53834533691406,
+      "learning_rate": 1.7970965987532947e-05,
+      "loss": 3.8524,
       "step": 2426
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 15.579310417175293,
-      "learning_rate": 1.5136814673749624e-05,
-      "loss": 2.8265,
+      "epoch": 0.3,
+      "grad_norm": 10.720416069030762,
+      "learning_rate": 1.797012927247626e-05,
+      "loss": 0.6583,
       "step": 2427
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 12.242110252380371,
-      "learning_rate": 1.5134810063145236e-05,
-      "loss": 2.2286,
+      "epoch": 0.3,
+      "grad_norm": 11.420807838439941,
+      "learning_rate": 1.7969292557419574e-05,
+      "loss": 2.1776,
       "step": 2428
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 30.915904998779297,
-      "learning_rate": 1.5132805452540844e-05,
-      "loss": 2.2951,
+      "epoch": 0.3,
+      "grad_norm": 22.484323501586914,
+      "learning_rate": 1.7968455842362885e-05,
+      "loss": 4.2214,
       "step": 2429
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 15.82615852355957,
-      "learning_rate": 1.5130800841936454e-05,
-      "loss": 2.0723,
+      "epoch": 0.3,
+      "grad_norm": 12.813691139221191,
+      "learning_rate": 1.79676191273062e-05,
+      "loss": 3.3768,
       "step": 2430
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 15.196529388427734,
-      "learning_rate": 1.5128796231332066e-05,
-      "loss": 2.2142,
+      "epoch": 0.31,
+      "grad_norm": 30.47164535522461,
+      "learning_rate": 1.7966782412249512e-05,
+      "loss": 2.4702,
       "step": 2431
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 43.0781135559082,
-      "learning_rate": 1.5126791620727674e-05,
-      "loss": 3.249,
+      "epoch": 0.31,
+      "grad_norm": 12.788331985473633,
+      "learning_rate": 1.7965945697192822e-05,
+      "loss": 2.0694,
       "step": 2432
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 10.74176025390625,
-      "learning_rate": 1.5124787010123284e-05,
-      "loss": 1.763,
+      "epoch": 0.31,
+      "grad_norm": 14.737342834472656,
+      "learning_rate": 1.7965108982136136e-05,
+      "loss": 1.2319,
       "step": 2433
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 14.15688705444336,
-      "learning_rate": 1.5122782399518895e-05,
-      "loss": 1.8488,
+      "epoch": 0.31,
+      "grad_norm": 16.493003845214844,
+      "learning_rate": 1.7964272267079446e-05,
+      "loss": 1.5493,
       "step": 2434
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 16.205501556396484,
-      "learning_rate": 1.5120777788914505e-05,
-      "loss": 2.8312,
+      "epoch": 0.31,
+      "grad_norm": 8.839437484741211,
+      "learning_rate": 1.796343555202276e-05,
+      "loss": 2.0695,
       "step": 2435
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 23.575292587280273,
-      "learning_rate": 1.5118773178310115e-05,
-      "loss": 1.8974,
+      "epoch": 0.31,
+      "grad_norm": 11.862302780151367,
+      "learning_rate": 1.7962598836966073e-05,
+      "loss": 2.0112,
       "step": 2436
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 12.507033348083496,
-      "learning_rate": 1.5116768567705725e-05,
-      "loss": 2.0892,
+      "epoch": 0.31,
+      "grad_norm": 9.451587677001953,
+      "learning_rate": 1.7961762121909384e-05,
+      "loss": 2.5243,
       "step": 2437
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 11.459385871887207,
-      "learning_rate": 1.5114763957101335e-05,
-      "loss": 2.232,
+      "epoch": 0.31,
+      "grad_norm": 17.41968536376953,
+      "learning_rate": 1.7960925406852697e-05,
+      "loss": 1.4975,
       "step": 2438
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 13.064870834350586,
-      "learning_rate": 1.5112759346496943e-05,
-      "loss": 2.9855,
+      "epoch": 0.31,
+      "grad_norm": 8.577614784240723,
+      "learning_rate": 1.796008869179601e-05,
+      "loss": 3.4461,
       "step": 2439
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 21.083112716674805,
-      "learning_rate": 1.5110754735892555e-05,
-      "loss": 2.7381,
+      "epoch": 0.31,
+      "grad_norm": 8.34268856048584,
+      "learning_rate": 1.795925197673932e-05,
+      "loss": 0.9853,
       "step": 2440
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 16.428953170776367,
-      "learning_rate": 1.5108750125288163e-05,
-      "loss": 2.8427,
+      "epoch": 0.31,
+      "grad_norm": 19.16912841796875,
+      "learning_rate": 1.7958415261682635e-05,
+      "loss": 2.7435,
       "step": 2441
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 13.690245628356934,
-      "learning_rate": 1.5106745514683774e-05,
-      "loss": 1.9179,
+      "epoch": 0.31,
+      "grad_norm": 14.721293449401855,
+      "learning_rate": 1.795757854662595e-05,
+      "loss": 4.1871,
       "step": 2442
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 13.37304401397705,
-      "learning_rate": 1.5104740904079385e-05,
-      "loss": 1.8193,
+      "epoch": 0.31,
+      "grad_norm": 18.573949813842773,
+      "learning_rate": 1.795674183156926e-05,
+      "loss": 2.5489,
       "step": 2443
     },
     {
-      "epoch": 0.73,
-      "grad_norm": 10.359079360961914,
-      "learning_rate": 1.5102736293474994e-05,
-      "loss": 2.5558,
+      "epoch": 0.31,
+      "grad_norm": 14.529594421386719,
+      "learning_rate": 1.7955905116512572e-05,
+      "loss": 3.8118,
       "step": 2444
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 12.4136381149292,
-      "learning_rate": 1.5100731682870604e-05,
-      "loss": 2.191,
+      "epoch": 0.31,
+      "grad_norm": 8.199951171875,
+      "learning_rate": 1.7955068401455886e-05,
+      "loss": 2.007,
       "step": 2445
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 9.771052360534668,
-      "learning_rate": 1.5098727072266212e-05,
-      "loss": 1.5917,
+      "epoch": 0.31,
+      "grad_norm": 196.73123168945312,
+      "learning_rate": 1.7954231686399196e-05,
+      "loss": 3.6715,
       "step": 2446
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 12.872698783874512,
-      "learning_rate": 1.5096722461661824e-05,
-      "loss": 1.6524,
+      "epoch": 0.31,
+      "grad_norm": 13.598657608032227,
+      "learning_rate": 1.795339497134251e-05,
+      "loss": 0.8439,
       "step": 2447
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 28.35697364807129,
-      "learning_rate": 1.5094717851057434e-05,
-      "loss": 2.569,
+      "epoch": 0.31,
+      "grad_norm": 15.713129997253418,
+      "learning_rate": 1.7952558256285824e-05,
+      "loss": 4.4424,
       "step": 2448
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 18.459850311279297,
-      "learning_rate": 1.5092713240453042e-05,
-      "loss": 2.1667,
+      "epoch": 0.31,
+      "grad_norm": 11.949092864990234,
+      "learning_rate": 1.7951721541229137e-05,
+      "loss": 2.1954,
       "step": 2449
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 12.4850492477417,
-      "learning_rate": 1.5090708629848654e-05,
-      "loss": 2.2356,
+      "epoch": 0.31,
+      "grad_norm": 17.65532112121582,
+      "learning_rate": 1.7950884826172448e-05,
+      "loss": 1.9608,
       "step": 2450
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 21.161081314086914,
-      "learning_rate": 1.5088704019244263e-05,
-      "loss": 2.1366,
+      "epoch": 0.31,
+      "grad_norm": 25.382020950317383,
+      "learning_rate": 1.795004811111576e-05,
+      "loss": 3.9136,
       "step": 2451
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 22.59266471862793,
-      "learning_rate": 1.5086699408639873e-05,
-      "loss": 2.4504,
+      "epoch": 0.31,
+      "grad_norm": 10.749935150146484,
+      "learning_rate": 1.7949211396059075e-05,
+      "loss": 2.5799,
       "step": 2452
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 11.43664836883545,
-      "learning_rate": 1.5084694798035481e-05,
-      "loss": 2.434,
+      "epoch": 0.31,
+      "grad_norm": 9.420044898986816,
+      "learning_rate": 1.7948374681002385e-05,
+      "loss": 0.2144,
       "step": 2453
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 12.218484878540039,
-      "learning_rate": 1.5082690187431093e-05,
-      "loss": 1.2119,
+      "epoch": 0.31,
+      "grad_norm": 20.088491439819336,
+      "learning_rate": 1.79475379659457e-05,
+      "loss": 3.2122,
       "step": 2454
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 10.82869815826416,
-      "learning_rate": 1.5080685576826703e-05,
-      "loss": 1.8022,
+      "epoch": 0.31,
+      "grad_norm": 10.93820858001709,
+      "learning_rate": 1.7946701250889012e-05,
+      "loss": 1.3192,
       "step": 2455
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 17.006765365600586,
-      "learning_rate": 1.5078680966222311e-05,
-      "loss": 2.1377,
+      "epoch": 0.31,
+      "grad_norm": 13.77704906463623,
+      "learning_rate": 1.7945864535832326e-05,
+      "loss": 1.2174,
       "step": 2456
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 21.112869262695312,
-      "learning_rate": 1.5076676355617923e-05,
-      "loss": 2.799,
+      "epoch": 0.31,
+      "grad_norm": 11.582486152648926,
+      "learning_rate": 1.7945027820775636e-05,
+      "loss": 2.6627,
       "step": 2457
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 24.335466384887695,
-      "learning_rate": 1.5074671745013532e-05,
-      "loss": 2.4533,
+      "epoch": 0.31,
+      "grad_norm": 13.67541790008545,
+      "learning_rate": 1.794419110571895e-05,
+      "loss": 2.7674,
       "step": 2458
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 20.010364532470703,
-      "learning_rate": 1.5072667134409142e-05,
-      "loss": 1.9907,
+      "epoch": 0.31,
+      "grad_norm": 14.400947570800781,
+      "learning_rate": 1.7943354390662264e-05,
+      "loss": 0.8326,
       "step": 2459
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 34.753047943115234,
-      "learning_rate": 1.5070662523804753e-05,
-      "loss": 2.0948,
+      "epoch": 0.31,
+      "grad_norm": 11.022433280944824,
+      "learning_rate": 1.7942517675605574e-05,
+      "loss": 2.6278,
       "step": 2460
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 19.572534561157227,
-      "learning_rate": 1.5068657913200362e-05,
-      "loss": 2.1058,
+      "epoch": 0.31,
+      "grad_norm": 8.584240913391113,
+      "learning_rate": 1.7941680960548888e-05,
+      "loss": 1.2164,
       "step": 2461
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 13.721003532409668,
-      "learning_rate": 1.5066653302595972e-05,
-      "loss": 1.4905,
+      "epoch": 0.31,
+      "grad_norm": 11.213518142700195,
+      "learning_rate": 1.7940844245492198e-05,
+      "loss": 1.2233,
       "step": 2462
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 14.475379943847656,
-      "learning_rate": 1.5064648691991582e-05,
-      "loss": 1.7658,
+      "epoch": 0.31,
+      "grad_norm": 17.76910400390625,
+      "learning_rate": 1.794000753043551e-05,
+      "loss": 2.8389,
       "step": 2463
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 24.027629852294922,
-      "learning_rate": 1.5062644081387192e-05,
-      "loss": 2.6237,
+      "epoch": 0.31,
+      "grad_norm": 10.86572265625,
+      "learning_rate": 1.7939170815378825e-05,
+      "loss": 1.7707,
       "step": 2464
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 15.826460838317871,
-      "learning_rate": 1.50606394707828e-05,
-      "loss": 2.6549,
+      "epoch": 0.31,
+      "grad_norm": 16.505216598510742,
+      "learning_rate": 1.7938334100322135e-05,
+      "loss": 1.9726,
       "step": 2465
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 43.245018005371094,
-      "learning_rate": 1.5058634860178412e-05,
-      "loss": 2.9669,
+      "epoch": 0.31,
+      "grad_norm": 18.240514755249023,
+      "learning_rate": 1.793749738526545e-05,
+      "loss": 4.2534,
       "step": 2466
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 13.400920867919922,
-      "learning_rate": 1.5056630249574022e-05,
-      "loss": 1.9453,
+      "epoch": 0.31,
+      "grad_norm": 11.698887825012207,
+      "learning_rate": 1.793666067020876e-05,
+      "loss": 2.9967,
       "step": 2467
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 15.512686729431152,
-      "learning_rate": 1.505462563896963e-05,
-      "loss": 1.9563,
+      "epoch": 0.31,
+      "grad_norm": 16.417949676513672,
+      "learning_rate": 1.7935823955152073e-05,
+      "loss": 3.3745,
       "step": 2468
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 18.510339736938477,
-      "learning_rate": 1.5052621028365242e-05,
-      "loss": 1.5118,
+      "epoch": 0.31,
+      "grad_norm": 9.704050064086914,
+      "learning_rate": 1.7934987240095387e-05,
+      "loss": 3.0882,
       "step": 2469
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 13.163358688354492,
-      "learning_rate": 1.505061641776085e-05,
-      "loss": 1.8231,
+      "epoch": 0.31,
+      "grad_norm": 12.788137435913086,
+      "learning_rate": 1.79341505250387e-05,
+      "loss": 3.1269,
       "step": 2470
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 8.8781156539917,
-      "learning_rate": 1.5048611807156461e-05,
-      "loss": 1.1656,
+      "epoch": 0.31,
+      "grad_norm": 12.552366256713867,
+      "learning_rate": 1.793331380998201e-05,
+      "loss": 2.8718,
       "step": 2471
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 25.674057006835938,
-      "learning_rate": 1.504660719655207e-05,
-      "loss": 2.7279,
+      "epoch": 0.31,
+      "grad_norm": 9.499432563781738,
+      "learning_rate": 1.7932477094925324e-05,
+      "loss": 3.5561,
       "step": 2472
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 16.811283111572266,
-      "learning_rate": 1.5044602585947681e-05,
-      "loss": 2.0451,
+      "epoch": 0.31,
+      "grad_norm": 7.19352388381958,
+      "learning_rate": 1.7931640379868638e-05,
+      "loss": 1.654,
       "step": 2473
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 15.99625301361084,
-      "learning_rate": 1.5042597975343291e-05,
-      "loss": 1.7554,
+      "epoch": 0.31,
+      "grad_norm": 9.706038475036621,
+      "learning_rate": 1.7930803664811948e-05,
+      "loss": 0.6615,
       "step": 2474
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 16.740371704101562,
-      "learning_rate": 1.50405933647389e-05,
-      "loss": 2.9046,
+      "epoch": 0.31,
+      "grad_norm": 22.5076847076416,
+      "learning_rate": 1.7929966949755262e-05,
+      "loss": 2.4768,
       "step": 2475
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 11.545669555664062,
-      "learning_rate": 1.5038588754134511e-05,
-      "loss": 1.1914,
+      "epoch": 0.31,
+      "grad_norm": 8.739575386047363,
+      "learning_rate": 1.7929130234698575e-05,
+      "loss": 0.5049,
       "step": 2476
     },
     {
-      "epoch": 0.74,
-      "grad_norm": 12.485255241394043,
-      "learning_rate": 1.503658414353012e-05,
-      "loss": 2.0212,
+      "epoch": 0.31,
+      "grad_norm": 12.416593551635742,
+      "learning_rate": 1.792829351964189e-05,
+      "loss": 2.2825,
       "step": 2477
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 28.34869956970215,
-      "learning_rate": 1.503457953292573e-05,
-      "loss": 1.4173,
+      "epoch": 0.31,
+      "grad_norm": 12.056394577026367,
+      "learning_rate": 1.79274568045852e-05,
+      "loss": 1.8361,
       "step": 2478
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 14.046238899230957,
-      "learning_rate": 1.5032574922321342e-05,
-      "loss": 2.3209,
+      "epoch": 0.31,
+      "grad_norm": 13.777711868286133,
+      "learning_rate": 1.7926620089528513e-05,
+      "loss": 2.752,
       "step": 2479
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 40.98884201049805,
-      "learning_rate": 1.503057031171695e-05,
-      "loss": 2.7137,
+      "epoch": 0.31,
+      "grad_norm": 8.878487586975098,
+      "learning_rate": 1.7925783374471827e-05,
+      "loss": 1.4853,
       "step": 2480
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 16.54846954345703,
-      "learning_rate": 1.502856570111256e-05,
-      "loss": 1.7074,
+      "epoch": 0.31,
+      "grad_norm": 16.995180130004883,
+      "learning_rate": 1.7924946659415137e-05,
+      "loss": 1.9121,
       "step": 2481
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 34.91267776489258,
-      "learning_rate": 1.502656109050817e-05,
-      "loss": 3.7392,
+      "epoch": 0.31,
+      "grad_norm": 22.008657455444336,
+      "learning_rate": 1.792410994435845e-05,
+      "loss": 2.4208,
       "step": 2482
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 11.860535621643066,
-      "learning_rate": 1.502455647990378e-05,
-      "loss": 2.6046,
+      "epoch": 0.31,
+      "grad_norm": 14.444175720214844,
+      "learning_rate": 1.7923273229301764e-05,
+      "loss": 2.655,
       "step": 2483
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 14.997228622436523,
-      "learning_rate": 1.5022551869299389e-05,
-      "loss": 2.761,
+      "epoch": 0.31,
+      "grad_norm": 10.106785774230957,
+      "learning_rate": 1.7922436514245078e-05,
+      "loss": 1.8945,
       "step": 2484
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 15.930320739746094,
-      "learning_rate": 1.5020547258695e-05,
-      "loss": 2.0798,
+      "epoch": 0.31,
+      "grad_norm": 7.286982536315918,
+      "learning_rate": 1.7921599799188388e-05,
+      "loss": 1.2408,
       "step": 2485
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 28.564027786254883,
-      "learning_rate": 1.501854264809061e-05,
-      "loss": 2.5476,
+      "epoch": 0.31,
+      "grad_norm": 19.98948097229004,
+      "learning_rate": 1.7920763084131702e-05,
+      "loss": 1.4448,
       "step": 2486
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 18.10663604736328,
-      "learning_rate": 1.5016538037486219e-05,
-      "loss": 2.6392,
+      "epoch": 0.31,
+      "grad_norm": 12.024795532226562,
+      "learning_rate": 1.7919926369075012e-05,
+      "loss": 2.46,
       "step": 2487
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 22.97527313232422,
-      "learning_rate": 1.501453342688183e-05,
-      "loss": 2.0554,
+      "epoch": 0.31,
+      "grad_norm": 20.353967666625977,
+      "learning_rate": 1.7919089654018326e-05,
+      "loss": 2.8392,
       "step": 2488
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 18.91280174255371,
-      "learning_rate": 1.5012528816277439e-05,
-      "loss": 2.5951,
+      "epoch": 0.31,
+      "grad_norm": 11.455394744873047,
+      "learning_rate": 1.791825293896164e-05,
+      "loss": 1.6463,
       "step": 2489
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 22.503339767456055,
-      "learning_rate": 1.5010524205673049e-05,
-      "loss": 1.7317,
+      "epoch": 0.31,
+      "grad_norm": 24.082109451293945,
+      "learning_rate": 1.791741622390495e-05,
+      "loss": 1.7333,
       "step": 2490
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 20.57047462463379,
-      "learning_rate": 1.5008519595068661e-05,
-      "loss": 2.3906,
+      "epoch": 0.31,
+      "grad_norm": 15.034388542175293,
+      "learning_rate": 1.7916579508848263e-05,
+      "loss": 3.2048,
       "step": 2491
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 17.44088363647461,
-      "learning_rate": 1.500651498446427e-05,
-      "loss": 2.3833,
+      "epoch": 0.31,
+      "grad_norm": 8.346967697143555,
+      "learning_rate": 1.7915742793791573e-05,
+      "loss": 0.8005,
       "step": 2492
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 11.56372356414795,
-      "learning_rate": 1.500451037385988e-05,
-      "loss": 2.2393,
+      "epoch": 0.31,
+      "grad_norm": 19.777572631835938,
+      "learning_rate": 1.7914906078734887e-05,
+      "loss": 3.237,
       "step": 2493
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 36.66616439819336,
-      "learning_rate": 1.5002505763255488e-05,
-      "loss": 2.0996,
+      "epoch": 0.31,
+      "grad_norm": 10.796478271484375,
+      "learning_rate": 1.79140693636782e-05,
+      "loss": 2.6851,
       "step": 2494
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 59.76615905761719,
-      "learning_rate": 1.50005011526511e-05,
-      "loss": 1.3206,
+      "epoch": 0.31,
+      "grad_norm": 28.78697967529297,
+      "learning_rate": 1.791323264862151e-05,
+      "loss": 2.892,
       "step": 2495
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 10.487852096557617,
-      "learning_rate": 1.4998496542046708e-05,
-      "loss": 1.2596,
+      "epoch": 0.31,
+      "grad_norm": 13.504036903381348,
+      "learning_rate": 1.7912395933564825e-05,
+      "loss": 2.1438,
       "step": 2496
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 17.123096466064453,
-      "learning_rate": 1.4996491931442318e-05,
-      "loss": 2.1562,
+      "epoch": 0.31,
+      "grad_norm": 16.800512313842773,
+      "learning_rate": 1.7911559218508138e-05,
+      "loss": 2.822,
       "step": 2497
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 13.120617866516113,
-      "learning_rate": 1.499448732083793e-05,
-      "loss": 1.2493,
+      "epoch": 0.31,
+      "grad_norm": 15.418956756591797,
+      "learning_rate": 1.7910722503451452e-05,
+      "loss": 2.4324,
       "step": 2498
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 11.029601097106934,
-      "learning_rate": 1.4992482710233538e-05,
-      "loss": 0.918,
+      "epoch": 0.31,
+      "grad_norm": 15.057035446166992,
+      "learning_rate": 1.7909885788394762e-05,
+      "loss": 2.5772,
       "step": 2499
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 38.185813903808594,
-      "learning_rate": 1.4990478099629148e-05,
-      "loss": 3.1123,
+      "epoch": 0.31,
+      "grad_norm": 19.443241119384766,
+      "learning_rate": 1.7909049073338076e-05,
+      "loss": 4.5712,
       "step": 2500
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 21.2143611907959,
-      "learning_rate": 1.4988473489024757e-05,
-      "loss": 1.9136,
+      "epoch": 0.31,
+      "grad_norm": 18.084739685058594,
+      "learning_rate": 1.790821235828139e-05,
+      "loss": 4.0203,
       "step": 2501
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 18.287214279174805,
-      "learning_rate": 1.4986468878420368e-05,
-      "loss": 2.3058,
+      "epoch": 0.31,
+      "grad_norm": 16.893775939941406,
+      "learning_rate": 1.79073756432247e-05,
+      "loss": 1.9783,
       "step": 2502
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 20.503644943237305,
-      "learning_rate": 1.4984464267815979e-05,
-      "loss": 3.3248,
+      "epoch": 0.31,
+      "grad_norm": 17.322256088256836,
+      "learning_rate": 1.7906538928168013e-05,
+      "loss": 2.7375,
       "step": 2503
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 18.117403030395508,
-      "learning_rate": 1.4982459657211587e-05,
-      "loss": 2.5097,
+      "epoch": 0.31,
+      "grad_norm": 14.303987503051758,
+      "learning_rate": 1.7905702213111327e-05,
+      "loss": 1.305,
       "step": 2504
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 16.31001853942871,
-      "learning_rate": 1.4980455046607199e-05,
-      "loss": 1.7266,
+      "epoch": 0.31,
+      "grad_norm": 9.188926696777344,
+      "learning_rate": 1.790486549805464e-05,
+      "loss": 1.6407,
       "step": 2505
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 12.371569633483887,
-      "learning_rate": 1.4978450436002807e-05,
-      "loss": 1.9252,
+      "epoch": 0.31,
+      "grad_norm": 12.038647651672363,
+      "learning_rate": 1.790402878299795e-05,
+      "loss": 2.5819,
       "step": 2506
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 11.64188289642334,
-      "learning_rate": 1.4976445825398417e-05,
-      "loss": 2.0164,
+      "epoch": 0.31,
+      "grad_norm": 12.675663948059082,
+      "learning_rate": 1.7903192067941265e-05,
+      "loss": 2.2682,
       "step": 2507
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 17.393896102905273,
-      "learning_rate": 1.4974441214794027e-05,
-      "loss": 2.4099,
+      "epoch": 0.31,
+      "grad_norm": 7.899895191192627,
+      "learning_rate": 1.7902355352884578e-05,
+      "loss": 0.2706,
       "step": 2508
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 16.229440689086914,
-      "learning_rate": 1.4972436604189637e-05,
-      "loss": 2.8192,
+      "epoch": 0.31,
+      "grad_norm": 14.513761520385742,
+      "learning_rate": 1.790151863782789e-05,
+      "loss": 3.308,
       "step": 2509
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 19.683637619018555,
-      "learning_rate": 1.4970431993585247e-05,
-      "loss": 2.0972,
+      "epoch": 0.32,
+      "grad_norm": 7.7307000160217285,
+      "learning_rate": 1.7900681922771202e-05,
+      "loss": 1.3174,
       "step": 2510
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 23.35310173034668,
-      "learning_rate": 1.4968427382980858e-05,
-      "loss": 2.0364,
+      "epoch": 0.32,
+      "grad_norm": 9.518709182739258,
+      "learning_rate": 1.7899845207714516e-05,
+      "loss": 0.598,
       "step": 2511
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 10.506977081298828,
-      "learning_rate": 1.4966422772376468e-05,
-      "loss": 1.575,
+      "epoch": 0.32,
+      "grad_norm": 23.85696029663086,
+      "learning_rate": 1.789900849265783e-05,
+      "loss": 3.9654,
       "step": 2512
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 24.63909339904785,
-      "learning_rate": 1.4964418161772076e-05,
-      "loss": 1.9545,
+      "epoch": 0.32,
+      "grad_norm": 6.402545928955078,
+      "learning_rate": 1.789817177760114e-05,
+      "loss": 0.2189,
       "step": 2513
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 18.600322723388672,
-      "learning_rate": 1.4962413551167688e-05,
-      "loss": 2.0238,
+      "epoch": 0.32,
+      "grad_norm": 8.172592163085938,
+      "learning_rate": 1.7897335062544453e-05,
+      "loss": 1.0552,
       "step": 2514
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 25.803688049316406,
-      "learning_rate": 1.4960408940563296e-05,
-      "loss": 1.7973,
+      "epoch": 0.32,
+      "grad_norm": 16.517786026000977,
+      "learning_rate": 1.7896498347487764e-05,
+      "loss": 3.0275,
       "step": 2515
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 13.325533866882324,
-      "learning_rate": 1.4958404329958906e-05,
-      "loss": 1.5489,
+      "epoch": 0.32,
+      "grad_norm": 17.168933868408203,
+      "learning_rate": 1.7895661632431077e-05,
+      "loss": 1.0541,
       "step": 2516
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 17.538631439208984,
-      "learning_rate": 1.4956399719354518e-05,
-      "loss": 1.6318,
+      "epoch": 0.32,
+      "grad_norm": 5.158097743988037,
+      "learning_rate": 1.789482491737439e-05,
+      "loss": 1.6071,
       "step": 2517
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 20.671701431274414,
-      "learning_rate": 1.4954395108750126e-05,
-      "loss": 2.2905,
+      "epoch": 0.32,
+      "grad_norm": 11.919731140136719,
+      "learning_rate": 1.78939882023177e-05,
+      "loss": 1.4199,
       "step": 2518
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 18.96885871887207,
-      "learning_rate": 1.4952390498145736e-05,
-      "loss": 1.2961,
+      "epoch": 0.32,
+      "grad_norm": 11.247430801391602,
+      "learning_rate": 1.7893151487261015e-05,
+      "loss": 2.6482,
       "step": 2519
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 17.124177932739258,
-      "learning_rate": 1.4950385887541345e-05,
-      "loss": 2.1604,
-      "step": 2520
-    },
-    {
-      "epoch": 0.76,
-      "eval_loss": 0.269967257976532,
-      "eval_runtime": 43.2806,
-      "eval_samples_per_second": 34.172,
-      "eval_steps_per_second": 34.172,
+      "epoch": 0.32,
+      "grad_norm": 36.31636428833008,
+      "learning_rate": 1.7892314772204325e-05,
+      "loss": 1.5908,
       "step": 2520
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 22.06162452697754,
-      "learning_rate": 1.4948381276936957e-05,
-      "loss": 2.1107,
+      "epoch": 0.32,
+      "grad_norm": 12.56596565246582,
+      "learning_rate": 1.789147805714764e-05,
+      "loss": 2.6619,
       "step": 2521
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 15.108476638793945,
-      "learning_rate": 1.4946376666332567e-05,
-      "loss": 1.7432,
+      "epoch": 0.32,
+      "grad_norm": 6.203104496002197,
+      "learning_rate": 1.7890641342090952e-05,
+      "loss": 0.7091,
       "step": 2522
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 13.295374870300293,
-      "learning_rate": 1.4944372055728175e-05,
-      "loss": 2.03,
+      "epoch": 0.32,
+      "grad_norm": 17.171630859375,
+      "learning_rate": 1.7889804627034263e-05,
+      "loss": 1.9813,
       "step": 2523
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 19.436033248901367,
-      "learning_rate": 1.4942367445123787e-05,
-      "loss": 1.6658,
+      "epoch": 0.32,
+      "grad_norm": 6.632760047912598,
+      "learning_rate": 1.7888967911977576e-05,
+      "loss": 1.1465,
       "step": 2524
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 36.37664031982422,
-      "learning_rate": 1.4940362834519395e-05,
-      "loss": 3.2737,
+      "epoch": 0.32,
+      "grad_norm": 24.890317916870117,
+      "learning_rate": 1.788813119692089e-05,
+      "loss": 2.9452,
       "step": 2525
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 11.738432884216309,
-      "learning_rate": 1.4938358223915005e-05,
-      "loss": 1.6208,
+      "epoch": 0.32,
+      "grad_norm": 10.765260696411133,
+      "learning_rate": 1.7887294481864204e-05,
+      "loss": 2.3937,
       "step": 2526
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 17.715194702148438,
-      "learning_rate": 1.4936353613310614e-05,
-      "loss": 2.3697,
+      "epoch": 0.32,
+      "grad_norm": 12.482851028442383,
+      "learning_rate": 1.7886457766807514e-05,
+      "loss": 2.7714,
       "step": 2527
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 21.08336639404297,
-      "learning_rate": 1.4934349002706226e-05,
-      "loss": 2.4212,
+      "epoch": 0.32,
+      "grad_norm": 11.03923511505127,
+      "learning_rate": 1.7885621051750828e-05,
+      "loss": 3.0968,
       "step": 2528
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 19.754703521728516,
-      "learning_rate": 1.4932344392101836e-05,
-      "loss": 1.71,
+      "epoch": 0.32,
+      "grad_norm": 13.121932029724121,
+      "learning_rate": 1.788478433669414e-05,
+      "loss": 1.7945,
       "step": 2529
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 21.412639617919922,
-      "learning_rate": 1.4930339781497446e-05,
-      "loss": 2.807,
+      "epoch": 0.32,
+      "grad_norm": 16.075044631958008,
+      "learning_rate": 1.788394762163745e-05,
+      "loss": 2.1452,
       "step": 2530
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 18.40741729736328,
-      "learning_rate": 1.4928335170893056e-05,
-      "loss": 2.2224,
+      "epoch": 0.32,
+      "grad_norm": 21.7185115814209,
+      "learning_rate": 1.7883110906580765e-05,
+      "loss": 3.6425,
       "step": 2531
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 33.87398147583008,
-      "learning_rate": 1.4926330560288664e-05,
-      "loss": 2.8005,
+      "epoch": 0.32,
+      "grad_norm": 22.87784767150879,
+      "learning_rate": 1.788227419152408e-05,
+      "loss": 1.8277,
       "step": 2532
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 14.195319175720215,
-      "learning_rate": 1.4924325949684276e-05,
-      "loss": 2.4552,
+      "epoch": 0.32,
+      "grad_norm": 21.376379013061523,
+      "learning_rate": 1.7881437476467392e-05,
+      "loss": 2.8571,
       "step": 2533
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 14.160918235778809,
-      "learning_rate": 1.4922321339079886e-05,
-      "loss": 2.4736,
+      "epoch": 0.32,
+      "grad_norm": 10.283143997192383,
+      "learning_rate": 1.7880600761410703e-05,
+      "loss": 1.6794,
       "step": 2534
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 31.502965927124023,
-      "learning_rate": 1.4920316728475494e-05,
-      "loss": 2.5237,
+      "epoch": 0.32,
+      "grad_norm": 20.66465187072754,
+      "learning_rate": 1.7879764046354016e-05,
+      "loss": 2.2436,
       "step": 2535
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 17.290319442749023,
-      "learning_rate": 1.4918312117871106e-05,
-      "loss": 2.2399,
+      "epoch": 0.32,
+      "grad_norm": 8.404132843017578,
+      "learning_rate": 1.787892733129733e-05,
+      "loss": 1.4817,
       "step": 2536
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 13.980632781982422,
-      "learning_rate": 1.4916307507266715e-05,
-      "loss": 2.3494,
+      "epoch": 0.32,
+      "grad_norm": 10.486577033996582,
+      "learning_rate": 1.787809061624064e-05,
+      "loss": 2.5069,
       "step": 2537
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 14.719244956970215,
-      "learning_rate": 1.4914302896662325e-05,
-      "loss": 2.5073,
+      "epoch": 0.32,
+      "grad_norm": 8.028034210205078,
+      "learning_rate": 1.7877253901183954e-05,
+      "loss": 1.0015,
       "step": 2538
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 19.83281707763672,
-      "learning_rate": 1.4912298286057933e-05,
-      "loss": 2.4892,
+      "epoch": 0.32,
+      "grad_norm": 23.063966751098633,
+      "learning_rate": 1.7876417186127267e-05,
+      "loss": 3.4067,
       "step": 2539
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 13.933186531066895,
-      "learning_rate": 1.4910293675453545e-05,
-      "loss": 2.6791,
+      "epoch": 0.32,
+      "grad_norm": 23.007061004638672,
+      "learning_rate": 1.7875580471070578e-05,
+      "loss": 2.9092,
       "step": 2540
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 20.118640899658203,
-      "learning_rate": 1.4908289064849155e-05,
-      "loss": 1.996,
+      "epoch": 0.32,
+      "grad_norm": 7.917148590087891,
+      "learning_rate": 1.787474375601389e-05,
+      "loss": 1.5232,
       "step": 2541
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 20.20134162902832,
-      "learning_rate": 1.4906284454244763e-05,
-      "loss": 2.2007,
+      "epoch": 0.32,
+      "grad_norm": 9.655314445495605,
+      "learning_rate": 1.7873907040957205e-05,
+      "loss": 1.5226,
       "step": 2542
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 14.23250961303711,
-      "learning_rate": 1.4904279843640375e-05,
-      "loss": 2.2191,
+      "epoch": 0.32,
+      "grad_norm": 11.38919734954834,
+      "learning_rate": 1.7873070325900515e-05,
+      "loss": 1.2438,
       "step": 2543
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 14.728852272033691,
-      "learning_rate": 1.4902275233035984e-05,
-      "loss": 1.9154,
+      "epoch": 0.32,
+      "grad_norm": 9.895431518554688,
+      "learning_rate": 1.787223361084383e-05,
+      "loss": 2.0275,
       "step": 2544
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 38.36711120605469,
-      "learning_rate": 1.4900270622431594e-05,
-      "loss": 2.367,
+      "epoch": 0.32,
+      "grad_norm": 11.29218864440918,
+      "learning_rate": 1.787139689578714e-05,
+      "loss": 2.4134,
       "step": 2545
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 17.614601135253906,
-      "learning_rate": 1.4898266011827205e-05,
-      "loss": 1.1838,
+      "epoch": 0.32,
+      "grad_norm": 39.96311569213867,
+      "learning_rate": 1.7870560180730453e-05,
+      "loss": 2.4428,
       "step": 2546
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 23.657424926757812,
-      "learning_rate": 1.4896261401222814e-05,
-      "loss": 2.6376,
+      "epoch": 0.32,
+      "grad_norm": 27.714492797851562,
+      "learning_rate": 1.7869723465673767e-05,
+      "loss": 3.6484,
       "step": 2547
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 10.91047191619873,
-      "learning_rate": 1.4894256790618424e-05,
-      "loss": 2.0975,
+      "epoch": 0.32,
+      "grad_norm": 7.043478012084961,
+      "learning_rate": 1.7868886750617077e-05,
+      "loss": 1.2048,
       "step": 2548
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 19.83192253112793,
-      "learning_rate": 1.4892252180014032e-05,
-      "loss": 2.7231,
+      "epoch": 0.32,
+      "grad_norm": 13.39302921295166,
+      "learning_rate": 1.786805003556039e-05,
+      "loss": 2.9228,
       "step": 2549
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 6.836325645446777,
-      "learning_rate": 1.4890247569409644e-05,
-      "loss": 1.3988,
+      "epoch": 0.32,
+      "grad_norm": 20.650859832763672,
+      "learning_rate": 1.7867213320503704e-05,
+      "loss": 3.4154,
       "step": 2550
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 16.332334518432617,
-      "learning_rate": 1.4888242958805252e-05,
-      "loss": 1.9877,
+      "epoch": 0.32,
+      "grad_norm": 17.851686477661133,
+      "learning_rate": 1.7866376605447014e-05,
+      "loss": 2.9386,
       "step": 2551
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 23.387727737426758,
-      "learning_rate": 1.4886238348200862e-05,
-      "loss": 2.8849,
+      "epoch": 0.32,
+      "grad_norm": 10.273046493530273,
+      "learning_rate": 1.7865539890390328e-05,
+      "loss": 1.0958,
       "step": 2552
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 33.43765640258789,
-      "learning_rate": 1.4884233737596474e-05,
-      "loss": 2.3284,
+      "epoch": 0.32,
+      "grad_norm": 25.07984161376953,
+      "learning_rate": 1.786470317533364e-05,
+      "loss": 4.0964,
       "step": 2553
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 10.308141708374023,
-      "learning_rate": 1.4882229126992083e-05,
-      "loss": 2.6766,
+      "epoch": 0.32,
+      "grad_norm": 21.487104415893555,
+      "learning_rate": 1.7863866460276955e-05,
+      "loss": 2.4228,
       "step": 2554
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 20.615131378173828,
-      "learning_rate": 1.4880224516387693e-05,
-      "loss": 2.4127,
+      "epoch": 0.32,
+      "grad_norm": 11.11598014831543,
+      "learning_rate": 1.7863029745220266e-05,
+      "loss": 1.9753,
       "step": 2555
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 18.643369674682617,
-      "learning_rate": 1.4878219905783303e-05,
-      "loss": 1.9807,
+      "epoch": 0.32,
+      "grad_norm": 13.04572868347168,
+      "learning_rate": 1.786219303016358e-05,
+      "loss": 2.3264,
       "step": 2556
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 29.241533279418945,
-      "learning_rate": 1.4876215295178913e-05,
-      "loss": 2.6441,
+      "epoch": 0.32,
+      "grad_norm": 8.850125312805176,
+      "learning_rate": 1.7861356315106893e-05,
+      "loss": 1.6477,
       "step": 2557
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 19.864036560058594,
-      "learning_rate": 1.4874210684574521e-05,
-      "loss": 2.237,
+      "epoch": 0.32,
+      "grad_norm": 14.255194664001465,
+      "learning_rate": 1.7860519600050203e-05,
+      "loss": 3.6128,
       "step": 2558
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 9.182751655578613,
-      "learning_rate": 1.4872206073970133e-05,
-      "loss": 1.6721,
+      "epoch": 0.32,
+      "grad_norm": 12.04518985748291,
+      "learning_rate": 1.7859682884993517e-05,
+      "loss": 2.2058,
       "step": 2559
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 17.696592330932617,
-      "learning_rate": 1.4870201463365743e-05,
-      "loss": 1.8645,
+      "epoch": 0.32,
+      "grad_norm": 9.668314933776855,
+      "learning_rate": 1.785884616993683e-05,
+      "loss": 1.3232,
       "step": 2560
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 10.09614086151123,
-      "learning_rate": 1.4868196852761352e-05,
-      "loss": 1.4887,
+      "epoch": 0.32,
+      "grad_norm": 13.472427368164062,
+      "learning_rate": 1.7858009454880144e-05,
+      "loss": 2.4424,
       "step": 2561
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 10.682119369506836,
-      "learning_rate": 1.4866192242156963e-05,
-      "loss": 1.296,
+      "epoch": 0.32,
+      "grad_norm": 15.245526313781738,
+      "learning_rate": 1.7857172739823454e-05,
+      "loss": 1.7416,
       "step": 2562
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 18.929485321044922,
-      "learning_rate": 1.4864187631552572e-05,
-      "loss": 2.7717,
+      "epoch": 0.32,
+      "grad_norm": 13.720373153686523,
+      "learning_rate": 1.7856336024766768e-05,
+      "loss": 1.6349,
       "step": 2563
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 21.349645614624023,
-      "learning_rate": 1.4862183020948182e-05,
-      "loss": 2.4609,
+      "epoch": 0.32,
+      "grad_norm": 15.144506454467773,
+      "learning_rate": 1.785549930971008e-05,
+      "loss": 2.4472,
       "step": 2564
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 7.982182502746582,
-      "learning_rate": 1.4860178410343794e-05,
-      "loss": 1.9265,
+      "epoch": 0.32,
+      "grad_norm": 14.76689624786377,
+      "learning_rate": 1.7854662594653392e-05,
+      "loss": 0.6717,
       "step": 2565
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 15.188501358032227,
-      "learning_rate": 1.4858173799739402e-05,
-      "loss": 3.0483,
+      "epoch": 0.32,
+      "grad_norm": 11.874075889587402,
+      "learning_rate": 1.7853825879596706e-05,
+      "loss": 1.989,
       "step": 2566
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 37.26816177368164,
-      "learning_rate": 1.4856169189135012e-05,
-      "loss": 2.585,
+      "epoch": 0.32,
+      "grad_norm": 11.744961738586426,
+      "learning_rate": 1.785298916454002e-05,
+      "loss": 1.938,
       "step": 2567
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 13.02489948272705,
-      "learning_rate": 1.485416457853062e-05,
-      "loss": 1.8137,
+      "epoch": 0.32,
+      "grad_norm": 7.1501145362854,
+      "learning_rate": 1.785215244948333e-05,
+      "loss": 0.8565,
       "step": 2568
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 15.654376029968262,
-      "learning_rate": 1.4852159967926232e-05,
-      "loss": 2.2203,
+      "epoch": 0.32,
+      "grad_norm": 11.965447425842285,
+      "learning_rate": 1.7851315734426643e-05,
+      "loss": 0.9023,
       "step": 2569
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 9.267987251281738,
-      "learning_rate": 1.485015535732184e-05,
-      "loss": 1.362,
+      "epoch": 0.32,
+      "grad_norm": 11.98586368560791,
+      "learning_rate": 1.7850479019369957e-05,
+      "loss": 0.6164,
       "step": 2570
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 31.388477325439453,
-      "learning_rate": 1.484815074671745e-05,
-      "loss": 2.3867,
+      "epoch": 0.32,
+      "grad_norm": 28.456218719482422,
+      "learning_rate": 1.7849642304313267e-05,
+      "loss": 4.24,
       "step": 2571
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 19.304162979125977,
-      "learning_rate": 1.4846146136113062e-05,
-      "loss": 2.6042,
+      "epoch": 0.32,
+      "grad_norm": 17.911380767822266,
+      "learning_rate": 1.784880558925658e-05,
+      "loss": 2.4132,
       "step": 2572
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 21.53330421447754,
-      "learning_rate": 1.4844141525508671e-05,
-      "loss": 2.2464,
+      "epoch": 0.32,
+      "grad_norm": 5.884377479553223,
+      "learning_rate": 1.784796887419989e-05,
+      "loss": 2.5055,
       "step": 2573
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 14.074399948120117,
-      "learning_rate": 1.4842136914904281e-05,
-      "loss": 1.6423,
+      "epoch": 0.32,
+      "grad_norm": 16.891359329223633,
+      "learning_rate": 1.7847132159143205e-05,
+      "loss": 2.5807,
       "step": 2574
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 30.000886917114258,
-      "learning_rate": 1.484013230429989e-05,
-      "loss": 1.9405,
+      "epoch": 0.32,
+      "grad_norm": 16.468059539794922,
+      "learning_rate": 1.7846295444086518e-05,
+      "loss": 1.6703,
       "step": 2575
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 12.38129711151123,
-      "learning_rate": 1.4838127693695501e-05,
-      "loss": 1.6343,
+      "epoch": 0.32,
+      "grad_norm": 10.617067337036133,
+      "learning_rate": 1.784545872902983e-05,
+      "loss": 1.2433,
       "step": 2576
     },
     {
-      "epoch": 0.77,
-      "grad_norm": 15.83830451965332,
-      "learning_rate": 1.4836123083091111e-05,
-      "loss": 1.4498,
+      "epoch": 0.32,
+      "grad_norm": 12.173160552978516,
+      "learning_rate": 1.7844622013973142e-05,
+      "loss": 3.2946,
       "step": 2577
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 20.503822326660156,
-      "learning_rate": 1.483411847248672e-05,
-      "loss": 2.8636,
+      "epoch": 0.32,
+      "grad_norm": 5.2667059898376465,
+      "learning_rate": 1.7843785298916456e-05,
+      "loss": 1.5948,
       "step": 2578
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 15.516641616821289,
-      "learning_rate": 1.4832113861882331e-05,
-      "loss": 2.0966,
+      "epoch": 0.32,
+      "grad_norm": 5.8900837898254395,
+      "learning_rate": 1.7842948583859766e-05,
+      "loss": 0.7239,
       "step": 2579
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 18.49781036376953,
-      "learning_rate": 1.483010925127794e-05,
-      "loss": 3.8588,
+      "epoch": 0.32,
+      "grad_norm": 14.397953033447266,
+      "learning_rate": 1.784211186880308e-05,
+      "loss": 0.5331,
       "step": 2580
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 11.766064643859863,
-      "learning_rate": 1.4828104640673552e-05,
-      "loss": 3.0766,
+      "epoch": 0.32,
+      "grad_norm": 18.99256706237793,
+      "learning_rate": 1.7841275153746393e-05,
+      "loss": 2.269,
       "step": 2581
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 28.008140563964844,
-      "learning_rate": 1.482610003006916e-05,
-      "loss": 2.8604,
+      "epoch": 0.32,
+      "grad_norm": 25.798974990844727,
+      "learning_rate": 1.7840438438689707e-05,
+      "loss": 2.0258,
       "step": 2582
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 11.355888366699219,
-      "learning_rate": 1.482409541946477e-05,
-      "loss": 1.8569,
+      "epoch": 0.32,
+      "grad_norm": 12.33133316040039,
+      "learning_rate": 1.7839601723633017e-05,
+      "loss": 1.9399,
       "step": 2583
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 24.921497344970703,
-      "learning_rate": 1.4822090808860382e-05,
-      "loss": 2.3311,
+      "epoch": 0.32,
+      "grad_norm": 10.246801376342773,
+      "learning_rate": 1.783876500857633e-05,
+      "loss": 2.757,
       "step": 2584
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 49.57196807861328,
-      "learning_rate": 1.482008619825599e-05,
-      "loss": 3.8008,
+      "epoch": 0.32,
+      "grad_norm": 23.0379581451416,
+      "learning_rate": 1.7837928293519645e-05,
+      "loss": 2.7737,
       "step": 2585
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 18.97112274169922,
-      "learning_rate": 1.48180815876516e-05,
-      "loss": 2.2731,
+      "epoch": 0.32,
+      "grad_norm": 6.544103145599365,
+      "learning_rate": 1.7837091578462955e-05,
+      "loss": 0.6896,
       "step": 2586
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 19.699682235717773,
-      "learning_rate": 1.4816076977047209e-05,
-      "loss": 3.2435,
+      "epoch": 0.32,
+      "grad_norm": 12.11794662475586,
+      "learning_rate": 1.783625486340627e-05,
+      "loss": 1.5552,
       "step": 2587
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 15.111373901367188,
-      "learning_rate": 1.481407236644282e-05,
-      "loss": 2.6261,
+      "epoch": 0.32,
+      "grad_norm": 17.367733001708984,
+      "learning_rate": 1.7835418148349582e-05,
+      "loss": 2.991,
       "step": 2588
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 39.739288330078125,
-      "learning_rate": 1.481206775583843e-05,
-      "loss": 3.0957,
+      "epoch": 0.32,
+      "grad_norm": 14.063064575195312,
+      "learning_rate": 1.7834581433292896e-05,
+      "loss": 2.0534,
       "step": 2589
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 7.469949245452881,
-      "learning_rate": 1.4810063145234039e-05,
-      "loss": 1.5037,
+      "epoch": 0.33,
+      "grad_norm": 30.218265533447266,
+      "learning_rate": 1.7833744718236206e-05,
+      "loss": 2.8303,
       "step": 2590
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 48.33216857910156,
-      "learning_rate": 1.480805853462965e-05,
-      "loss": 1.5074,
+      "epoch": 0.33,
+      "grad_norm": 41.22004318237305,
+      "learning_rate": 1.783290800317952e-05,
+      "loss": 2.892,
       "step": 2591
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 14.457548141479492,
-      "learning_rate": 1.4806053924025259e-05,
-      "loss": 2.0293,
+      "epoch": 0.33,
+      "grad_norm": 9.430306434631348,
+      "learning_rate": 1.7832071288122833e-05,
+      "loss": 2.3149,
       "step": 2592
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 9.994394302368164,
-      "learning_rate": 1.480404931342087e-05,
-      "loss": 1.6135,
+      "epoch": 0.33,
+      "grad_norm": 12.786920547485352,
+      "learning_rate": 1.7831234573066144e-05,
+      "loss": 1.0034,
       "step": 2593
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 16.707799911499023,
-      "learning_rate": 1.4802044702816478e-05,
-      "loss": 2.0803,
+      "epoch": 0.33,
+      "grad_norm": 20.871122360229492,
+      "learning_rate": 1.7830397858009457e-05,
+      "loss": 2.1801,
       "step": 2594
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 11.45969009399414,
-      "learning_rate": 1.480004009221209e-05,
-      "loss": 2.0687,
+      "epoch": 0.33,
+      "grad_norm": 9.6037015914917,
+      "learning_rate": 1.782956114295277e-05,
+      "loss": 0.4981,
       "step": 2595
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 17.346237182617188,
-      "learning_rate": 1.47980354816077e-05,
-      "loss": 1.968,
+      "epoch": 0.33,
+      "grad_norm": 25.444738388061523,
+      "learning_rate": 1.782872442789608e-05,
+      "loss": 3.3543,
       "step": 2596
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 16.296354293823242,
-      "learning_rate": 1.4796030871003308e-05,
-      "loss": 2.7071,
+      "epoch": 0.33,
+      "grad_norm": 7.93801736831665,
+      "learning_rate": 1.7827887712839395e-05,
+      "loss": 0.1098,
       "step": 2597
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 24.699426651000977,
-      "learning_rate": 1.479402626039892e-05,
-      "loss": 2.575,
+      "epoch": 0.33,
+      "grad_norm": 15.348746299743652,
+      "learning_rate": 1.7827050997782705e-05,
+      "loss": 2.0703,
       "step": 2598
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 15.403938293457031,
-      "learning_rate": 1.4792021649794528e-05,
-      "loss": 2.3392,
+      "epoch": 0.33,
+      "grad_norm": 21.247926712036133,
+      "learning_rate": 1.782621428272602e-05,
+      "loss": 2.3706,
       "step": 2599
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 13.177908897399902,
-      "learning_rate": 1.4790017039190138e-05,
-      "loss": 1.4653,
+      "epoch": 0.33,
+      "grad_norm": 20.806386947631836,
+      "learning_rate": 1.7825377567669332e-05,
+      "loss": 1.9695,
       "step": 2600
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 34.016021728515625,
-      "learning_rate": 1.4788012428585748e-05,
-      "loss": 2.3153,
+      "epoch": 0.33,
+      "grad_norm": 14.542155265808105,
+      "learning_rate": 1.7824540852612643e-05,
+      "loss": 2.949,
       "step": 2601
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 15.840597152709961,
-      "learning_rate": 1.4786007817981358e-05,
-      "loss": 2.5034,
+      "epoch": 0.33,
+      "grad_norm": 9.618982315063477,
+      "learning_rate": 1.7823704137555956e-05,
+      "loss": 0.7286,
       "step": 2602
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 31.41590118408203,
-      "learning_rate": 1.4784003207376968e-05,
-      "loss": 2.0677,
+      "epoch": 0.33,
+      "grad_norm": 21.41895294189453,
+      "learning_rate": 1.782286742249927e-05,
+      "loss": 2.5352,
       "step": 2603
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 12.184178352355957,
-      "learning_rate": 1.4781998596772578e-05,
-      "loss": 2.0374,
+      "epoch": 0.33,
+      "grad_norm": 14.966450691223145,
+      "learning_rate": 1.782203070744258e-05,
+      "loss": 1.6908,
       "step": 2604
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 16.264497756958008,
-      "learning_rate": 1.4779993986168188e-05,
-      "loss": 2.4106,
+      "epoch": 0.33,
+      "grad_norm": 21.944358825683594,
+      "learning_rate": 1.7821193992385894e-05,
+      "loss": 1.6259,
       "step": 2605
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 16.872928619384766,
-      "learning_rate": 1.4777989375563797e-05,
-      "loss": 2.3471,
+      "epoch": 0.33,
+      "grad_norm": 33.289207458496094,
+      "learning_rate": 1.7820357277329207e-05,
+      "loss": 3.2477,
       "step": 2606
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 47.74468994140625,
-      "learning_rate": 1.4775984764959409e-05,
-      "loss": 3.2204,
+      "epoch": 0.33,
+      "grad_norm": 26.20473289489746,
+      "learning_rate": 1.7819520562272518e-05,
+      "loss": 1.8603,
       "step": 2607
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 20.217355728149414,
-      "learning_rate": 1.4773980154355019e-05,
-      "loss": 1.9179,
+      "epoch": 0.33,
+      "grad_norm": 27.471969604492188,
+      "learning_rate": 1.781868384721583e-05,
+      "loss": 2.7483,
       "step": 2608
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 13.135704040527344,
-      "learning_rate": 1.4771975543750627e-05,
-      "loss": 2.2236,
+      "epoch": 0.33,
+      "grad_norm": 33.70694351196289,
+      "learning_rate": 1.7817847132159145e-05,
+      "loss": 3.7981,
       "step": 2609
     },
     {
-      "epoch": 0.78,
-      "grad_norm": 10.10952377319336,
-      "learning_rate": 1.4769970933146239e-05,
-      "loss": 1.2541,
+      "epoch": 0.33,
+      "grad_norm": 11.390097618103027,
+      "learning_rate": 1.781701041710246e-05,
+      "loss": 2.6712,
       "step": 2610
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 14.991072654724121,
-      "learning_rate": 1.4767966322541847e-05,
-      "loss": 1.6626,
+      "epoch": 0.33,
+      "grad_norm": 116.96737670898438,
+      "learning_rate": 1.781617370204577e-05,
+      "loss": 2.0294,
       "step": 2611
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 18.55371856689453,
-      "learning_rate": 1.4765961711937457e-05,
-      "loss": 2.59,
+      "epoch": 0.33,
+      "grad_norm": 23.15081787109375,
+      "learning_rate": 1.7815336986989083e-05,
+      "loss": 3.2805,
       "step": 2612
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 153.79713439941406,
-      "learning_rate": 1.4763957101333066e-05,
-      "loss": 1.6441,
+      "epoch": 0.33,
+      "grad_norm": 25.70244598388672,
+      "learning_rate": 1.7814500271932396e-05,
+      "loss": 2.0328,
       "step": 2613
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 20.791709899902344,
-      "learning_rate": 1.4761952490728678e-05,
-      "loss": 2.2488,
+      "epoch": 0.33,
+      "grad_norm": 19.2539119720459,
+      "learning_rate": 1.7813663556875706e-05,
+      "loss": 2.9757,
       "step": 2614
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 26.96784019470215,
-      "learning_rate": 1.4759947880124288e-05,
-      "loss": 2.4947,
+      "epoch": 0.33,
+      "grad_norm": 23.953353881835938,
+      "learning_rate": 1.781282684181902e-05,
+      "loss": 1.2766,
       "step": 2615
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 31.95100975036621,
-      "learning_rate": 1.4757943269519896e-05,
-      "loss": 2.5303,
+      "epoch": 0.33,
+      "grad_norm": 11.397101402282715,
+      "learning_rate": 1.7811990126762334e-05,
+      "loss": 2.3818,
       "step": 2616
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 16.064725875854492,
-      "learning_rate": 1.4755938658915508e-05,
-      "loss": 2.2598,
+      "epoch": 0.33,
+      "grad_norm": 20.783130645751953,
+      "learning_rate": 1.7811153411705647e-05,
+      "loss": 2.5762,
       "step": 2617
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 19.35765266418457,
-      "learning_rate": 1.4753934048311116e-05,
-      "loss": 2.0552,
+      "epoch": 0.33,
+      "grad_norm": 7.296566009521484,
+      "learning_rate": 1.7810316696648958e-05,
+      "loss": 0.759,
       "step": 2618
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 11.326220512390137,
-      "learning_rate": 1.4751929437706726e-05,
-      "loss": 2.1031,
+      "epoch": 0.33,
+      "grad_norm": 12.399662971496582,
+      "learning_rate": 1.780947998159227e-05,
+      "loss": 1.9589,
       "step": 2619
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 8.919992446899414,
-      "learning_rate": 1.4749924827102338e-05,
-      "loss": 1.8168,
+      "epoch": 0.33,
+      "grad_norm": 35.2669677734375,
+      "learning_rate": 1.7808643266535585e-05,
+      "loss": 3.8683,
       "step": 2620
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 14.81724739074707,
-      "learning_rate": 1.4747920216497946e-05,
-      "loss": 1.9064,
+      "epoch": 0.33,
+      "grad_norm": 23.687824249267578,
+      "learning_rate": 1.7807806551478895e-05,
+      "loss": 2.1461,
       "step": 2621
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 31.04239845275879,
-      "learning_rate": 1.4745915605893557e-05,
-      "loss": 2.7644,
+      "epoch": 0.33,
+      "grad_norm": 21.20037078857422,
+      "learning_rate": 1.780696983642221e-05,
+      "loss": 3.3364,
       "step": 2622
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 16.15270233154297,
-      "learning_rate": 1.4743910995289165e-05,
-      "loss": 2.0384,
+      "epoch": 0.33,
+      "grad_norm": 16.72637176513672,
+      "learning_rate": 1.7806133121365523e-05,
+      "loss": 2.9599,
       "step": 2623
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 20.225746154785156,
-      "learning_rate": 1.4741906384684777e-05,
-      "loss": 2.7575,
+      "epoch": 0.33,
+      "grad_norm": 13.701391220092773,
+      "learning_rate": 1.7805296406308833e-05,
+      "loss": 4.3635,
       "step": 2624
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 29.362613677978516,
-      "learning_rate": 1.4739901774080385e-05,
-      "loss": 2.0288,
+      "epoch": 0.33,
+      "grad_norm": 15.569754600524902,
+      "learning_rate": 1.7804459691252146e-05,
+      "loss": 3.1822,
       "step": 2625
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 18.233638763427734,
-      "learning_rate": 1.4737897163475995e-05,
-      "loss": 2.4319,
+      "epoch": 0.33,
+      "grad_norm": 9.121126174926758,
+      "learning_rate": 1.7803622976195457e-05,
+      "loss": 1.5491,
       "step": 2626
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 15.207660675048828,
-      "learning_rate": 1.4735892552871607e-05,
-      "loss": 2.1672,
+      "epoch": 0.33,
+      "grad_norm": 12.470527648925781,
+      "learning_rate": 1.780278626113877e-05,
+      "loss": 1.9018,
       "step": 2627
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 21.931516647338867,
-      "learning_rate": 1.4733887942267215e-05,
-      "loss": 2.7001,
+      "epoch": 0.33,
+      "grad_norm": 9.061731338500977,
+      "learning_rate": 1.7801949546082084e-05,
+      "loss": 1.7634,
       "step": 2628
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 11.918092727661133,
-      "learning_rate": 1.4731883331662825e-05,
-      "loss": 2.1923,
+      "epoch": 0.33,
+      "grad_norm": 20.732572555541992,
+      "learning_rate": 1.7801112831025394e-05,
+      "loss": 1.6575,
       "step": 2629
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 24.173839569091797,
-      "learning_rate": 1.4729878721058436e-05,
-      "loss": 2.7606,
+      "epoch": 0.33,
+      "grad_norm": 18.07527732849121,
+      "learning_rate": 1.7800276115968708e-05,
+      "loss": 3.7961,
       "step": 2630
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 18.89200782775879,
-      "learning_rate": 1.4727874110454046e-05,
-      "loss": 2.3708,
+      "epoch": 0.33,
+      "grad_norm": 12.62188720703125,
+      "learning_rate": 1.779943940091202e-05,
+      "loss": 1.3397,
       "step": 2631
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 18.636999130249023,
-      "learning_rate": 1.4725869499849654e-05,
-      "loss": 2.3505,
+      "epoch": 0.33,
+      "grad_norm": 6.171538352966309,
+      "learning_rate": 1.7798602685855332e-05,
+      "loss": 0.4798,
       "step": 2632
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 24.666654586791992,
-      "learning_rate": 1.4723864889245266e-05,
-      "loss": 2.4304,
+      "epoch": 0.33,
+      "grad_norm": 8.710653305053711,
+      "learning_rate": 1.7797765970798645e-05,
+      "loss": 1.3855,
       "step": 2633
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 10.553362846374512,
-      "learning_rate": 1.4721860278640876e-05,
-      "loss": 1.8944,
+      "epoch": 0.33,
+      "grad_norm": 9.380043029785156,
+      "learning_rate": 1.779692925574196e-05,
+      "loss": 0.9739,
       "step": 2634
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 21.21739387512207,
-      "learning_rate": 1.4719855668036484e-05,
-      "loss": 2.6751,
+      "epoch": 0.33,
+      "grad_norm": 17.16081428527832,
+      "learning_rate": 1.779609254068527e-05,
+      "loss": 1.7828,
       "step": 2635
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 38.01958465576172,
-      "learning_rate": 1.4717851057432096e-05,
-      "loss": 1.8805,
+      "epoch": 0.33,
+      "grad_norm": 16.555482864379883,
+      "learning_rate": 1.7795255825628583e-05,
+      "loss": 3.708,
       "step": 2636
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 16.428316116333008,
-      "learning_rate": 1.4715846446827704e-05,
-      "loss": 2.9402,
+      "epoch": 0.33,
+      "grad_norm": 13.493598937988281,
+      "learning_rate": 1.7794419110571897e-05,
+      "loss": 1.899,
       "step": 2637
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 22.54930305480957,
-      "learning_rate": 1.4713841836223315e-05,
-      "loss": 1.9364,
+      "epoch": 0.33,
+      "grad_norm": 8.096471786499023,
+      "learning_rate": 1.779358239551521e-05,
+      "loss": 1.2152,
       "step": 2638
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 14.97050952911377,
-      "learning_rate": 1.4711837225618926e-05,
-      "loss": 1.7904,
+      "epoch": 0.33,
+      "grad_norm": 8.873205184936523,
+      "learning_rate": 1.779274568045852e-05,
+      "loss": 2.4998,
       "step": 2639
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 16.823516845703125,
-      "learning_rate": 1.4709832615014535e-05,
-      "loss": 2.2004,
-      "step": 2640
-    },
-    {
-      "epoch": 0.79,
-      "eval_loss": 0.28386011719703674,
-      "eval_runtime": 44.8703,
-      "eval_samples_per_second": 32.962,
-      "eval_steps_per_second": 32.962,
+      "epoch": 0.33,
+      "grad_norm": 7.485636234283447,
+      "learning_rate": 1.7791908965401834e-05,
+      "loss": 1.4641,
       "step": 2640
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 14.531737327575684,
-      "learning_rate": 1.4707828004410145e-05,
-      "loss": 2.1963,
+      "epoch": 0.33,
+      "grad_norm": 13.725582122802734,
+      "learning_rate": 1.7791072250345148e-05,
+      "loss": 1.7835,
       "step": 2641
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 24.638132095336914,
-      "learning_rate": 1.4705823393805753e-05,
-      "loss": 2.2464,
+      "epoch": 0.33,
+      "grad_norm": 16.9346923828125,
+      "learning_rate": 1.7790235535288458e-05,
+      "loss": 2.8712,
       "step": 2642
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 17.104135513305664,
-      "learning_rate": 1.4703818783201365e-05,
-      "loss": 2.0654,
+      "epoch": 0.33,
+      "grad_norm": 25.24087905883789,
+      "learning_rate": 1.7789398820231772e-05,
+      "loss": 3.9757,
       "step": 2643
     },
     {
-      "epoch": 0.79,
-      "grad_norm": 7.9517621994018555,
-      "learning_rate": 1.4701814172596973e-05,
-      "loss": 0.8721,
+      "epoch": 0.33,
+      "grad_norm": 24.99532127380371,
+      "learning_rate": 1.7788562105175085e-05,
+      "loss": 1.4441,
       "step": 2644
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 23.127620697021484,
-      "learning_rate": 1.4699809561992583e-05,
-      "loss": 2.0121,
+      "epoch": 0.33,
+      "grad_norm": 10.38292121887207,
+      "learning_rate": 1.77877253901184e-05,
+      "loss": 1.3967,
       "step": 2645
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 18.799203872680664,
-      "learning_rate": 1.4697804951388195e-05,
-      "loss": 2.0454,
+      "epoch": 0.33,
+      "grad_norm": 18.216909408569336,
+      "learning_rate": 1.778688867506171e-05,
+      "loss": 2.1929,
       "step": 2646
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 22.311729431152344,
-      "learning_rate": 1.4695800340783804e-05,
-      "loss": 2.1801,
+      "epoch": 0.33,
+      "grad_norm": 9.85275936126709,
+      "learning_rate": 1.7786051960005023e-05,
+      "loss": 2.047,
       "step": 2647
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 17.162185668945312,
-      "learning_rate": 1.4693795730179414e-05,
-      "loss": 1.814,
+      "epoch": 0.33,
+      "grad_norm": 12.772254943847656,
+      "learning_rate": 1.7785215244948337e-05,
+      "loss": 1.4923,
       "step": 2648
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 9.871721267700195,
-      "learning_rate": 1.4691791119575024e-05,
-      "loss": 1.6796,
+      "epoch": 0.33,
+      "grad_norm": 13.479397773742676,
+      "learning_rate": 1.7784378529891647e-05,
+      "loss": 2.3109,
       "step": 2649
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 18.128135681152344,
-      "learning_rate": 1.4689786508970634e-05,
-      "loss": 2.3865,
+      "epoch": 0.33,
+      "grad_norm": 28.885215759277344,
+      "learning_rate": 1.778354181483496e-05,
+      "loss": 2.2379,
       "step": 2650
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 15.663517951965332,
-      "learning_rate": 1.4687781898366244e-05,
-      "loss": 1.9406,
+      "epoch": 0.33,
+      "grad_norm": 12.536259651184082,
+      "learning_rate": 1.778270509977827e-05,
+      "loss": 2.3095,
       "step": 2651
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 37.457672119140625,
-      "learning_rate": 1.4685777287761854e-05,
-      "loss": 2.7683,
+      "epoch": 0.33,
+      "grad_norm": 12.961633682250977,
+      "learning_rate": 1.7781868384721584e-05,
+      "loss": 1.6991,
       "step": 2652
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 19.439468383789062,
-      "learning_rate": 1.4683772677157464e-05,
-      "loss": 2.354,
+      "epoch": 0.33,
+      "grad_norm": 10.719265937805176,
+      "learning_rate": 1.7781031669664898e-05,
+      "loss": 1.0747,
       "step": 2653
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 29.902244567871094,
-      "learning_rate": 1.4681768066553072e-05,
-      "loss": 1.993,
+      "epoch": 0.33,
+      "grad_norm": 31.90835952758789,
+      "learning_rate": 1.778019495460821e-05,
+      "loss": 3.2469,
       "step": 2654
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 11.236685752868652,
-      "learning_rate": 1.4679763455948684e-05,
-      "loss": 1.8682,
+      "epoch": 0.33,
+      "grad_norm": 20.774629592895508,
+      "learning_rate": 1.7779358239551522e-05,
+      "loss": 1.533,
       "step": 2655
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 11.110152244567871,
-      "learning_rate": 1.4677758845344293e-05,
-      "loss": 1.378,
+      "epoch": 0.33,
+      "grad_norm": 7.061487674713135,
+      "learning_rate": 1.7778521524494832e-05,
+      "loss": 3.181,
       "step": 2656
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 15.20004653930664,
-      "learning_rate": 1.4675754234739903e-05,
-      "loss": 2.4361,
+      "epoch": 0.33,
+      "grad_norm": 11.634260177612305,
+      "learning_rate": 1.7777684809438146e-05,
+      "loss": 1.6974,
       "step": 2657
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 44.016544342041016,
-      "learning_rate": 1.4673749624135514e-05,
-      "loss": 2.6364,
+      "epoch": 0.33,
+      "grad_norm": 15.743753433227539,
+      "learning_rate": 1.777684809438146e-05,
+      "loss": 2.3629,
       "step": 2658
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 18.134666442871094,
-      "learning_rate": 1.4671745013531123e-05,
-      "loss": 2.1308,
+      "epoch": 0.33,
+      "grad_norm": 10.4588041305542,
+      "learning_rate": 1.7776011379324773e-05,
+      "loss": 1.0229,
       "step": 2659
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 83.64218139648438,
-      "learning_rate": 1.4669740402926733e-05,
-      "loss": 3.793,
+      "epoch": 0.33,
+      "grad_norm": 23.603330612182617,
+      "learning_rate": 1.7775174664268083e-05,
+      "loss": 2.5311,
       "step": 2660
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 18.634946823120117,
-      "learning_rate": 1.4667735792322341e-05,
-      "loss": 2.0182,
+      "epoch": 0.33,
+      "grad_norm": 12.474515914916992,
+      "learning_rate": 1.7774337949211397e-05,
+      "loss": 2.6368,
       "step": 2661
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 46.72085189819336,
-      "learning_rate": 1.4665731181717953e-05,
-      "loss": 2.1095,
+      "epoch": 0.33,
+      "grad_norm": 15.076018333435059,
+      "learning_rate": 1.777350123415471e-05,
+      "loss": 2.7853,
       "step": 2662
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 18.656938552856445,
-      "learning_rate": 1.4663726571113563e-05,
-      "loss": 2.4388,
+      "epoch": 0.33,
+      "grad_norm": 11.735377311706543,
+      "learning_rate": 1.777266451909802e-05,
+      "loss": 3.4325,
       "step": 2663
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 16.998310089111328,
-      "learning_rate": 1.4661721960509172e-05,
-      "loss": 1.9669,
+      "epoch": 0.33,
+      "grad_norm": 12.364471435546875,
+      "learning_rate": 1.7771827804041335e-05,
+      "loss": 1.7619,
       "step": 2664
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 25.790451049804688,
-      "learning_rate": 1.4659717349904783e-05,
-      "loss": 2.6492,
+      "epoch": 0.33,
+      "grad_norm": 12.087916374206543,
+      "learning_rate": 1.777099108898465e-05,
+      "loss": 0.9063,
       "step": 2665
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 11.008011817932129,
-      "learning_rate": 1.4657712739300392e-05,
-      "loss": 1.926,
+      "epoch": 0.33,
+      "grad_norm": 9.966253280639648,
+      "learning_rate": 1.7770154373927962e-05,
+      "loss": 2.5133,
       "step": 2666
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 8.969502449035645,
-      "learning_rate": 1.4655708128696002e-05,
-      "loss": 0.8433,
+      "epoch": 0.33,
+      "grad_norm": 13.330294609069824,
+      "learning_rate": 1.7769317658871272e-05,
+      "loss": 2.1951,
       "step": 2667
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 24.146800994873047,
-      "learning_rate": 1.465370351809161e-05,
-      "loss": 1.7039,
+      "epoch": 0.33,
+      "grad_norm": 10.899292945861816,
+      "learning_rate": 1.7768480943814586e-05,
+      "loss": 2.7435,
       "step": 2668
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 32.524452209472656,
-      "learning_rate": 1.4651698907487222e-05,
-      "loss": 2.2074,
+      "epoch": 0.33,
+      "grad_norm": 17.162519454956055,
+      "learning_rate": 1.77676442287579e-05,
+      "loss": 2.4719,
       "step": 2669
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 17.53997802734375,
-      "learning_rate": 1.4649694296882832e-05,
-      "loss": 2.1067,
+      "epoch": 0.34,
+      "grad_norm": 8.715301513671875,
+      "learning_rate": 1.776680751370121e-05,
+      "loss": 0.6385,
       "step": 2670
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 70.99577331542969,
-      "learning_rate": 1.464768968627844e-05,
-      "loss": 2.5416,
+      "epoch": 0.34,
+      "grad_norm": 20.172853469848633,
+      "learning_rate": 1.7765970798644523e-05,
+      "loss": 2.5166,
       "step": 2671
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 15.582018852233887,
-      "learning_rate": 1.4645685075674052e-05,
-      "loss": 2.4462,
+      "epoch": 0.34,
+      "grad_norm": 11.463748931884766,
+      "learning_rate": 1.7765134083587837e-05,
+      "loss": 2.1591,
       "step": 2672
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 26.792186737060547,
-      "learning_rate": 1.464368046506966e-05,
-      "loss": 1.7403,
+      "epoch": 0.34,
+      "grad_norm": 15.173055648803711,
+      "learning_rate": 1.776429736853115e-05,
+      "loss": 1.7884,
       "step": 2673
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 13.64623737335205,
-      "learning_rate": 1.464167585446527e-05,
-      "loss": 2.7516,
+      "epoch": 0.34,
+      "grad_norm": 21.31320571899414,
+      "learning_rate": 1.776346065347446e-05,
+      "loss": 3.175,
       "step": 2674
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 17.885374069213867,
-      "learning_rate": 1.463967124386088e-05,
-      "loss": 2.9638,
+      "epoch": 0.34,
+      "grad_norm": 12.708723068237305,
+      "learning_rate": 1.7762623938417775e-05,
+      "loss": 1.8536,
       "step": 2675
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 11.002893447875977,
-      "learning_rate": 1.4637666633256491e-05,
-      "loss": 1.5192,
+      "epoch": 0.34,
+      "grad_norm": 9.75171947479248,
+      "learning_rate": 1.7761787223361088e-05,
+      "loss": 1.2302,
       "step": 2676
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 18.869712829589844,
-      "learning_rate": 1.4635662022652101e-05,
-      "loss": 2.5806,
+      "epoch": 0.34,
+      "grad_norm": 12.269949913024902,
+      "learning_rate": 1.77609505083044e-05,
+      "loss": 2.3773,
       "step": 2677
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 13.818095207214355,
-      "learning_rate": 1.4633657412047711e-05,
-      "loss": 2.1488,
+      "epoch": 0.34,
+      "grad_norm": 14.616438865661621,
+      "learning_rate": 1.7760113793247712e-05,
+      "loss": 2.2528,
       "step": 2678
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 28.199312210083008,
-      "learning_rate": 1.4631652801443321e-05,
-      "loss": 1.8242,
+      "epoch": 0.34,
+      "grad_norm": 10.644914627075195,
+      "learning_rate": 1.7759277078191022e-05,
+      "loss": 1.71,
       "step": 2679
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 19.68671417236328,
-      "learning_rate": 1.462964819083893e-05,
-      "loss": 1.5373,
+      "epoch": 0.34,
+      "grad_norm": 13.962441444396973,
+      "learning_rate": 1.7758440363134336e-05,
+      "loss": 1.6699,
       "step": 2680
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 36.305686950683594,
-      "learning_rate": 1.4627643580234541e-05,
-      "loss": 3.0527,
+      "epoch": 0.34,
+      "grad_norm": 33.41127395629883,
+      "learning_rate": 1.775760364807765e-05,
+      "loss": 1.6907,
       "step": 2681
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 22.186141967773438,
-      "learning_rate": 1.4625638969630151e-05,
-      "loss": 3.4701,
+      "epoch": 0.34,
+      "grad_norm": 11.684005737304688,
+      "learning_rate": 1.775676693302096e-05,
+      "loss": 0.837,
       "step": 2682
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 17.44498634338379,
-      "learning_rate": 1.462363435902576e-05,
-      "loss": 2.0777,
+      "epoch": 0.34,
+      "grad_norm": 11.661955833435059,
+      "learning_rate": 1.7755930217964274e-05,
+      "loss": 1.3837,
       "step": 2683
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 25.54764747619629,
-      "learning_rate": 1.4621629748421372e-05,
-      "loss": 1.9166,
+      "epoch": 0.34,
+      "grad_norm": 20.234384536743164,
+      "learning_rate": 1.7755093502907584e-05,
+      "loss": 3.4401,
       "step": 2684
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 15.869314193725586,
-      "learning_rate": 1.461962513781698e-05,
-      "loss": 2.0511,
+      "epoch": 0.34,
+      "grad_norm": 11.128459930419922,
+      "learning_rate": 1.7754256787850898e-05,
+      "loss": 1.632,
       "step": 2685
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 16.541519165039062,
-      "learning_rate": 1.461762052721259e-05,
-      "loss": 2.2221,
+      "epoch": 0.34,
+      "grad_norm": 7.553885459899902,
+      "learning_rate": 1.775342007279421e-05,
+      "loss": 1.9309,
       "step": 2686
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 12.096132278442383,
-      "learning_rate": 1.4615615916608198e-05,
-      "loss": 2.2477,
+      "epoch": 0.34,
+      "grad_norm": 17.167726516723633,
+      "learning_rate": 1.775258335773752e-05,
+      "loss": 0.35,
       "step": 2687
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 32.8545036315918,
-      "learning_rate": 1.461361130600381e-05,
-      "loss": 2.4819,
+      "epoch": 0.34,
+      "grad_norm": 16.623626708984375,
+      "learning_rate": 1.7751746642680835e-05,
+      "loss": 2.3347,
       "step": 2688
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 17.078535079956055,
-      "learning_rate": 1.461160669539942e-05,
-      "loss": 2.5493,
+      "epoch": 0.34,
+      "grad_norm": 15.077774047851562,
+      "learning_rate": 1.775090992762415e-05,
+      "loss": 2.5027,
       "step": 2689
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 20.856630325317383,
-      "learning_rate": 1.4609602084795029e-05,
-      "loss": 2.9245,
+      "epoch": 0.34,
+      "grad_norm": 17.26755142211914,
+      "learning_rate": 1.7750073212567462e-05,
+      "loss": 1.7269,
       "step": 2690
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 30.489768981933594,
-      "learning_rate": 1.460759747419064e-05,
-      "loss": 2.2884,
+      "epoch": 0.34,
+      "grad_norm": 9.179740905761719,
+      "learning_rate": 1.7749236497510773e-05,
+      "loss": 1.4656,
       "step": 2691
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 13.105022430419922,
-      "learning_rate": 1.4605592863586249e-05,
-      "loss": 2.3658,
+      "epoch": 0.34,
+      "grad_norm": 13.58669376373291,
+      "learning_rate": 1.7748399782454086e-05,
+      "loss": 3.1928,
       "step": 2692
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 14.78238296508789,
-      "learning_rate": 1.4603588252981859e-05,
-      "loss": 2.5098,
+      "epoch": 0.34,
+      "grad_norm": 14.194979667663574,
+      "learning_rate": 1.77475630673974e-05,
+      "loss": 3.3585,
       "step": 2693
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 17.489654541015625,
-      "learning_rate": 1.460158364237747e-05,
-      "loss": 1.8979,
+      "epoch": 0.34,
+      "grad_norm": 24.152509689331055,
+      "learning_rate": 1.774672635234071e-05,
+      "loss": 1.9617,
       "step": 2694
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 15.65456771850586,
-      "learning_rate": 1.4599579031773079e-05,
-      "loss": 2.2142,
+      "epoch": 0.34,
+      "grad_norm": 20.574277877807617,
+      "learning_rate": 1.7745889637284024e-05,
+      "loss": 1.93,
       "step": 2695
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 24.52695655822754,
-      "learning_rate": 1.459757442116869e-05,
-      "loss": 2.7726,
+      "epoch": 0.34,
+      "grad_norm": 15.949616432189941,
+      "learning_rate": 1.7745052922227338e-05,
+      "loss": 1.8353,
       "step": 2696
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 33.155067443847656,
-      "learning_rate": 1.4595569810564298e-05,
-      "loss": 2.2196,
+      "epoch": 0.34,
+      "grad_norm": 14.588492393493652,
+      "learning_rate": 1.774421620717065e-05,
+      "loss": 1.7303,
       "step": 2697
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 9.755550384521484,
-      "learning_rate": 1.459356519995991e-05,
-      "loss": 1.7204,
+      "epoch": 0.34,
+      "grad_norm": 11.943470001220703,
+      "learning_rate": 1.774337949211396e-05,
+      "loss": 2.2723,
       "step": 2698
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 19.83466911315918,
-      "learning_rate": 1.4591560589355518e-05,
-      "loss": 2.5923,
+      "epoch": 0.34,
+      "grad_norm": 10.931513786315918,
+      "learning_rate": 1.7742542777057275e-05,
+      "loss": 1.7117,
       "step": 2699
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 21.58110809326172,
-      "learning_rate": 1.458955597875113e-05,
-      "loss": 1.7443,
+      "epoch": 0.34,
+      "grad_norm": 11.123241424560547,
+      "learning_rate": 1.774170606200059e-05,
+      "loss": 1.2408,
       "step": 2700
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 20.373117446899414,
-      "learning_rate": 1.458755136814674e-05,
-      "loss": 2.949,
+      "epoch": 0.34,
+      "grad_norm": 24.709917068481445,
+      "learning_rate": 1.77408693469439e-05,
+      "loss": 2.841,
       "step": 2701
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 12.08033275604248,
-      "learning_rate": 1.4585546757542348e-05,
-      "loss": 1.7417,
+      "epoch": 0.34,
+      "grad_norm": 75.23438262939453,
+      "learning_rate": 1.7740032631887213e-05,
+      "loss": 1.7636,
       "step": 2702
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 16.984411239624023,
-      "learning_rate": 1.458354214693796e-05,
-      "loss": 1.973,
+      "epoch": 0.34,
+      "grad_norm": 19.17953872680664,
+      "learning_rate": 1.7739195916830526e-05,
+      "loss": 3.1879,
       "step": 2703
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 24.945363998413086,
-      "learning_rate": 1.4581537536333568e-05,
-      "loss": 1.7648,
+      "epoch": 0.34,
+      "grad_norm": 12.488292694091797,
+      "learning_rate": 1.7738359201773837e-05,
+      "loss": 2.1714,
       "step": 2704
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 18.574451446533203,
-      "learning_rate": 1.4579532925729178e-05,
-      "loss": 1.326,
+      "epoch": 0.34,
+      "grad_norm": 18.393566131591797,
+      "learning_rate": 1.773752248671715e-05,
+      "loss": 3.5136,
       "step": 2705
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 32.63951873779297,
-      "learning_rate": 1.457752831512479e-05,
-      "loss": 2.0216,
+      "epoch": 0.34,
+      "grad_norm": 20.458900451660156,
+      "learning_rate": 1.7736685771660464e-05,
+      "loss": 1.256,
       "step": 2706
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 20.653427124023438,
-      "learning_rate": 1.4575523704520398e-05,
-      "loss": 2.0913,
+      "epoch": 0.34,
+      "grad_norm": 10.179489135742188,
+      "learning_rate": 1.7735849056603774e-05,
+      "loss": 0.9635,
       "step": 2707
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 41.5728645324707,
-      "learning_rate": 1.4573519093916009e-05,
-      "loss": 3.4579,
+      "epoch": 0.34,
+      "grad_norm": 15.560358047485352,
+      "learning_rate": 1.7735012341547088e-05,
+      "loss": 2.8496,
       "step": 2708
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 12.26309585571289,
-      "learning_rate": 1.4571514483311617e-05,
-      "loss": 1.4278,
+      "epoch": 0.34,
+      "grad_norm": 14.694575309753418,
+      "learning_rate": 1.7734175626490398e-05,
+      "loss": 1.7613,
       "step": 2709
     },
     {
-      "epoch": 0.81,
-      "grad_norm": 17.044031143188477,
-      "learning_rate": 1.4569509872707229e-05,
-      "loss": 1.7787,
+      "epoch": 0.34,
+      "grad_norm": 11.86589527130127,
+      "learning_rate": 1.7733338911433712e-05,
+      "loss": 1.9254,
       "step": 2710
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 22.376020431518555,
-      "learning_rate": 1.4567505262102837e-05,
-      "loss": 2.2045,
+      "epoch": 0.34,
+      "grad_norm": 7.738234519958496,
+      "learning_rate": 1.7732502196377025e-05,
+      "loss": 0.7106,
       "step": 2711
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 15.766566276550293,
-      "learning_rate": 1.4565500651498447e-05,
-      "loss": 2.087,
+      "epoch": 0.34,
+      "grad_norm": 35.35801315307617,
+      "learning_rate": 1.7731665481320336e-05,
+      "loss": 2.4502,
       "step": 2712
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 10.13654899597168,
-      "learning_rate": 1.4563496040894059e-05,
-      "loss": 2.7784,
+      "epoch": 0.34,
+      "grad_norm": 15.700124740600586,
+      "learning_rate": 1.773082876626365e-05,
+      "loss": 2.1184,
       "step": 2713
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 12.07814884185791,
-      "learning_rate": 1.4561491430289667e-05,
-      "loss": 2.2884,
+      "epoch": 0.34,
+      "grad_norm": 4.762723445892334,
+      "learning_rate": 1.7729992051206963e-05,
+      "loss": 0.1881,
       "step": 2714
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 15.864092826843262,
-      "learning_rate": 1.4559486819685277e-05,
-      "loss": 2.1431,
+      "epoch": 0.34,
+      "grad_norm": 16.25357437133789,
+      "learning_rate": 1.7729155336150273e-05,
+      "loss": 1.4104,
       "step": 2715
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 18.540136337280273,
-      "learning_rate": 1.4557482209080886e-05,
-      "loss": 2.4553,
+      "epoch": 0.34,
+      "grad_norm": 19.92592430114746,
+      "learning_rate": 1.7728318621093587e-05,
+      "loss": 3.8268,
       "step": 2716
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 9.907877922058105,
-      "learning_rate": 1.4555477598476498e-05,
-      "loss": 1.6814,
+      "epoch": 0.34,
+      "grad_norm": 12.886872291564941,
+      "learning_rate": 1.77274819060369e-05,
+      "loss": 3.1179,
       "step": 2717
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 9.542145729064941,
-      "learning_rate": 1.4553472987872106e-05,
-      "loss": 1.1338,
+      "epoch": 0.34,
+      "grad_norm": 14.823153495788574,
+      "learning_rate": 1.7726645190980214e-05,
+      "loss": 0.8029,
       "step": 2718
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 18.332304000854492,
-      "learning_rate": 1.4551468377267716e-05,
-      "loss": 1.7549,
+      "epoch": 0.34,
+      "grad_norm": 14.305251121520996,
+      "learning_rate": 1.7725808475923524e-05,
+      "loss": 3.3145,
       "step": 2719
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 9.094773292541504,
-      "learning_rate": 1.4549463766663328e-05,
-      "loss": 1.1251,
+      "epoch": 0.34,
+      "grad_norm": 14.948331832885742,
+      "learning_rate": 1.7724971760866838e-05,
+      "loss": 2.4797,
       "step": 2720
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 15.948875427246094,
-      "learning_rate": 1.4547459156058936e-05,
-      "loss": 1.5578,
+      "epoch": 0.34,
+      "grad_norm": 6.8798699378967285,
+      "learning_rate": 1.772413504581015e-05,
+      "loss": 0.9005,
       "step": 2721
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 13.107511520385742,
-      "learning_rate": 1.4545454545454546e-05,
-      "loss": 1.8671,
+      "epoch": 0.34,
+      "grad_norm": 18.058286666870117,
+      "learning_rate": 1.7723298330753462e-05,
+      "loss": 1.6352,
       "step": 2722
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 17.999923706054688,
-      "learning_rate": 1.4543449934850156e-05,
-      "loss": 2.1096,
+      "epoch": 0.34,
+      "grad_norm": 11.84815502166748,
+      "learning_rate": 1.7722461615696776e-05,
+      "loss": 1.6202,
       "step": 2723
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 22.923246383666992,
-      "learning_rate": 1.4541445324245767e-05,
-      "loss": 2.1625,
+      "epoch": 0.34,
+      "grad_norm": 12.707499504089355,
+      "learning_rate": 1.772162490064009e-05,
+      "loss": 1.5351,
       "step": 2724
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 12.2493314743042,
-      "learning_rate": 1.4539440713641377e-05,
-      "loss": 2.9069,
+      "epoch": 0.34,
+      "grad_norm": 12.71505355834961,
+      "learning_rate": 1.7720788185583403e-05,
+      "loss": 2.962,
       "step": 2725
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 20.940412521362305,
-      "learning_rate": 1.4537436103036987e-05,
-      "loss": 2.6205,
+      "epoch": 0.34,
+      "grad_norm": 14.472443580627441,
+      "learning_rate": 1.7719951470526713e-05,
+      "loss": 2.7893,
       "step": 2726
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 28.031394958496094,
-      "learning_rate": 1.4535431492432597e-05,
-      "loss": 2.167,
+      "epoch": 0.34,
+      "grad_norm": 11.948232650756836,
+      "learning_rate": 1.7719114755470027e-05,
+      "loss": 1.3851,
       "step": 2727
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 12.496899604797363,
-      "learning_rate": 1.4533426881828205e-05,
-      "loss": 2.8417,
+      "epoch": 0.34,
+      "grad_norm": 26.376602172851562,
+      "learning_rate": 1.771827804041334e-05,
+      "loss": 2.3619,
       "step": 2728
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 15.856078147888184,
-      "learning_rate": 1.4531422271223817e-05,
-      "loss": 2.5802,
+      "epoch": 0.34,
+      "grad_norm": 14.06189250946045,
+      "learning_rate": 1.771744132535665e-05,
+      "loss": 1.9028,
       "step": 2729
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 19.049468994140625,
-      "learning_rate": 1.4529417660619425e-05,
-      "loss": 2.2789,
+      "epoch": 0.34,
+      "grad_norm": 13.420397758483887,
+      "learning_rate": 1.7716604610299964e-05,
+      "loss": 2.0853,
       "step": 2730
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 16.88075828552246,
-      "learning_rate": 1.4527413050015035e-05,
-      "loss": 1.6496,
+      "epoch": 0.34,
+      "grad_norm": 12.996766090393066,
+      "learning_rate": 1.7715767895243278e-05,
+      "loss": 2.9761,
       "step": 2731
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 23.054603576660156,
-      "learning_rate": 1.4525408439410647e-05,
-      "loss": 1.8223,
+      "epoch": 0.34,
+      "grad_norm": 11.15153980255127,
+      "learning_rate": 1.7714931180186588e-05,
+      "loss": 1.5978,
       "step": 2732
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 27.310646057128906,
-      "learning_rate": 1.4523403828806256e-05,
-      "loss": 2.5239,
+      "epoch": 0.34,
+      "grad_norm": 24.499765396118164,
+      "learning_rate": 1.7714094465129902e-05,
+      "loss": 2.8331,
       "step": 2733
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 20.423734664916992,
-      "learning_rate": 1.4521399218201866e-05,
-      "loss": 2.2311,
+      "epoch": 0.34,
+      "grad_norm": 125.93241882324219,
+      "learning_rate": 1.7713257750073216e-05,
+      "loss": 1.9642,
       "step": 2734
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 13.929975509643555,
-      "learning_rate": 1.4519394607597474e-05,
-      "loss": 2.5546,
+      "epoch": 0.34,
+      "grad_norm": 21.821949005126953,
+      "learning_rate": 1.7712421035016526e-05,
+      "loss": 2.0355,
       "step": 2735
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 12.425309181213379,
-      "learning_rate": 1.4517389996993086e-05,
-      "loss": 1.4081,
+      "epoch": 0.34,
+      "grad_norm": 12.31302261352539,
+      "learning_rate": 1.771158431995984e-05,
+      "loss": 1.9599,
       "step": 2736
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 11.25351333618164,
-      "learning_rate": 1.4515385386388696e-05,
-      "loss": 1.7953,
+      "epoch": 0.34,
+      "grad_norm": 12.336325645446777,
+      "learning_rate": 1.771074760490315e-05,
+      "loss": 2.4827,
       "step": 2737
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 19.53854751586914,
-      "learning_rate": 1.4513380775784304e-05,
-      "loss": 1.207,
+      "epoch": 0.34,
+      "grad_norm": 11.652437210083008,
+      "learning_rate": 1.7709910889846463e-05,
+      "loss": 2.3775,
       "step": 2738
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 22.41812515258789,
-      "learning_rate": 1.4511376165179916e-05,
-      "loss": 2.4787,
+      "epoch": 0.34,
+      "grad_norm": 7.911839008331299,
+      "learning_rate": 1.7709074174789777e-05,
+      "loss": 1.0462,
       "step": 2739
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 21.234926223754883,
-      "learning_rate": 1.4509371554575524e-05,
-      "loss": 2.2675,
+      "epoch": 0.34,
+      "grad_norm": 11.181539535522461,
+      "learning_rate": 1.7708237459733087e-05,
+      "loss": 1.6236,
       "step": 2740
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 19.57403564453125,
-      "learning_rate": 1.4507366943971135e-05,
-      "loss": 2.3916,
+      "epoch": 0.34,
+      "grad_norm": 21.37649154663086,
+      "learning_rate": 1.77074007446764e-05,
+      "loss": 2.3584,
       "step": 2741
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 25.845134735107422,
-      "learning_rate": 1.4505362333366743e-05,
-      "loss": 2.4502,
+      "epoch": 0.34,
+      "grad_norm": 8.938224792480469,
+      "learning_rate": 1.7706564029619715e-05,
+      "loss": 2.2325,
       "step": 2742
     },
     {
-      "epoch": 0.82,
-      "grad_norm": 42.33112716674805,
-      "learning_rate": 1.4503357722762355e-05,
-      "loss": 2.0345,
+      "epoch": 0.34,
+      "grad_norm": 17.55942153930664,
+      "learning_rate": 1.7705727314563025e-05,
+      "loss": 1.9935,
       "step": 2743
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 15.049954414367676,
-      "learning_rate": 1.4501353112157965e-05,
-      "loss": 2.8193,
+      "epoch": 0.34,
+      "grad_norm": 31.897920608520508,
+      "learning_rate": 1.770489059950634e-05,
+      "loss": 1.6684,
       "step": 2744
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 25.87766456604004,
-      "learning_rate": 1.4499348501553573e-05,
-      "loss": 2.2518,
+      "epoch": 0.34,
+      "grad_norm": 23.712093353271484,
+      "learning_rate": 1.7704053884449652e-05,
+      "loss": 1.8706,
       "step": 2745
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 21.152681350708008,
-      "learning_rate": 1.4497343890949185e-05,
-      "loss": 2.3211,
+      "epoch": 0.34,
+      "grad_norm": 14.985076904296875,
+      "learning_rate": 1.7703217169392966e-05,
+      "loss": 2.9731,
       "step": 2746
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 36.79154586791992,
-      "learning_rate": 1.4495339280344793e-05,
-      "loss": 2.1298,
+      "epoch": 0.34,
+      "grad_norm": 25.487062454223633,
+      "learning_rate": 1.7702380454336276e-05,
+      "loss": 2.1705,
       "step": 2747
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 9.398030281066895,
-      "learning_rate": 1.4493334669740403e-05,
-      "loss": 1.8969,
+      "epoch": 0.34,
+      "grad_norm": 11.097169876098633,
+      "learning_rate": 1.770154373927959e-05,
+      "loss": 1.2433,
       "step": 2748
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 20.946956634521484,
-      "learning_rate": 1.4491330059136015e-05,
-      "loss": 2.3464,
+      "epoch": 0.34,
+      "grad_norm": 7.978656768798828,
+      "learning_rate": 1.7700707024222903e-05,
+      "loss": 1.1207,
       "step": 2749
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 9.130115509033203,
-      "learning_rate": 1.4489325448531624e-05,
-      "loss": 1.7633,
+      "epoch": 0.35,
+      "grad_norm": 7.823447227478027,
+      "learning_rate": 1.7699870309166214e-05,
+      "loss": 1.0118,
       "step": 2750
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 14.199234962463379,
-      "learning_rate": 1.4487320837927234e-05,
-      "loss": 1.8352,
+      "epoch": 0.35,
+      "grad_norm": 15.047630310058594,
+      "learning_rate": 1.7699033594109527e-05,
+      "loss": 1.2199,
       "step": 2751
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 8.78415298461914,
-      "learning_rate": 1.4485316227322844e-05,
-      "loss": 1.6194,
+      "epoch": 0.35,
+      "grad_norm": 14.528481483459473,
+      "learning_rate": 1.769819687905284e-05,
+      "loss": 1.4999,
       "step": 2752
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 32.84090805053711,
-      "learning_rate": 1.4483311616718454e-05,
-      "loss": 1.7946,
+      "epoch": 0.35,
+      "grad_norm": 21.73208999633789,
+      "learning_rate": 1.7697360163996155e-05,
+      "loss": 2.8917,
       "step": 2753
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 18.42530632019043,
-      "learning_rate": 1.4481307006114062e-05,
-      "loss": 2.4003,
+      "epoch": 0.35,
+      "grad_norm": 17.519514083862305,
+      "learning_rate": 1.7696523448939465e-05,
+      "loss": 2.5031,
       "step": 2754
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 10.56641960144043,
-      "learning_rate": 1.4479302395509674e-05,
-      "loss": 1.698,
+      "epoch": 0.35,
+      "grad_norm": 13.164900779724121,
+      "learning_rate": 1.769568673388278e-05,
+      "loss": 1.3442,
       "step": 2755
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 13.460083961486816,
-      "learning_rate": 1.4477297784905284e-05,
-      "loss": 2.7324,
+      "epoch": 0.35,
+      "grad_norm": 18.803659439086914,
+      "learning_rate": 1.7694850018826092e-05,
+      "loss": 1.4444,
       "step": 2756
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 20.335878372192383,
-      "learning_rate": 1.4475293174300893e-05,
-      "loss": 1.5947,
+      "epoch": 0.35,
+      "grad_norm": 38.67995834350586,
+      "learning_rate": 1.7694013303769402e-05,
+      "loss": 0.2441,
       "step": 2757
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 44.001834869384766,
-      "learning_rate": 1.4473288563696504e-05,
-      "loss": 1.8942,
+      "epoch": 0.35,
+      "grad_norm": 13.010946273803711,
+      "learning_rate": 1.7693176588712716e-05,
+      "loss": 1.1844,
       "step": 2758
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 13.544523239135742,
-      "learning_rate": 1.4471283953092113e-05,
-      "loss": 2.016,
+      "epoch": 0.35,
+      "grad_norm": 17.245418548583984,
+      "learning_rate": 1.769233987365603e-05,
+      "loss": 2.793,
       "step": 2759
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 30.979524612426758,
-      "learning_rate": 1.4469279342487723e-05,
-      "loss": 3.1479,
-      "step": 2760
-    },
-    {
-      "epoch": 0.83,
-      "eval_loss": 0.2657427191734314,
-      "eval_runtime": 43.5175,
-      "eval_samples_per_second": 33.986,
-      "eval_steps_per_second": 33.986,
+      "epoch": 0.35,
+      "grad_norm": 8.228273391723633,
+      "learning_rate": 1.769150315859934e-05,
+      "loss": 1.1293,
       "step": 2760
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 23.16549301147461,
-      "learning_rate": 1.4467274731883331e-05,
-      "loss": 2.2599,
+      "epoch": 0.35,
+      "grad_norm": 15.159379959106445,
+      "learning_rate": 1.7690666443542654e-05,
+      "loss": 2.7442,
       "step": 2761
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 18.973628997802734,
-      "learning_rate": 1.4465270121278943e-05,
-      "loss": 2.1258,
+      "epoch": 0.35,
+      "grad_norm": 7.7474894523620605,
+      "learning_rate": 1.7689829728485964e-05,
+      "loss": 0.6898,
       "step": 2762
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 26.273101806640625,
-      "learning_rate": 1.4463265510674553e-05,
-      "loss": 3.4268,
+      "epoch": 0.35,
+      "grad_norm": 22.496170043945312,
+      "learning_rate": 1.7688993013429277e-05,
+      "loss": 2.9833,
       "step": 2763
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 14.230957984924316,
-      "learning_rate": 1.4461260900070161e-05,
-      "loss": 1.4143,
+      "epoch": 0.35,
+      "grad_norm": 15.267127990722656,
+      "learning_rate": 1.768815629837259e-05,
+      "loss": 3.6275,
       "step": 2764
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 14.468798637390137,
-      "learning_rate": 1.4459256289465773e-05,
-      "loss": 2.3477,
+      "epoch": 0.35,
+      "grad_norm": 10.416468620300293,
+      "learning_rate": 1.76873195833159e-05,
+      "loss": 2.7587,
       "step": 2765
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 16.977825164794922,
-      "learning_rate": 1.4457251678861382e-05,
-      "loss": 2.3843,
+      "epoch": 0.35,
+      "grad_norm": 12.335618019104004,
+      "learning_rate": 1.7686482868259215e-05,
+      "loss": 2.8012,
       "step": 2766
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 14.977238655090332,
-      "learning_rate": 1.4455247068256992e-05,
-      "loss": 2.0197,
+      "epoch": 0.35,
+      "grad_norm": 9.918840408325195,
+      "learning_rate": 1.768564615320253e-05,
+      "loss": 1.3928,
       "step": 2767
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 14.926851272583008,
-      "learning_rate": 1.4453242457652603e-05,
-      "loss": 2.9558,
+      "epoch": 0.35,
+      "grad_norm": 36.22608184814453,
+      "learning_rate": 1.768480943814584e-05,
+      "loss": 2.8527,
       "step": 2768
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 23.74448013305664,
-      "learning_rate": 1.4451237847048212e-05,
-      "loss": 2.1472,
+      "epoch": 0.35,
+      "grad_norm": 37.150150299072266,
+      "learning_rate": 1.7683972723089153e-05,
+      "loss": 2.1441,
       "step": 2769
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 41.35685729980469,
-      "learning_rate": 1.4449233236443822e-05,
-      "loss": 2.7331,
+      "epoch": 0.35,
+      "grad_norm": 6.690373420715332,
+      "learning_rate": 1.7683136008032466e-05,
+      "loss": 1.4193,
       "step": 2770
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 13.470633506774902,
-      "learning_rate": 1.4447228625839432e-05,
-      "loss": 1.9928,
+      "epoch": 0.35,
+      "grad_norm": 13.47512149810791,
+      "learning_rate": 1.7682299292975777e-05,
+      "loss": 2.6323,
       "step": 2771
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 12.274412155151367,
-      "learning_rate": 1.4445224015235042e-05,
-      "loss": 1.7747,
+      "epoch": 0.35,
+      "grad_norm": 18.517271041870117,
+      "learning_rate": 1.768146257791909e-05,
+      "loss": 2.4489,
       "step": 2772
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 12.649324417114258,
-      "learning_rate": 1.444321940463065e-05,
-      "loss": 1.599,
+      "epoch": 0.35,
+      "grad_norm": 13.417510032653809,
+      "learning_rate": 1.7680625862862404e-05,
+      "loss": 2.7828,
       "step": 2773
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 13.914393424987793,
-      "learning_rate": 1.4441214794026262e-05,
-      "loss": 1.7285,
+      "epoch": 0.35,
+      "grad_norm": 19.947721481323242,
+      "learning_rate": 1.7679789147805717e-05,
+      "loss": 3.3091,
       "step": 2774
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 9.780823707580566,
-      "learning_rate": 1.4439210183421872e-05,
-      "loss": 1.3613,
+      "epoch": 0.35,
+      "grad_norm": 8.461196899414062,
+      "learning_rate": 1.7678952432749028e-05,
+      "loss": 3.1268,
       "step": 2775
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 13.444343566894531,
-      "learning_rate": 1.443720557281748e-05,
-      "loss": 1.4692,
+      "epoch": 0.35,
+      "grad_norm": 16.18375015258789,
+      "learning_rate": 1.767811571769234e-05,
+      "loss": 1.1768,
       "step": 2776
     },
     {
-      "epoch": 0.83,
-      "grad_norm": 11.597440719604492,
-      "learning_rate": 1.4435200962213093e-05,
-      "loss": 2.5432,
+      "epoch": 0.35,
+      "grad_norm": 20.837011337280273,
+      "learning_rate": 1.7677279002635655e-05,
+      "loss": 1.2293,
       "step": 2777
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 66.30996704101562,
-      "learning_rate": 1.4433196351608701e-05,
-      "loss": 1.9169,
+      "epoch": 0.35,
+      "grad_norm": 16.653564453125,
+      "learning_rate": 1.7676442287578965e-05,
+      "loss": 1.8402,
       "step": 2778
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 42.77153778076172,
-      "learning_rate": 1.4431191741004311e-05,
-      "loss": 2.1056,
+      "epoch": 0.35,
+      "grad_norm": 7.685312271118164,
+      "learning_rate": 1.767560557252228e-05,
+      "loss": 2.022,
       "step": 2779
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 17.99850082397461,
-      "learning_rate": 1.4429187130399923e-05,
-      "loss": 1.6885,
+      "epoch": 0.35,
+      "grad_norm": 23.538360595703125,
+      "learning_rate": 1.7674768857465593e-05,
+      "loss": 1.1104,
       "step": 2780
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 11.443147659301758,
-      "learning_rate": 1.4427182519795531e-05,
-      "loss": 2.2285,
+      "epoch": 0.35,
+      "grad_norm": 15.101953506469727,
+      "learning_rate": 1.7673932142408906e-05,
+      "loss": 3.7181,
       "step": 2781
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 18.145082473754883,
-      "learning_rate": 1.4425177909191141e-05,
-      "loss": 2.14,
+      "epoch": 0.35,
+      "grad_norm": 20.23198890686035,
+      "learning_rate": 1.7673095427352216e-05,
+      "loss": 3.4027,
       "step": 2782
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 12.900897026062012,
-      "learning_rate": 1.442317329858675e-05,
-      "loss": 1.8419,
+      "epoch": 0.35,
+      "grad_norm": 13.295400619506836,
+      "learning_rate": 1.767225871229553e-05,
+      "loss": 3.3314,
       "step": 2783
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 12.71401596069336,
-      "learning_rate": 1.4421168687982361e-05,
-      "loss": 1.6758,
+      "epoch": 0.35,
+      "grad_norm": 16.027250289916992,
+      "learning_rate": 1.7671421997238844e-05,
+      "loss": 1.3427,
       "step": 2784
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 10.644526481628418,
-      "learning_rate": 1.441916407737797e-05,
-      "loss": 1.5368,
+      "epoch": 0.35,
+      "grad_norm": 12.23305892944336,
+      "learning_rate": 1.7670585282182154e-05,
+      "loss": 1.5726,
       "step": 2785
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 11.878511428833008,
-      "learning_rate": 1.441715946677358e-05,
-      "loss": 2.8906,
+      "epoch": 0.35,
+      "grad_norm": 11.057424545288086,
+      "learning_rate": 1.7669748567125468e-05,
+      "loss": 2.4224,
       "step": 2786
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 23.1770076751709,
-      "learning_rate": 1.4415154856169192e-05,
-      "loss": 2.9468,
+      "epoch": 0.35,
+      "grad_norm": 18.01150894165039,
+      "learning_rate": 1.766891185206878e-05,
+      "loss": 3.5597,
       "step": 2787
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 13.855196952819824,
-      "learning_rate": 1.44131502455648e-05,
-      "loss": 2.0071,
+      "epoch": 0.35,
+      "grad_norm": 19.285484313964844,
+      "learning_rate": 1.766807513701209e-05,
+      "loss": 1.7419,
       "step": 2788
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 10.715313911437988,
-      "learning_rate": 1.441114563496041e-05,
-      "loss": 1.3851,
+      "epoch": 0.35,
+      "grad_norm": 14.585424423217773,
+      "learning_rate": 1.7667238421955405e-05,
+      "loss": 3.3739,
       "step": 2789
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 11.7968168258667,
-      "learning_rate": 1.4409141024356019e-05,
-      "loss": 1.7609,
+      "epoch": 0.35,
+      "grad_norm": 9.299799919128418,
+      "learning_rate": 1.7666401706898716e-05,
+      "loss": 1.6106,
       "step": 2790
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 15.877503395080566,
-      "learning_rate": 1.440713641375163e-05,
-      "loss": 3.124,
+      "epoch": 0.35,
+      "grad_norm": 16.445552825927734,
+      "learning_rate": 1.766556499184203e-05,
+      "loss": 1.2977,
       "step": 2791
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 19.398963928222656,
-      "learning_rate": 1.440513180314724e-05,
-      "loss": 2.2776,
+      "epoch": 0.35,
+      "grad_norm": 10.708333969116211,
+      "learning_rate": 1.7664728276785343e-05,
+      "loss": 3.0598,
       "step": 2792
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 13.756909370422363,
-      "learning_rate": 1.4403127192542849e-05,
-      "loss": 2.429,
+      "epoch": 0.35,
+      "grad_norm": 15.353691101074219,
+      "learning_rate": 1.7663891561728653e-05,
+      "loss": 3.7235,
       "step": 2793
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 8.20438289642334,
-      "learning_rate": 1.440112258193846e-05,
-      "loss": 1.3021,
+      "epoch": 0.35,
+      "grad_norm": 22.779550552368164,
+      "learning_rate": 1.7663054846671967e-05,
+      "loss": 2.2107,
       "step": 2794
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 35.246524810791016,
-      "learning_rate": 1.4399117971334069e-05,
-      "loss": 3.4113,
+      "epoch": 0.35,
+      "grad_norm": 5.81117057800293,
+      "learning_rate": 1.766221813161528e-05,
+      "loss": 0.5669,
       "step": 2795
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 15.750861167907715,
-      "learning_rate": 1.4397113360729679e-05,
-      "loss": 2.1455,
+      "epoch": 0.35,
+      "grad_norm": 4.142109394073486,
+      "learning_rate": 1.766138141655859e-05,
+      "loss": 0.2277,
       "step": 2796
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 12.513425827026367,
-      "learning_rate": 1.4395108750125289e-05,
-      "loss": 2.1875,
+      "epoch": 0.35,
+      "grad_norm": 15.641167640686035,
+      "learning_rate": 1.7660544701501904e-05,
+      "loss": 2.264,
       "step": 2797
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 20.23858070373535,
-      "learning_rate": 1.43931041395209e-05,
-      "loss": 2.686,
+      "epoch": 0.35,
+      "grad_norm": 12.36007022857666,
+      "learning_rate": 1.7659707986445218e-05,
+      "loss": 3.5559,
       "step": 2798
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 14.214842796325684,
-      "learning_rate": 1.439109952891651e-05,
-      "loss": 1.5095,
+      "epoch": 0.35,
+      "grad_norm": 17.426403045654297,
+      "learning_rate": 1.7658871271388528e-05,
+      "loss": 3.0523,
       "step": 2799
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 13.132567405700684,
-      "learning_rate": 1.438909491831212e-05,
-      "loss": 1.4686,
+      "epoch": 0.35,
+      "grad_norm": 13.43343448638916,
+      "learning_rate": 1.7658034556331842e-05,
+      "loss": 3.2828,
       "step": 2800
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 12.941914558410645,
-      "learning_rate": 1.438709030770773e-05,
-      "loss": 1.6349,
+      "epoch": 0.35,
+      "eval_loss": 0.15001514554023743,
+      "eval_runtime": 94.9595,
+      "eval_samples_per_second": 37.3,
+      "eval_steps_per_second": 37.3,
+      "step": 2800
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 20.58778953552246,
+      "learning_rate": 1.7657197841275155e-05,
+      "loss": 3.473,
       "step": 2801
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 23.00050926208496,
-      "learning_rate": 1.4385085697103338e-05,
-      "loss": 2.3867,
+      "epoch": 0.35,
+      "grad_norm": 8.765314102172852,
+      "learning_rate": 1.765636112621847e-05,
+      "loss": 1.4409,
       "step": 2802
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 27.743297576904297,
-      "learning_rate": 1.438308108649895e-05,
-      "loss": 2.8904,
+      "epoch": 0.35,
+      "grad_norm": 32.125858306884766,
+      "learning_rate": 1.765552441116178e-05,
+      "loss": 2.5083,
       "step": 2803
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 28.265438079833984,
-      "learning_rate": 1.4381076475894558e-05,
-      "loss": 2.0912,
+      "epoch": 0.35,
+      "grad_norm": 13.41010570526123,
+      "learning_rate": 1.7654687696105093e-05,
+      "loss": 1.4009,
       "step": 2804
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 42.51979064941406,
-      "learning_rate": 1.4379071865290168e-05,
-      "loss": 2.9312,
+      "epoch": 0.35,
+      "grad_norm": 10.478439331054688,
+      "learning_rate": 1.7653850981048407e-05,
+      "loss": 0.8087,
       "step": 2805
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 13.225244522094727,
-      "learning_rate": 1.437706725468578e-05,
-      "loss": 2.3977,
+      "epoch": 0.35,
+      "grad_norm": 13.714133262634277,
+      "learning_rate": 1.7653014265991717e-05,
+      "loss": 1.5322,
       "step": 2806
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 21.915138244628906,
-      "learning_rate": 1.4375062644081388e-05,
-      "loss": 2.3075,
+      "epoch": 0.35,
+      "grad_norm": 12.272222518920898,
+      "learning_rate": 1.765217755093503e-05,
+      "loss": 1.9538,
       "step": 2807
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 27.076114654541016,
-      "learning_rate": 1.4373058033476998e-05,
-      "loss": 2.9827,
+      "epoch": 0.35,
+      "grad_norm": 29.58174705505371,
+      "learning_rate": 1.7651340835878344e-05,
+      "loss": 2.5173,
       "step": 2808
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 29.377344131469727,
-      "learning_rate": 1.4371053422872607e-05,
-      "loss": 1.7643,
+      "epoch": 0.35,
+      "grad_norm": 16.548465728759766,
+      "learning_rate": 1.7650504120821658e-05,
+      "loss": 2.0753,
       "step": 2809
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 16.541988372802734,
-      "learning_rate": 1.4369048812268219e-05,
-      "loss": 2.0169,
+      "epoch": 0.35,
+      "grad_norm": 12.644722938537598,
+      "learning_rate": 1.7649667405764968e-05,
+      "loss": 2.7874,
       "step": 2810
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 11.847968101501465,
-      "learning_rate": 1.4367044201663829e-05,
-      "loss": 1.9688,
+      "epoch": 0.35,
+      "grad_norm": 7.718317985534668,
+      "learning_rate": 1.7648830690708282e-05,
+      "loss": 1.015,
       "step": 2811
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 19.333833694458008,
-      "learning_rate": 1.4365039591059437e-05,
-      "loss": 1.6196,
+      "epoch": 0.35,
+      "grad_norm": 9.271620750427246,
+      "learning_rate": 1.7647993975651595e-05,
+      "loss": 1.8111,
       "step": 2812
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 11.35824966430664,
-      "learning_rate": 1.4363034980455049e-05,
-      "loss": 2.2096,
+      "epoch": 0.35,
+      "grad_norm": 13.277092933654785,
+      "learning_rate": 1.7647157260594906e-05,
+      "loss": 1.4301,
       "step": 2813
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 21.909666061401367,
-      "learning_rate": 1.4361030369850657e-05,
-      "loss": 2.0165,
+      "epoch": 0.35,
+      "grad_norm": 10.31788158416748,
+      "learning_rate": 1.764632054553822e-05,
+      "loss": 1.5806,
       "step": 2814
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 12.537689208984375,
-      "learning_rate": 1.4359025759246267e-05,
-      "loss": 1.4654,
+      "epoch": 0.35,
+      "grad_norm": 13.863436698913574,
+      "learning_rate": 1.764548383048153e-05,
+      "loss": 2.6027,
       "step": 2815
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 16.89116096496582,
-      "learning_rate": 1.4357021148641876e-05,
-      "loss": 2.2603,
+      "epoch": 0.35,
+      "grad_norm": 13.447190284729004,
+      "learning_rate": 1.7644647115424843e-05,
+      "loss": 3.0108,
       "step": 2816
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 29.52833366394043,
-      "learning_rate": 1.4355016538037487e-05,
-      "loss": 2.1104,
+      "epoch": 0.35,
+      "grad_norm": 11.306679725646973,
+      "learning_rate": 1.7643810400368157e-05,
+      "loss": 1.6569,
       "step": 2817
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 31.932817459106445,
-      "learning_rate": 1.4353011927433097e-05,
-      "loss": 3.2788,
+      "epoch": 0.35,
+      "grad_norm": 18.264440536499023,
+      "learning_rate": 1.7642973685311467e-05,
+      "loss": 4.5724,
       "step": 2818
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 13.009839057922363,
-      "learning_rate": 1.4351007316828708e-05,
-      "loss": 1.8396,
+      "epoch": 0.35,
+      "grad_norm": 12.04774284362793,
+      "learning_rate": 1.764213697025478e-05,
+      "loss": 2.782,
       "step": 2819
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 20.032115936279297,
-      "learning_rate": 1.4349002706224318e-05,
-      "loss": 2.5377,
+      "epoch": 0.35,
+      "grad_norm": 15.501517295837402,
+      "learning_rate": 1.764130025519809e-05,
+      "loss": 1.5684,
       "step": 2820
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 30.521154403686523,
-      "learning_rate": 1.4346998095619926e-05,
-      "loss": 2.9662,
+      "epoch": 0.35,
+      "grad_norm": 9.706591606140137,
+      "learning_rate": 1.7640463540141405e-05,
+      "loss": 1.604,
       "step": 2821
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 25.22045135498047,
-      "learning_rate": 1.4344993485015538e-05,
-      "loss": 2.4709,
+      "epoch": 0.35,
+      "grad_norm": 14.179362297058105,
+      "learning_rate": 1.763962682508472e-05,
+      "loss": 1.4216,
       "step": 2822
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 36.31182861328125,
-      "learning_rate": 1.4342988874411148e-05,
-      "loss": 2.1305,
+      "epoch": 0.35,
+      "grad_norm": 7.879173755645752,
+      "learning_rate": 1.7638790110028032e-05,
+      "loss": 1.9733,
       "step": 2823
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 10.13243293762207,
-      "learning_rate": 1.4340984263806756e-05,
-      "loss": 1.2709,
+      "epoch": 0.35,
+      "grad_norm": 17.222301483154297,
+      "learning_rate": 1.7637953394971342e-05,
+      "loss": 3.1487,
       "step": 2824
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 13.83365535736084,
-      "learning_rate": 1.4338979653202368e-05,
-      "loss": 2.4132,
+      "epoch": 0.35,
+      "grad_norm": 12.826745986938477,
+      "learning_rate": 1.7637116679914656e-05,
+      "loss": 2.612,
       "step": 2825
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 13.130005836486816,
-      "learning_rate": 1.4336975042597976e-05,
-      "loss": 1.4074,
+      "epoch": 0.35,
+      "grad_norm": 13.919160842895508,
+      "learning_rate": 1.763627996485797e-05,
+      "loss": 2.6432,
       "step": 2826
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 20.482786178588867,
-      "learning_rate": 1.4334970431993587e-05,
-      "loss": 2.3476,
+      "epoch": 0.35,
+      "grad_norm": 11.945897102355957,
+      "learning_rate": 1.763544324980128e-05,
+      "loss": 1.0819,
       "step": 2827
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 9.617236137390137,
-      "learning_rate": 1.4332965821389195e-05,
-      "loss": 1.595,
+      "epoch": 0.35,
+      "grad_norm": 18.103195190429688,
+      "learning_rate": 1.7634606534744594e-05,
+      "loss": 3.3143,
       "step": 2828
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 11.414641380310059,
-      "learning_rate": 1.4330961210784807e-05,
-      "loss": 1.5817,
+      "epoch": 0.36,
+      "grad_norm": 7.475184917449951,
+      "learning_rate": 1.7633769819687907e-05,
+      "loss": 1.0552,
       "step": 2829
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 16.799848556518555,
-      "learning_rate": 1.4328956600180417e-05,
-      "loss": 2.1956,
+      "epoch": 0.36,
+      "grad_norm": 14.897072792053223,
+      "learning_rate": 1.763293310463122e-05,
+      "loss": 2.4431,
       "step": 2830
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 10.216323852539062,
-      "learning_rate": 1.4326951989576025e-05,
-      "loss": 1.4594,
+      "epoch": 0.36,
+      "grad_norm": 13.615031242370605,
+      "learning_rate": 1.763209638957453e-05,
+      "loss": 2.111,
       "step": 2831
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 28.910261154174805,
-      "learning_rate": 1.4324947378971637e-05,
-      "loss": 2.556,
+      "epoch": 0.36,
+      "grad_norm": 11.770406723022461,
+      "learning_rate": 1.7631259674517845e-05,
+      "loss": 1.1007,
       "step": 2832
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 26.716176986694336,
-      "learning_rate": 1.4322942768367245e-05,
-      "loss": 2.3864,
+      "epoch": 0.36,
+      "grad_norm": 27.718374252319336,
+      "learning_rate": 1.763042295946116e-05,
+      "loss": 2.6574,
       "step": 2833
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 16.635345458984375,
-      "learning_rate": 1.4320938157762855e-05,
-      "loss": 2.3808,
+      "epoch": 0.36,
+      "grad_norm": 37.83617401123047,
+      "learning_rate": 1.762958624440447e-05,
+      "loss": 2.9518,
       "step": 2834
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 21.195531845092773,
-      "learning_rate": 1.4318933547158467e-05,
-      "loss": 1.7784,
+      "epoch": 0.36,
+      "grad_norm": 6.720913887023926,
+      "learning_rate": 1.7628749529347782e-05,
+      "loss": 1.9323,
       "step": 2835
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 16.338502883911133,
-      "learning_rate": 1.4316928936554076e-05,
-      "loss": 2.1817,
+      "epoch": 0.36,
+      "grad_norm": 13.271183013916016,
+      "learning_rate": 1.7627912814291096e-05,
+      "loss": 3.0903,
       "step": 2836
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 10.698309898376465,
-      "learning_rate": 1.4314924325949686e-05,
-      "loss": 1.8357,
+      "epoch": 0.36,
+      "grad_norm": 14.988275527954102,
+      "learning_rate": 1.762707609923441e-05,
+      "loss": 1.5745,
       "step": 2837
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 22.48438262939453,
-      "learning_rate": 1.4312919715345294e-05,
-      "loss": 1.9495,
+      "epoch": 0.36,
+      "grad_norm": 13.079087257385254,
+      "learning_rate": 1.762623938417772e-05,
+      "loss": 2.4665,
       "step": 2838
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 72.1753921508789,
-      "learning_rate": 1.4310915104740906e-05,
-      "loss": 1.8956,
+      "epoch": 0.36,
+      "grad_norm": 15.878899574279785,
+      "learning_rate": 1.7625402669121033e-05,
+      "loss": 1.0119,
       "step": 2839
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 14.804839134216309,
-      "learning_rate": 1.4308910494136514e-05,
-      "loss": 2.0498,
+      "epoch": 0.36,
+      "grad_norm": 12.98774528503418,
+      "learning_rate": 1.7624565954064347e-05,
+      "loss": 1.7453,
       "step": 2840
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 46.27671432495117,
-      "learning_rate": 1.4306905883532124e-05,
-      "loss": 2.3295,
+      "epoch": 0.36,
+      "grad_norm": 37.145545959472656,
+      "learning_rate": 1.7623729239007657e-05,
+      "loss": 1.7648,
       "step": 2841
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 20.91038703918457,
-      "learning_rate": 1.4304901272927736e-05,
-      "loss": 1.7985,
+      "epoch": 0.36,
+      "grad_norm": 17.651174545288086,
+      "learning_rate": 1.762289252395097e-05,
+      "loss": 2.2728,
       "step": 2842
     },
     {
-      "epoch": 0.85,
-      "grad_norm": 15.907576560974121,
-      "learning_rate": 1.4302896662323345e-05,
-      "loss": 1.945,
+      "epoch": 0.36,
+      "grad_norm": 91.55896759033203,
+      "learning_rate": 1.762205580889428e-05,
+      "loss": 1.4332,
       "step": 2843
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 11.249744415283203,
-      "learning_rate": 1.4300892051718955e-05,
-      "loss": 2.0269,
+      "epoch": 0.36,
+      "grad_norm": 19.58599281311035,
+      "learning_rate": 1.7621219093837595e-05,
+      "loss": 2.1461,
       "step": 2844
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 31.293607711791992,
-      "learning_rate": 1.4298887441114565e-05,
-      "loss": 2.0211,
+      "epoch": 0.36,
+      "grad_norm": 40.6429557800293,
+      "learning_rate": 1.762038237878091e-05,
+      "loss": 3.9308,
       "step": 2845
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 15.472600936889648,
-      "learning_rate": 1.4296882830510175e-05,
-      "loss": 2.2907,
+      "epoch": 0.36,
+      "grad_norm": 27.435991287231445,
+      "learning_rate": 1.761954566372422e-05,
+      "loss": 3.3426,
       "step": 2846
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 15.497085571289062,
-      "learning_rate": 1.4294878219905783e-05,
-      "loss": 2.4152,
+      "epoch": 0.36,
+      "grad_norm": 23.73316192626953,
+      "learning_rate": 1.7618708948667533e-05,
+      "loss": 3.6674,
       "step": 2847
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 16.131122589111328,
-      "learning_rate": 1.4292873609301395e-05,
-      "loss": 1.7781,
+      "epoch": 0.36,
+      "grad_norm": 18.161706924438477,
+      "learning_rate": 1.7617872233610843e-05,
+      "loss": 2.6338,
       "step": 2848
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 20.631099700927734,
-      "learning_rate": 1.4290868998697005e-05,
-      "loss": 2.6782,
+      "epoch": 0.36,
+      "grad_norm": 4.604570388793945,
+      "learning_rate": 1.7617035518554156e-05,
+      "loss": 1.7168,
       "step": 2849
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 28.31133460998535,
-      "learning_rate": 1.4288864388092613e-05,
-      "loss": 2.7064,
+      "epoch": 0.36,
+      "grad_norm": 10.27620792388916,
+      "learning_rate": 1.761619880349747e-05,
+      "loss": 1.9358,
       "step": 2850
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 22.287992477416992,
-      "learning_rate": 1.4286859777488225e-05,
-      "loss": 1.5085,
+      "epoch": 0.36,
+      "grad_norm": 9.236021041870117,
+      "learning_rate": 1.7615362088440784e-05,
+      "loss": 3.0787,
       "step": 2851
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 14.099394798278809,
-      "learning_rate": 1.4284855166883834e-05,
-      "loss": 2.1791,
+      "epoch": 0.36,
+      "grad_norm": 12.75297737121582,
+      "learning_rate": 1.7614525373384094e-05,
+      "loss": 2.4493,
       "step": 2852
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 23.74030876159668,
-      "learning_rate": 1.4282850556279444e-05,
-      "loss": 1.5725,
+      "epoch": 0.36,
+      "grad_norm": 8.181326866149902,
+      "learning_rate": 1.7613688658327408e-05,
+      "loss": 1.5055,
       "step": 2853
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 18.97103500366211,
-      "learning_rate": 1.4280845945675055e-05,
-      "loss": 2.8069,
+      "epoch": 0.36,
+      "grad_norm": 20.349233627319336,
+      "learning_rate": 1.761285194327072e-05,
+      "loss": 2.5806,
       "step": 2854
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 15.047412872314453,
-      "learning_rate": 1.4278841335070664e-05,
-      "loss": 2.3392,
+      "epoch": 0.36,
+      "grad_norm": 10.25844955444336,
+      "learning_rate": 1.761201522821403e-05,
+      "loss": 0.9415,
       "step": 2855
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 11.085140228271484,
-      "learning_rate": 1.4276836724466274e-05,
-      "loss": 1.8397,
+      "epoch": 0.36,
+      "grad_norm": 14.34598445892334,
+      "learning_rate": 1.7611178513157345e-05,
+      "loss": 1.1714,
       "step": 2856
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 25.895097732543945,
-      "learning_rate": 1.4274832113861882e-05,
-      "loss": 2.0397,
+      "epoch": 0.36,
+      "grad_norm": 26.95628547668457,
+      "learning_rate": 1.761034179810066e-05,
+      "loss": 3.4365,
       "step": 2857
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 15.016706466674805,
-      "learning_rate": 1.4272827503257494e-05,
-      "loss": 2.7215,
+      "epoch": 0.36,
+      "grad_norm": 26.015779495239258,
+      "learning_rate": 1.7609505083043972e-05,
+      "loss": 1.4385,
       "step": 2858
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 30.092390060424805,
-      "learning_rate": 1.4270822892653102e-05,
-      "loss": 2.8657,
+      "epoch": 0.36,
+      "grad_norm": 15.12834358215332,
+      "learning_rate": 1.7608668367987283e-05,
+      "loss": 3.0311,
       "step": 2859
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 16.425628662109375,
-      "learning_rate": 1.4268818282048713e-05,
-      "loss": 1.7927,
+      "epoch": 0.36,
+      "grad_norm": 40.686676025390625,
+      "learning_rate": 1.7607831652930596e-05,
+      "loss": 1.7407,
       "step": 2860
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 39.504478454589844,
-      "learning_rate": 1.4266813671444324e-05,
-      "loss": 2.4321,
+      "epoch": 0.36,
+      "grad_norm": 8.17596435546875,
+      "learning_rate": 1.760699493787391e-05,
+      "loss": 1.5953,
       "step": 2861
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 15.988898277282715,
-      "learning_rate": 1.4264809060839933e-05,
-      "loss": 3.2789,
+      "epoch": 0.36,
+      "grad_norm": 12.739534378051758,
+      "learning_rate": 1.760615822281722e-05,
+      "loss": 2.7701,
       "step": 2862
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 25.57854461669922,
-      "learning_rate": 1.4262804450235543e-05,
-      "loss": 2.2735,
+      "epoch": 0.36,
+      "grad_norm": 10.815932273864746,
+      "learning_rate": 1.7605321507760534e-05,
+      "loss": 1.4069,
       "step": 2863
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 24.288047790527344,
-      "learning_rate": 1.4260799839631151e-05,
-      "loss": 2.4726,
+      "epoch": 0.36,
+      "grad_norm": 14.07378101348877,
+      "learning_rate": 1.7604484792703848e-05,
+      "loss": 1.7353,
       "step": 2864
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 9.727755546569824,
-      "learning_rate": 1.4258795229026763e-05,
-      "loss": 1.2214,
+      "epoch": 0.36,
+      "grad_norm": 10.749878883361816,
+      "learning_rate": 1.760364807764716e-05,
+      "loss": 3.3234,
       "step": 2865
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 11.615317344665527,
-      "learning_rate": 1.4256790618422373e-05,
-      "loss": 1.5758,
+      "epoch": 0.36,
+      "grad_norm": 15.514153480529785,
+      "learning_rate": 1.760281136259047e-05,
+      "loss": 2.0563,
       "step": 2866
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 18.63646697998047,
-      "learning_rate": 1.4254786007817981e-05,
-      "loss": 3.0492,
+      "epoch": 0.36,
+      "grad_norm": 9.187424659729004,
+      "learning_rate": 1.7601974647533785e-05,
+      "loss": 1.5415,
       "step": 2867
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 18.21438217163086,
-      "learning_rate": 1.4252781397213593e-05,
-      "loss": 2.2775,
+      "epoch": 0.36,
+      "grad_norm": 9.863618850708008,
+      "learning_rate": 1.7601137932477095e-05,
+      "loss": 1.2027,
       "step": 2868
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 11.930097579956055,
-      "learning_rate": 1.4250776786609202e-05,
-      "loss": 1.8362,
+      "epoch": 0.36,
+      "grad_norm": 13.988532066345215,
+      "learning_rate": 1.760030121742041e-05,
+      "loss": 1.6365,
       "step": 2869
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 18.075613021850586,
-      "learning_rate": 1.4248772176004812e-05,
-      "loss": 3.2771,
+      "epoch": 0.36,
+      "grad_norm": 10.842628479003906,
+      "learning_rate": 1.7599464502363723e-05,
+      "loss": 1.1997,
       "step": 2870
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 17.12186050415039,
-      "learning_rate": 1.4246767565400422e-05,
-      "loss": 1.9388,
+      "epoch": 0.36,
+      "grad_norm": 12.993040084838867,
+      "learning_rate": 1.7598627787307033e-05,
+      "loss": 3.3648,
       "step": 2871
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 20.949424743652344,
-      "learning_rate": 1.4244762954796032e-05,
-      "loss": 2.5683,
+      "epoch": 0.36,
+      "grad_norm": 15.58774185180664,
+      "learning_rate": 1.7597791072250347e-05,
+      "loss": 2.8855,
       "step": 2872
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 36.75067901611328,
-      "learning_rate": 1.4242758344191644e-05,
-      "loss": 1.8602,
+      "epoch": 0.36,
+      "grad_norm": 7.670743942260742,
+      "learning_rate": 1.7596954357193657e-05,
+      "loss": 1.8101,
       "step": 2873
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 12.314196586608887,
-      "learning_rate": 1.4240753733587252e-05,
-      "loss": 1.7533,
+      "epoch": 0.36,
+      "grad_norm": 18.823976516723633,
+      "learning_rate": 1.759611764213697e-05,
+      "loss": 1.8527,
       "step": 2874
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 9.40181827545166,
-      "learning_rate": 1.4238749122982862e-05,
-      "loss": 1.0538,
+      "epoch": 0.36,
+      "grad_norm": 3.7253715991973877,
+      "learning_rate": 1.7595280927080284e-05,
+      "loss": 0.1347,
       "step": 2875
     },
     {
-      "epoch": 0.86,
-      "grad_norm": 18.72924041748047,
-      "learning_rate": 1.423674451237847e-05,
-      "loss": 1.741,
+      "epoch": 0.36,
+      "grad_norm": 10.255762100219727,
+      "learning_rate": 1.7594444212023594e-05,
+      "loss": 1.8739,
       "step": 2876
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 86.12344360351562,
-      "learning_rate": 1.4234739901774082e-05,
-      "loss": 1.8543,
+      "epoch": 0.36,
+      "grad_norm": 6.958003520965576,
+      "learning_rate": 1.7593607496966908e-05,
+      "loss": 1.1171,
       "step": 2877
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 11.64871883392334,
-      "learning_rate": 1.4232735291169692e-05,
-      "loss": 1.8377,
+      "epoch": 0.36,
+      "grad_norm": 3.8992414474487305,
+      "learning_rate": 1.7592770781910222e-05,
+      "loss": 0.1972,
       "step": 2878
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 20.858604431152344,
-      "learning_rate": 1.42307306805653e-05,
-      "loss": 2.407,
+      "epoch": 0.36,
+      "grad_norm": 8.176952362060547,
+      "learning_rate": 1.7591934066853535e-05,
+      "loss": 1.4856,
       "step": 2879
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 16.45133399963379,
-      "learning_rate": 1.4228726069960913e-05,
-      "loss": 2.1642,
-      "step": 2880
-    },
-    {
-      "epoch": 0.87,
-      "eval_loss": 0.27140170335769653,
-      "eval_runtime": 43.4249,
-      "eval_samples_per_second": 34.059,
-      "eval_steps_per_second": 34.059,
+      "epoch": 0.36,
+      "grad_norm": 19.96003532409668,
+      "learning_rate": 1.7591097351796846e-05,
+      "loss": 2.7468,
       "step": 2880
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 16.983951568603516,
-      "learning_rate": 1.4226721459356521e-05,
-      "loss": 2.1911,
+      "epoch": 0.36,
+      "grad_norm": 15.29173755645752,
+      "learning_rate": 1.759026063674016e-05,
+      "loss": 2.4146,
       "step": 2881
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 12.091941833496094,
-      "learning_rate": 1.4224716848752131e-05,
-      "loss": 1.6567,
+      "epoch": 0.36,
+      "grad_norm": 7.0636305809021,
+      "learning_rate": 1.7589423921683473e-05,
+      "loss": 0.9918,
       "step": 2882
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 22.655092239379883,
-      "learning_rate": 1.422271223814774e-05,
-      "loss": 1.7178,
+      "epoch": 0.36,
+      "grad_norm": 46.704769134521484,
+      "learning_rate": 1.7588587206626783e-05,
+      "loss": 2.6682,
       "step": 2883
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 13.33409595489502,
-      "learning_rate": 1.4220707627543351e-05,
-      "loss": 1.42,
+      "epoch": 0.36,
+      "grad_norm": 12.229507446289062,
+      "learning_rate": 1.7587750491570097e-05,
+      "loss": 3.7473,
       "step": 2884
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 41.79966735839844,
-      "learning_rate": 1.4218703016938961e-05,
-      "loss": 2.5635,
+      "epoch": 0.36,
+      "grad_norm": 29.765493392944336,
+      "learning_rate": 1.758691377651341e-05,
+      "loss": 2.2255,
       "step": 2885
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 9.73848819732666,
-      "learning_rate": 1.421669840633457e-05,
-      "loss": 1.5657,
+      "epoch": 0.36,
+      "grad_norm": 14.75451374053955,
+      "learning_rate": 1.7586077061456724e-05,
+      "loss": 2.9142,
       "step": 2886
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 13.572857856750488,
-      "learning_rate": 1.4214693795730181e-05,
-      "loss": 1.8109,
+      "epoch": 0.36,
+      "grad_norm": 10.729453086853027,
+      "learning_rate": 1.7585240346400034e-05,
+      "loss": 2.4328,
       "step": 2887
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 19.471956253051758,
-      "learning_rate": 1.421268918512579e-05,
-      "loss": 1.7335,
+      "epoch": 0.36,
+      "grad_norm": 8.80831527709961,
+      "learning_rate": 1.7584403631343348e-05,
+      "loss": 2.3335,
       "step": 2888
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 16.479246139526367,
-      "learning_rate": 1.42106845745214e-05,
-      "loss": 2.1827,
+      "epoch": 0.36,
+      "grad_norm": 15.32617473602295,
+      "learning_rate": 1.7583566916286662e-05,
+      "loss": 2.5147,
       "step": 2889
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 42.60647964477539,
-      "learning_rate": 1.420867996391701e-05,
-      "loss": 4.2229,
+      "epoch": 0.36,
+      "grad_norm": 16.763200759887695,
+      "learning_rate": 1.7582730201229972e-05,
+      "loss": 2.105,
       "step": 2890
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 29.324525833129883,
-      "learning_rate": 1.420667535331262e-05,
-      "loss": 2.2616,
+      "epoch": 0.36,
+      "grad_norm": 9.007331848144531,
+      "learning_rate": 1.7581893486173286e-05,
+      "loss": 1.1816,
       "step": 2891
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 17.365251541137695,
-      "learning_rate": 1.420467074270823e-05,
-      "loss": 1.9783,
+      "epoch": 0.36,
+      "grad_norm": 11.328272819519043,
+      "learning_rate": 1.75810567711166e-05,
+      "loss": 2.079,
       "step": 2892
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 14.282939910888672,
-      "learning_rate": 1.420266613210384e-05,
-      "loss": 1.9463,
+      "epoch": 0.36,
+      "grad_norm": 8.322172164916992,
+      "learning_rate": 1.758022005605991e-05,
+      "loss": 1.808,
       "step": 2893
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 18.807802200317383,
-      "learning_rate": 1.420066152149945e-05,
-      "loss": 1.9243,
+      "epoch": 0.36,
+      "grad_norm": 22.682048797607422,
+      "learning_rate": 1.7579383341003223e-05,
+      "loss": 2.3697,
       "step": 2894
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 12.65291976928711,
-      "learning_rate": 1.4198656910895059e-05,
-      "loss": 1.5824,
+      "epoch": 0.36,
+      "grad_norm": 12.664169311523438,
+      "learning_rate": 1.7578546625946537e-05,
+      "loss": 2.7809,
       "step": 2895
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 24.225215911865234,
-      "learning_rate": 1.419665230029067e-05,
-      "loss": 1.4894,
+      "epoch": 0.36,
+      "grad_norm": 23.70795440673828,
+      "learning_rate": 1.7577709910889847e-05,
+      "loss": 2.8973,
       "step": 2896
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 19.319774627685547,
-      "learning_rate": 1.419464768968628e-05,
-      "loss": 2.8796,
+      "epoch": 0.36,
+      "grad_norm": 12.631494522094727,
+      "learning_rate": 1.757687319583316e-05,
+      "loss": 1.4789,
       "step": 2897
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 14.177779197692871,
-      "learning_rate": 1.4192643079081889e-05,
-      "loss": 1.7554,
+      "epoch": 0.36,
+      "grad_norm": 17.631160736083984,
+      "learning_rate": 1.7576036480776474e-05,
+      "loss": 2.8649,
       "step": 2898
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 17.083988189697266,
-      "learning_rate": 1.41906384684775e-05,
-      "loss": 2.0681,
+      "epoch": 0.36,
+      "grad_norm": 11.38050365447998,
+      "learning_rate": 1.7575199765719785e-05,
+      "loss": 2.7911,
       "step": 2899
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 40.30973434448242,
-      "learning_rate": 1.418863385787311e-05,
-      "loss": 2.4183,
+      "epoch": 0.36,
+      "grad_norm": 10.819679260253906,
+      "learning_rate": 1.7574363050663098e-05,
+      "loss": 3.1221,
       "step": 2900
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 9.512898445129395,
-      "learning_rate": 1.418662924726872e-05,
-      "loss": 1.5707,
+      "epoch": 0.36,
+      "grad_norm": 9.505936622619629,
+      "learning_rate": 1.757352633560641e-05,
+      "loss": 1.217,
       "step": 2901
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 52.70304870605469,
-      "learning_rate": 1.4184624636664328e-05,
-      "loss": 2.9803,
+      "epoch": 0.36,
+      "grad_norm": 11.83896255493164,
+      "learning_rate": 1.7572689620549722e-05,
+      "loss": 0.502,
       "step": 2902
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 16.734039306640625,
-      "learning_rate": 1.418262002605994e-05,
-      "loss": 2.037,
+      "epoch": 0.36,
+      "grad_norm": 7.863925933837891,
+      "learning_rate": 1.7571852905493036e-05,
+      "loss": 1.6534,
       "step": 2903
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 19.740087509155273,
-      "learning_rate": 1.418061541545555e-05,
-      "loss": 1.8485,
+      "epoch": 0.36,
+      "grad_norm": 10.372111320495605,
+      "learning_rate": 1.7571016190436346e-05,
+      "loss": 1.9913,
       "step": 2904
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 14.386117935180664,
-      "learning_rate": 1.4178610804851158e-05,
-      "loss": 1.0815,
+      "epoch": 0.36,
+      "grad_norm": 27.196186065673828,
+      "learning_rate": 1.757017947537966e-05,
+      "loss": 3.9937,
       "step": 2905
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 16.817588806152344,
-      "learning_rate": 1.417660619424677e-05,
-      "loss": 1.6195,
+      "epoch": 0.36,
+      "grad_norm": 16.792673110961914,
+      "learning_rate": 1.7569342760322973e-05,
+      "loss": 2.6868,
       "step": 2906
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 27.766712188720703,
-      "learning_rate": 1.4174601583642378e-05,
-      "loss": 1.9603,
+      "epoch": 0.36,
+      "grad_norm": 16.74106216430664,
+      "learning_rate": 1.7568506045266287e-05,
+      "loss": 2.2425,
       "step": 2907
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 33.24577713012695,
-      "learning_rate": 1.4172596973037988e-05,
-      "loss": 1.6909,
+      "epoch": 0.36,
+      "grad_norm": 10.993690490722656,
+      "learning_rate": 1.7567669330209597e-05,
+      "loss": 1.5335,
       "step": 2908
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 11.028932571411133,
-      "learning_rate": 1.41705923624336e-05,
-      "loss": 1.3208,
+      "epoch": 0.37,
+      "grad_norm": 10.971261978149414,
+      "learning_rate": 1.756683261515291e-05,
+      "loss": 3.1004,
       "step": 2909
     },
     {
-      "epoch": 0.87,
-      "grad_norm": 12.61074447631836,
-      "learning_rate": 1.4168587751829208e-05,
-      "loss": 1.9588,
+      "epoch": 0.37,
+      "grad_norm": 20.94425392150879,
+      "learning_rate": 1.7565995900096225e-05,
+      "loss": 3.9462,
       "step": 2910
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 11.678028106689453,
-      "learning_rate": 1.4166583141224818e-05,
-      "loss": 0.8482,
+      "epoch": 0.37,
+      "grad_norm": 17.663053512573242,
+      "learning_rate": 1.7565159185039535e-05,
+      "loss": 2.0604,
       "step": 2911
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 12.001538276672363,
-      "learning_rate": 1.4164578530620427e-05,
-      "loss": 1.9327,
+      "epoch": 0.37,
+      "grad_norm": 35.784873962402344,
+      "learning_rate": 1.756432246998285e-05,
+      "loss": 3.1819,
       "step": 2912
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 16.76682472229004,
-      "learning_rate": 1.4162573920016039e-05,
-      "loss": 2.0711,
+      "epoch": 0.37,
+      "grad_norm": 18.560909271240234,
+      "learning_rate": 1.7563485754926162e-05,
+      "loss": 1.7441,
       "step": 2913
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 27.799612045288086,
-      "learning_rate": 1.4160569309411647e-05,
-      "loss": 2.1891,
+      "epoch": 0.37,
+      "grad_norm": 16.16483497619629,
+      "learning_rate": 1.7562649039869476e-05,
+      "loss": 2.3441,
       "step": 2914
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 24.618627548217773,
-      "learning_rate": 1.4158564698807257e-05,
-      "loss": 1.8592,
+      "epoch": 0.37,
+      "grad_norm": 19.676025390625,
+      "learning_rate": 1.7561812324812786e-05,
+      "loss": 3.5454,
       "step": 2915
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 24.141916275024414,
-      "learning_rate": 1.4156560088202869e-05,
-      "loss": 2.2455,
+      "epoch": 0.37,
+      "grad_norm": 14.388813018798828,
+      "learning_rate": 1.75609756097561e-05,
+      "loss": 1.9982,
       "step": 2916
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 10.506229400634766,
-      "learning_rate": 1.4154555477598477e-05,
-      "loss": 1.8823,
+      "epoch": 0.37,
+      "grad_norm": 14.46896743774414,
+      "learning_rate": 1.7560138894699413e-05,
+      "loss": 2.5543,
       "step": 2917
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 14.979269981384277,
-      "learning_rate": 1.4152550866994087e-05,
-      "loss": 2.0007,
+      "epoch": 0.37,
+      "grad_norm": 13.310667991638184,
+      "learning_rate": 1.7559302179642724e-05,
+      "loss": 3.2821,
       "step": 2918
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 34.168128967285156,
-      "learning_rate": 1.4150546256389697e-05,
-      "loss": 2.2599,
+      "epoch": 0.37,
+      "grad_norm": 14.989908218383789,
+      "learning_rate": 1.7558465464586037e-05,
+      "loss": 1.7186,
       "step": 2919
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 17.805347442626953,
-      "learning_rate": 1.4148541645785307e-05,
-      "loss": 2.8172,
+      "epoch": 0.37,
+      "grad_norm": 7.282630920410156,
+      "learning_rate": 1.755762874952935e-05,
+      "loss": 1.1641,
       "step": 2920
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 18.875598907470703,
-      "learning_rate": 1.4146537035180918e-05,
-      "loss": 1.7194,
+      "epoch": 0.37,
+      "grad_norm": 25.337600708007812,
+      "learning_rate": 1.755679203447266e-05,
+      "loss": 1.0596,
       "step": 2921
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 44.858428955078125,
-      "learning_rate": 1.4144532424576528e-05,
-      "loss": 1.6975,
+      "epoch": 0.37,
+      "grad_norm": 12.670037269592285,
+      "learning_rate": 1.7555955319415975e-05,
+      "loss": 2.7756,
       "step": 2922
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 10.500758171081543,
-      "learning_rate": 1.4142527813972138e-05,
-      "loss": 1.8214,
+      "epoch": 0.37,
+      "grad_norm": 12.989263534545898,
+      "learning_rate": 1.755511860435929e-05,
+      "loss": 1.7594,
       "step": 2923
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 19.17152214050293,
-      "learning_rate": 1.4140523203367746e-05,
-      "loss": 1.9509,
+      "epoch": 0.37,
+      "grad_norm": 9.042985916137695,
+      "learning_rate": 1.75542818893026e-05,
+      "loss": 1.5087,
       "step": 2924
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 19.903955459594727,
-      "learning_rate": 1.4138518592763358e-05,
-      "loss": 2.9421,
+      "epoch": 0.37,
+      "grad_norm": 14.791471481323242,
+      "learning_rate": 1.7553445174245912e-05,
+      "loss": 2.6902,
       "step": 2925
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 17.388938903808594,
-      "learning_rate": 1.4136513982158966e-05,
-      "loss": 1.8245,
+      "epoch": 0.37,
+      "grad_norm": 10.423829078674316,
+      "learning_rate": 1.7552608459189223e-05,
+      "loss": 1.6936,
       "step": 2926
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 15.076804161071777,
-      "learning_rate": 1.4134509371554576e-05,
-      "loss": 1.5359,
+      "epoch": 0.37,
+      "grad_norm": 11.191834449768066,
+      "learning_rate": 1.7551771744132536e-05,
+      "loss": 1.3695,
       "step": 2927
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 11.603935241699219,
-      "learning_rate": 1.4132504760950188e-05,
-      "loss": 1.5563,
+      "epoch": 0.37,
+      "grad_norm": 11.632593154907227,
+      "learning_rate": 1.755093502907585e-05,
+      "loss": 0.9178,
       "step": 2928
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 25.9620418548584,
-      "learning_rate": 1.4130500150345797e-05,
-      "loss": 3.2783,
+      "epoch": 0.37,
+      "grad_norm": 63.4360237121582,
+      "learning_rate": 1.755009831401916e-05,
+      "loss": 1.2587,
       "step": 2929
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 16.795143127441406,
-      "learning_rate": 1.4128495539741407e-05,
-      "loss": 1.5015,
+      "epoch": 0.37,
+      "grad_norm": 10.623050689697266,
+      "learning_rate": 1.7549261598962474e-05,
+      "loss": 4.4241,
       "step": 2930
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 11.010591506958008,
-      "learning_rate": 1.4126490929137015e-05,
-      "loss": 1.8603,
+      "epoch": 0.37,
+      "grad_norm": 28.13072395324707,
+      "learning_rate": 1.7548424883905788e-05,
+      "loss": 1.4253,
       "step": 2931
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 12.306628227233887,
-      "learning_rate": 1.4124486318532627e-05,
-      "loss": 1.4446,
+      "epoch": 0.37,
+      "grad_norm": 15.974139213562012,
+      "learning_rate": 1.7547588168849098e-05,
+      "loss": 3.0009,
       "step": 2932
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 14.734245300292969,
-      "learning_rate": 1.4122481707928235e-05,
-      "loss": 2.5827,
+      "epoch": 0.37,
+      "grad_norm": 10.374823570251465,
+      "learning_rate": 1.754675145379241e-05,
+      "loss": 2.2726,
       "step": 2933
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 23.01751708984375,
-      "learning_rate": 1.4120477097323845e-05,
-      "loss": 2.9274,
+      "epoch": 0.37,
+      "grad_norm": 10.936897277832031,
+      "learning_rate": 1.7545914738735725e-05,
+      "loss": 1.3559,
       "step": 2934
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 20.132675170898438,
-      "learning_rate": 1.4118472486719457e-05,
-      "loss": 2.0589,
+      "epoch": 0.37,
+      "grad_norm": 7.858895301818848,
+      "learning_rate": 1.754507802367904e-05,
+      "loss": 1.4125,
       "step": 2935
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 17.275482177734375,
-      "learning_rate": 1.4116467876115065e-05,
-      "loss": 2.1773,
+      "epoch": 0.37,
+      "grad_norm": 11.782905578613281,
+      "learning_rate": 1.754424130862235e-05,
+      "loss": 2.6438,
       "step": 2936
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 40.08796310424805,
-      "learning_rate": 1.4114463265510675e-05,
-      "loss": 2.3652,
+      "epoch": 0.37,
+      "grad_norm": 13.125471115112305,
+      "learning_rate": 1.7543404593565663e-05,
+      "loss": 2.3761,
       "step": 2937
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 22.47896957397461,
-      "learning_rate": 1.4112458654906286e-05,
-      "loss": 2.069,
+      "epoch": 0.37,
+      "grad_norm": 7.501157760620117,
+      "learning_rate": 1.7542567878508976e-05,
+      "loss": 2.2839,
       "step": 2938
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 21.368188858032227,
-      "learning_rate": 1.4110454044301896e-05,
-      "loss": 2.5254,
+      "epoch": 0.37,
+      "grad_norm": 25.19307518005371,
+      "learning_rate": 1.7541731163452287e-05,
+      "loss": 2.704,
       "step": 2939
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 8.838847160339355,
-      "learning_rate": 1.4108449433697506e-05,
-      "loss": 1.4187,
+      "epoch": 0.37,
+      "grad_norm": 13.05322551727295,
+      "learning_rate": 1.75408944483956e-05,
+      "loss": 1.9,
       "step": 2940
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 19.025794982910156,
-      "learning_rate": 1.4106444823093116e-05,
-      "loss": 1.7403,
+      "epoch": 0.37,
+      "grad_norm": 26.257415771484375,
+      "learning_rate": 1.7540057733338914e-05,
+      "loss": 2.7862,
       "step": 2941
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 26.310054779052734,
-      "learning_rate": 1.4104440212488726e-05,
-      "loss": 2.0741,
+      "epoch": 0.37,
+      "grad_norm": 6.446777820587158,
+      "learning_rate": 1.7539221018282228e-05,
+      "loss": 0.5044,
       "step": 2942
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 14.412237167358398,
-      "learning_rate": 1.4102435601884334e-05,
-      "loss": 1.6891,
+      "epoch": 0.37,
+      "grad_norm": 17.562593460083008,
+      "learning_rate": 1.7538384303225538e-05,
+      "loss": 3.0915,
       "step": 2943
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 15.92914867401123,
-      "learning_rate": 1.4100430991279946e-05,
-      "loss": 2.542,
+      "epoch": 0.37,
+      "grad_norm": 14.979905128479004,
+      "learning_rate": 1.753754758816885e-05,
+      "loss": 3.0751,
       "step": 2944
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 26.77935028076172,
-      "learning_rate": 1.4098426380675554e-05,
-      "loss": 3.2923,
+      "epoch": 0.37,
+      "grad_norm": 41.20693588256836,
+      "learning_rate": 1.7536710873112165e-05,
+      "loss": 1.6509,
       "step": 2945
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 61.19276428222656,
-      "learning_rate": 1.4096421770071165e-05,
-      "loss": 2.9397,
+      "epoch": 0.37,
+      "grad_norm": 35.70596694946289,
+      "learning_rate": 1.7535874158055475e-05,
+      "loss": 2.5653,
       "step": 2946
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 15.064374923706055,
-      "learning_rate": 1.4094417159466776e-05,
-      "loss": 1.757,
+      "epoch": 0.37,
+      "grad_norm": 34.14892578125,
+      "learning_rate": 1.753503744299879e-05,
+      "loss": 2.738,
       "step": 2947
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 14.561102867126465,
-      "learning_rate": 1.4092412548862385e-05,
-      "loss": 1.7057,
+      "epoch": 0.37,
+      "grad_norm": 10.848490715026855,
+      "learning_rate": 1.7534200727942103e-05,
+      "loss": 1.7576,
       "step": 2948
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 17.088661193847656,
-      "learning_rate": 1.4090407938257995e-05,
-      "loss": 2.2723,
+      "epoch": 0.37,
+      "grad_norm": 7.745413303375244,
+      "learning_rate": 1.7533364012885413e-05,
+      "loss": 0.7855,
       "step": 2949
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 7.7882256507873535,
-      "learning_rate": 1.4088403327653603e-05,
-      "loss": 1.8397,
+      "epoch": 0.37,
+      "grad_norm": 10.996760368347168,
+      "learning_rate": 1.7532527297828727e-05,
+      "loss": 1.5187,
       "step": 2950
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 11.358787536621094,
-      "learning_rate": 1.4086398717049215e-05,
-      "loss": 2.2978,
+      "epoch": 0.37,
+      "grad_norm": 15.464694023132324,
+      "learning_rate": 1.753169058277204e-05,
+      "loss": 2.7653,
       "step": 2951
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 28.74677276611328,
-      "learning_rate": 1.4084394106444825e-05,
-      "loss": 2.7206,
+      "epoch": 0.37,
+      "grad_norm": 9.640546798706055,
+      "learning_rate": 1.753085386771535e-05,
+      "loss": 1.0689,
       "step": 2952
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 16.747051239013672,
-      "learning_rate": 1.4082389495840433e-05,
-      "loss": 1.9734,
+      "epoch": 0.37,
+      "grad_norm": 28.65819549560547,
+      "learning_rate": 1.7530017152658664e-05,
+      "loss": 1.2512,
       "step": 2953
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 12.262884140014648,
-      "learning_rate": 1.4080384885236045e-05,
-      "loss": 2.0553,
+      "epoch": 0.37,
+      "grad_norm": 14.472492218017578,
+      "learning_rate": 1.7529180437601974e-05,
+      "loss": 1.1207,
       "step": 2954
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 20.407411575317383,
-      "learning_rate": 1.4078380274631654e-05,
-      "loss": 1.3774,
+      "epoch": 0.37,
+      "grad_norm": 10.543621063232422,
+      "learning_rate": 1.7528343722545288e-05,
+      "loss": 1.9395,
       "step": 2955
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 25.87983512878418,
-      "learning_rate": 1.4076375664027264e-05,
-      "loss": 2.7685,
+      "epoch": 0.37,
+      "grad_norm": 33.20399475097656,
+      "learning_rate": 1.75275070074886e-05,
+      "loss": 2.0916,
       "step": 2956
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 14.667790412902832,
-      "learning_rate": 1.4074371053422872e-05,
-      "loss": 2.1229,
+      "epoch": 0.37,
+      "grad_norm": 10.970297813415527,
+      "learning_rate": 1.7526670292431912e-05,
+      "loss": 1.5082,
       "step": 2957
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 10.724139213562012,
-      "learning_rate": 1.4072366442818484e-05,
-      "loss": 1.6101,
+      "epoch": 0.37,
+      "grad_norm": 13.990462303161621,
+      "learning_rate": 1.7525833577375226e-05,
+      "loss": 2.0058,
       "step": 2958
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 25.648866653442383,
-      "learning_rate": 1.4070361832214094e-05,
-      "loss": 2.4417,
+      "epoch": 0.37,
+      "grad_norm": 9.842002868652344,
+      "learning_rate": 1.752499686231854e-05,
+      "loss": 1.9603,
       "step": 2959
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 24.464935302734375,
-      "learning_rate": 1.4068357221609702e-05,
-      "loss": 1.071,
+      "epoch": 0.37,
+      "grad_norm": 7.550311088562012,
+      "learning_rate": 1.752416014726185e-05,
+      "loss": 1.5242,
       "step": 2960
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 22.294677734375,
-      "learning_rate": 1.4066352611005314e-05,
-      "loss": 2.0512,
+      "epoch": 0.37,
+      "grad_norm": 14.05145263671875,
+      "learning_rate": 1.7523323432205163e-05,
+      "loss": 2.3942,
       "step": 2961
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 20.227296829223633,
-      "learning_rate": 1.4064348000400923e-05,
-      "loss": 1.4506,
+      "epoch": 0.37,
+      "grad_norm": 8.621101379394531,
+      "learning_rate": 1.7522486717148477e-05,
+      "loss": 1.4718,
       "step": 2962
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 35.15935134887695,
-      "learning_rate": 1.4062343389796533e-05,
-      "loss": 2.224,
+      "epoch": 0.37,
+      "grad_norm": 8.392345428466797,
+      "learning_rate": 1.752165000209179e-05,
+      "loss": 0.8059,
       "step": 2963
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 15.452341079711914,
-      "learning_rate": 1.4060338779192144e-05,
-      "loss": 3.0408,
+      "epoch": 0.37,
+      "grad_norm": 20.275230407714844,
+      "learning_rate": 1.75208132870351e-05,
+      "loss": 2.6383,
       "step": 2964
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 13.116132736206055,
-      "learning_rate": 1.4058334168587753e-05,
-      "loss": 2.6705,
+      "epoch": 0.37,
+      "grad_norm": 26.309600830078125,
+      "learning_rate": 1.7519976571978414e-05,
+      "loss": 1.837,
       "step": 2965
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 22.434358596801758,
-      "learning_rate": 1.4056329557983363e-05,
-      "loss": 2.6559,
+      "epoch": 0.37,
+      "grad_norm": 21.659997940063477,
+      "learning_rate": 1.7519139856921728e-05,
+      "loss": 3.1818,
       "step": 2966
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 27.88434600830078,
-      "learning_rate": 1.4054324947378973e-05,
-      "loss": 2.2654,
+      "epoch": 0.37,
+      "grad_norm": 26.298851013183594,
+      "learning_rate": 1.7518303141865038e-05,
+      "loss": 3.1853,
       "step": 2967
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 29.59733772277832,
-      "learning_rate": 1.4052320336774583e-05,
-      "loss": 1.7697,
+      "epoch": 0.37,
+      "grad_norm": 9.970256805419922,
+      "learning_rate": 1.7517466426808352e-05,
+      "loss": 2.1248,
       "step": 2968
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 27.069503784179688,
-      "learning_rate": 1.4050315726170191e-05,
-      "loss": 2.1905,
+      "epoch": 0.37,
+      "grad_norm": 21.72904396057129,
+      "learning_rate": 1.7516629711751666e-05,
+      "loss": 1.4126,
       "step": 2969
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 29.36873435974121,
-      "learning_rate": 1.4048311115565803e-05,
-      "loss": 2.3816,
+      "epoch": 0.37,
+      "grad_norm": 26.062219619750977,
+      "learning_rate": 1.751579299669498e-05,
+      "loss": 1.6812,
       "step": 2970
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 10.980319023132324,
-      "learning_rate": 1.4046306504961413e-05,
-      "loss": 1.6791,
+      "epoch": 0.37,
+      "grad_norm": 15.527777671813965,
+      "learning_rate": 1.751495628163829e-05,
+      "loss": 1.2272,
       "step": 2971
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 13.023073196411133,
-      "learning_rate": 1.4044301894357022e-05,
-      "loss": 1.9126,
+      "epoch": 0.37,
+      "grad_norm": 14.073158264160156,
+      "learning_rate": 1.7514119566581603e-05,
+      "loss": 1.8272,
       "step": 2972
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 38.787940979003906,
-      "learning_rate": 1.4042297283752633e-05,
-      "loss": 2.6036,
+      "epoch": 0.37,
+      "grad_norm": 17.060989379882812,
+      "learning_rate": 1.7513282851524917e-05,
+      "loss": 1.0744,
       "step": 2973
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 29.560956954956055,
-      "learning_rate": 1.4040292673148242e-05,
-      "loss": 2.8441,
+      "epoch": 0.37,
+      "grad_norm": 7.3651838302612305,
+      "learning_rate": 1.7512446136468227e-05,
+      "loss": 1.3256,
       "step": 2974
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 8.834044456481934,
-      "learning_rate": 1.4038288062543852e-05,
-      "loss": 1.4787,
+      "epoch": 0.37,
+      "grad_norm": 13.686807632446289,
+      "learning_rate": 1.751160942141154e-05,
+      "loss": 3.2167,
       "step": 2975
     },
     {
-      "epoch": 0.89,
-      "grad_norm": 11.226600646972656,
-      "learning_rate": 1.403628345193946e-05,
-      "loss": 1.5387,
+      "epoch": 0.37,
+      "grad_norm": 14.580486297607422,
+      "learning_rate": 1.7510772706354854e-05,
+      "loss": 2.9224,
       "step": 2976
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 16.720735549926758,
-      "learning_rate": 1.4034278841335072e-05,
-      "loss": 2.2926,
+      "epoch": 0.37,
+      "grad_norm": 15.495469093322754,
+      "learning_rate": 1.7509935991298165e-05,
+      "loss": 0.9868,
       "step": 2977
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 20.5881290435791,
-      "learning_rate": 1.4032274230730682e-05,
-      "loss": 2.5574,
+      "epoch": 0.37,
+      "grad_norm": 11.540633201599121,
+      "learning_rate": 1.7509099276241478e-05,
+      "loss": 1.0614,
       "step": 2978
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 31.438135147094727,
-      "learning_rate": 1.403026962012629e-05,
-      "loss": 1.8076,
+      "epoch": 0.37,
+      "grad_norm": 8.092855453491211,
+      "learning_rate": 1.750826256118479e-05,
+      "loss": 1.0501,
       "step": 2979
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 12.068266868591309,
-      "learning_rate": 1.4028265009521902e-05,
-      "loss": 1.2816,
+      "epoch": 0.37,
+      "grad_norm": 12.961438179016113,
+      "learning_rate": 1.7507425846128102e-05,
+      "loss": 2.6531,
       "step": 2980
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 25.184410095214844,
-      "learning_rate": 1.402626039891751e-05,
-      "loss": 2.4355,
+      "epoch": 0.37,
+      "grad_norm": 8.35956859588623,
+      "learning_rate": 1.7506589131071416e-05,
+      "loss": 2.8781,
       "step": 2981
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 11.62043285369873,
-      "learning_rate": 1.402425578831312e-05,
-      "loss": 1.7278,
+      "epoch": 0.37,
+      "grad_norm": 16.814393997192383,
+      "learning_rate": 1.7505752416014726e-05,
+      "loss": 2.518,
       "step": 2982
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 10.73161506652832,
-      "learning_rate": 1.4022251177708733e-05,
-      "loss": 1.5009,
+      "epoch": 0.37,
+      "grad_norm": 9.360370635986328,
+      "learning_rate": 1.750491570095804e-05,
+      "loss": 1.1125,
       "step": 2983
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 18.191015243530273,
-      "learning_rate": 1.4020246567104341e-05,
-      "loss": 1.7434,
+      "epoch": 0.37,
+      "grad_norm": 13.405939102172852,
+      "learning_rate": 1.7504078985901353e-05,
+      "loss": 1.3022,
       "step": 2984
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 35.12432861328125,
-      "learning_rate": 1.4018241956499951e-05,
-      "loss": 1.6909,
+      "epoch": 0.37,
+      "grad_norm": 28.36098861694336,
+      "learning_rate": 1.7503242270844664e-05,
+      "loss": 3.2326,
       "step": 2985
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 11.526291847229004,
-      "learning_rate": 1.401623734589556e-05,
-      "loss": 1.4617,
+      "epoch": 0.37,
+      "grad_norm": 19.087400436401367,
+      "learning_rate": 1.7502405555787977e-05,
+      "loss": 2.0111,
       "step": 2986
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 61.24599075317383,
-      "learning_rate": 1.4014232735291171e-05,
-      "loss": 1.9207,
+      "epoch": 0.37,
+      "grad_norm": 32.10877227783203,
+      "learning_rate": 1.750156884073129e-05,
+      "loss": 2.5922,
       "step": 2987
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 21.56028175354004,
-      "learning_rate": 1.401222812468678e-05,
-      "loss": 2.4295,
+      "epoch": 0.37,
+      "grad_norm": 10.107169151306152,
+      "learning_rate": 1.75007321256746e-05,
+      "loss": 2.5872,
       "step": 2988
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 15.971736907958984,
-      "learning_rate": 1.4010223514082391e-05,
-      "loss": 1.9247,
+      "epoch": 0.38,
+      "grad_norm": 18.61122703552246,
+      "learning_rate": 1.7499895410617915e-05,
+      "loss": 3.6177,
       "step": 2989
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 18.692066192626953,
-      "learning_rate": 1.4008218903478001e-05,
-      "loss": 2.1351,
+      "epoch": 0.38,
+      "grad_norm": 7.76904296875,
+      "learning_rate": 1.749905869556123e-05,
+      "loss": 1.7388,
       "step": 2990
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 19.446582794189453,
-      "learning_rate": 1.400621429287361e-05,
-      "loss": 2.2271,
+      "epoch": 0.38,
+      "grad_norm": 19.504474639892578,
+      "learning_rate": 1.7498221980504542e-05,
+      "loss": 3.4638,
       "step": 2991
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 14.894234657287598,
-      "learning_rate": 1.4004209682269222e-05,
-      "loss": 2.0351,
+      "epoch": 0.38,
+      "grad_norm": 16.34959602355957,
+      "learning_rate": 1.7497385265447852e-05,
+      "loss": 2.5737,
       "step": 2992
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 10.384711265563965,
-      "learning_rate": 1.400220507166483e-05,
-      "loss": 1.3777,
+      "epoch": 0.38,
+      "grad_norm": 6.932929515838623,
+      "learning_rate": 1.7496548550391166e-05,
+      "loss": 0.9677,
       "step": 2993
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 15.35462760925293,
-      "learning_rate": 1.400020046106044e-05,
-      "loss": 1.4528,
+      "epoch": 0.38,
+      "grad_norm": 13.935084342956543,
+      "learning_rate": 1.749571183533448e-05,
+      "loss": 1.6441,
       "step": 2994
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 9.53518009185791,
-      "learning_rate": 1.3998195850456052e-05,
-      "loss": 2.0396,
+      "epoch": 0.38,
+      "grad_norm": 10.431344985961914,
+      "learning_rate": 1.749487512027779e-05,
+      "loss": 2.1761,
       "step": 2995
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 12.702194213867188,
-      "learning_rate": 1.399619123985166e-05,
-      "loss": 3.5201,
+      "epoch": 0.38,
+      "grad_norm": 5.776036739349365,
+      "learning_rate": 1.7494038405221104e-05,
+      "loss": 0.7499,
       "step": 2996
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 16.689231872558594,
-      "learning_rate": 1.399418662924727e-05,
-      "loss": 2.5204,
+      "epoch": 0.38,
+      "grad_norm": 14.878742218017578,
+      "learning_rate": 1.7493201690164417e-05,
+      "loss": 2.2408,
       "step": 2997
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 16.812484741210938,
-      "learning_rate": 1.3992182018642879e-05,
-      "loss": 1.823,
+      "epoch": 0.38,
+      "grad_norm": 9.810598373413086,
+      "learning_rate": 1.749236497510773e-05,
+      "loss": 1.5583,
       "step": 2998
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 27.859556198120117,
-      "learning_rate": 1.399017740803849e-05,
-      "loss": 2.3145,
+      "epoch": 0.38,
+      "grad_norm": 16.407333374023438,
+      "learning_rate": 1.749152826005104e-05,
+      "loss": 1.1398,
       "step": 2999
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 40.26885223388672,
-      "learning_rate": 1.3988172797434099e-05,
-      "loss": 1.981,
-      "step": 3000
-    },
-    {
-      "epoch": 0.9,
-      "eval_loss": 0.290047824382782,
-      "eval_runtime": 43.5045,
-      "eval_samples_per_second": 33.996,
-      "eval_steps_per_second": 33.996,
+      "epoch": 0.38,
+      "grad_norm": 10.976561546325684,
+      "learning_rate": 1.7490691544994355e-05,
+      "loss": 2.8406,
       "step": 3000
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 16.576269149780273,
-      "learning_rate": 1.3986168186829709e-05,
-      "loss": 1.7576,
+      "epoch": 0.38,
+      "grad_norm": 11.879322052001953,
+      "learning_rate": 1.748985482993767e-05,
+      "loss": 1.451,
       "step": 3001
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 11.136350631713867,
-      "learning_rate": 1.398416357622532e-05,
-      "loss": 1.1301,
+      "epoch": 0.38,
+      "grad_norm": 16.093420028686523,
+      "learning_rate": 1.748901811488098e-05,
+      "loss": 2.4345,
       "step": 3002
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 7.8271684646606445,
-      "learning_rate": 1.398215896562093e-05,
-      "loss": 1.6797,
+      "epoch": 0.38,
+      "grad_norm": 15.955079078674316,
+      "learning_rate": 1.7488181399824292e-05,
+      "loss": 3.383,
       "step": 3003
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 18.726409912109375,
-      "learning_rate": 1.398015435501654e-05,
-      "loss": 2.3383,
+      "epoch": 0.38,
+      "grad_norm": 14.334516525268555,
+      "learning_rate": 1.7487344684767603e-05,
+      "loss": 3.2137,
       "step": 3004
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 13.888818740844727,
-      "learning_rate": 1.3978149744412148e-05,
-      "loss": 1.7,
+      "epoch": 0.38,
+      "grad_norm": 9.579103469848633,
+      "learning_rate": 1.7486507969710916e-05,
+      "loss": 3.0365,
       "step": 3005
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 17.36934471130371,
-      "learning_rate": 1.397614513380776e-05,
-      "loss": 1.9836,
+      "epoch": 0.38,
+      "grad_norm": 8.867793083190918,
+      "learning_rate": 1.748567125465423e-05,
+      "loss": 1.9244,
       "step": 3006
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 34.03765869140625,
-      "learning_rate": 1.3974140523203368e-05,
-      "loss": 2.0497,
+      "epoch": 0.38,
+      "grad_norm": 41.3364372253418,
+      "learning_rate": 1.748483453959754e-05,
+      "loss": 3.6235,
       "step": 3007
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 24.520832061767578,
-      "learning_rate": 1.3972135912598978e-05,
-      "loss": 3.1885,
+      "epoch": 0.38,
+      "grad_norm": 53.81862258911133,
+      "learning_rate": 1.7483997824540854e-05,
+      "loss": 2.396,
       "step": 3008
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 13.99000072479248,
-      "learning_rate": 1.397013130199459e-05,
-      "loss": 2.3549,
+      "epoch": 0.38,
+      "grad_norm": 9.474702835083008,
+      "learning_rate": 1.7483161109484167e-05,
+      "loss": 1.8655,
       "step": 3009
     },
     {
-      "epoch": 0.9,
-      "grad_norm": 14.597472190856934,
-      "learning_rate": 1.3968126691390198e-05,
-      "loss": 2.2443,
+      "epoch": 0.38,
+      "grad_norm": 17.841140747070312,
+      "learning_rate": 1.7482324394427478e-05,
+      "loss": 1.9738,
       "step": 3010
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 10.168909072875977,
-      "learning_rate": 1.3966122080785808e-05,
-      "loss": 2.1224,
+      "epoch": 0.38,
+      "grad_norm": 11.464025497436523,
+      "learning_rate": 1.748148767937079e-05,
+      "loss": 2.1443,
       "step": 3011
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 21.07402229309082,
-      "learning_rate": 1.3964117470181418e-05,
-      "loss": 2.2544,
+      "epoch": 0.38,
+      "grad_norm": 24.358856201171875,
+      "learning_rate": 1.7480650964314105e-05,
+      "loss": 1.866,
       "step": 3012
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 35.29275894165039,
-      "learning_rate": 1.3962112859577028e-05,
-      "loss": 4.147,
+      "epoch": 0.38,
+      "grad_norm": 8.274801254272461,
+      "learning_rate": 1.7479814249257415e-05,
+      "loss": 0.9875,
       "step": 3013
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 19.924930572509766,
-      "learning_rate": 1.3960108248972638e-05,
-      "loss": 2.3426,
+      "epoch": 0.38,
+      "grad_norm": 16.024633407592773,
+      "learning_rate": 1.747897753420073e-05,
+      "loss": 1.8152,
       "step": 3014
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 18.175373077392578,
-      "learning_rate": 1.3958103638368249e-05,
-      "loss": 1.6238,
+      "epoch": 0.38,
+      "grad_norm": 22.47686195373535,
+      "learning_rate": 1.7478140819144043e-05,
+      "loss": 3.7221,
       "step": 3015
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 16.749662399291992,
-      "learning_rate": 1.3956099027763859e-05,
-      "loss": 1.9444,
+      "epoch": 0.38,
+      "grad_norm": 10.875396728515625,
+      "learning_rate": 1.7477304104087353e-05,
+      "loss": 2.6813,
       "step": 3016
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 32.66911697387695,
-      "learning_rate": 1.3954094417159467e-05,
-      "loss": 2.5848,
+      "epoch": 0.38,
+      "grad_norm": 10.279182434082031,
+      "learning_rate": 1.7476467389030666e-05,
+      "loss": 3.5712,
       "step": 3017
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 10.309561729431152,
-      "learning_rate": 1.3952089806555079e-05,
-      "loss": 1.2661,
+      "epoch": 0.38,
+      "grad_norm": 34.010494232177734,
+      "learning_rate": 1.747563067397398e-05,
+      "loss": 2.5344,
       "step": 3018
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 32.622840881347656,
-      "learning_rate": 1.3950085195950687e-05,
-      "loss": 3.1314,
+      "epoch": 0.38,
+      "grad_norm": 9.324440956115723,
+      "learning_rate": 1.7474793958917294e-05,
+      "loss": 1.9682,
       "step": 3019
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 51.04716491699219,
-      "learning_rate": 1.3948080585346297e-05,
-      "loss": 2.5912,
+      "epoch": 0.38,
+      "grad_norm": 9.498290061950684,
+      "learning_rate": 1.7473957243860604e-05,
+      "loss": 2.3948,
       "step": 3020
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 9.933348655700684,
-      "learning_rate": 1.3946075974741909e-05,
-      "loss": 1.6906,
+      "epoch": 0.38,
+      "grad_norm": 33.985198974609375,
+      "learning_rate": 1.7473120528803918e-05,
+      "loss": 3.7208,
       "step": 3021
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 11.241128921508789,
-      "learning_rate": 1.3944071364137517e-05,
-      "loss": 2.0141,
+      "epoch": 0.38,
+      "grad_norm": 19.785444259643555,
+      "learning_rate": 1.747228381374723e-05,
+      "loss": 2.1617,
       "step": 3022
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 20.985105514526367,
-      "learning_rate": 1.3942066753533127e-05,
-      "loss": 2.3962,
+      "epoch": 0.38,
+      "grad_norm": 23.751007080078125,
+      "learning_rate": 1.747144709869054e-05,
+      "loss": 2.6466,
       "step": 3023
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 14.399240493774414,
-      "learning_rate": 1.3940062142928736e-05,
-      "loss": 2.4056,
+      "epoch": 0.38,
+      "grad_norm": 28.25634765625,
+      "learning_rate": 1.7470610383633855e-05,
+      "loss": 2.7269,
       "step": 3024
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 23.836753845214844,
-      "learning_rate": 1.3938057532324348e-05,
-      "loss": 1.492,
+      "epoch": 0.38,
+      "grad_norm": 15.494318962097168,
+      "learning_rate": 1.746977366857717e-05,
+      "loss": 1.9662,
       "step": 3025
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 18.37982940673828,
-      "learning_rate": 1.3936052921719958e-05,
-      "loss": 1.8183,
+      "epoch": 0.38,
+      "grad_norm": 40.99577713012695,
+      "learning_rate": 1.7468936953520483e-05,
+      "loss": 3.8287,
       "step": 3026
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 13.537169456481934,
-      "learning_rate": 1.3934048311115566e-05,
-      "loss": 2.1759,
+      "epoch": 0.38,
+      "grad_norm": 15.126395225524902,
+      "learning_rate": 1.7468100238463793e-05,
+      "loss": 2.892,
       "step": 3027
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 23.467910766601562,
-      "learning_rate": 1.3932043700511178e-05,
-      "loss": 2.6208,
+      "epoch": 0.38,
+      "grad_norm": 15.43254280090332,
+      "learning_rate": 1.7467263523407106e-05,
+      "loss": 1.6425,
       "step": 3028
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 17.59671401977539,
-      "learning_rate": 1.3930039089906786e-05,
-      "loss": 2.3887,
+      "epoch": 0.38,
+      "grad_norm": 22.664308547973633,
+      "learning_rate": 1.746642680835042e-05,
+      "loss": 2.8555,
       "step": 3029
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 12.384592056274414,
-      "learning_rate": 1.3928034479302396e-05,
-      "loss": 2.0166,
+      "epoch": 0.38,
+      "grad_norm": 22.935243606567383,
+      "learning_rate": 1.746559009329373e-05,
+      "loss": 3.0264,
       "step": 3030
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 17.471967697143555,
-      "learning_rate": 1.3926029868698005e-05,
-      "loss": 1.875,
+      "epoch": 0.38,
+      "grad_norm": 8.225278854370117,
+      "learning_rate": 1.7464753378237044e-05,
+      "loss": 4.3243,
       "step": 3031
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 12.046953201293945,
-      "learning_rate": 1.3924025258093617e-05,
-      "loss": 1.3814,
+      "epoch": 0.38,
+      "grad_norm": 6.5009613037109375,
+      "learning_rate": 1.7463916663180354e-05,
+      "loss": 3.2662,
       "step": 3032
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 15.179215431213379,
-      "learning_rate": 1.3922020647489227e-05,
-      "loss": 1.6944,
+      "epoch": 0.38,
+      "grad_norm": 10.92690372467041,
+      "learning_rate": 1.7463079948123668e-05,
+      "loss": 1.8278,
       "step": 3033
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 10.748503684997559,
-      "learning_rate": 1.3920016036884835e-05,
-      "loss": 2.1539,
+      "epoch": 0.38,
+      "grad_norm": 17.307109832763672,
+      "learning_rate": 1.746224323306698e-05,
+      "loss": 3.076,
       "step": 3034
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 20.441787719726562,
-      "learning_rate": 1.3918011426280447e-05,
-      "loss": 2.9749,
+      "epoch": 0.38,
+      "grad_norm": 12.11257266998291,
+      "learning_rate": 1.7461406518010292e-05,
+      "loss": 2.4566,
       "step": 3035
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 23.096572875976562,
-      "learning_rate": 1.3916006815676055e-05,
-      "loss": 1.9606,
+      "epoch": 0.38,
+      "grad_norm": 21.447715759277344,
+      "learning_rate": 1.7460569802953605e-05,
+      "loss": 1.4901,
       "step": 3036
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 27.173370361328125,
-      "learning_rate": 1.3914002205071665e-05,
-      "loss": 2.5916,
+      "epoch": 0.38,
+      "grad_norm": 8.973800659179688,
+      "learning_rate": 1.7459733087896916e-05,
+      "loss": 2.4119,
       "step": 3037
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 11.456189155578613,
-      "learning_rate": 1.3911997594467277e-05,
-      "loss": 1.7153,
+      "epoch": 0.38,
+      "grad_norm": 12.067655563354492,
+      "learning_rate": 1.745889637284023e-05,
+      "loss": 1.4386,
       "step": 3038
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 36.13789367675781,
-      "learning_rate": 1.3909992983862885e-05,
-      "loss": 2.2955,
+      "epoch": 0.38,
+      "grad_norm": 10.95497989654541,
+      "learning_rate": 1.7458059657783543e-05,
+      "loss": 2.5098,
       "step": 3039
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 25.46799659729004,
-      "learning_rate": 1.3907988373258496e-05,
-      "loss": 2.2377,
+      "epoch": 0.38,
+      "grad_norm": 27.85697364807129,
+      "learning_rate": 1.7457222942726853e-05,
+      "loss": 3.0961,
       "step": 3040
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 18.01449203491211,
-      "learning_rate": 1.3905983762654106e-05,
-      "loss": 2.5831,
+      "epoch": 0.38,
+      "grad_norm": 23.892608642578125,
+      "learning_rate": 1.7456386227670167e-05,
+      "loss": 2.4394,
       "step": 3041
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 25.396291732788086,
-      "learning_rate": 1.3903979152049716e-05,
-      "loss": 2.2237,
+      "epoch": 0.38,
+      "grad_norm": 26.527515411376953,
+      "learning_rate": 1.745554951261348e-05,
+      "loss": 3.225,
       "step": 3042
     },
     {
-      "epoch": 0.91,
-      "grad_norm": 15.773884773254395,
-      "learning_rate": 1.3901974541445324e-05,
-      "loss": 2.6421,
+      "epoch": 0.38,
+      "grad_norm": 13.135971069335938,
+      "learning_rate": 1.7454712797556794e-05,
+      "loss": 1.3418,
       "step": 3043
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 16.961776733398438,
-      "learning_rate": 1.3899969930840936e-05,
-      "loss": 1.8037,
+      "epoch": 0.38,
+      "grad_norm": 7.4857072830200195,
+      "learning_rate": 1.7453876082500104e-05,
+      "loss": 2.2289,
       "step": 3044
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 46.810585021972656,
-      "learning_rate": 1.3897965320236546e-05,
-      "loss": 1.4078,
+      "epoch": 0.38,
+      "grad_norm": 20.1530818939209,
+      "learning_rate": 1.7453039367443418e-05,
+      "loss": 3.5615,
       "step": 3045
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 19.385047912597656,
-      "learning_rate": 1.3895960709632154e-05,
-      "loss": 1.9391,
+      "epoch": 0.38,
+      "grad_norm": 12.91379165649414,
+      "learning_rate": 1.7452202652386732e-05,
+      "loss": 4.9487,
       "step": 3046
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 14.272333145141602,
-      "learning_rate": 1.3893956099027766e-05,
-      "loss": 1.7094,
+      "epoch": 0.38,
+      "grad_norm": 6.44926118850708,
+      "learning_rate": 1.7451365937330042e-05,
+      "loss": 0.9218,
       "step": 3047
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 12.611112594604492,
-      "learning_rate": 1.3891951488423375e-05,
-      "loss": 2.0959,
+      "epoch": 0.38,
+      "grad_norm": 14.764775276184082,
+      "learning_rate": 1.7450529222273356e-05,
+      "loss": 2.2744,
       "step": 3048
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 17.307199478149414,
-      "learning_rate": 1.3889946877818985e-05,
-      "loss": 2.1561,
+      "epoch": 0.38,
+      "grad_norm": 10.772073745727539,
+      "learning_rate": 1.744969250721667e-05,
+      "loss": 2.0143,
       "step": 3049
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 20.245845794677734,
-      "learning_rate": 1.3887942267214593e-05,
-      "loss": 2.8128,
+      "epoch": 0.38,
+      "grad_norm": 34.17717361450195,
+      "learning_rate": 1.7448855792159983e-05,
+      "loss": 4.1729,
       "step": 3050
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 17.531375885009766,
-      "learning_rate": 1.3885937656610205e-05,
-      "loss": 1.6429,
+      "epoch": 0.38,
+      "grad_norm": 20.677505493164062,
+      "learning_rate": 1.7448019077103293e-05,
+      "loss": 2.2328,
       "step": 3051
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 12.451351165771484,
-      "learning_rate": 1.3883933046005815e-05,
-      "loss": 2.0644,
+      "epoch": 0.38,
+      "grad_norm": 13.210722923278809,
+      "learning_rate": 1.7447182362046607e-05,
+      "loss": 1.8375,
       "step": 3052
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 16.824329376220703,
-      "learning_rate": 1.3881928435401423e-05,
-      "loss": 2.1751,
+      "epoch": 0.38,
+      "grad_norm": 7.626525402069092,
+      "learning_rate": 1.744634564698992e-05,
+      "loss": 1.3073,
       "step": 3053
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 19.493457794189453,
-      "learning_rate": 1.3879923824797035e-05,
-      "loss": 1.8751,
+      "epoch": 0.38,
+      "grad_norm": 15.71025276184082,
+      "learning_rate": 1.744550893193323e-05,
+      "loss": 1.94,
       "step": 3054
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 16.869348526000977,
-      "learning_rate": 1.3877919214192643e-05,
-      "loss": 1.9817,
+      "epoch": 0.38,
+      "grad_norm": 10.431207656860352,
+      "learning_rate": 1.7444672216876544e-05,
+      "loss": 1.1058,
       "step": 3055
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 29.737041473388672,
-      "learning_rate": 1.3875914603588254e-05,
-      "loss": 2.4454,
+      "epoch": 0.38,
+      "grad_norm": 50.677146911621094,
+      "learning_rate": 1.7443835501819858e-05,
+      "loss": 1.7067,
       "step": 3056
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 22.09736442565918,
-      "learning_rate": 1.3873909992983865e-05,
-      "loss": 1.7898,
+      "epoch": 0.38,
+      "grad_norm": 17.605249404907227,
+      "learning_rate": 1.744299878676317e-05,
+      "loss": 2.4274,
       "step": 3057
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 17.416990280151367,
-      "learning_rate": 1.3871905382379474e-05,
-      "loss": 2.9486,
+      "epoch": 0.38,
+      "grad_norm": 18.438739776611328,
+      "learning_rate": 1.7442162071706482e-05,
+      "loss": 1.4258,
       "step": 3058
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 13.20331859588623,
-      "learning_rate": 1.3869900771775084e-05,
-      "loss": 2.2757,
+      "epoch": 0.38,
+      "grad_norm": 37.58784484863281,
+      "learning_rate": 1.7441325356649796e-05,
+      "loss": 3.915,
       "step": 3059
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 7.922008514404297,
-      "learning_rate": 1.3867896161170694e-05,
-      "loss": 2.0063,
+      "epoch": 0.38,
+      "grad_norm": 18.80876350402832,
+      "learning_rate": 1.7440488641593106e-05,
+      "loss": 2.3819,
       "step": 3060
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 14.735453605651855,
-      "learning_rate": 1.3865891550566304e-05,
-      "loss": 2.4153,
+      "epoch": 0.38,
+      "grad_norm": 12.009927749633789,
+      "learning_rate": 1.743965192653642e-05,
+      "loss": 2.8513,
       "step": 3061
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 13.794514656066895,
-      "learning_rate": 1.3863886939961912e-05,
-      "loss": 1.8875,
+      "epoch": 0.38,
+      "grad_norm": 28.326574325561523,
+      "learning_rate": 1.7438815211479733e-05,
+      "loss": 2.6981,
       "step": 3062
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 11.291130065917969,
-      "learning_rate": 1.3861882329357524e-05,
-      "loss": 1.3319,
+      "epoch": 0.38,
+      "grad_norm": 10.06318473815918,
+      "learning_rate": 1.7437978496423043e-05,
+      "loss": 3.7585,
       "step": 3063
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 19.629484176635742,
-      "learning_rate": 1.3859877718753134e-05,
-      "loss": 1.4707,
+      "epoch": 0.38,
+      "grad_norm": 13.952555656433105,
+      "learning_rate": 1.7437141781366357e-05,
+      "loss": 1.7787,
       "step": 3064
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 15.42160415649414,
-      "learning_rate": 1.3857873108148743e-05,
-      "loss": 1.5667,
+      "epoch": 0.38,
+      "grad_norm": 39.10877227783203,
+      "learning_rate": 1.7436305066309667e-05,
+      "loss": 2.7993,
       "step": 3065
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 14.26404094696045,
-      "learning_rate": 1.3855868497544354e-05,
-      "loss": 2.0599,
+      "epoch": 0.38,
+      "grad_norm": 20.320301055908203,
+      "learning_rate": 1.743546835125298e-05,
+      "loss": 1.2898,
       "step": 3066
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 8.3535795211792,
-      "learning_rate": 1.3853863886939963e-05,
-      "loss": 1.3769,
+      "epoch": 0.38,
+      "grad_norm": 23.73604965209961,
+      "learning_rate": 1.7434631636196295e-05,
+      "loss": 3.5141,
       "step": 3067
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 13.800082206726074,
-      "learning_rate": 1.3851859276335573e-05,
-      "loss": 1.5491,
+      "epoch": 0.39,
+      "grad_norm": 10.633537292480469,
+      "learning_rate": 1.7433794921139605e-05,
+      "loss": 2.189,
       "step": 3068
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 16.2960262298584,
-      "learning_rate": 1.3849854665731185e-05,
-      "loss": 1.8461,
+      "epoch": 0.39,
+      "grad_norm": 13.728384971618652,
+      "learning_rate": 1.743295820608292e-05,
+      "loss": 1.515,
       "step": 3069
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 12.411531448364258,
-      "learning_rate": 1.3847850055126793e-05,
-      "loss": 2.1916,
+      "epoch": 0.39,
+      "grad_norm": 11.651106834411621,
+      "learning_rate": 1.7432121491026232e-05,
+      "loss": 1.1459,
       "step": 3070
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 10.797046661376953,
-      "learning_rate": 1.3845845444522403e-05,
-      "loss": 1.2897,
+      "epoch": 0.39,
+      "grad_norm": 9.04185962677002,
+      "learning_rate": 1.7431284775969546e-05,
+      "loss": 0.6994,
       "step": 3071
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 17.1668643951416,
-      "learning_rate": 1.3843840833918011e-05,
-      "loss": 2.0636,
+      "epoch": 0.39,
+      "grad_norm": 10.415605545043945,
+      "learning_rate": 1.7430448060912856e-05,
+      "loss": 2.3379,
       "step": 3072
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 12.581513404846191,
-      "learning_rate": 1.3841836223313623e-05,
-      "loss": 2.0915,
+      "epoch": 0.39,
+      "grad_norm": 14.66109848022461,
+      "learning_rate": 1.742961134585617e-05,
+      "loss": 2.6971,
       "step": 3073
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 18.048904418945312,
-      "learning_rate": 1.3839831612709232e-05,
-      "loss": 1.8037,
+      "epoch": 0.39,
+      "grad_norm": 19.801557540893555,
+      "learning_rate": 1.7428774630799483e-05,
+      "loss": 2.4251,
       "step": 3074
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 25.745256423950195,
-      "learning_rate": 1.3837827002104842e-05,
-      "loss": 2.4617,
+      "epoch": 0.39,
+      "grad_norm": 12.651113510131836,
+      "learning_rate": 1.7427937915742794e-05,
+      "loss": 2.507,
       "step": 3075
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 23.290611267089844,
-      "learning_rate": 1.3835822391500453e-05,
-      "loss": 2.3306,
+      "epoch": 0.39,
+      "grad_norm": 14.438727378845215,
+      "learning_rate": 1.7427101200686107e-05,
+      "loss": 1.5898,
       "step": 3076
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 18.817665100097656,
-      "learning_rate": 1.3833817780896062e-05,
-      "loss": 2.0089,
+      "epoch": 0.39,
+      "grad_norm": 21.17618179321289,
+      "learning_rate": 1.742626448562942e-05,
+      "loss": 3.7627,
       "step": 3077
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 16.671045303344727,
-      "learning_rate": 1.3831813170291672e-05,
-      "loss": 2.0306,
+      "epoch": 0.39,
+      "grad_norm": 20.78024673461914,
+      "learning_rate": 1.7425427770572735e-05,
+      "loss": 2.8387,
       "step": 3078
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 17.590496063232422,
-      "learning_rate": 1.382980855968728e-05,
-      "loss": 2.5062,
+      "epoch": 0.39,
+      "grad_norm": 8.485142707824707,
+      "learning_rate": 1.7424591055516045e-05,
+      "loss": 2.4435,
       "step": 3079
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 15.332666397094727,
-      "learning_rate": 1.3827803949082892e-05,
-      "loss": 1.5195,
+      "epoch": 0.39,
+      "grad_norm": 7.811776161193848,
+      "learning_rate": 1.742375434045936e-05,
+      "loss": 2.0698,
       "step": 3080
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 28.62594985961914,
-      "learning_rate": 1.3825799338478502e-05,
-      "loss": 2.3762,
+      "epoch": 0.39,
+      "grad_norm": 16.553388595581055,
+      "learning_rate": 1.7422917625402672e-05,
+      "loss": 1.2277,
       "step": 3081
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 15.981874465942383,
-      "learning_rate": 1.382379472787411e-05,
-      "loss": 2.0642,
+      "epoch": 0.39,
+      "grad_norm": 23.243209838867188,
+      "learning_rate": 1.7422080910345982e-05,
+      "loss": 3.6227,
       "step": 3082
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 16.672204971313477,
-      "learning_rate": 1.3821790117269722e-05,
-      "loss": 2.0906,
+      "epoch": 0.39,
+      "grad_norm": 11.811983108520508,
+      "learning_rate": 1.7421244195289296e-05,
+      "loss": 1.2775,
       "step": 3083
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 15.016446113586426,
-      "learning_rate": 1.381978550666533e-05,
-      "loss": 1.7059,
+      "epoch": 0.39,
+      "grad_norm": 14.592747688293457,
+      "learning_rate": 1.742040748023261e-05,
+      "loss": 1.3105,
       "step": 3084
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 9.689031600952148,
-      "learning_rate": 1.3817780896060941e-05,
-      "loss": 1.625,
+      "epoch": 0.39,
+      "grad_norm": 15.486154556274414,
+      "learning_rate": 1.741957076517592e-05,
+      "loss": 2.1047,
       "step": 3085
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 8.496919631958008,
-      "learning_rate": 1.3815776285456551e-05,
-      "loss": 0.9668,
+      "epoch": 0.39,
+      "grad_norm": 28.537649154663086,
+      "learning_rate": 1.7418734050119234e-05,
+      "loss": 3.2185,
       "step": 3086
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 20.545669555664062,
-      "learning_rate": 1.3813771674852161e-05,
-      "loss": 1.827,
+      "epoch": 0.39,
+      "grad_norm": 10.975678443908691,
+      "learning_rate": 1.7417897335062547e-05,
+      "loss": 1.8193,
       "step": 3087
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 14.81933307647705,
-      "learning_rate": 1.3811767064247771e-05,
-      "loss": 2.3411,
+      "epoch": 0.39,
+      "grad_norm": 21.351566314697266,
+      "learning_rate": 1.7417060620005858e-05,
+      "loss": 1.0022,
       "step": 3088
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 11.812262535095215,
-      "learning_rate": 1.3809762453643381e-05,
-      "loss": 3.2352,
+      "epoch": 0.39,
+      "grad_norm": 16.831201553344727,
+      "learning_rate": 1.741622390494917e-05,
+      "loss": 2.9873,
       "step": 3089
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 50.63083267211914,
-      "learning_rate": 1.3807757843038991e-05,
-      "loss": 1.491,
+      "epoch": 0.39,
+      "grad_norm": 12.240836143493652,
+      "learning_rate": 1.741538718989248e-05,
+      "loss": 2.8131,
       "step": 3090
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 17.194198608398438,
-      "learning_rate": 1.38057532324346e-05,
-      "loss": 1.45,
+      "epoch": 0.39,
+      "grad_norm": 12.75204086303711,
+      "learning_rate": 1.7414550474835795e-05,
+      "loss": 1.3815,
       "step": 3091
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 25.95490074157715,
-      "learning_rate": 1.3803748621830211e-05,
-      "loss": 3.7476,
+      "epoch": 0.39,
+      "grad_norm": 17.798721313476562,
+      "learning_rate": 1.741371375977911e-05,
+      "loss": 1.5151,
       "step": 3092
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 12.458473205566406,
-      "learning_rate": 1.380174401122582e-05,
-      "loss": 2.0199,
+      "epoch": 0.39,
+      "grad_norm": 12.037940979003906,
+      "learning_rate": 1.741287704472242e-05,
+      "loss": 2.6141,
       "step": 3093
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 17.711647033691406,
-      "learning_rate": 1.379973940062143e-05,
-      "loss": 2.1909,
+      "epoch": 0.39,
+      "grad_norm": 6.18284273147583,
+      "learning_rate": 1.7412040329665733e-05,
+      "loss": 0.9533,
       "step": 3094
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 21.15656852722168,
-      "learning_rate": 1.3797734790017042e-05,
-      "loss": 2.6897,
+      "epoch": 0.39,
+      "grad_norm": 9.289584159851074,
+      "learning_rate": 1.7411203614609046e-05,
+      "loss": 1.0435,
       "step": 3095
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 14.197000503540039,
-      "learning_rate": 1.379573017941265e-05,
-      "loss": 2.5904,
+      "epoch": 0.39,
+      "grad_norm": 10.22266674041748,
+      "learning_rate": 1.7410366899552357e-05,
+      "loss": 1.583,
       "step": 3096
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 24.11342430114746,
-      "learning_rate": 1.379372556880826e-05,
-      "loss": 2.4864,
+      "epoch": 0.39,
+      "grad_norm": 24.6475887298584,
+      "learning_rate": 1.740953018449567e-05,
+      "loss": 1.5329,
       "step": 3097
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 18.00290298461914,
-      "learning_rate": 1.3791720958203869e-05,
-      "loss": 2.6676,
+      "epoch": 0.39,
+      "grad_norm": 14.650251388549805,
+      "learning_rate": 1.7408693469438984e-05,
+      "loss": 2.0842,
       "step": 3098
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 10.474432945251465,
-      "learning_rate": 1.378971634759948e-05,
-      "loss": 1.9275,
+      "epoch": 0.39,
+      "grad_norm": 12.11793327331543,
+      "learning_rate": 1.7407856754382298e-05,
+      "loss": 2.7101,
       "step": 3099
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 15.15256118774414,
-      "learning_rate": 1.378771173699509e-05,
-      "loss": 2.0671,
+      "epoch": 0.39,
+      "grad_norm": 28.798845291137695,
+      "learning_rate": 1.7407020039325608e-05,
+      "loss": 4.1897,
       "step": 3100
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 16.475624084472656,
-      "learning_rate": 1.3785707126390699e-05,
-      "loss": 2.206,
+      "epoch": 0.39,
+      "grad_norm": 22.567594528198242,
+      "learning_rate": 1.740618332426892e-05,
+      "loss": 2.1251,
       "step": 3101
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 32.260719299316406,
-      "learning_rate": 1.378370251578631e-05,
-      "loss": 1.7881,
+      "epoch": 0.39,
+      "grad_norm": 10.604387283325195,
+      "learning_rate": 1.7405346609212235e-05,
+      "loss": 1.1912,
       "step": 3102
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 15.899197578430176,
-      "learning_rate": 1.3781697905181919e-05,
-      "loss": 1.865,
+      "epoch": 0.39,
+      "grad_norm": 23.615659713745117,
+      "learning_rate": 1.7404509894155545e-05,
+      "loss": 2.5832,
       "step": 3103
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 16.820289611816406,
-      "learning_rate": 1.3779693294577529e-05,
-      "loss": 1.9125,
+      "epoch": 0.39,
+      "grad_norm": 14.243099212646484,
+      "learning_rate": 1.740367317909886e-05,
+      "loss": 2.0185,
       "step": 3104
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 24.38895606994629,
-      "learning_rate": 1.3777688683973137e-05,
-      "loss": 1.5985,
+      "epoch": 0.39,
+      "grad_norm": 25.76270866394043,
+      "learning_rate": 1.7402836464042173e-05,
+      "loss": 1.8266,
       "step": 3105
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 34.085914611816406,
-      "learning_rate": 1.377568407336875e-05,
-      "loss": 2.2492,
+      "epoch": 0.39,
+      "grad_norm": 14.212577819824219,
+      "learning_rate": 1.7401999748985486e-05,
+      "loss": 1.8881,
       "step": 3106
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 14.358930587768555,
-      "learning_rate": 1.377367946276436e-05,
-      "loss": 1.9382,
+      "epoch": 0.39,
+      "grad_norm": 11.802764892578125,
+      "learning_rate": 1.7401163033928797e-05,
+      "loss": 2.3556,
       "step": 3107
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 25.895401000976562,
-      "learning_rate": 1.377167485215997e-05,
-      "loss": 1.1291,
+      "epoch": 0.39,
+      "grad_norm": 8.763890266418457,
+      "learning_rate": 1.740032631887211e-05,
+      "loss": 2.9164,
       "step": 3108
     },
     {
-      "epoch": 0.93,
-      "grad_norm": 17.190685272216797,
-      "learning_rate": 1.376967024155558e-05,
-      "loss": 1.4162,
+      "epoch": 0.39,
+      "grad_norm": 16.387601852416992,
+      "learning_rate": 1.7399489603815424e-05,
+      "loss": 2.4992,
       "step": 3109
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 13.299299240112305,
-      "learning_rate": 1.3767665630951188e-05,
-      "loss": 1.6711,
+      "epoch": 0.39,
+      "grad_norm": 18.163043975830078,
+      "learning_rate": 1.7398652888758734e-05,
+      "loss": 1.8894,
       "step": 3110
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 14.426549911499023,
-      "learning_rate": 1.37656610203468e-05,
-      "loss": 1.9699,
+      "epoch": 0.39,
+      "grad_norm": 13.54871654510498,
+      "learning_rate": 1.7397816173702048e-05,
+      "loss": 1.6424,
       "step": 3111
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 25.35927963256836,
-      "learning_rate": 1.376365640974241e-05,
-      "loss": 2.6309,
+      "epoch": 0.39,
+      "grad_norm": 18.142717361450195,
+      "learning_rate": 1.739697945864536e-05,
+      "loss": 1.5697,
       "step": 3112
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 9.705031394958496,
-      "learning_rate": 1.3761651799138018e-05,
-      "loss": 2.4938,
+      "epoch": 0.39,
+      "grad_norm": 9.927594184875488,
+      "learning_rate": 1.7396142743588672e-05,
+      "loss": 1.808,
       "step": 3113
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 12.233379364013672,
-      "learning_rate": 1.375964718853363e-05,
-      "loss": 1.8221,
+      "epoch": 0.39,
+      "grad_norm": 12.471274375915527,
+      "learning_rate": 1.7395306028531985e-05,
+      "loss": 2.2684,
       "step": 3114
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 15.394095420837402,
-      "learning_rate": 1.3757642577929238e-05,
-      "loss": 1.7775,
+      "epoch": 0.39,
+      "grad_norm": 45.707679748535156,
+      "learning_rate": 1.73944693134753e-05,
+      "loss": 2.5269,
       "step": 3115
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 13.07430362701416,
-      "learning_rate": 1.3755637967324848e-05,
-      "loss": 1.0702,
+      "epoch": 0.39,
+      "grad_norm": 10.306867599487305,
+      "learning_rate": 1.739363259841861e-05,
+      "loss": 2.6692,
       "step": 3116
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 12.21866226196289,
-      "learning_rate": 1.3753633356720457e-05,
-      "loss": 1.6868,
+      "epoch": 0.39,
+      "grad_norm": 9.580382347106934,
+      "learning_rate": 1.7392795883361923e-05,
+      "loss": 1.0849,
       "step": 3117
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 10.357650756835938,
-      "learning_rate": 1.3751628746116069e-05,
-      "loss": 2.2551,
+      "epoch": 0.39,
+      "grad_norm": 9.687782287597656,
+      "learning_rate": 1.7391959168305233e-05,
+      "loss": 1.3304,
       "step": 3118
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 18.206022262573242,
-      "learning_rate": 1.3749624135511679e-05,
-      "loss": 1.8297,
+      "epoch": 0.39,
+      "grad_norm": 13.099113464355469,
+      "learning_rate": 1.7391122453248547e-05,
+      "loss": 1.5095,
       "step": 3119
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 15.580822944641113,
-      "learning_rate": 1.3747619524907287e-05,
-      "loss": 1.9925,
-      "step": 3120
-    },
-    {
-      "epoch": 0.94,
-      "eval_loss": 0.2486942857503891,
-      "eval_runtime": 43.5686,
-      "eval_samples_per_second": 33.946,
-      "eval_steps_per_second": 33.946,
+      "epoch": 0.39,
+      "grad_norm": 25.176481246948242,
+      "learning_rate": 1.739028573819186e-05,
+      "loss": 1.6254,
       "step": 3120
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 10.770172119140625,
-      "learning_rate": 1.3745614914302899e-05,
-      "loss": 2.3822,
+      "epoch": 0.39,
+      "grad_norm": 8.78892707824707,
+      "learning_rate": 1.738944902313517e-05,
+      "loss": 3.0991,
       "step": 3121
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 25.76287078857422,
-      "learning_rate": 1.3743610303698507e-05,
-      "loss": 2.8554,
+      "epoch": 0.39,
+      "grad_norm": 12.70933723449707,
+      "learning_rate": 1.7388612308078484e-05,
+      "loss": 2.7174,
       "step": 3122
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 11.431665420532227,
-      "learning_rate": 1.3741605693094117e-05,
-      "loss": 2.0408,
+      "epoch": 0.39,
+      "grad_norm": 21.841047286987305,
+      "learning_rate": 1.7387775593021798e-05,
+      "loss": 2.8923,
       "step": 3123
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 12.141244888305664,
-      "learning_rate": 1.3739601082489729e-05,
-      "loss": 1.5182,
+      "epoch": 0.39,
+      "grad_norm": 9.589460372924805,
+      "learning_rate": 1.7386938877965108e-05,
+      "loss": 2.7387,
       "step": 3124
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 12.981203079223633,
-      "learning_rate": 1.3737596471885337e-05,
-      "loss": 1.3628,
+      "epoch": 0.39,
+      "grad_norm": 30.16423225402832,
+      "learning_rate": 1.7386102162908422e-05,
+      "loss": 1.0212,
       "step": 3125
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 12.477705955505371,
-      "learning_rate": 1.3735591861280948e-05,
-      "loss": 1.9388,
+      "epoch": 0.39,
+      "grad_norm": 19.877349853515625,
+      "learning_rate": 1.7385265447851736e-05,
+      "loss": 2.6417,
       "step": 3126
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 21.606971740722656,
-      "learning_rate": 1.3733587250676556e-05,
-      "loss": 1.7253,
+      "epoch": 0.39,
+      "grad_norm": 10.562746047973633,
+      "learning_rate": 1.738442873279505e-05,
+      "loss": 0.9116,
       "step": 3127
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 15.517189025878906,
-      "learning_rate": 1.3731582640072168e-05,
-      "loss": 2.0146,
+      "epoch": 0.39,
+      "grad_norm": 13.513136863708496,
+      "learning_rate": 1.738359201773836e-05,
+      "loss": 1.7079,
       "step": 3128
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 30.338869094848633,
-      "learning_rate": 1.3729578029467776e-05,
-      "loss": 1.8816,
+      "epoch": 0.39,
+      "grad_norm": 12.24991512298584,
+      "learning_rate": 1.7382755302681673e-05,
+      "loss": 1.3581,
       "step": 3129
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 34.77288818359375,
-      "learning_rate": 1.3727573418863386e-05,
-      "loss": 3.4102,
+      "epoch": 0.39,
+      "grad_norm": 23.374011993408203,
+      "learning_rate": 1.7381918587624987e-05,
+      "loss": 2.057,
       "step": 3130
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 29.494319915771484,
-      "learning_rate": 1.3725568808258998e-05,
-      "loss": 1.3876,
+      "epoch": 0.39,
+      "grad_norm": 14.807626724243164,
+      "learning_rate": 1.7381081872568297e-05,
+      "loss": 2.0256,
       "step": 3131
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 15.238880157470703,
-      "learning_rate": 1.3723564197654606e-05,
-      "loss": 2.0381,
+      "epoch": 0.39,
+      "grad_norm": 9.561110496520996,
+      "learning_rate": 1.738024515751161e-05,
+      "loss": 1.4404,
       "step": 3132
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 13.510786056518555,
-      "learning_rate": 1.3721559587050216e-05,
-      "loss": 2.5507,
+      "epoch": 0.39,
+      "grad_norm": 20.877609252929688,
+      "learning_rate": 1.7379408442454924e-05,
+      "loss": 2.6945,
       "step": 3133
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 24.444250106811523,
-      "learning_rate": 1.3719554976445827e-05,
-      "loss": 1.7889,
+      "epoch": 0.39,
+      "grad_norm": 14.954484939575195,
+      "learning_rate": 1.7378571727398238e-05,
+      "loss": 3.6158,
       "step": 3134
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 12.126347541809082,
-      "learning_rate": 1.3717550365841437e-05,
-      "loss": 1.2326,
+      "epoch": 0.39,
+      "grad_norm": 17.07823944091797,
+      "learning_rate": 1.7377735012341548e-05,
+      "loss": 1.9067,
       "step": 3135
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 10.488152503967285,
-      "learning_rate": 1.3715545755237045e-05,
-      "loss": 1.2664,
+      "epoch": 0.39,
+      "grad_norm": 18.759994506835938,
+      "learning_rate": 1.7376898297284862e-05,
+      "loss": 1.8789,
       "step": 3136
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 30.025480270385742,
-      "learning_rate": 1.3713541144632657e-05,
-      "loss": 1.5859,
+      "epoch": 0.39,
+      "grad_norm": 13.437284469604492,
+      "learning_rate": 1.7376061582228176e-05,
+      "loss": 3.1343,
       "step": 3137
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 13.164788246154785,
-      "learning_rate": 1.3711536534028267e-05,
-      "loss": 2.1662,
+      "epoch": 0.39,
+      "grad_norm": 10.817352294921875,
+      "learning_rate": 1.7375224867171486e-05,
+      "loss": 2.9662,
       "step": 3138
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 18.30267333984375,
-      "learning_rate": 1.3709531923423875e-05,
-      "loss": 2.1693,
+      "epoch": 0.39,
+      "grad_norm": 15.099756240844727,
+      "learning_rate": 1.73743881521148e-05,
+      "loss": 1.8025,
       "step": 3139
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 16.636390686035156,
-      "learning_rate": 1.3707527312819487e-05,
-      "loss": 2.3369,
+      "epoch": 0.39,
+      "grad_norm": 31.67547607421875,
+      "learning_rate": 1.7373551437058113e-05,
+      "loss": 1.9846,
       "step": 3140
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 13.153757095336914,
-      "learning_rate": 1.3705522702215095e-05,
-      "loss": 2.3091,
+      "epoch": 0.39,
+      "grad_norm": 7.465260028839111,
+      "learning_rate": 1.7372714722001423e-05,
+      "loss": 0.8328,
       "step": 3141
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 30.390827178955078,
-      "learning_rate": 1.3703518091610706e-05,
-      "loss": 1.786,
+      "epoch": 0.39,
+      "grad_norm": 8.22437858581543,
+      "learning_rate": 1.7371878006944737e-05,
+      "loss": 0.545,
       "step": 3142
     },
     {
-      "epoch": 0.94,
-      "grad_norm": 41.70283889770508,
-      "learning_rate": 1.3701513481006317e-05,
-      "loss": 3.5995,
+      "epoch": 0.39,
+      "grad_norm": 20.757461547851562,
+      "learning_rate": 1.7371041291888047e-05,
+      "loss": 3.8331,
       "step": 3143
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 32.86235809326172,
-      "learning_rate": 1.3699508870401926e-05,
-      "loss": 2.4132,
+      "epoch": 0.39,
+      "grad_norm": 23.26932716369629,
+      "learning_rate": 1.737020457683136e-05,
+      "loss": 2.3774,
       "step": 3144
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 14.464203834533691,
-      "learning_rate": 1.3697504259797536e-05,
-      "loss": 2.0061,
+      "epoch": 0.39,
+      "grad_norm": 11.285542488098145,
+      "learning_rate": 1.7369367861774675e-05,
+      "loss": 1.4571,
       "step": 3145
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 11.750015258789062,
-      "learning_rate": 1.3695499649193144e-05,
-      "loss": 2.1382,
+      "epoch": 0.39,
+      "grad_norm": 38.73393630981445,
+      "learning_rate": 1.7368531146717985e-05,
+      "loss": 1.7928,
       "step": 3146
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 13.602632522583008,
-      "learning_rate": 1.3693495038588756e-05,
-      "loss": 2.4135,
+      "epoch": 0.39,
+      "grad_norm": 20.771852493286133,
+      "learning_rate": 1.73676944316613e-05,
+      "loss": 1.804,
       "step": 3147
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 15.727476119995117,
-      "learning_rate": 1.3691490427984364e-05,
-      "loss": 2.9052,
+      "epoch": 0.4,
+      "grad_norm": 24.225017547607422,
+      "learning_rate": 1.7366857716604612e-05,
+      "loss": 3.4957,
       "step": 3148
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 17.43680763244629,
-      "learning_rate": 1.3689485817379974e-05,
-      "loss": 1.3355,
+      "epoch": 0.4,
+      "grad_norm": 7.701132774353027,
+      "learning_rate": 1.7366021001547922e-05,
+      "loss": 1.9459,
       "step": 3149
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 14.329429626464844,
-      "learning_rate": 1.3687481206775586e-05,
-      "loss": 3.1391,
+      "epoch": 0.4,
+      "grad_norm": 14.99876594543457,
+      "learning_rate": 1.7365184286491236e-05,
+      "loss": 1.73,
       "step": 3150
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 25.729965209960938,
-      "learning_rate": 1.3685476596171195e-05,
-      "loss": 1.6923,
+      "epoch": 0.4,
+      "grad_norm": 20.34157943725586,
+      "learning_rate": 1.736434757143455e-05,
+      "loss": 1.83,
       "step": 3151
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 95.43783569335938,
-      "learning_rate": 1.3683471985566805e-05,
-      "loss": 2.0273,
+      "epoch": 0.4,
+      "grad_norm": 24.359678268432617,
+      "learning_rate": 1.736351085637786e-05,
+      "loss": 2.2814,
       "step": 3152
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 17.75636100769043,
-      "learning_rate": 1.3681467374962413e-05,
-      "loss": 1.5341,
+      "epoch": 0.4,
+      "grad_norm": 8.147969245910645,
+      "learning_rate": 1.7362674141321174e-05,
+      "loss": 1.2628,
       "step": 3153
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 19.18214988708496,
-      "learning_rate": 1.3679462764358025e-05,
-      "loss": 1.8777,
+      "epoch": 0.4,
+      "grad_norm": 15.603983879089355,
+      "learning_rate": 1.7361837426264487e-05,
+      "loss": 2.999,
       "step": 3154
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 15.892659187316895,
-      "learning_rate": 1.3677458153753635e-05,
-      "loss": 1.9435,
+      "epoch": 0.4,
+      "grad_norm": 7.318797588348389,
+      "learning_rate": 1.73610007112078e-05,
+      "loss": 1.5075,
       "step": 3155
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 20.043859481811523,
-      "learning_rate": 1.3675453543149243e-05,
-      "loss": 2.1202,
+      "epoch": 0.4,
+      "grad_norm": 27.904884338378906,
+      "learning_rate": 1.736016399615111e-05,
+      "loss": 2.882,
       "step": 3156
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 13.109265327453613,
-      "learning_rate": 1.3673448932544855e-05,
-      "loss": 2.429,
+      "epoch": 0.4,
+      "grad_norm": 14.651124954223633,
+      "learning_rate": 1.7359327281094425e-05,
+      "loss": 1.8119,
       "step": 3157
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 15.493414878845215,
-      "learning_rate": 1.3671444321940463e-05,
-      "loss": 1.768,
+      "epoch": 0.4,
+      "grad_norm": 15.951963424682617,
+      "learning_rate": 1.735849056603774e-05,
+      "loss": 1.1661,
       "step": 3158
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 36.695735931396484,
-      "learning_rate": 1.3669439711336074e-05,
-      "loss": 2.0781,
+      "epoch": 0.4,
+      "grad_norm": 23.046314239501953,
+      "learning_rate": 1.735765385098105e-05,
+      "loss": 2.3287,
       "step": 3159
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 14.086923599243164,
-      "learning_rate": 1.3667435100731684e-05,
-      "loss": 0.8197,
+      "epoch": 0.4,
+      "grad_norm": 18.682222366333008,
+      "learning_rate": 1.7356817135924362e-05,
+      "loss": 2.2169,
       "step": 3160
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 13.209424018859863,
-      "learning_rate": 1.3665430490127294e-05,
-      "loss": 1.9733,
+      "epoch": 0.4,
+      "grad_norm": 11.814753532409668,
+      "learning_rate": 1.7355980420867676e-05,
+      "loss": 2.3565,
       "step": 3161
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 17.077037811279297,
-      "learning_rate": 1.3663425879522905e-05,
-      "loss": 2.5923,
+      "epoch": 0.4,
+      "grad_norm": 17.43982696533203,
+      "learning_rate": 1.735514370581099e-05,
+      "loss": 1.3372,
       "step": 3162
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 49.46474075317383,
-      "learning_rate": 1.3661421268918514e-05,
-      "loss": 2.0692,
+      "epoch": 0.4,
+      "grad_norm": 10.957281112670898,
+      "learning_rate": 1.73543069907543e-05,
+      "loss": 1.3389,
       "step": 3163
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 14.314860343933105,
-      "learning_rate": 1.3659416658314124e-05,
-      "loss": 2.8229,
+      "epoch": 0.4,
+      "grad_norm": 6.849418640136719,
+      "learning_rate": 1.7353470275697614e-05,
+      "loss": 1.378,
       "step": 3164
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 8.489893913269043,
-      "learning_rate": 1.3657412047709732e-05,
-      "loss": 1.3168,
+      "epoch": 0.4,
+      "grad_norm": 13.42451286315918,
+      "learning_rate": 1.7352633560640927e-05,
+      "loss": 2.8483,
       "step": 3165
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 32.04702377319336,
-      "learning_rate": 1.3655407437105344e-05,
-      "loss": 1.6409,
+      "epoch": 0.4,
+      "grad_norm": 8.09776496887207,
+      "learning_rate": 1.7351796845584238e-05,
+      "loss": 0.6772,
       "step": 3166
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 16.088096618652344,
-      "learning_rate": 1.3653402826500954e-05,
-      "loss": 1.1645,
+      "epoch": 0.4,
+      "grad_norm": 7.820061683654785,
+      "learning_rate": 1.735096013052755e-05,
+      "loss": 0.4904,
       "step": 3167
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 12.829047203063965,
-      "learning_rate": 1.3651398215896563e-05,
-      "loss": 1.3659,
+      "epoch": 0.4,
+      "grad_norm": 15.628246307373047,
+      "learning_rate": 1.735012341547086e-05,
+      "loss": 2.5608,
       "step": 3168
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 19.811403274536133,
-      "learning_rate": 1.3649393605292174e-05,
-      "loss": 2.6096,
+      "epoch": 0.4,
+      "grad_norm": 13.724005699157715,
+      "learning_rate": 1.7349286700414175e-05,
+      "loss": 2.8304,
       "step": 3169
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 15.218836784362793,
-      "learning_rate": 1.3647388994687783e-05,
-      "loss": 1.4327,
+      "epoch": 0.4,
+      "grad_norm": 10.125463485717773,
+      "learning_rate": 1.734844998535749e-05,
+      "loss": 2.0583,
       "step": 3170
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 11.305002212524414,
-      "learning_rate": 1.3645384384083393e-05,
-      "loss": 1.1141,
+      "epoch": 0.4,
+      "grad_norm": 9.629894256591797,
+      "learning_rate": 1.73476132703008e-05,
+      "loss": 1.896,
       "step": 3171
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 12.872888565063477,
-      "learning_rate": 1.3643379773479001e-05,
-      "loss": 1.6382,
+      "epoch": 0.4,
+      "grad_norm": 7.258272647857666,
+      "learning_rate": 1.7346776555244113e-05,
+      "loss": 0.7141,
       "step": 3172
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 18.41777801513672,
-      "learning_rate": 1.3641375162874613e-05,
-      "loss": 2.1711,
+      "epoch": 0.4,
+      "grad_norm": 15.432266235351562,
+      "learning_rate": 1.7345939840187426e-05,
+      "loss": 2.3333,
       "step": 3173
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 14.97694206237793,
-      "learning_rate": 1.3639370552270223e-05,
-      "loss": 2.1335,
+      "epoch": 0.4,
+      "grad_norm": 17.209692001342773,
+      "learning_rate": 1.7345103125130737e-05,
+      "loss": 1.8321,
       "step": 3174
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 17.031770706176758,
-      "learning_rate": 1.3637365941665832e-05,
-      "loss": 2.1181,
+      "epoch": 0.4,
+      "grad_norm": 28.579994201660156,
+      "learning_rate": 1.734426641007405e-05,
+      "loss": 2.4984,
       "step": 3175
     },
     {
-      "epoch": 0.95,
-      "grad_norm": 15.743291854858398,
-      "learning_rate": 1.3635361331061443e-05,
-      "loss": 2.4674,
+      "epoch": 0.4,
+      "grad_norm": 12.683948516845703,
+      "learning_rate": 1.7343429695017364e-05,
+      "loss": 0.7704,
       "step": 3176
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 9.414546966552734,
-      "learning_rate": 1.3633356720457052e-05,
-      "loss": 2.1942,
+      "epoch": 0.4,
+      "grad_norm": 15.408038139343262,
+      "learning_rate": 1.7342592979960674e-05,
+      "loss": 3.2682,
       "step": 3177
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 10.637990951538086,
-      "learning_rate": 1.3631352109852662e-05,
-      "loss": 1.5859,
+      "epoch": 0.4,
+      "grad_norm": 8.45233154296875,
+      "learning_rate": 1.7341756264903988e-05,
+      "loss": 1.3961,
       "step": 3178
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 13.900732040405273,
-      "learning_rate": 1.3629347499248272e-05,
-      "loss": 1.6751,
+      "epoch": 0.4,
+      "grad_norm": 35.033164978027344,
+      "learning_rate": 1.73409195498473e-05,
+      "loss": 1.7532,
       "step": 3179
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 20.666290283203125,
-      "learning_rate": 1.3627342888643882e-05,
-      "loss": 2.6414,
+      "epoch": 0.4,
+      "grad_norm": 23.8881893157959,
+      "learning_rate": 1.734008283479061e-05,
+      "loss": 2.8757,
       "step": 3180
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 35.015647888183594,
-      "learning_rate": 1.3625338278039492e-05,
-      "loss": 2.7838,
+      "epoch": 0.4,
+      "grad_norm": 11.64614200592041,
+      "learning_rate": 1.7339246119733925e-05,
+      "loss": 2.8886,
       "step": 3181
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 15.191834449768066,
-      "learning_rate": 1.3623333667435102e-05,
-      "loss": 1.831,
+      "epoch": 0.4,
+      "grad_norm": 23.38692283630371,
+      "learning_rate": 1.733840940467724e-05,
+      "loss": 4.9053,
       "step": 3182
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 34.37846755981445,
-      "learning_rate": 1.3621329056830712e-05,
-      "loss": 2.7531,
+      "epoch": 0.4,
+      "grad_norm": 9.945205688476562,
+      "learning_rate": 1.7337572689620553e-05,
+      "loss": 0.6558,
       "step": 3183
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 12.335997581481934,
-      "learning_rate": 1.361932444622632e-05,
-      "loss": 1.0529,
+      "epoch": 0.4,
+      "grad_norm": 8.835298538208008,
+      "learning_rate": 1.7336735974563863e-05,
+      "loss": 2.348,
       "step": 3184
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 25.120954513549805,
-      "learning_rate": 1.3617319835621932e-05,
-      "loss": 2.2635,
+      "epoch": 0.4,
+      "grad_norm": 15.17750072479248,
+      "learning_rate": 1.7335899259507177e-05,
+      "loss": 1.2624,
       "step": 3185
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 11.803996086120605,
-      "learning_rate": 1.3615315225017542e-05,
-      "loss": 1.8602,
+      "epoch": 0.4,
+      "grad_norm": 14.124608993530273,
+      "learning_rate": 1.733506254445049e-05,
+      "loss": 1.0089,
       "step": 3186
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 14.749687194824219,
-      "learning_rate": 1.361331061441315e-05,
-      "loss": 1.0415,
+      "epoch": 0.4,
+      "grad_norm": 14.128517150878906,
+      "learning_rate": 1.73342258293938e-05,
+      "loss": 3.2298,
       "step": 3187
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 15.90971851348877,
-      "learning_rate": 1.3611306003808763e-05,
-      "loss": 1.9429,
+      "epoch": 0.4,
+      "grad_norm": 10.97527027130127,
+      "learning_rate": 1.7333389114337114e-05,
+      "loss": 3.3839,
       "step": 3188
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 42.85434341430664,
-      "learning_rate": 1.3609301393204371e-05,
-      "loss": 2.2651,
+      "epoch": 0.4,
+      "grad_norm": 8.833992004394531,
+      "learning_rate": 1.7332552399280428e-05,
+      "loss": 1.568,
       "step": 3189
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 22.296979904174805,
-      "learning_rate": 1.3607296782599981e-05,
-      "loss": 2.2514,
+      "epoch": 0.4,
+      "grad_norm": 12.724773406982422,
+      "learning_rate": 1.733171568422374e-05,
+      "loss": 2.8502,
       "step": 3190
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 16.979684829711914,
-      "learning_rate": 1.360529217199559e-05,
-      "loss": 1.9144,
+      "epoch": 0.4,
+      "grad_norm": 14.78669548034668,
+      "learning_rate": 1.733087896916705e-05,
+      "loss": 3.0281,
       "step": 3191
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 51.25554656982422,
-      "learning_rate": 1.3603287561391201e-05,
-      "loss": 1.4266,
+      "epoch": 0.4,
+      "grad_norm": 8.72022819519043,
+      "learning_rate": 1.7330042254110365e-05,
+      "loss": 3.1723,
       "step": 3192
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 32.474002838134766,
-      "learning_rate": 1.3601282950786811e-05,
-      "loss": 2.3617,
+      "epoch": 0.4,
+      "grad_norm": 12.51976490020752,
+      "learning_rate": 1.732920553905368e-05,
+      "loss": 1.8388,
       "step": 3193
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 12.489779472351074,
-      "learning_rate": 1.359927834018242e-05,
-      "loss": 1.4284,
+      "epoch": 0.4,
+      "grad_norm": 27.790912628173828,
+      "learning_rate": 1.732836882399699e-05,
+      "loss": 2.4745,
       "step": 3194
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 14.795775413513184,
-      "learning_rate": 1.3597273729578032e-05,
-      "loss": 0.9304,
+      "epoch": 0.4,
+      "grad_norm": 12.896369934082031,
+      "learning_rate": 1.7327532108940303e-05,
+      "loss": 2.2489,
       "step": 3195
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 35.45854568481445,
-      "learning_rate": 1.359526911897364e-05,
-      "loss": 3.7488,
+      "epoch": 0.4,
+      "grad_norm": 23.709392547607422,
+      "learning_rate": 1.7326695393883613e-05,
+      "loss": 2.3466,
       "step": 3196
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 6.524545669555664,
-      "learning_rate": 1.359326450836925e-05,
-      "loss": 1.08,
+      "epoch": 0.4,
+      "grad_norm": 13.660783767700195,
+      "learning_rate": 1.7325858678826927e-05,
+      "loss": 1.4347,
       "step": 3197
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 13.13879680633545,
-      "learning_rate": 1.3591259897764862e-05,
-      "loss": 2.1342,
+      "epoch": 0.4,
+      "grad_norm": 10.898387908935547,
+      "learning_rate": 1.732502196377024e-05,
+      "loss": 2.9651,
       "step": 3198
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 8.252395629882812,
-      "learning_rate": 1.358925528716047e-05,
-      "loss": 1.6381,
+      "epoch": 0.4,
+      "grad_norm": 15.468527793884277,
+      "learning_rate": 1.732418524871355e-05,
+      "loss": 1.7693,
       "step": 3199
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 12.134532928466797,
-      "learning_rate": 1.358725067655608e-05,
-      "loss": 1.7298,
+      "epoch": 0.4,
+      "grad_norm": 12.056096076965332,
+      "learning_rate": 1.7323348533656864e-05,
+      "loss": 3.39,
       "step": 3200
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 17.264564514160156,
-      "learning_rate": 1.3585246065951689e-05,
-      "loss": 1.6213,
+      "epoch": 0.4,
+      "eval_loss": 0.15101495385169983,
+      "eval_runtime": 94.045,
+      "eval_samples_per_second": 37.663,
+      "eval_steps_per_second": 37.663,
+      "step": 3200
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 12.105559349060059,
+      "learning_rate": 1.7322511818600175e-05,
+      "loss": 3.396,
       "step": 3201
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 17.299379348754883,
-      "learning_rate": 1.35832414553473e-05,
-      "loss": 1.9645,
+      "epoch": 0.4,
+      "grad_norm": 7.145201683044434,
+      "learning_rate": 1.7321675103543488e-05,
+      "loss": 1.2291,
       "step": 3202
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 9.386181831359863,
-      "learning_rate": 1.3581236844742909e-05,
-      "loss": 1.0434,
+      "epoch": 0.4,
+      "grad_norm": 10.953343391418457,
+      "learning_rate": 1.7320838388486802e-05,
+      "loss": 1.6902,
       "step": 3203
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 10.785475730895996,
-      "learning_rate": 1.3579232234138519e-05,
-      "loss": 1.387,
+      "epoch": 0.4,
+      "grad_norm": 15.755420684814453,
+      "learning_rate": 1.7320001673430116e-05,
+      "loss": 2.4279,
       "step": 3204
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 18.141578674316406,
-      "learning_rate": 1.357722762353413e-05,
-      "loss": 1.816,
+      "epoch": 0.4,
+      "grad_norm": 14.094854354858398,
+      "learning_rate": 1.7319164958373426e-05,
+      "loss": 2.5785,
       "step": 3205
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 10.100055694580078,
-      "learning_rate": 1.3575223012929739e-05,
-      "loss": 2.0753,
+      "epoch": 0.4,
+      "grad_norm": 14.587738990783691,
+      "learning_rate": 1.731832824331674e-05,
+      "loss": 1.3048,
       "step": 3206
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 13.844127655029297,
-      "learning_rate": 1.3573218402325349e-05,
-      "loss": 1.4975,
+      "epoch": 0.4,
+      "grad_norm": 23.59645652770996,
+      "learning_rate": 1.7317491528260053e-05,
+      "loss": 2.2371,
       "step": 3207
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 13.211896896362305,
-      "learning_rate": 1.357121379172096e-05,
-      "loss": 2.011,
+      "epoch": 0.4,
+      "grad_norm": 15.983088493347168,
+      "learning_rate": 1.7316654813203363e-05,
+      "loss": 1.6247,
       "step": 3208
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 16.434125900268555,
-      "learning_rate": 1.356920918111657e-05,
-      "loss": 1.3007,
+      "epoch": 0.4,
+      "grad_norm": 9.358094215393066,
+      "learning_rate": 1.7315818098146677e-05,
+      "loss": 2.544,
       "step": 3209
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 25.20293426513672,
-      "learning_rate": 1.356720457051218e-05,
-      "loss": 1.4808,
+      "epoch": 0.4,
+      "grad_norm": 18.30937957763672,
+      "learning_rate": 1.731498138308999e-05,
+      "loss": 2.3604,
       "step": 3210
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 17.422204971313477,
-      "learning_rate": 1.356519995990779e-05,
-      "loss": 2.0575,
+      "epoch": 0.4,
+      "grad_norm": 9.7167329788208,
+      "learning_rate": 1.7314144668033304e-05,
+      "loss": 2.4223,
       "step": 3211
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 22.5146484375,
-      "learning_rate": 1.35631953493034e-05,
-      "loss": 2.0943,
+      "epoch": 0.4,
+      "grad_norm": 21.414541244506836,
+      "learning_rate": 1.7313307952976615e-05,
+      "loss": 2.3931,
       "step": 3212
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 46.014678955078125,
-      "learning_rate": 1.3561190738699008e-05,
-      "loss": 2.7204,
+      "epoch": 0.4,
+      "grad_norm": 12.109614372253418,
+      "learning_rate": 1.7312471237919928e-05,
+      "loss": 0.2461,
       "step": 3213
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 16.03790283203125,
-      "learning_rate": 1.355918612809462e-05,
-      "loss": 2.5307,
+      "epoch": 0.4,
+      "grad_norm": 22.76210594177246,
+      "learning_rate": 1.7311634522863242e-05,
+      "loss": 3.2436,
       "step": 3214
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 17.003280639648438,
-      "learning_rate": 1.3557181517490228e-05,
-      "loss": 2.766,
+      "epoch": 0.4,
+      "grad_norm": 7.287879467010498,
+      "learning_rate": 1.7310797807806552e-05,
+      "loss": 0.9333,
       "step": 3215
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 21.383115768432617,
-      "learning_rate": 1.3555176906885838e-05,
-      "loss": 1.8888,
+      "epoch": 0.4,
+      "grad_norm": 10.937824249267578,
+      "learning_rate": 1.7309961092749866e-05,
+      "loss": 2.7631,
       "step": 3216
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 29.028945922851562,
-      "learning_rate": 1.355317229628145e-05,
-      "loss": 1.2596,
+      "epoch": 0.4,
+      "grad_norm": 18.63129425048828,
+      "learning_rate": 1.730912437769318e-05,
+      "loss": 3.2587,
       "step": 3217
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 14.80164623260498,
-      "learning_rate": 1.3551167685677058e-05,
-      "loss": 2.1869,
+      "epoch": 0.4,
+      "grad_norm": 12.023300170898438,
+      "learning_rate": 1.7308287662636493e-05,
+      "loss": 1.2662,
       "step": 3218
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 35.633094787597656,
-      "learning_rate": 1.3549163075072668e-05,
-      "loss": 3.2196,
+      "epoch": 0.4,
+      "grad_norm": 8.753905296325684,
+      "learning_rate": 1.7307450947579803e-05,
+      "loss": 1.474,
       "step": 3219
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 17.463098526000977,
-      "learning_rate": 1.3547158464468277e-05,
-      "loss": 2.077,
+      "epoch": 0.4,
+      "grad_norm": 26.231277465820312,
+      "learning_rate": 1.7306614232523117e-05,
+      "loss": 2.2326,
       "step": 3220
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 15.604183197021484,
-      "learning_rate": 1.3545153853863889e-05,
-      "loss": 1.6204,
+      "epoch": 0.4,
+      "grad_norm": 19.625652313232422,
+      "learning_rate": 1.7305777517466427e-05,
+      "loss": 2.0372,
       "step": 3221
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 28.1143798828125,
-      "learning_rate": 1.3543149243259497e-05,
-      "loss": 2.9363,
+      "epoch": 0.4,
+      "grad_norm": 23.23263931274414,
+      "learning_rate": 1.730494080240974e-05,
+      "loss": 3.344,
       "step": 3222
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 18.119604110717773,
-      "learning_rate": 1.3541144632655107e-05,
-      "loss": 1.3793,
+      "epoch": 0.4,
+      "grad_norm": 14.629230499267578,
+      "learning_rate": 1.7304104087353055e-05,
+      "loss": 2.0983,
       "step": 3223
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 16.877845764160156,
-      "learning_rate": 1.3539140022050719e-05,
-      "loss": 2.0204,
+      "epoch": 0.4,
+      "grad_norm": 16.69709014892578,
+      "learning_rate": 1.7303267372296365e-05,
+      "loss": 1.9195,
       "step": 3224
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 20.121807098388672,
-      "learning_rate": 1.3537135411446327e-05,
-      "loss": 2.4456,
+      "epoch": 0.4,
+      "grad_norm": 14.589938163757324,
+      "learning_rate": 1.730243065723968e-05,
+      "loss": 2.1431,
       "step": 3225
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 15.562324523925781,
-      "learning_rate": 1.3535130800841937e-05,
-      "loss": 1.8663,
+      "epoch": 0.4,
+      "grad_norm": 25.879173278808594,
+      "learning_rate": 1.7301593942182992e-05,
+      "loss": 3.4365,
       "step": 3226
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 16.265233993530273,
-      "learning_rate": 1.3533126190237547e-05,
-      "loss": 1.7686,
+      "epoch": 0.4,
+      "grad_norm": 16.38176727294922,
+      "learning_rate": 1.7300757227126302e-05,
+      "loss": 2.1582,
       "step": 3227
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 24.303163528442383,
-      "learning_rate": 1.3531121579633158e-05,
-      "loss": 2.5792,
+      "epoch": 0.41,
+      "grad_norm": 8.85407829284668,
+      "learning_rate": 1.7299920512069616e-05,
+      "loss": 1.8366,
       "step": 3228
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 19.46893310546875,
-      "learning_rate": 1.3529116969028768e-05,
-      "loss": 1.8132,
+      "epoch": 0.41,
+      "grad_norm": 21.186267852783203,
+      "learning_rate": 1.7299083797012926e-05,
+      "loss": 0.9108,
       "step": 3229
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 39.980438232421875,
-      "learning_rate": 1.3527112358424378e-05,
-      "loss": 2.3918,
+      "epoch": 0.41,
+      "grad_norm": 38.015869140625,
+      "learning_rate": 1.729824708195624e-05,
+      "loss": 3.4078,
       "step": 3230
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 18.492206573486328,
-      "learning_rate": 1.3525107747819988e-05,
-      "loss": 1.6302,
+      "epoch": 0.41,
+      "grad_norm": 10.098426818847656,
+      "learning_rate": 1.7297410366899554e-05,
+      "loss": 1.6497,
       "step": 3231
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 11.828949928283691,
-      "learning_rate": 1.3523103137215596e-05,
-      "loss": 1.7631,
+      "epoch": 0.41,
+      "grad_norm": 5.503562927246094,
+      "learning_rate": 1.7296573651842867e-05,
+      "loss": 0.1918,
       "step": 3232
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 13.994562149047852,
-      "learning_rate": 1.3521098526611208e-05,
-      "loss": 1.9207,
+      "epoch": 0.41,
+      "grad_norm": 16.33560562133789,
+      "learning_rate": 1.7295736936786177e-05,
+      "loss": 1.874,
       "step": 3233
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 22.90715789794922,
-      "learning_rate": 1.3519093916006816e-05,
-      "loss": 3.6763,
+      "epoch": 0.41,
+      "grad_norm": 12.88305377960205,
+      "learning_rate": 1.729490022172949e-05,
+      "loss": 2.6811,
       "step": 3234
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 45.03823471069336,
-      "learning_rate": 1.3517089305402426e-05,
-      "loss": 2.1425,
+      "epoch": 0.41,
+      "grad_norm": 14.501304626464844,
+      "learning_rate": 1.7294063506672805e-05,
+      "loss": 0.6626,
       "step": 3235
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 12.193424224853516,
-      "learning_rate": 1.3515084694798038e-05,
-      "loss": 1.4887,
+      "epoch": 0.41,
+      "grad_norm": 23.03480339050293,
+      "learning_rate": 1.7293226791616115e-05,
+      "loss": 2.6398,
       "step": 3236
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 21.530868530273438,
-      "learning_rate": 1.3513080084193647e-05,
-      "loss": 2.9298,
+      "epoch": 0.41,
+      "grad_norm": 20.49584197998047,
+      "learning_rate": 1.729239007655943e-05,
+      "loss": 1.5989,
       "step": 3237
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 30.308269500732422,
-      "learning_rate": 1.3511075473589257e-05,
-      "loss": 2.4928,
+      "epoch": 0.41,
+      "grad_norm": 38.74869918823242,
+      "learning_rate": 1.7291553361502742e-05,
+      "loss": 2.7613,
       "step": 3238
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 11.13912582397461,
-      "learning_rate": 1.3509070862984865e-05,
-      "loss": 1.7033,
+      "epoch": 0.41,
+      "grad_norm": 19.7424259185791,
+      "learning_rate": 1.7290716646446056e-05,
+      "loss": 3.1999,
       "step": 3239
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 11.577181816101074,
-      "learning_rate": 1.3507066252380477e-05,
-      "loss": 1.4173,
-      "step": 3240
-    },
-    {
-      "epoch": 0.97,
-      "eval_loss": 0.22535113990306854,
-      "eval_runtime": 43.4738,
-      "eval_samples_per_second": 34.021,
-      "eval_steps_per_second": 34.021,
+      "epoch": 0.41,
+      "grad_norm": 13.532513618469238,
+      "learning_rate": 1.7289879931389366e-05,
+      "loss": 2.4183,
       "step": 3240
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 27.40611457824707,
-      "learning_rate": 1.3505061641776087e-05,
-      "loss": 1.9135,
+      "epoch": 0.41,
+      "grad_norm": 19.050241470336914,
+      "learning_rate": 1.728904321633268e-05,
+      "loss": 3.016,
       "step": 3241
     },
     {
-      "epoch": 0.97,
-      "grad_norm": 19.149837493896484,
-      "learning_rate": 1.3503057031171695e-05,
-      "loss": 1.7921,
+      "epoch": 0.41,
+      "grad_norm": 13.003695487976074,
+      "learning_rate": 1.7288206501275994e-05,
+      "loss": 2.5176,
       "step": 3242
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 15.974568367004395,
-      "learning_rate": 1.3501052420567307e-05,
-      "loss": 2.5266,
+      "epoch": 0.41,
+      "grad_norm": 7.656522750854492,
+      "learning_rate": 1.7287369786219304e-05,
+      "loss": 2.0928,
       "step": 3243
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 19.187986373901367,
-      "learning_rate": 1.3499047809962915e-05,
-      "loss": 1.9754,
+      "epoch": 0.41,
+      "grad_norm": 39.75627517700195,
+      "learning_rate": 1.7286533071162617e-05,
+      "loss": 1.4883,
       "step": 3244
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 10.218606948852539,
-      "learning_rate": 1.3497043199358526e-05,
-      "loss": 0.9435,
+      "epoch": 0.41,
+      "grad_norm": 11.915572166442871,
+      "learning_rate": 1.728569635610593e-05,
+      "loss": 1.5882,
       "step": 3245
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 16.653217315673828,
-      "learning_rate": 1.3495038588754134e-05,
-      "loss": 1.8804,
+      "epoch": 0.41,
+      "grad_norm": 11.25812816619873,
+      "learning_rate": 1.7284859641049245e-05,
+      "loss": 1.5589,
       "step": 3246
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 19.304424285888672,
-      "learning_rate": 1.3493033978149746e-05,
-      "loss": 2.038,
+      "epoch": 0.41,
+      "grad_norm": 12.67131519317627,
+      "learning_rate": 1.7284022925992555e-05,
+      "loss": 3.2036,
       "step": 3247
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 17.04471206665039,
-      "learning_rate": 1.3491029367545356e-05,
-      "loss": 1.9324,
+      "epoch": 0.41,
+      "grad_norm": 22.16834259033203,
+      "learning_rate": 1.728318621093587e-05,
+      "loss": 3.1681,
       "step": 3248
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 16.119550704956055,
-      "learning_rate": 1.3489024756940964e-05,
-      "loss": 2.1206,
+      "epoch": 0.41,
+      "grad_norm": 18.441844940185547,
+      "learning_rate": 1.728234949587918e-05,
+      "loss": 4.0314,
       "step": 3249
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 16.321861267089844,
-      "learning_rate": 1.3487020146336576e-05,
-      "loss": 2.3402,
+      "epoch": 0.41,
+      "grad_norm": 8.221848487854004,
+      "learning_rate": 1.7281512780822493e-05,
+      "loss": 1.0007,
       "step": 3250
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 30.2368221282959,
-      "learning_rate": 1.3485015535732184e-05,
-      "loss": 1.683,
+      "epoch": 0.41,
+      "grad_norm": 13.110788345336914,
+      "learning_rate": 1.7280676065765806e-05,
+      "loss": 1.006,
       "step": 3251
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 18.461002349853516,
-      "learning_rate": 1.3483010925127794e-05,
-      "loss": 2.1962,
+      "epoch": 0.41,
+      "grad_norm": 21.545730590820312,
+      "learning_rate": 1.7279839350709116e-05,
+      "loss": 2.5467,
       "step": 3252
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 20.721721649169922,
-      "learning_rate": 1.3481006314523406e-05,
-      "loss": 2.1013,
+      "epoch": 0.41,
+      "grad_norm": 11.939325332641602,
+      "learning_rate": 1.727900263565243e-05,
+      "loss": 0.92,
       "step": 3253
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 31.358562469482422,
-      "learning_rate": 1.3479001703919015e-05,
-      "loss": 2.7242,
+      "epoch": 0.41,
+      "grad_norm": 8.76098918914795,
+      "learning_rate": 1.727816592059574e-05,
+      "loss": 3.1085,
       "step": 3254
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 20.12774085998535,
-      "learning_rate": 1.3476997093314625e-05,
-      "loss": 1.5558,
+      "epoch": 0.41,
+      "grad_norm": 19.767866134643555,
+      "learning_rate": 1.7277329205539054e-05,
+      "loss": 2.4461,
       "step": 3255
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 28.31780242919922,
-      "learning_rate": 1.3474992482710235e-05,
-      "loss": 2.1186,
+      "epoch": 0.41,
+      "grad_norm": 12.018153190612793,
+      "learning_rate": 1.7276492490482368e-05,
+      "loss": 1.4399,
       "step": 3256
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 11.361303329467773,
-      "learning_rate": 1.3472987872105845e-05,
-      "loss": 1.4945,
+      "epoch": 0.41,
+      "grad_norm": 5.61497163772583,
+      "learning_rate": 1.7275655775425678e-05,
+      "loss": 0.6077,
       "step": 3257
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 112.40892028808594,
-      "learning_rate": 1.3470983261501453e-05,
-      "loss": 1.87,
+      "epoch": 0.41,
+      "grad_norm": 20.77367401123047,
+      "learning_rate": 1.727481906036899e-05,
+      "loss": 2.672,
       "step": 3258
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 48.792320251464844,
-      "learning_rate": 1.3468978650897065e-05,
-      "loss": 2.829,
+      "epoch": 0.41,
+      "grad_norm": 21.33584976196289,
+      "learning_rate": 1.7273982345312305e-05,
+      "loss": 0.8965,
       "step": 3259
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 15.994609832763672,
-      "learning_rate": 1.3466974040292675e-05,
-      "loss": 2.33,
+      "epoch": 0.41,
+      "grad_norm": 11.513010025024414,
+      "learning_rate": 1.727314563025562e-05,
+      "loss": 1.8187,
       "step": 3260
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 14.680255889892578,
-      "learning_rate": 1.3464969429688284e-05,
-      "loss": 1.5227,
+      "epoch": 0.41,
+      "grad_norm": 19.68485450744629,
+      "learning_rate": 1.727230891519893e-05,
+      "loss": 1.6764,
       "step": 3261
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 20.413484573364258,
-      "learning_rate": 1.3462964819083895e-05,
-      "loss": 1.3934,
+      "epoch": 0.41,
+      "grad_norm": 4.071218013763428,
+      "learning_rate": 1.7271472200142243e-05,
+      "loss": 0.5521,
       "step": 3262
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 10.785961151123047,
-      "learning_rate": 1.3460960208479504e-05,
-      "loss": 1.6991,
+      "epoch": 0.41,
+      "grad_norm": 11.459487915039062,
+      "learning_rate": 1.7270635485085556e-05,
+      "loss": 2.8706,
       "step": 3263
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 23.43621063232422,
-      "learning_rate": 1.3458955597875114e-05,
-      "loss": 2.1189,
+      "epoch": 0.41,
+      "grad_norm": 16.798728942871094,
+      "learning_rate": 1.7269798770028867e-05,
+      "loss": 2.188,
       "step": 3264
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 18.55811882019043,
-      "learning_rate": 1.3456950987270722e-05,
-      "loss": 1.6837,
+      "epoch": 0.41,
+      "grad_norm": 38.427467346191406,
+      "learning_rate": 1.726896205497218e-05,
+      "loss": 3.6072,
       "step": 3265
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 25.324520111083984,
-      "learning_rate": 1.3454946376666334e-05,
-      "loss": 2.5444,
+      "epoch": 0.41,
+      "grad_norm": 11.476984977722168,
+      "learning_rate": 1.7268125339915494e-05,
+      "loss": 1.1716,
       "step": 3266
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 29.4528751373291,
-      "learning_rate": 1.3452941766061944e-05,
-      "loss": 1.7111,
+      "epoch": 0.41,
+      "grad_norm": 10.207488059997559,
+      "learning_rate": 1.7267288624858808e-05,
+      "loss": 1.4146,
       "step": 3267
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 10.598390579223633,
-      "learning_rate": 1.3450937155457552e-05,
-      "loss": 1.1832,
+      "epoch": 0.41,
+      "grad_norm": 8.6456298828125,
+      "learning_rate": 1.7266451909802118e-05,
+      "loss": 2.5119,
       "step": 3268
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 16.23030662536621,
-      "learning_rate": 1.3448932544853164e-05,
-      "loss": 1.845,
+      "epoch": 0.41,
+      "grad_norm": 9.207403182983398,
+      "learning_rate": 1.726561519474543e-05,
+      "loss": 1.3832,
       "step": 3269
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 16.992780685424805,
-      "learning_rate": 1.3446927934248773e-05,
-      "loss": 1.5488,
+      "epoch": 0.41,
+      "grad_norm": 12.22926139831543,
+      "learning_rate": 1.7264778479688745e-05,
+      "loss": 2.1751,
       "step": 3270
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 16.57977294921875,
-      "learning_rate": 1.3444923323644383e-05,
-      "loss": 2.4538,
+      "epoch": 0.41,
+      "grad_norm": 15.51880931854248,
+      "learning_rate": 1.7263941764632055e-05,
+      "loss": 3.7102,
       "step": 3271
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 17.394916534423828,
-      "learning_rate": 1.3442918713039994e-05,
-      "loss": 2.4952,
+      "epoch": 0.41,
+      "grad_norm": 15.530855178833008,
+      "learning_rate": 1.726310504957537e-05,
+      "loss": 1.5472,
       "step": 3272
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 10.907772064208984,
-      "learning_rate": 1.3440914102435603e-05,
-      "loss": 2.2693,
+      "epoch": 0.41,
+      "grad_norm": 9.8801908493042,
+      "learning_rate": 1.7262268334518683e-05,
+      "loss": 0.8076,
       "step": 3273
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 15.381023406982422,
-      "learning_rate": 1.3438909491831213e-05,
-      "loss": 1.9541,
+      "epoch": 0.41,
+      "grad_norm": 16.780431747436523,
+      "learning_rate": 1.7261431619461993e-05,
+      "loss": 2.3568,
       "step": 3274
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 13.303698539733887,
-      "learning_rate": 1.3436904881226821e-05,
-      "loss": 1.7353,
+      "epoch": 0.41,
+      "grad_norm": 22.34392547607422,
+      "learning_rate": 1.7260594904405307e-05,
+      "loss": 2.1313,
       "step": 3275
     },
     {
-      "epoch": 0.98,
-      "grad_norm": 22.471403121948242,
-      "learning_rate": 1.3434900270622433e-05,
-      "loss": 1.8707,
+      "epoch": 0.41,
+      "grad_norm": 10.950980186462402,
+      "learning_rate": 1.725975818934862e-05,
+      "loss": 1.5318,
       "step": 3276
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 12.483773231506348,
-      "learning_rate": 1.3432895660018041e-05,
-      "loss": 1.862,
+      "epoch": 0.41,
+      "grad_norm": 15.691773414611816,
+      "learning_rate": 1.725892147429193e-05,
+      "loss": 4.148,
       "step": 3277
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 24.261032104492188,
-      "learning_rate": 1.3430891049413652e-05,
-      "loss": 3.0355,
+      "epoch": 0.41,
+      "grad_norm": 19.82890510559082,
+      "learning_rate": 1.7258084759235244e-05,
+      "loss": 2.2056,
       "step": 3278
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 27.064979553222656,
-      "learning_rate": 1.3428886438809263e-05,
-      "loss": 2.6927,
+      "epoch": 0.41,
+      "grad_norm": 13.338221549987793,
+      "learning_rate": 1.7257248044178554e-05,
+      "loss": 1.574,
       "step": 3279
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 23.447118759155273,
-      "learning_rate": 1.3426881828204872e-05,
-      "loss": 2.2596,
+      "epoch": 0.41,
+      "grad_norm": 12.845236778259277,
+      "learning_rate": 1.7256411329121868e-05,
+      "loss": 1.8639,
       "step": 3280
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 19.78642463684082,
-      "learning_rate": 1.3424877217600484e-05,
-      "loss": 2.1217,
+      "epoch": 0.41,
+      "grad_norm": 21.712717056274414,
+      "learning_rate": 1.7255574614065182e-05,
+      "loss": 1.6217,
       "step": 3281
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 11.758055686950684,
-      "learning_rate": 1.3422872606996092e-05,
-      "loss": 1.0749,
+      "epoch": 0.41,
+      "grad_norm": 7.393682956695557,
+      "learning_rate": 1.7254737899008492e-05,
+      "loss": 0.6024,
       "step": 3282
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 15.738198280334473,
-      "learning_rate": 1.3420867996391702e-05,
-      "loss": 2.25,
+      "epoch": 0.41,
+      "grad_norm": 8.766348838806152,
+      "learning_rate": 1.7253901183951806e-05,
+      "loss": 0.8492,
       "step": 3283
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 21.461877822875977,
-      "learning_rate": 1.3418863385787314e-05,
-      "loss": 2.1378,
+      "epoch": 0.41,
+      "grad_norm": 17.923397064208984,
+      "learning_rate": 1.725306446889512e-05,
+      "loss": 3.0523,
       "step": 3284
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 39.54304504394531,
-      "learning_rate": 1.3416858775182922e-05,
-      "loss": 3.1118,
+      "epoch": 0.41,
+      "grad_norm": 17.999114990234375,
+      "learning_rate": 1.725222775383843e-05,
+      "loss": 2.9979,
       "step": 3285
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 14.904977798461914,
-      "learning_rate": 1.3414854164578532e-05,
-      "loss": 1.2001,
+      "epoch": 0.41,
+      "grad_norm": 7.008960247039795,
+      "learning_rate": 1.7251391038781743e-05,
+      "loss": 1.8619,
       "step": 3286
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 19.5516414642334,
-      "learning_rate": 1.341284955397414e-05,
-      "loss": 2.2941,
+      "epoch": 0.41,
+      "grad_norm": 22.732624053955078,
+      "learning_rate": 1.7250554323725057e-05,
+      "loss": 1.7138,
       "step": 3287
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 18.030540466308594,
-      "learning_rate": 1.3410844943369752e-05,
-      "loss": 2.1409,
+      "epoch": 0.41,
+      "grad_norm": 15.985183715820312,
+      "learning_rate": 1.724971760866837e-05,
+      "loss": 2.6133,
       "step": 3288
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 17.5920352935791,
-      "learning_rate": 1.340884033276536e-05,
-      "loss": 3.0416,
+      "epoch": 0.41,
+      "grad_norm": 8.475821495056152,
+      "learning_rate": 1.724888089361168e-05,
+      "loss": 1.7767,
       "step": 3289
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 18.530750274658203,
-      "learning_rate": 1.3406835722160971e-05,
-      "loss": 1.1762,
+      "epoch": 0.41,
+      "grad_norm": 15.1087064743042,
+      "learning_rate": 1.7248044178554994e-05,
+      "loss": 1.368,
       "step": 3290
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 15.765362739562988,
-      "learning_rate": 1.3404831111556583e-05,
-      "loss": 2.0459,
+      "epoch": 0.41,
+      "grad_norm": 13.120559692382812,
+      "learning_rate": 1.7247207463498308e-05,
+      "loss": 2.2243,
       "step": 3291
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 27.75279998779297,
-      "learning_rate": 1.3402826500952191e-05,
-      "loss": 2.1783,
+      "epoch": 0.41,
+      "grad_norm": 12.287973403930664,
+      "learning_rate": 1.724637074844162e-05,
+      "loss": 2.1131,
       "step": 3292
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 16.64308738708496,
-      "learning_rate": 1.3400821890347801e-05,
-      "loss": 2.4156,
+      "epoch": 0.41,
+      "grad_norm": 17.94985580444336,
+      "learning_rate": 1.7245534033384932e-05,
+      "loss": 3.8318,
       "step": 3293
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 19.890602111816406,
-      "learning_rate": 1.339881727974341e-05,
-      "loss": 1.0157,
+      "epoch": 0.41,
+      "grad_norm": 22.18570327758789,
+      "learning_rate": 1.7244697318328246e-05,
+      "loss": 1.9355,
       "step": 3294
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 15.657976150512695,
-      "learning_rate": 1.3396812669139021e-05,
-      "loss": 1.3915,
+      "epoch": 0.41,
+      "grad_norm": 10.512239456176758,
+      "learning_rate": 1.724386060327156e-05,
+      "loss": 1.9186,
       "step": 3295
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 14.942137718200684,
-      "learning_rate": 1.3394808058534631e-05,
-      "loss": 2.0334,
+      "epoch": 0.41,
+      "grad_norm": 47.57612991333008,
+      "learning_rate": 1.724302388821487e-05,
+      "loss": 1.8597,
       "step": 3296
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 10.705955505371094,
-      "learning_rate": 1.339280344793024e-05,
-      "loss": 1.7803,
+      "epoch": 0.41,
+      "grad_norm": 14.717118263244629,
+      "learning_rate": 1.7242187173158183e-05,
+      "loss": 3.2491,
       "step": 3297
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 11.542771339416504,
-      "learning_rate": 1.3390798837325852e-05,
-      "loss": 1.2149,
+      "epoch": 0.41,
+      "grad_norm": 12.336624145507812,
+      "learning_rate": 1.7241350458101497e-05,
+      "loss": 1.408,
       "step": 3298
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 16.316530227661133,
-      "learning_rate": 1.338879422672146e-05,
-      "loss": 2.4369,
+      "epoch": 0.41,
+      "grad_norm": 22.74224853515625,
+      "learning_rate": 1.7240513743044807e-05,
+      "loss": 1.4211,
       "step": 3299
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 23.6247615814209,
-      "learning_rate": 1.338678961611707e-05,
-      "loss": 2.1207,
+      "epoch": 0.41,
+      "grad_norm": 18.7371883392334,
+      "learning_rate": 1.723967702798812e-05,
+      "loss": 4.5971,
       "step": 3300
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 17.557518005371094,
-      "learning_rate": 1.338478500551268e-05,
-      "loss": 1.5633,
+      "epoch": 0.41,
+      "grad_norm": 19.930030822753906,
+      "learning_rate": 1.7238840312931434e-05,
+      "loss": 1.3958,
       "step": 3301
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 13.458683967590332,
-      "learning_rate": 1.338278039490829e-05,
-      "loss": 2.0549,
+      "epoch": 0.41,
+      "grad_norm": 17.725854873657227,
+      "learning_rate": 1.7238003597874745e-05,
+      "loss": 1.8913,
       "step": 3302
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 10.493450164794922,
-      "learning_rate": 1.33807757843039e-05,
-      "loss": 2.0356,
+      "epoch": 0.41,
+      "grad_norm": 9.200337409973145,
+      "learning_rate": 1.723716688281806e-05,
+      "loss": 2.1126,
       "step": 3303
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 16.16809844970703,
-      "learning_rate": 1.337877117369951e-05,
-      "loss": 1.918,
+      "epoch": 0.41,
+      "grad_norm": 19.342815399169922,
+      "learning_rate": 1.7236330167761372e-05,
+      "loss": 2.3606,
       "step": 3304
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 15.41076374053955,
-      "learning_rate": 1.337676656309512e-05,
-      "loss": 2.8471,
+      "epoch": 0.41,
+      "grad_norm": 11.29781436920166,
+      "learning_rate": 1.7235493452704682e-05,
+      "loss": 2.2782,
       "step": 3305
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 18.323535919189453,
-      "learning_rate": 1.3374761952490729e-05,
-      "loss": 1.6805,
+      "epoch": 0.41,
+      "grad_norm": 40.30754089355469,
+      "learning_rate": 1.7234656737647996e-05,
+      "loss": 2.1958,
       "step": 3306
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 10.888114929199219,
-      "learning_rate": 1.337275734188634e-05,
-      "loss": 1.4534,
+      "epoch": 0.42,
+      "grad_norm": 10.85988712310791,
+      "learning_rate": 1.7233820022591306e-05,
+      "loss": 3.3253,
       "step": 3307
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 23.824207305908203,
-      "learning_rate": 1.3370752731281949e-05,
-      "loss": 2.1954,
+      "epoch": 0.42,
+      "grad_norm": 8.996675491333008,
+      "learning_rate": 1.723298330753462e-05,
+      "loss": 2.2486,
       "step": 3308
     },
     {
-      "epoch": 0.99,
-      "grad_norm": 12.056889533996582,
-      "learning_rate": 1.3368748120677559e-05,
-      "loss": 1.9769,
+      "epoch": 0.42,
+      "grad_norm": 9.021422386169434,
+      "learning_rate": 1.7232146592477933e-05,
+      "loss": 1.5477,
       "step": 3309
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 20.176801681518555,
-      "learning_rate": 1.3366743510073171e-05,
-      "loss": 2.3137,
+      "epoch": 0.42,
+      "grad_norm": 12.186882972717285,
+      "learning_rate": 1.7231309877421244e-05,
+      "loss": 1.0307,
       "step": 3310
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 12.740705490112305,
-      "learning_rate": 1.336473889946878e-05,
-      "loss": 2.1326,
+      "epoch": 0.42,
+      "grad_norm": 16.202449798583984,
+      "learning_rate": 1.7230473162364557e-05,
+      "loss": 2.3638,
       "step": 3311
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 13.5363130569458,
-      "learning_rate": 1.336273428886439e-05,
-      "loss": 1.5278,
+      "epoch": 0.42,
+      "grad_norm": 9.610941886901855,
+      "learning_rate": 1.722963644730787e-05,
+      "loss": 2.0287,
       "step": 3312
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 11.068243026733398,
-      "learning_rate": 1.3360729678259998e-05,
-      "loss": 1.186,
+      "epoch": 0.42,
+      "grad_norm": 20.206573486328125,
+      "learning_rate": 1.722879973225118e-05,
+      "loss": 2.0326,
       "step": 3313
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 16.938138961791992,
-      "learning_rate": 1.335872506765561e-05,
-      "loss": 1.0064,
+      "epoch": 0.42,
+      "grad_norm": 8.457555770874023,
+      "learning_rate": 1.7227963017194495e-05,
+      "loss": 3.5017,
       "step": 3314
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 13.62263298034668,
-      "learning_rate": 1.335672045705122e-05,
-      "loss": 1.909,
+      "epoch": 0.42,
+      "grad_norm": 45.251060485839844,
+      "learning_rate": 1.722712630213781e-05,
+      "loss": 2.96,
       "step": 3315
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 11.283188819885254,
-      "learning_rate": 1.3354715846446828e-05,
-      "loss": 1.6886,
+      "epoch": 0.42,
+      "grad_norm": 8.923137664794922,
+      "learning_rate": 1.7226289587081122e-05,
+      "loss": 0.499,
       "step": 3316
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 14.703868865966797,
-      "learning_rate": 1.335271123584244e-05,
-      "loss": 1.9791,
+      "epoch": 0.42,
+      "grad_norm": 15.219002723693848,
+      "learning_rate": 1.7225452872024432e-05,
+      "loss": 1.5465,
       "step": 3317
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 13.806490898132324,
-      "learning_rate": 1.3350706625238048e-05,
-      "loss": 1.516,
+      "epoch": 0.42,
+      "grad_norm": 92.25977325439453,
+      "learning_rate": 1.7224616156967746e-05,
+      "loss": 1.0988,
       "step": 3318
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 14.787043571472168,
-      "learning_rate": 1.3348702014633658e-05,
-      "loss": 1.8667,
+      "epoch": 0.42,
+      "grad_norm": 15.196372985839844,
+      "learning_rate": 1.722377944191106e-05,
+      "loss": 2.8061,
       "step": 3319
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 17.8718318939209,
-      "learning_rate": 1.3346697404029267e-05,
-      "loss": 2.0014,
+      "epoch": 0.42,
+      "grad_norm": 10.119935989379883,
+      "learning_rate": 1.722294272685437e-05,
+      "loss": 0.7383,
       "step": 3320
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 23.68500518798828,
-      "learning_rate": 1.3344692793424878e-05,
-      "loss": 1.7045,
+      "epoch": 0.42,
+      "grad_norm": 18.1783390045166,
+      "learning_rate": 1.7222106011797684e-05,
+      "loss": 0.8793,
       "step": 3321
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 10.535503387451172,
-      "learning_rate": 1.3342688182820488e-05,
-      "loss": 1.183,
+      "epoch": 0.42,
+      "grad_norm": 24.50297737121582,
+      "learning_rate": 1.7221269296740997e-05,
+      "loss": 3.5501,
       "step": 3322
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 25.851089477539062,
-      "learning_rate": 1.3340683572216097e-05,
-      "loss": 2.7104,
+      "epoch": 0.42,
+      "grad_norm": 9.117147445678711,
+      "learning_rate": 1.722043258168431e-05,
+      "loss": 1.0697,
       "step": 3323
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 10.395030975341797,
-      "learning_rate": 1.3338678961611709e-05,
-      "loss": 1.7227,
+      "epoch": 0.42,
+      "grad_norm": 12.340163230895996,
+      "learning_rate": 1.721959586662762e-05,
+      "loss": 1.7272,
       "step": 3324
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 14.099095344543457,
-      "learning_rate": 1.3336674351007317e-05,
-      "loss": 1.7941,
+      "epoch": 0.42,
+      "grad_norm": 35.53330612182617,
+      "learning_rate": 1.7218759151570935e-05,
+      "loss": 2.7344,
       "step": 3325
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 15.447168350219727,
-      "learning_rate": 1.3334669740402927e-05,
-      "loss": 2.0854,
+      "epoch": 0.42,
+      "grad_norm": 14.975069046020508,
+      "learning_rate": 1.721792243651425e-05,
+      "loss": 2.8875,
       "step": 3326
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 20.244766235351562,
-      "learning_rate": 1.3332665129798539e-05,
-      "loss": 3.0187,
+      "epoch": 0.42,
+      "grad_norm": 10.984192848205566,
+      "learning_rate": 1.721708572145756e-05,
+      "loss": 2.2025,
       "step": 3327
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 56.30267333984375,
-      "learning_rate": 1.3330660519194147e-05,
-      "loss": 2.1849,
+      "epoch": 0.42,
+      "grad_norm": 8.138824462890625,
+      "learning_rate": 1.7216249006400872e-05,
+      "loss": 1.3284,
       "step": 3328
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 28.362627029418945,
-      "learning_rate": 1.3328655908589757e-05,
-      "loss": 1.7998,
+      "epoch": 0.42,
+      "grad_norm": 8.697530746459961,
+      "learning_rate": 1.7215412291344186e-05,
+      "loss": 3.7476,
       "step": 3329
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 12.778199195861816,
-      "learning_rate": 1.3326651297985367e-05,
-      "loss": 1.5023,
+      "epoch": 0.42,
+      "grad_norm": 16.129322052001953,
+      "learning_rate": 1.7214575576287496e-05,
+      "loss": 2.1301,
       "step": 3330
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 9.352411270141602,
-      "learning_rate": 1.3324646687380978e-05,
-      "loss": 1.3491,
+      "epoch": 0.42,
+      "grad_norm": 16.356788635253906,
+      "learning_rate": 1.721373886123081e-05,
+      "loss": 3.1456,
       "step": 3331
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 32.7597770690918,
-      "learning_rate": 1.3322642076776586e-05,
-      "loss": 2.2412,
+      "epoch": 0.42,
+      "grad_norm": 11.671011924743652,
+      "learning_rate": 1.721290214617412e-05,
+      "loss": 2.0738,
       "step": 3332
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 14.371981620788574,
-      "learning_rate": 1.3320637466172198e-05,
-      "loss": 2.2198,
+      "epoch": 0.42,
+      "grad_norm": 12.250069618225098,
+      "learning_rate": 1.7212065431117434e-05,
+      "loss": 3.1575,
       "step": 3333
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 15.051918029785156,
-      "learning_rate": 1.3318632855567808e-05,
-      "loss": 1.5761,
+      "epoch": 0.42,
+      "grad_norm": 15.652741432189941,
+      "learning_rate": 1.7211228716060748e-05,
+      "loss": 3.4465,
       "step": 3334
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 18.553064346313477,
-      "learning_rate": 1.3316628244963416e-05,
-      "loss": 2.2955,
+      "epoch": 0.42,
+      "grad_norm": 8.946577072143555,
+      "learning_rate": 1.7210392001004058e-05,
+      "loss": 1.8496,
       "step": 3335
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 13.871806144714355,
-      "learning_rate": 1.3314623634359028e-05,
-      "loss": 1.7798,
+      "epoch": 0.42,
+      "grad_norm": 20.72052764892578,
+      "learning_rate": 1.720955528594737e-05,
+      "loss": 1.8472,
       "step": 3336
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 24.853288650512695,
-      "learning_rate": 1.3312619023754636e-05,
-      "loss": 2.5074,
+      "epoch": 0.42,
+      "grad_norm": 11.61407470703125,
+      "learning_rate": 1.7208718570890685e-05,
+      "loss": 1.032,
       "step": 3337
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 7.72559928894043,
-      "learning_rate": 1.3310614413150246e-05,
-      "loss": 2.0241,
+      "epoch": 0.42,
+      "grad_norm": 16.900367736816406,
+      "learning_rate": 1.7207881855833995e-05,
+      "loss": 1.6605,
       "step": 3338
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 67.96630096435547,
-      "learning_rate": 1.3308609802545858e-05,
-      "loss": 2.9062,
+      "epoch": 0.42,
+      "grad_norm": 10.358806610107422,
+      "learning_rate": 1.720704514077731e-05,
+      "loss": 0.8799,
       "step": 3339
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 15.404458045959473,
-      "learning_rate": 1.3306605191941467e-05,
-      "loss": 1.2167,
+      "epoch": 0.42,
+      "grad_norm": 13.098993301391602,
+      "learning_rate": 1.7206208425720623e-05,
+      "loss": 1.7514,
       "step": 3340
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 20.57508087158203,
-      "learning_rate": 1.3304600581337077e-05,
-      "loss": 2.3393,
+      "epoch": 0.42,
+      "grad_norm": 14.889695167541504,
+      "learning_rate": 1.7205371710663933e-05,
+      "loss": 2.7566,
       "step": 3341
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 13.644542694091797,
-      "learning_rate": 1.3302595970732685e-05,
-      "loss": 2.7504,
+      "epoch": 0.42,
+      "grad_norm": 17.534114837646484,
+      "learning_rate": 1.7204534995607247e-05,
+      "loss": 1.4067,
       "step": 3342
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 27.4990291595459,
-      "learning_rate": 1.3300591360128297e-05,
-      "loss": 2.4179,
+      "epoch": 0.42,
+      "grad_norm": 20.281309127807617,
+      "learning_rate": 1.720369828055056e-05,
+      "loss": 2.097,
       "step": 3343
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 18.17327308654785,
-      "learning_rate": 1.3298586749523905e-05,
-      "loss": 1.8711,
+      "epoch": 0.42,
+      "grad_norm": 22.249191284179688,
+      "learning_rate": 1.7202861565493874e-05,
+      "loss": 1.0598,
       "step": 3344
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 11.454172134399414,
-      "learning_rate": 1.3296582138919515e-05,
-      "loss": 2.2247,
+      "epoch": 0.42,
+      "grad_norm": 4.688612461090088,
+      "learning_rate": 1.7202024850437184e-05,
+      "loss": 1.986,
       "step": 3345
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 14.537096977233887,
-      "learning_rate": 1.3294577528315127e-05,
-      "loss": 1.3509,
+      "epoch": 0.42,
+      "grad_norm": 11.584707260131836,
+      "learning_rate": 1.7201188135380498e-05,
+      "loss": 1.064,
       "step": 3346
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 17.769039154052734,
-      "learning_rate": 1.3292572917710736e-05,
-      "loss": 1.6675,
+      "epoch": 0.42,
+      "grad_norm": 13.541443824768066,
+      "learning_rate": 1.720035142032381e-05,
+      "loss": 2.173,
       "step": 3347
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 20.219247817993164,
-      "learning_rate": 1.3290568307106346e-05,
-      "loss": 2.1718,
+      "epoch": 0.42,
+      "grad_norm": 10.384590148925781,
+      "learning_rate": 1.7199514705267122e-05,
+      "loss": 2.0759,
       "step": 3348
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 12.704767227172852,
-      "learning_rate": 1.3288563696501956e-05,
-      "loss": 1.2209,
+      "epoch": 0.42,
+      "grad_norm": 10.545215606689453,
+      "learning_rate": 1.7198677990210435e-05,
+      "loss": 1.4586,
       "step": 3349
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 15.621160507202148,
-      "learning_rate": 1.3286559085897566e-05,
-      "loss": 1.5298,
+      "epoch": 0.42,
+      "grad_norm": 19.95136260986328,
+      "learning_rate": 1.719784127515375e-05,
+      "loss": 2.455,
       "step": 3350
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 17.41335678100586,
-      "learning_rate": 1.3284554475293174e-05,
-      "loss": 2.7119,
+      "epoch": 0.42,
+      "grad_norm": 22.57049560546875,
+      "learning_rate": 1.7197004560097063e-05,
+      "loss": 2.7227,
       "step": 3351
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 27.469499588012695,
-      "learning_rate": 1.3282549864688786e-05,
-      "loss": 2.821,
+      "epoch": 0.42,
+      "grad_norm": 26.080463409423828,
+      "learning_rate": 1.7196167845040373e-05,
+      "loss": 1.8036,
       "step": 3352
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 47.88648986816406,
-      "learning_rate": 1.3280545254084396e-05,
-      "loss": 2.4446,
+      "epoch": 0.42,
+      "grad_norm": 7.422852039337158,
+      "learning_rate": 1.7195331129983687e-05,
+      "loss": 1.4551,
       "step": 3353
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 9.525798797607422,
-      "learning_rate": 1.3278540643480004e-05,
-      "loss": 0.9091,
+      "epoch": 0.42,
+      "grad_norm": 5.2106733322143555,
+      "learning_rate": 1.7194494414927e-05,
+      "loss": 2.3569,
       "step": 3354
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 22.738584518432617,
-      "learning_rate": 1.3276536032875616e-05,
-      "loss": 2.3079,
+      "epoch": 0.42,
+      "grad_norm": 12.15693187713623,
+      "learning_rate": 1.719365769987031e-05,
+      "loss": 3.4812,
       "step": 3355
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 11.072502136230469,
-      "learning_rate": 1.3274531422271225e-05,
-      "loss": 2.7678,
+      "epoch": 0.42,
+      "grad_norm": 12.908676147460938,
+      "learning_rate": 1.7192820984813624e-05,
+      "loss": 0.6298,
       "step": 3356
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 12.42871379852295,
-      "learning_rate": 1.3272526811666835e-05,
-      "loss": 1.1835,
+      "epoch": 0.42,
+      "grad_norm": 11.007267951965332,
+      "learning_rate": 1.7191984269756938e-05,
+      "loss": 2.6467,
       "step": 3357
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 10.291549682617188,
-      "learning_rate": 1.3270522201062446e-05,
-      "loss": 1.3435,
+      "epoch": 0.42,
+      "grad_norm": 12.836434364318848,
+      "learning_rate": 1.7191147554700248e-05,
+      "loss": 2.1427,
       "step": 3358
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 18.563772201538086,
-      "learning_rate": 1.3268517590458055e-05,
-      "loss": 1.2068,
+      "epoch": 0.42,
+      "grad_norm": 5.820464134216309,
+      "learning_rate": 1.719031083964356e-05,
+      "loss": 0.6328,
       "step": 3359
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 16.957143783569336,
-      "learning_rate": 1.3266512979853665e-05,
-      "loss": 2.197,
-      "step": 3360
-    },
-    {
-      "epoch": 1.01,
-      "eval_loss": 0.22327160835266113,
-      "eval_runtime": 43.5349,
-      "eval_samples_per_second": 33.973,
-      "eval_steps_per_second": 33.973,
+      "epoch": 0.42,
+      "grad_norm": 13.101532936096191,
+      "learning_rate": 1.7189474124586872e-05,
+      "loss": 1.8192,
       "step": 3360
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 12.542174339294434,
-      "learning_rate": 1.3264508369249273e-05,
-      "loss": 1.7032,
+      "epoch": 0.42,
+      "grad_norm": 8.012151718139648,
+      "learning_rate": 1.7188637409530186e-05,
+      "loss": 3.1705,
       "step": 3361
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 12.516151428222656,
-      "learning_rate": 1.3262503758644885e-05,
-      "loss": 1.4823,
+      "epoch": 0.42,
+      "grad_norm": 14.383709907531738,
+      "learning_rate": 1.71878006944735e-05,
+      "loss": 2.7463,
       "step": 3362
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 13.449585914611816,
-      "learning_rate": 1.3260499148040493e-05,
-      "loss": 1.9671,
+      "epoch": 0.42,
+      "grad_norm": 12.688018798828125,
+      "learning_rate": 1.718696397941681e-05,
+      "loss": 2.3427,
       "step": 3363
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 20.43147850036621,
-      "learning_rate": 1.3258494537436104e-05,
-      "loss": 1.5756,
+      "epoch": 0.42,
+      "grad_norm": 28.22138214111328,
+      "learning_rate": 1.7186127264360123e-05,
+      "loss": 2.986,
       "step": 3364
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 9.373948097229004,
-      "learning_rate": 1.3256489926831715e-05,
-      "loss": 2.1166,
+      "epoch": 0.42,
+      "grad_norm": 7.574335098266602,
+      "learning_rate": 1.7185290549303437e-05,
+      "loss": 2.1322,
       "step": 3365
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 28.55181312561035,
-      "learning_rate": 1.3254485316227324e-05,
-      "loss": 2.7417,
+      "epoch": 0.42,
+      "grad_norm": 8.840447425842285,
+      "learning_rate": 1.7184453834246747e-05,
+      "loss": 1.3887,
       "step": 3366
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 23.936809539794922,
-      "learning_rate": 1.3252480705622934e-05,
-      "loss": 1.5005,
+      "epoch": 0.42,
+      "grad_norm": 23.318866729736328,
+      "learning_rate": 1.718361711919006e-05,
+      "loss": 3.8719,
       "step": 3367
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 14.939528465270996,
-      "learning_rate": 1.3250476095018542e-05,
-      "loss": 2.3812,
+      "epoch": 0.42,
+      "grad_norm": 33.8458137512207,
+      "learning_rate": 1.7182780404133374e-05,
+      "loss": 2.5061,
       "step": 3368
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 23.387401580810547,
-      "learning_rate": 1.3248471484414154e-05,
-      "loss": 2.5629,
+      "epoch": 0.42,
+      "grad_norm": 41.34958267211914,
+      "learning_rate": 1.7181943689076685e-05,
+      "loss": 2.972,
       "step": 3369
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 17.187917709350586,
-      "learning_rate": 1.3246466873809764e-05,
-      "loss": 1.8393,
+      "epoch": 0.42,
+      "grad_norm": 12.544574737548828,
+      "learning_rate": 1.7181106974019998e-05,
+      "loss": 0.7907,
       "step": 3370
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 15.523499488830566,
-      "learning_rate": 1.3244462263205372e-05,
-      "loss": 2.066,
+      "epoch": 0.42,
+      "grad_norm": 21.2841796875,
+      "learning_rate": 1.7180270258963312e-05,
+      "loss": 3.0655,
       "step": 3371
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 8.489646911621094,
-      "learning_rate": 1.3242457652600984e-05,
-      "loss": 1.5775,
+      "epoch": 0.42,
+      "grad_norm": 11.424372673034668,
+      "learning_rate": 1.7179433543906626e-05,
+      "loss": 2.7168,
       "step": 3372
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 12.514199256896973,
-      "learning_rate": 1.3240453041996593e-05,
-      "loss": 1.3532,
+      "epoch": 0.42,
+      "grad_norm": 24.937965393066406,
+      "learning_rate": 1.7178596828849936e-05,
+      "loss": 2.7555,
       "step": 3373
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 39.9799919128418,
-      "learning_rate": 1.3238448431392203e-05,
-      "loss": 1.7734,
+      "epoch": 0.42,
+      "grad_norm": 20.2894229888916,
+      "learning_rate": 1.717776011379325e-05,
+      "loss": 2.5918,
       "step": 3374
     },
     {
-      "epoch": 1.01,
-      "grad_norm": 12.430253028869629,
-      "learning_rate": 1.3236443820787813e-05,
-      "loss": 2.0699,
+      "epoch": 0.42,
+      "grad_norm": 16.203311920166016,
+      "learning_rate": 1.7176923398736563e-05,
+      "loss": 1.8023,
       "step": 3375
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 32.21778106689453,
-      "learning_rate": 1.3234439210183423e-05,
-      "loss": 2.4161,
+      "epoch": 0.42,
+      "grad_norm": 15.349499702453613,
+      "learning_rate": 1.7176086683679873e-05,
+      "loss": 2.4695,
       "step": 3376
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 10.491954803466797,
-      "learning_rate": 1.3232434599579033e-05,
-      "loss": 2.0041,
+      "epoch": 0.42,
+      "grad_norm": 37.25876998901367,
+      "learning_rate": 1.7175249968623187e-05,
+      "loss": 2.3071,
       "step": 3377
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 24.133899688720703,
-      "learning_rate": 1.3230429988974643e-05,
-      "loss": 2.884,
+      "epoch": 0.42,
+      "grad_norm": 9.62696647644043,
+      "learning_rate": 1.71744132535665e-05,
+      "loss": 1.7059,
       "step": 3378
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 13.717731475830078,
-      "learning_rate": 1.3228425378370253e-05,
-      "loss": 2.4387,
+      "epoch": 0.42,
+      "grad_norm": 13.929464340209961,
+      "learning_rate": 1.7173576538509814e-05,
+      "loss": 2.8409,
       "step": 3379
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 20.04202651977539,
-      "learning_rate": 1.3226420767765862e-05,
-      "loss": 1.764,
+      "epoch": 0.42,
+      "grad_norm": 10.758846282958984,
+      "learning_rate": 1.7172739823453125e-05,
+      "loss": 1.1154,
       "step": 3380
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 16.394018173217773,
-      "learning_rate": 1.3224416157161473e-05,
-      "loss": 1.6656,
+      "epoch": 0.42,
+      "grad_norm": 6.401970863342285,
+      "learning_rate": 1.7171903108396438e-05,
+      "loss": 1.3731,
       "step": 3381
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 6.424349308013916,
-      "learning_rate": 1.3222411546557083e-05,
-      "loss": 1.2659,
+      "epoch": 0.42,
+      "grad_norm": 15.83933162689209,
+      "learning_rate": 1.7171066393339752e-05,
+      "loss": 3.2017,
       "step": 3382
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 12.825157165527344,
-      "learning_rate": 1.3220406935952692e-05,
-      "loss": 1.5172,
+      "epoch": 0.42,
+      "grad_norm": 8.87679386138916,
+      "learning_rate": 1.7170229678283062e-05,
+      "loss": 0.891,
       "step": 3383
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 32.225006103515625,
-      "learning_rate": 1.3218402325348304e-05,
-      "loss": 2.4318,
+      "epoch": 0.42,
+      "grad_norm": 15.874035835266113,
+      "learning_rate": 1.7169392963226376e-05,
+      "loss": 3.5279,
       "step": 3384
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 10.880355834960938,
-      "learning_rate": 1.3216397714743912e-05,
-      "loss": 1.5311,
+      "epoch": 0.42,
+      "grad_norm": 9.200181007385254,
+      "learning_rate": 1.7168556248169686e-05,
+      "loss": 1.3639,
       "step": 3385
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 23.808944702148438,
-      "learning_rate": 1.3214393104139522e-05,
-      "loss": 2.809,
+      "epoch": 0.42,
+      "grad_norm": 12.776347160339355,
+      "learning_rate": 1.7167719533113e-05,
+      "loss": 1.8228,
       "step": 3386
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 18.621950149536133,
-      "learning_rate": 1.321238849353513e-05,
-      "loss": 1.5185,
+      "epoch": 0.43,
+      "grad_norm": 19.877140045166016,
+      "learning_rate": 1.7166882818056313e-05,
+      "loss": 2.9879,
       "step": 3387
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 13.235946655273438,
-      "learning_rate": 1.3210383882930742e-05,
-      "loss": 1.5264,
+      "epoch": 0.43,
+      "grad_norm": 6.658298969268799,
+      "learning_rate": 1.7166046102999624e-05,
+      "loss": 0.5819,
       "step": 3388
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 12.991135597229004,
-      "learning_rate": 1.3208379272326352e-05,
-      "loss": 1.8659,
+      "epoch": 0.43,
+      "grad_norm": 51.25400161743164,
+      "learning_rate": 1.7165209387942937e-05,
+      "loss": 4.3137,
       "step": 3389
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 38.41598129272461,
-      "learning_rate": 1.320637466172196e-05,
-      "loss": 2.2308,
+      "epoch": 0.43,
+      "grad_norm": 12.011516571044922,
+      "learning_rate": 1.7164372672886248e-05,
+      "loss": 2.2849,
       "step": 3390
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 16.32402801513672,
-      "learning_rate": 1.3204370051117572e-05,
-      "loss": 1.7613,
+      "epoch": 0.43,
+      "grad_norm": 9.610902786254883,
+      "learning_rate": 1.716353595782956e-05,
+      "loss": 3.3038,
       "step": 3391
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 18.9952335357666,
-      "learning_rate": 1.320236544051318e-05,
-      "loss": 1.6083,
+      "epoch": 0.43,
+      "grad_norm": 16.154939651489258,
+      "learning_rate": 1.7162699242772875e-05,
+      "loss": 1.5948,
       "step": 3392
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 20.658859252929688,
-      "learning_rate": 1.3200360829908791e-05,
-      "loss": 1.7974,
+      "epoch": 0.43,
+      "grad_norm": 21.398250579833984,
+      "learning_rate": 1.7161862527716185e-05,
+      "loss": 2.7455,
       "step": 3393
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 15.612691879272461,
-      "learning_rate": 1.31983562193044e-05,
-      "loss": 2.5692,
+      "epoch": 0.43,
+      "grad_norm": 10.393528938293457,
+      "learning_rate": 1.71610258126595e-05,
+      "loss": 1.631,
       "step": 3394
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 49.652915954589844,
-      "learning_rate": 1.3196351608700011e-05,
-      "loss": 2.6861,
+      "epoch": 0.43,
+      "grad_norm": 16.182029724121094,
+      "learning_rate": 1.7160189097602812e-05,
+      "loss": 1.3885,
       "step": 3395
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 10.916786193847656,
-      "learning_rate": 1.3194346998095621e-05,
-      "loss": 1.4085,
+      "epoch": 0.43,
+      "grad_norm": 13.672966003417969,
+      "learning_rate": 1.7159352382546126e-05,
+      "loss": 1.5113,
       "step": 3396
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 21.12449073791504,
-      "learning_rate": 1.3192342387491231e-05,
-      "loss": 2.6789,
+      "epoch": 0.43,
+      "grad_norm": 16.824934005737305,
+      "learning_rate": 1.7158515667489436e-05,
+      "loss": 2.0353,
       "step": 3397
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 18.018606185913086,
-      "learning_rate": 1.3190337776886841e-05,
-      "loss": 1.6874,
+      "epoch": 0.43,
+      "grad_norm": 24.244441986083984,
+      "learning_rate": 1.715767895243275e-05,
+      "loss": 3.1288,
       "step": 3398
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 22.249736785888672,
-      "learning_rate": 1.318833316628245e-05,
-      "loss": 2.0898,
+      "epoch": 0.43,
+      "grad_norm": 13.222444534301758,
+      "learning_rate": 1.7156842237376064e-05,
+      "loss": 1.6778,
       "step": 3399
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 14.536703109741211,
-      "learning_rate": 1.3186328555678062e-05,
-      "loss": 2.2679,
+      "epoch": 0.43,
+      "grad_norm": 19.14320182800293,
+      "learning_rate": 1.7156005522319374e-05,
+      "loss": 3.5595,
       "step": 3400
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 16.38414764404297,
-      "learning_rate": 1.3184323945073672e-05,
-      "loss": 1.3831,
+      "epoch": 0.43,
+      "grad_norm": 9.551873207092285,
+      "learning_rate": 1.7155168807262687e-05,
+      "loss": 1.6328,
       "step": 3401
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 16.283510208129883,
-      "learning_rate": 1.318231933446928e-05,
-      "loss": 1.622,
+      "epoch": 0.43,
+      "grad_norm": 15.074509620666504,
+      "learning_rate": 1.7154332092206e-05,
+      "loss": 3.466,
       "step": 3402
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 12.426440238952637,
-      "learning_rate": 1.3180314723864892e-05,
-      "loss": 1.574,
+      "epoch": 0.43,
+      "grad_norm": 16.020980834960938,
+      "learning_rate": 1.7153495377149315e-05,
+      "loss": 1.7498,
       "step": 3403
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 10.712419509887695,
-      "learning_rate": 1.31783101132605e-05,
-      "loss": 1.8414,
+      "epoch": 0.43,
+      "grad_norm": 13.791266441345215,
+      "learning_rate": 1.7152658662092625e-05,
+      "loss": 1.2065,
       "step": 3404
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 12.0217924118042,
-      "learning_rate": 1.317630550265611e-05,
-      "loss": 1.817,
+      "epoch": 0.43,
+      "grad_norm": 14.800347328186035,
+      "learning_rate": 1.715182194703594e-05,
+      "loss": 1.5002,
       "step": 3405
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 10.142142295837402,
-      "learning_rate": 1.3174300892051719e-05,
-      "loss": 1.7129,
+      "epoch": 0.43,
+      "grad_norm": 15.264331817626953,
+      "learning_rate": 1.7150985231979252e-05,
+      "loss": 1.3831,
       "step": 3406
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 17.107044219970703,
-      "learning_rate": 1.317229628144733e-05,
-      "loss": 2.2242,
+      "epoch": 0.43,
+      "grad_norm": 5.368287086486816,
+      "learning_rate": 1.7150148516922563e-05,
+      "loss": 0.4597,
       "step": 3407
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 30.633485794067383,
-      "learning_rate": 1.317029167084294e-05,
-      "loss": 2.3054,
+      "epoch": 0.43,
+      "grad_norm": 15.589975357055664,
+      "learning_rate": 1.7149311801865876e-05,
+      "loss": 2.8628,
       "step": 3408
     },
     {
-      "epoch": 1.02,
-      "grad_norm": 10.093416213989258,
-      "learning_rate": 1.3168287060238549e-05,
-      "loss": 1.6337,
+      "epoch": 0.43,
+      "grad_norm": 7.814126968383789,
+      "learning_rate": 1.714847508680919e-05,
+      "loss": 0.5112,
       "step": 3409
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 16.132856369018555,
-      "learning_rate": 1.316628244963416e-05,
-      "loss": 1.8572,
+      "epoch": 0.43,
+      "grad_norm": 8.66985034942627,
+      "learning_rate": 1.7147638371752504e-05,
+      "loss": 0.4126,
       "step": 3410
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 26.83838653564453,
-      "learning_rate": 1.3164277839029769e-05,
-      "loss": 1.6212,
+      "epoch": 0.43,
+      "grad_norm": 10.87762451171875,
+      "learning_rate": 1.7146801656695814e-05,
+      "loss": 0.5436,
       "step": 3411
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 18.7099609375,
-      "learning_rate": 1.3162273228425379e-05,
-      "loss": 2.4694,
+      "epoch": 0.43,
+      "grad_norm": 10.160980224609375,
+      "learning_rate": 1.7145964941639127e-05,
+      "loss": 2.5114,
       "step": 3412
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 15.113785743713379,
-      "learning_rate": 1.3160268617820991e-05,
-      "loss": 1.8307,
+      "epoch": 0.43,
+      "grad_norm": 20.5477294921875,
+      "learning_rate": 1.7145128226582438e-05,
+      "loss": 4.0677,
       "step": 3413
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 13.659159660339355,
-      "learning_rate": 1.31582640072166e-05,
-      "loss": 2.0593,
+      "epoch": 0.43,
+      "grad_norm": 15.280623435974121,
+      "learning_rate": 1.714429151152575e-05,
+      "loss": 4.0942,
       "step": 3414
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 25.388036727905273,
-      "learning_rate": 1.315625939661221e-05,
-      "loss": 2.2341,
+      "epoch": 0.43,
+      "grad_norm": 17.571733474731445,
+      "learning_rate": 1.7143454796469065e-05,
+      "loss": 2.6576,
       "step": 3415
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 7.588253021240234,
-      "learning_rate": 1.3154254786007818e-05,
-      "loss": 1.8049,
+      "epoch": 0.43,
+      "grad_norm": 11.830584526062012,
+      "learning_rate": 1.7142618081412375e-05,
+      "loss": 2.3105,
       "step": 3416
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 18.153966903686523,
-      "learning_rate": 1.315225017540343e-05,
-      "loss": 2.1183,
+      "epoch": 0.43,
+      "grad_norm": 15.482280731201172,
+      "learning_rate": 1.714178136635569e-05,
+      "loss": 3.3672,
       "step": 3417
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 12.177376747131348,
-      "learning_rate": 1.3150245564799038e-05,
-      "loss": 1.611,
+      "epoch": 0.43,
+      "grad_norm": 12.5259370803833,
+      "learning_rate": 1.7140944651299e-05,
+      "loss": 1.351,
       "step": 3418
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 15.74305534362793,
-      "learning_rate": 1.3148240954194648e-05,
-      "loss": 2.5822,
+      "epoch": 0.43,
+      "grad_norm": 5.563350677490234,
+      "learning_rate": 1.7140107936242313e-05,
+      "loss": 0.5526,
       "step": 3419
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 13.683473587036133,
-      "learning_rate": 1.314623634359026e-05,
-      "loss": 1.4373,
+      "epoch": 0.43,
+      "grad_norm": 8.950671195983887,
+      "learning_rate": 1.7139271221185626e-05,
+      "loss": 2.2944,
       "step": 3420
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 17.658405303955078,
-      "learning_rate": 1.3144231732985868e-05,
-      "loss": 1.9642,
+      "epoch": 0.43,
+      "grad_norm": 19.60506248474121,
+      "learning_rate": 1.7138434506128937e-05,
+      "loss": 4.1311,
       "step": 3421
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 10.713934898376465,
-      "learning_rate": 1.3142227122381478e-05,
-      "loss": 1.7652,
+      "epoch": 0.43,
+      "grad_norm": 9.252056121826172,
+      "learning_rate": 1.713759779107225e-05,
+      "loss": 0.5094,
       "step": 3422
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 10.191012382507324,
-      "learning_rate": 1.3140222511777088e-05,
-      "loss": 1.7921,
+      "epoch": 0.43,
+      "grad_norm": 19.239261627197266,
+      "learning_rate": 1.7136761076015564e-05,
+      "loss": 2.1528,
       "step": 3423
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 9.874408721923828,
-      "learning_rate": 1.3138217901172698e-05,
-      "loss": 0.9375,
+      "epoch": 0.43,
+      "grad_norm": 6.057199478149414,
+      "learning_rate": 1.7135924360958878e-05,
+      "loss": 0.1486,
       "step": 3424
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 18.47734260559082,
-      "learning_rate": 1.3136213290568307e-05,
-      "loss": 2.3192,
+      "epoch": 0.43,
+      "grad_norm": 9.986576080322266,
+      "learning_rate": 1.7135087645902188e-05,
+      "loss": 0.3078,
       "step": 3425
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 16.030399322509766,
-      "learning_rate": 1.3134208679963919e-05,
-      "loss": 1.957,
+      "epoch": 0.43,
+      "grad_norm": 13.740504264831543,
+      "learning_rate": 1.71342509308455e-05,
+      "loss": 2.2573,
       "step": 3426
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 8.997057914733887,
-      "learning_rate": 1.3132204069359529e-05,
-      "loss": 1.5734,
+      "epoch": 0.43,
+      "grad_norm": 11.266700744628906,
+      "learning_rate": 1.7133414215788815e-05,
+      "loss": 1.1035,
       "step": 3427
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 174.50979614257812,
-      "learning_rate": 1.3130199458755137e-05,
-      "loss": 1.8404,
+      "epoch": 0.43,
+      "grad_norm": 24.04230499267578,
+      "learning_rate": 1.7132577500732126e-05,
+      "loss": 2.5449,
       "step": 3428
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 15.005576133728027,
-      "learning_rate": 1.3128194848150749e-05,
-      "loss": 1.4347,
+      "epoch": 0.43,
+      "grad_norm": 14.336504936218262,
+      "learning_rate": 1.713174078567544e-05,
+      "loss": 2.1929,
       "step": 3429
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 13.528023719787598,
-      "learning_rate": 1.3126190237546357e-05,
-      "loss": 2.0983,
+      "epoch": 0.43,
+      "grad_norm": 4.596050262451172,
+      "learning_rate": 1.7130904070618753e-05,
+      "loss": 1.9653,
       "step": 3430
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 61.08424377441406,
-      "learning_rate": 1.3124185626941967e-05,
-      "loss": 1.3849,
+      "epoch": 0.43,
+      "grad_norm": 19.6068115234375,
+      "learning_rate": 1.7130067355562066e-05,
+      "loss": 3.055,
       "step": 3431
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 19.432321548461914,
-      "learning_rate": 1.3122181016337579e-05,
-      "loss": 2.5755,
+      "epoch": 0.43,
+      "grad_norm": 22.83165740966797,
+      "learning_rate": 1.7129230640505377e-05,
+      "loss": 1.9638,
       "step": 3432
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 10.8148775100708,
-      "learning_rate": 1.3120176405733188e-05,
-      "loss": 2.1681,
+      "epoch": 0.43,
+      "grad_norm": 13.930540084838867,
+      "learning_rate": 1.712839392544869e-05,
+      "loss": 3.5265,
       "step": 3433
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 19.197267532348633,
-      "learning_rate": 1.3118171795128798e-05,
-      "loss": 1.7698,
+      "epoch": 0.43,
+      "grad_norm": 12.648981094360352,
+      "learning_rate": 1.7127557210392004e-05,
+      "loss": 0.4018,
       "step": 3434
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 16.004776000976562,
-      "learning_rate": 1.3116167184524406e-05,
-      "loss": 2.1169,
+      "epoch": 0.43,
+      "grad_norm": 30.49363136291504,
+      "learning_rate": 1.7126720495335314e-05,
+      "loss": 2.0054,
       "step": 3435
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 36.55406188964844,
-      "learning_rate": 1.3114162573920018e-05,
-      "loss": 2.0371,
+      "epoch": 0.43,
+      "grad_norm": 18.023563385009766,
+      "learning_rate": 1.7125883780278628e-05,
+      "loss": 3.2501,
       "step": 3436
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 20.354867935180664,
-      "learning_rate": 1.3112157963315626e-05,
-      "loss": 1.5596,
+      "epoch": 0.43,
+      "grad_norm": 11.947152137756348,
+      "learning_rate": 1.712504706522194e-05,
+      "loss": 3.3163,
       "step": 3437
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 13.395331382751465,
-      "learning_rate": 1.3110153352711236e-05,
-      "loss": 1.3172,
+      "epoch": 0.43,
+      "grad_norm": 22.248554229736328,
+      "learning_rate": 1.7124210350165252e-05,
+      "loss": 3.5797,
       "step": 3438
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 17.100326538085938,
-      "learning_rate": 1.3108148742106848e-05,
-      "loss": 1.9774,
+      "epoch": 0.43,
+      "grad_norm": 22.091278076171875,
+      "learning_rate": 1.7123373635108565e-05,
+      "loss": 1.3709,
       "step": 3439
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 13.835416793823242,
-      "learning_rate": 1.3106144131502456e-05,
-      "loss": 1.8643,
+      "epoch": 0.43,
+      "grad_norm": 15.646050453186035,
+      "learning_rate": 1.712253692005188e-05,
+      "loss": 2.442,
       "step": 3440
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 15.755050659179688,
-      "learning_rate": 1.3104139520898067e-05,
-      "loss": 3.0466,
+      "epoch": 0.43,
+      "grad_norm": 11.564681053161621,
+      "learning_rate": 1.712170020499519e-05,
+      "loss": 2.0503,
       "step": 3441
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 19.994091033935547,
-      "learning_rate": 1.3102134910293675e-05,
-      "loss": 1.9489,
+      "epoch": 0.43,
+      "grad_norm": 10.883163452148438,
+      "learning_rate": 1.7120863489938503e-05,
+      "loss": 3.0713,
       "step": 3442
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 10.072383880615234,
-      "learning_rate": 1.3100130299689287e-05,
-      "loss": 2.0764,
+      "epoch": 0.43,
+      "grad_norm": 17.00910758972168,
+      "learning_rate": 1.7120026774881813e-05,
+      "loss": 2.0418,
       "step": 3443
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 10.777514457702637,
-      "learning_rate": 1.3098125689084897e-05,
-      "loss": 1.2584,
+      "epoch": 0.43,
+      "grad_norm": 6.5379958152771,
+      "learning_rate": 1.7119190059825127e-05,
+      "loss": 0.9993,
       "step": 3444
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 24.992101669311523,
-      "learning_rate": 1.3096121078480505e-05,
-      "loss": 2.3589,
+      "epoch": 0.43,
+      "grad_norm": 25.576587677001953,
+      "learning_rate": 1.711835334476844e-05,
+      "loss": 4.4599,
       "step": 3445
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 13.744710922241211,
-      "learning_rate": 1.3094116467876117e-05,
-      "loss": 1.6164,
+      "epoch": 0.43,
+      "grad_norm": 14.60910415649414,
+      "learning_rate": 1.711751662971175e-05,
+      "loss": 2.8837,
       "step": 3446
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 9.5302095413208,
-      "learning_rate": 1.3092111857271725e-05,
-      "loss": 1.4418,
+      "epoch": 0.43,
+      "grad_norm": 15.904330253601074,
+      "learning_rate": 1.7116679914655065e-05,
+      "loss": 2.9878,
       "step": 3447
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 16.148855209350586,
-      "learning_rate": 1.3090107246667335e-05,
-      "loss": 1.4431,
+      "epoch": 0.43,
+      "grad_norm": 13.964025497436523,
+      "learning_rate": 1.7115843199598378e-05,
+      "loss": 2.119,
       "step": 3448
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 31.971755981445312,
-      "learning_rate": 1.3088102636062945e-05,
-      "loss": 1.506,
+      "epoch": 0.43,
+      "grad_norm": 11.501665115356445,
+      "learning_rate": 1.711500648454169e-05,
+      "loss": 1.7467,
       "step": 3449
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 18.346967697143555,
-      "learning_rate": 1.3086098025458556e-05,
-      "loss": 1.6131,
+      "epoch": 0.43,
+      "grad_norm": 13.98098087310791,
+      "learning_rate": 1.7114169769485002e-05,
+      "loss": 2.5961,
       "step": 3450
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 29.8282527923584,
-      "learning_rate": 1.3084093414854167e-05,
-      "loss": 2.3438,
+      "epoch": 0.43,
+      "grad_norm": 32.3702278137207,
+      "learning_rate": 1.7113333054428316e-05,
+      "loss": 1.9346,
       "step": 3451
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 19.195152282714844,
-      "learning_rate": 1.3082088804249776e-05,
-      "loss": 1.9727,
+      "epoch": 0.43,
+      "grad_norm": 21.77657127380371,
+      "learning_rate": 1.711249633937163e-05,
+      "loss": 2.1284,
       "step": 3452
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 17.298931121826172,
-      "learning_rate": 1.3080084193645386e-05,
-      "loss": 1.5604,
+      "epoch": 0.43,
+      "grad_norm": 14.308854103088379,
+      "learning_rate": 1.711165962431494e-05,
+      "loss": 1.0683,
       "step": 3453
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 10.650527954101562,
-      "learning_rate": 1.3078079583040994e-05,
-      "loss": 1.0794,
+      "epoch": 0.43,
+      "grad_norm": 15.169837951660156,
+      "learning_rate": 1.7110822909258253e-05,
+      "loss": 1.1417,
       "step": 3454
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 16.008625030517578,
-      "learning_rate": 1.3076074972436606e-05,
-      "loss": 1.7114,
+      "epoch": 0.43,
+      "grad_norm": 8.071978569030762,
+      "learning_rate": 1.7109986194201567e-05,
+      "loss": 1.4408,
       "step": 3455
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 15.615730285644531,
-      "learning_rate": 1.3074070361832216e-05,
-      "loss": 2.0135,
+      "epoch": 0.43,
+      "grad_norm": 17.008346557617188,
+      "learning_rate": 1.7109149479144877e-05,
+      "loss": 1.1823,
       "step": 3456
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 17.91425132751465,
-      "learning_rate": 1.3072065751227824e-05,
-      "loss": 1.8132,
+      "epoch": 0.43,
+      "grad_norm": 29.824209213256836,
+      "learning_rate": 1.710831276408819e-05,
+      "loss": 2.3075,
       "step": 3457
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 16.860435485839844,
-      "learning_rate": 1.3070061140623436e-05,
-      "loss": 1.9834,
+      "epoch": 0.43,
+      "grad_norm": 16.685344696044922,
+      "learning_rate": 1.7107476049031504e-05,
+      "loss": 3.901,
       "step": 3458
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 12.117476463317871,
-      "learning_rate": 1.3068056530019045e-05,
-      "loss": 1.62,
+      "epoch": 0.43,
+      "grad_norm": 6.225277423858643,
+      "learning_rate": 1.7106639333974818e-05,
+      "loss": 0.1791,
       "step": 3459
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 18.71920394897461,
-      "learning_rate": 1.3066051919414655e-05,
-      "loss": 2.2109,
+      "epoch": 0.43,
+      "grad_norm": 25.91197967529297,
+      "learning_rate": 1.710580261891813e-05,
+      "loss": 1.9825,
       "step": 3460
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 18.306224822998047,
-      "learning_rate": 1.3064047308810263e-05,
-      "loss": 2.2506,
+      "epoch": 0.43,
+      "grad_norm": 19.51018714904785,
+      "learning_rate": 1.7104965903861442e-05,
+      "loss": 2.7875,
       "step": 3461
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 27.221975326538086,
-      "learning_rate": 1.3062042698205875e-05,
-      "loss": 2.4873,
+      "epoch": 0.43,
+      "grad_norm": 14.964982986450195,
+      "learning_rate": 1.7104129188804756e-05,
+      "loss": 1.3888,
       "step": 3462
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 31.973318099975586,
-      "learning_rate": 1.3060038087601485e-05,
-      "loss": 2.6493,
+      "epoch": 0.43,
+      "grad_norm": 12.790801048278809,
+      "learning_rate": 1.7103292473748066e-05,
+      "loss": 3.0152,
       "step": 3463
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 10.039262771606445,
-      "learning_rate": 1.3058033476997093e-05,
-      "loss": 1.6217,
+      "epoch": 0.43,
+      "grad_norm": 22.963237762451172,
+      "learning_rate": 1.710245575869138e-05,
+      "loss": 2.8846,
       "step": 3464
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 17.775131225585938,
-      "learning_rate": 1.3056028866392705e-05,
-      "loss": 1.6111,
+      "epoch": 0.43,
+      "grad_norm": 5.108229160308838,
+      "learning_rate": 1.7101619043634693e-05,
+      "loss": 2.2731,
       "step": 3465
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 41.91878890991211,
-      "learning_rate": 1.3054024255788314e-05,
-      "loss": 2.4916,
+      "epoch": 0.43,
+      "grad_norm": 15.958724021911621,
+      "learning_rate": 1.7100782328578004e-05,
+      "loss": 0.1816,
       "step": 3466
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 19.899276733398438,
-      "learning_rate": 1.3052019645183924e-05,
-      "loss": 2.2337,
+      "epoch": 0.44,
+      "grad_norm": 4.365947246551514,
+      "learning_rate": 1.7099945613521317e-05,
+      "loss": 1.2604,
       "step": 3467
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 12.127092361450195,
-      "learning_rate": 1.3050015034579534e-05,
-      "loss": 1.9265,
+      "epoch": 0.44,
+      "grad_norm": 10.918221473693848,
+      "learning_rate": 1.709910889846463e-05,
+      "loss": 1.8209,
       "step": 3468
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 28.322265625,
-      "learning_rate": 1.3048010423975144e-05,
-      "loss": 1.8237,
+      "epoch": 0.44,
+      "grad_norm": 26.80780601501465,
+      "learning_rate": 1.709827218340794e-05,
+      "loss": 2.0377,
       "step": 3469
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 11.918156623840332,
-      "learning_rate": 1.3046005813370754e-05,
-      "loss": 2.1019,
+      "epoch": 0.44,
+      "grad_norm": 11.148392677307129,
+      "learning_rate": 1.7097435468351255e-05,
+      "loss": 1.6302,
       "step": 3470
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 10.887943267822266,
-      "learning_rate": 1.3044001202766364e-05,
-      "loss": 2.118,
+      "epoch": 0.44,
+      "grad_norm": 11.951037406921387,
+      "learning_rate": 1.7096598753294565e-05,
+      "loss": 1.1111,
       "step": 3471
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 14.735898971557617,
-      "learning_rate": 1.3041996592161974e-05,
-      "loss": 1.5392,
+      "epoch": 0.44,
+      "grad_norm": 5.3847975730896,
+      "learning_rate": 1.709576203823788e-05,
+      "loss": 0.1638,
       "step": 3472
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 22.728878021240234,
-      "learning_rate": 1.3039991981557582e-05,
-      "loss": 1.9459,
+      "epoch": 0.44,
+      "grad_norm": 28.452150344848633,
+      "learning_rate": 1.7094925323181192e-05,
+      "loss": 5.5813,
       "step": 3473
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 9.619507789611816,
-      "learning_rate": 1.3037987370953194e-05,
-      "loss": 1.1765,
+      "epoch": 0.44,
+      "grad_norm": 16.50628662109375,
+      "learning_rate": 1.7094088608124503e-05,
+      "loss": 2.004,
       "step": 3474
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 13.704413414001465,
-      "learning_rate": 1.3035982760348804e-05,
-      "loss": 1.5395,
+      "epoch": 0.44,
+      "grad_norm": 27.793264389038086,
+      "learning_rate": 1.7093251893067816e-05,
+      "loss": 3.2009,
       "step": 3475
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 13.936915397644043,
-      "learning_rate": 1.3033978149744413e-05,
-      "loss": 1.2055,
+      "epoch": 0.44,
+      "grad_norm": 16.817726135253906,
+      "learning_rate": 1.709241517801113e-05,
+      "loss": 1.409,
       "step": 3476
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 23.567424774169922,
-      "learning_rate": 1.3031973539140024e-05,
-      "loss": 1.7821,
+      "epoch": 0.44,
+      "grad_norm": 17.985727310180664,
+      "learning_rate": 1.709157846295444e-05,
+      "loss": 1.8906,
       "step": 3477
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 17.09320640563965,
-      "learning_rate": 1.3029968928535633e-05,
-      "loss": 2.4582,
+      "epoch": 0.44,
+      "grad_norm": 13.15288257598877,
+      "learning_rate": 1.7090741747897754e-05,
+      "loss": 2.555,
       "step": 3478
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 12.296806335449219,
-      "learning_rate": 1.3027964317931243e-05,
-      "loss": 1.4382,
+      "epoch": 0.44,
+      "grad_norm": 24.595491409301758,
+      "learning_rate": 1.7089905032841067e-05,
+      "loss": 1.6678,
       "step": 3479
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 12.535650253295898,
-      "learning_rate": 1.3025959707326851e-05,
-      "loss": 1.2141,
-      "step": 3480
-    },
-    {
-      "epoch": 1.05,
-      "eval_loss": 0.22333772480487823,
-      "eval_runtime": 43.6637,
-      "eval_samples_per_second": 33.873,
-      "eval_steps_per_second": 33.873,
+      "epoch": 0.44,
+      "grad_norm": 12.645052909851074,
+      "learning_rate": 1.708906831778438e-05,
+      "loss": 2.2637,
       "step": 3480
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 8.423393249511719,
-      "learning_rate": 1.3023955096722463e-05,
-      "loss": 1.2467,
+      "epoch": 0.44,
+      "grad_norm": 9.597436904907227,
+      "learning_rate": 1.708823160272769e-05,
+      "loss": 2.2271,
       "step": 3481
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 13.054819107055664,
-      "learning_rate": 1.3021950486118073e-05,
-      "loss": 1.2692,
+      "epoch": 0.44,
+      "grad_norm": 10.511176109313965,
+      "learning_rate": 1.7087394887671005e-05,
+      "loss": 1.7628,
       "step": 3482
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 67.5218734741211,
-      "learning_rate": 1.3019945875513682e-05,
-      "loss": 2.0009,
+      "epoch": 0.44,
+      "grad_norm": 7.529833793640137,
+      "learning_rate": 1.708655817261432e-05,
+      "loss": 1.633,
       "step": 3483
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 51.58919143676758,
-      "learning_rate": 1.3017941264909293e-05,
-      "loss": 3.2618,
+      "epoch": 0.44,
+      "grad_norm": 24.71387481689453,
+      "learning_rate": 1.708572145755763e-05,
+      "loss": 0.5247,
       "step": 3484
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 24.077619552612305,
-      "learning_rate": 1.3015936654304902e-05,
-      "loss": 2.4753,
+      "epoch": 0.44,
+      "grad_norm": 28.796314239501953,
+      "learning_rate": 1.7084884742500943e-05,
+      "loss": 2.1186,
       "step": 3485
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 25.09623146057129,
-      "learning_rate": 1.3013932043700512e-05,
-      "loss": 1.9661,
+      "epoch": 0.44,
+      "grad_norm": 17.669151306152344,
+      "learning_rate": 1.7084048027444256e-05,
+      "loss": 3.1068,
       "step": 3486
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 33.02434539794922,
-      "learning_rate": 1.3011927433096124e-05,
-      "loss": 2.3949,
+      "epoch": 0.44,
+      "grad_norm": 4.942477703094482,
+      "learning_rate": 1.708321131238757e-05,
+      "loss": 0.2292,
       "step": 3487
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 16.488039016723633,
-      "learning_rate": 1.3009922822491732e-05,
-      "loss": 2.4064,
+      "epoch": 0.44,
+      "grad_norm": 37.090293884277344,
+      "learning_rate": 1.708237459733088e-05,
+      "loss": 3.1592,
       "step": 3488
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 24.7977237701416,
-      "learning_rate": 1.3007918211887342e-05,
-      "loss": 1.2102,
+      "epoch": 0.44,
+      "grad_norm": 12.799302101135254,
+      "learning_rate": 1.7081537882274194e-05,
+      "loss": 2.6949,
       "step": 3489
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 16.36675262451172,
-      "learning_rate": 1.300591360128295e-05,
-      "loss": 1.4976,
+      "epoch": 0.44,
+      "grad_norm": 11.44233512878418,
+      "learning_rate": 1.7080701167217507e-05,
+      "loss": 1.5554,
       "step": 3490
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 16.940837860107422,
-      "learning_rate": 1.3003908990678562e-05,
-      "loss": 1.9573,
+      "epoch": 0.44,
+      "grad_norm": 16.598114013671875,
+      "learning_rate": 1.7079864452160818e-05,
+      "loss": 2.6653,
       "step": 3491
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 17.5327091217041,
-      "learning_rate": 1.300190438007417e-05,
-      "loss": 1.9867,
+      "epoch": 0.44,
+      "grad_norm": 19.33988380432129,
+      "learning_rate": 1.707902773710413e-05,
+      "loss": 3.1783,
       "step": 3492
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 14.204058647155762,
-      "learning_rate": 1.299989976946978e-05,
-      "loss": 1.8624,
+      "epoch": 0.44,
+      "grad_norm": 12.424285888671875,
+      "learning_rate": 1.7078191022047445e-05,
+      "loss": 0.5065,
       "step": 3493
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 41.586021423339844,
-      "learning_rate": 1.2997895158865392e-05,
-      "loss": 3.3605,
+      "epoch": 0.44,
+      "grad_norm": 26.837005615234375,
+      "learning_rate": 1.7077354306990755e-05,
+      "loss": 2.4939,
       "step": 3494
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 18.94367218017578,
-      "learning_rate": 1.2995890548261001e-05,
-      "loss": 1.3739,
+      "epoch": 0.44,
+      "grad_norm": 28.414806365966797,
+      "learning_rate": 1.707651759193407e-05,
+      "loss": 2.4368,
       "step": 3495
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 23.99241828918457,
-      "learning_rate": 1.2993885937656611e-05,
-      "loss": 2.0589,
+      "epoch": 0.44,
+      "grad_norm": 23.595781326293945,
+      "learning_rate": 1.707568087687738e-05,
+      "loss": 1.5891,
       "step": 3496
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 11.752195358276367,
-      "learning_rate": 1.2991881327052221e-05,
-      "loss": 1.3307,
+      "epoch": 0.44,
+      "grad_norm": 11.007417678833008,
+      "learning_rate": 1.7074844161820693e-05,
+      "loss": 2.7005,
       "step": 3497
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 16.153244018554688,
-      "learning_rate": 1.2989876716447831e-05,
-      "loss": 1.6504,
+      "epoch": 0.44,
+      "grad_norm": 14.353242874145508,
+      "learning_rate": 1.7074007446764006e-05,
+      "loss": 2.0073,
       "step": 3498
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 15.57225513458252,
-      "learning_rate": 1.2987872105843441e-05,
-      "loss": 2.247,
+      "epoch": 0.44,
+      "grad_norm": 15.713659286499023,
+      "learning_rate": 1.7073170731707317e-05,
+      "loss": 1.9655,
       "step": 3499
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 14.485071182250977,
-      "learning_rate": 1.2985867495239051e-05,
-      "loss": 1.7557,
+      "epoch": 0.44,
+      "grad_norm": 40.209938049316406,
+      "learning_rate": 1.707233401665063e-05,
+      "loss": 2.9224,
       "step": 3500
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 8.054557800292969,
-      "learning_rate": 1.2983862884634661e-05,
-      "loss": 1.3367,
+      "epoch": 0.44,
+      "grad_norm": 11.3770170211792,
+      "learning_rate": 1.7071497301593944e-05,
+      "loss": 2.0867,
       "step": 3501
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 19.42696189880371,
-      "learning_rate": 1.298185827403027e-05,
-      "loss": 1.7359,
+      "epoch": 0.44,
+      "grad_norm": 7.715713977813721,
+      "learning_rate": 1.7070660586537254e-05,
+      "loss": 0.7953,
       "step": 3502
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 21.340444564819336,
-      "learning_rate": 1.2979853663425882e-05,
-      "loss": 2.8674,
+      "epoch": 0.44,
+      "grad_norm": 21.278911590576172,
+      "learning_rate": 1.7069823871480568e-05,
+      "loss": 1.9252,
       "step": 3503
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 16.97898292541504,
-      "learning_rate": 1.297784905282149e-05,
-      "loss": 2.4501,
+      "epoch": 0.44,
+      "grad_norm": 20.147659301757812,
+      "learning_rate": 1.706898715642388e-05,
+      "loss": 2.7377,
       "step": 3504
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 17.556812286376953,
-      "learning_rate": 1.29758444422171e-05,
-      "loss": 1.421,
+      "epoch": 0.44,
+      "grad_norm": 42.69921112060547,
+      "learning_rate": 1.7068150441367192e-05,
+      "loss": 2.561,
       "step": 3505
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 12.702077865600586,
-      "learning_rate": 1.2973839831612712e-05,
-      "loss": 2.1779,
+      "epoch": 0.44,
+      "grad_norm": 8.338047981262207,
+      "learning_rate": 1.7067313726310505e-05,
+      "loss": 1.3805,
       "step": 3506
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 17.422500610351562,
-      "learning_rate": 1.297183522100832e-05,
-      "loss": 2.1388,
+      "epoch": 0.44,
+      "grad_norm": 7.393989562988281,
+      "learning_rate": 1.706647701125382e-05,
+      "loss": 0.876,
       "step": 3507
     },
     {
-      "epoch": 1.05,
-      "grad_norm": 27.59862518310547,
-      "learning_rate": 1.296983061040393e-05,
-      "loss": 3.23,
+      "epoch": 0.44,
+      "grad_norm": 11.099953651428223,
+      "learning_rate": 1.7065640296197133e-05,
+      "loss": 1.3944,
       "step": 3508
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 11.814319610595703,
-      "learning_rate": 1.2967825999799539e-05,
-      "loss": 1.5021,
+      "epoch": 0.44,
+      "grad_norm": 12.176742553710938,
+      "learning_rate": 1.7064803581140443e-05,
+      "loss": 2.0733,
       "step": 3509
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 10.947737693786621,
-      "learning_rate": 1.296582138919515e-05,
-      "loss": 1.298,
+      "epoch": 0.44,
+      "grad_norm": 15.4348783493042,
+      "learning_rate": 1.7063966866083757e-05,
+      "loss": 2.6984,
       "step": 3510
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 14.834168434143066,
-      "learning_rate": 1.2963816778590759e-05,
-      "loss": 1.6569,
+      "epoch": 0.44,
+      "grad_norm": 6.117305278778076,
+      "learning_rate": 1.706313015102707e-05,
+      "loss": 1.8807,
       "step": 3511
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 27.23177719116211,
-      "learning_rate": 1.2961812167986369e-05,
-      "loss": 1.6755,
+      "epoch": 0.44,
+      "grad_norm": 11.953407287597656,
+      "learning_rate": 1.706229343597038e-05,
+      "loss": 1.2245,
       "step": 3512
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 16.08486557006836,
-      "learning_rate": 1.295980755738198e-05,
-      "loss": 1.8572,
+      "epoch": 0.44,
+      "grad_norm": 29.934341430664062,
+      "learning_rate": 1.7061456720913694e-05,
+      "loss": 2.3307,
       "step": 3513
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 23.723173141479492,
-      "learning_rate": 1.2957802946777589e-05,
-      "loss": 2.7123,
+      "epoch": 0.44,
+      "grad_norm": 29.04018783569336,
+      "learning_rate": 1.7060620005857008e-05,
+      "loss": 2.1927,
       "step": 3514
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 15.125565528869629,
-      "learning_rate": 1.29557983361732e-05,
-      "loss": 2.5594,
+      "epoch": 0.44,
+      "grad_norm": 16.454509735107422,
+      "learning_rate": 1.705978329080032e-05,
+      "loss": 2.3173,
       "step": 3515
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 25.76398468017578,
-      "learning_rate": 1.295379372556881e-05,
-      "loss": 1.7984,
+      "epoch": 0.44,
+      "grad_norm": 16.549463272094727,
+      "learning_rate": 1.7058946575743632e-05,
+      "loss": 1.8457,
       "step": 3516
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 18.840059280395508,
-      "learning_rate": 1.295178911496442e-05,
-      "loss": 2.232,
+      "epoch": 0.44,
+      "grad_norm": 9.819560050964355,
+      "learning_rate": 1.7058109860686945e-05,
+      "loss": 2.0945,
       "step": 3517
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 35.28987503051758,
-      "learning_rate": 1.294978450436003e-05,
-      "loss": 2.267,
+      "epoch": 0.44,
+      "grad_norm": 14.210806846618652,
+      "learning_rate": 1.705727314563026e-05,
+      "loss": 2.162,
       "step": 3518
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 7.588108062744141,
-      "learning_rate": 1.294777989375564e-05,
-      "loss": 1.2528,
+      "epoch": 0.44,
+      "grad_norm": 37.674293518066406,
+      "learning_rate": 1.705643643057357e-05,
+      "loss": 1.5869,
       "step": 3519
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 13.926599502563477,
-      "learning_rate": 1.294577528315125e-05,
-      "loss": 1.7358,
+      "epoch": 0.44,
+      "grad_norm": 11.009414672851562,
+      "learning_rate": 1.7055599715516883e-05,
+      "loss": 1.1545,
       "step": 3520
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 16.25490951538086,
-      "learning_rate": 1.2943770672546858e-05,
-      "loss": 2.2304,
+      "epoch": 0.44,
+      "grad_norm": 21.234272003173828,
+      "learning_rate": 1.7054763000460197e-05,
+      "loss": 2.4492,
       "step": 3521
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 30.820024490356445,
-      "learning_rate": 1.294176606194247e-05,
-      "loss": 2.0343,
+      "epoch": 0.44,
+      "grad_norm": 19.242658615112305,
+      "learning_rate": 1.7053926285403507e-05,
+      "loss": 2.7148,
       "step": 3522
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 15.902755737304688,
-      "learning_rate": 1.2939761451338078e-05,
-      "loss": 1.5815,
+      "epoch": 0.44,
+      "grad_norm": 16.382221221923828,
+      "learning_rate": 1.705308957034682e-05,
+      "loss": 2.6571,
       "step": 3523
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 22.500484466552734,
-      "learning_rate": 1.2937756840733688e-05,
-      "loss": 2.0932,
+      "epoch": 0.44,
+      "grad_norm": 5.982744216918945,
+      "learning_rate": 1.705225285529013e-05,
+      "loss": 2.6794,
       "step": 3524
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 11.80160903930664,
-      "learning_rate": 1.29357522301293e-05,
-      "loss": 1.8902,
+      "epoch": 0.44,
+      "grad_norm": 15.731637001037598,
+      "learning_rate": 1.7051416140233444e-05,
+      "loss": 1.7883,
       "step": 3525
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 24.187225341796875,
-      "learning_rate": 1.2933747619524908e-05,
-      "loss": 1.9731,
+      "epoch": 0.44,
+      "grad_norm": 15.043095588684082,
+      "learning_rate": 1.7050579425176758e-05,
+      "loss": 1.3555,
       "step": 3526
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 13.519028663635254,
-      "learning_rate": 1.2931743008920519e-05,
-      "loss": 1.9801,
+      "epoch": 0.44,
+      "grad_norm": 15.139749526977539,
+      "learning_rate": 1.704974271012007e-05,
+      "loss": 4.0904,
       "step": 3527
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 54.62118911743164,
-      "learning_rate": 1.2929738398316127e-05,
-      "loss": 3.4219,
+      "epoch": 0.44,
+      "grad_norm": 13.007729530334473,
+      "learning_rate": 1.7048905995063382e-05,
+      "loss": 1.2369,
       "step": 3528
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 12.734724044799805,
-      "learning_rate": 1.2927733787711739e-05,
-      "loss": 2.3175,
+      "epoch": 0.44,
+      "grad_norm": 9.70779037475586,
+      "learning_rate": 1.7048069280006696e-05,
+      "loss": 1.2312,
       "step": 3529
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 15.261761665344238,
-      "learning_rate": 1.2925729177107349e-05,
-      "loss": 2.6037,
+      "epoch": 0.44,
+      "grad_norm": 12.419896125793457,
+      "learning_rate": 1.7047232564950006e-05,
+      "loss": 2.3041,
       "step": 3530
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 19.809280395507812,
-      "learning_rate": 1.2923724566502957e-05,
-      "loss": 1.9901,
+      "epoch": 0.44,
+      "grad_norm": 6.6420698165893555,
+      "learning_rate": 1.704639584989332e-05,
+      "loss": 0.5756,
       "step": 3531
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 18.573335647583008,
-      "learning_rate": 1.2921719955898569e-05,
-      "loss": 2.4825,
+      "epoch": 0.44,
+      "grad_norm": 9.08359146118164,
+      "learning_rate": 1.7045559134836633e-05,
+      "loss": 1.7639,
       "step": 3532
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 17.13604736328125,
-      "learning_rate": 1.2919715345294177e-05,
-      "loss": 1.3552,
+      "epoch": 0.44,
+      "grad_norm": 11.831747055053711,
+      "learning_rate": 1.7044722419779943e-05,
+      "loss": 0.6694,
       "step": 3533
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 14.92817497253418,
-      "learning_rate": 1.2917710734689787e-05,
-      "loss": 2.1547,
+      "epoch": 0.44,
+      "grad_norm": 29.910619735717773,
+      "learning_rate": 1.7043885704723257e-05,
+      "loss": 2.3868,
       "step": 3534
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 26.906400680541992,
-      "learning_rate": 1.2915706124085396e-05,
-      "loss": 1.741,
+      "epoch": 0.44,
+      "grad_norm": 11.793986320495605,
+      "learning_rate": 1.704304898966657e-05,
+      "loss": 2.742,
       "step": 3535
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 17.950481414794922,
-      "learning_rate": 1.2913701513481008e-05,
-      "loss": 1.7533,
+      "epoch": 0.44,
+      "grad_norm": 15.16977310180664,
+      "learning_rate": 1.7042212274609884e-05,
+      "loss": 3.2701,
       "step": 3536
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 18.598140716552734,
-      "learning_rate": 1.2911696902876618e-05,
-      "loss": 1.9817,
+      "epoch": 0.44,
+      "grad_norm": 25.637859344482422,
+      "learning_rate": 1.7041375559553195e-05,
+      "loss": 2.5424,
       "step": 3537
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 20.901409149169922,
-      "learning_rate": 1.2909692292272226e-05,
-      "loss": 1.8785,
+      "epoch": 0.44,
+      "grad_norm": 15.422876358032227,
+      "learning_rate": 1.7040538844496508e-05,
+      "loss": 2.5337,
       "step": 3538
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 13.569160461425781,
-      "learning_rate": 1.2907687681667838e-05,
-      "loss": 2.0731,
+      "epoch": 0.44,
+      "grad_norm": 24.96222496032715,
+      "learning_rate": 1.7039702129439822e-05,
+      "loss": 1.9194,
       "step": 3539
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 16.909326553344727,
-      "learning_rate": 1.2905683071063446e-05,
-      "loss": 1.912,
+      "epoch": 0.44,
+      "grad_norm": 26.966215133666992,
+      "learning_rate": 1.7038865414383132e-05,
+      "loss": 1.4356,
       "step": 3540
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 15.05936050415039,
-      "learning_rate": 1.2903678460459056e-05,
-      "loss": 3.6215,
+      "epoch": 0.44,
+      "grad_norm": 9.480104446411133,
+      "learning_rate": 1.7038028699326446e-05,
+      "loss": 1.5081,
       "step": 3541
     },
     {
-      "epoch": 1.06,
-      "grad_norm": 12.6626615524292,
-      "learning_rate": 1.2901673849854668e-05,
-      "loss": 1.6078,
+      "epoch": 0.44,
+      "grad_norm": 6.7902445793151855,
+      "learning_rate": 1.703719198426976e-05,
+      "loss": 0.9859,
       "step": 3542
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 12.250699043273926,
-      "learning_rate": 1.2899669239250276e-05,
-      "loss": 1.9456,
+      "epoch": 0.44,
+      "grad_norm": 7.768956184387207,
+      "learning_rate": 1.7036355269213073e-05,
+      "loss": 1.11,
       "step": 3543
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 28.121313095092773,
-      "learning_rate": 1.2897664628645887e-05,
-      "loss": 2.6878,
+      "epoch": 0.44,
+      "grad_norm": 11.356424331665039,
+      "learning_rate": 1.7035518554156383e-05,
+      "loss": 2.138,
       "step": 3544
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 21.93025779724121,
-      "learning_rate": 1.2895660018041497e-05,
-      "loss": 2.364,
+      "epoch": 0.44,
+      "grad_norm": 16.411840438842773,
+      "learning_rate": 1.7034681839099697e-05,
+      "loss": 1.1534,
       "step": 3545
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 11.087515830993652,
-      "learning_rate": 1.2893655407437107e-05,
-      "loss": 2.0794,
+      "epoch": 0.45,
+      "grad_norm": 17.7725830078125,
+      "learning_rate": 1.703384512404301e-05,
+      "loss": 2.2293,
       "step": 3546
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 38.5289421081543,
-      "learning_rate": 1.2891650796832715e-05,
-      "loss": 2.4577,
+      "epoch": 0.45,
+      "grad_norm": 12.674851417541504,
+      "learning_rate": 1.703300840898632e-05,
+      "loss": 1.0665,
       "step": 3547
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 29.22125816345215,
-      "learning_rate": 1.2889646186228327e-05,
-      "loss": 2.4881,
+      "epoch": 0.45,
+      "grad_norm": 6.166348457336426,
+      "learning_rate": 1.7032171693929635e-05,
+      "loss": 1.0158,
       "step": 3548
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 25.916982650756836,
-      "learning_rate": 1.2887641575623937e-05,
-      "loss": 1.4721,
+      "epoch": 0.45,
+      "grad_norm": 4.511837482452393,
+      "learning_rate": 1.7031334978872945e-05,
+      "loss": 0.0695,
       "step": 3549
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 20.641904830932617,
-      "learning_rate": 1.2885636965019545e-05,
-      "loss": 1.6015,
+      "epoch": 0.45,
+      "grad_norm": 3.8872060775756836,
+      "learning_rate": 1.703049826381626e-05,
+      "loss": 0.1033,
       "step": 3550
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 14.579924583435059,
-      "learning_rate": 1.2883632354415157e-05,
-      "loss": 1.638,
+      "epoch": 0.45,
+      "grad_norm": 19.17310905456543,
+      "learning_rate": 1.7029661548759572e-05,
+      "loss": 3.2545,
       "step": 3551
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 36.379661560058594,
-      "learning_rate": 1.2881627743810766e-05,
-      "loss": 2.2669,
+      "epoch": 0.45,
+      "grad_norm": 3.6682071685791016,
+      "learning_rate": 1.7028824833702882e-05,
+      "loss": 0.087,
       "step": 3552
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 16.51966094970703,
-      "learning_rate": 1.2879623133206376e-05,
-      "loss": 1.7893,
+      "epoch": 0.45,
+      "grad_norm": 5.926275730133057,
+      "learning_rate": 1.7027988118646196e-05,
+      "loss": 0.7951,
       "step": 3553
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 23.127784729003906,
-      "learning_rate": 1.2877618522601984e-05,
-      "loss": 2.266,
+      "epoch": 0.45,
+      "grad_norm": 6.377835273742676,
+      "learning_rate": 1.7027151403589506e-05,
+      "loss": 1.3846,
       "step": 3554
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 17.028398513793945,
-      "learning_rate": 1.2875613911997596e-05,
-      "loss": 1.6085,
+      "epoch": 0.45,
+      "grad_norm": 6.7998127937316895,
+      "learning_rate": 1.702631468853282e-05,
+      "loss": 1.9433,
       "step": 3555
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 9.327510833740234,
-      "learning_rate": 1.2873609301393206e-05,
-      "loss": 1.155,
+      "epoch": 0.45,
+      "grad_norm": 10.768909454345703,
+      "learning_rate": 1.7025477973476134e-05,
+      "loss": 1.9161,
       "step": 3556
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 9.496240615844727,
-      "learning_rate": 1.2871604690788814e-05,
-      "loss": 1.3495,
+      "epoch": 0.45,
+      "grad_norm": 19.759462356567383,
+      "learning_rate": 1.7024641258419447e-05,
+      "loss": 2.5175,
       "step": 3557
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 10.282184600830078,
-      "learning_rate": 1.2869600080184426e-05,
-      "loss": 1.4117,
+      "epoch": 0.45,
+      "grad_norm": 15.046670913696289,
+      "learning_rate": 1.7023804543362758e-05,
+      "loss": 3.0608,
       "step": 3558
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 15.490508079528809,
-      "learning_rate": 1.2867595469580034e-05,
-      "loss": 1.5863,
+      "epoch": 0.45,
+      "grad_norm": 14.481589317321777,
+      "learning_rate": 1.702296782830607e-05,
+      "loss": 1.6803,
       "step": 3559
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 17.8187198638916,
-      "learning_rate": 1.2865590858975645e-05,
-      "loss": 1.2819,
+      "epoch": 0.45,
+      "grad_norm": 12.190134048461914,
+      "learning_rate": 1.7022131113249385e-05,
+      "loss": 1.4591,
       "step": 3560
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 17.21719741821289,
-      "learning_rate": 1.2863586248371256e-05,
-      "loss": 2.0228,
+      "epoch": 0.45,
+      "grad_norm": 3.3954739570617676,
+      "learning_rate": 1.7021294398192695e-05,
+      "loss": 1.704,
       "step": 3561
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 23.23194694519043,
-      "learning_rate": 1.2861581637766865e-05,
-      "loss": 1.858,
+      "epoch": 0.45,
+      "grad_norm": 10.886055946350098,
+      "learning_rate": 1.702045768313601e-05,
+      "loss": 1.1718,
       "step": 3562
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 25.972341537475586,
-      "learning_rate": 1.2859577027162475e-05,
-      "loss": 1.3268,
+      "epoch": 0.45,
+      "grad_norm": 14.209492683410645,
+      "learning_rate": 1.7019620968079322e-05,
+      "loss": 2.3727,
       "step": 3563
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 27.290193557739258,
-      "learning_rate": 1.2857572416558083e-05,
-      "loss": 2.174,
+      "epoch": 0.45,
+      "grad_norm": 15.119983673095703,
+      "learning_rate": 1.7018784253022636e-05,
+      "loss": 1.0797,
       "step": 3564
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 14.838099479675293,
-      "learning_rate": 1.2855567805953695e-05,
-      "loss": 1.9506,
+      "epoch": 0.45,
+      "grad_norm": 20.525146484375,
+      "learning_rate": 1.7017947537965946e-05,
+      "loss": 2.7043,
       "step": 3565
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 8.05460262298584,
-      "learning_rate": 1.2853563195349303e-05,
-      "loss": 1.2423,
+      "epoch": 0.45,
+      "grad_norm": 20.406822204589844,
+      "learning_rate": 1.701711082290926e-05,
+      "loss": 2.5528,
       "step": 3566
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 10.019596099853516,
-      "learning_rate": 1.2851558584744913e-05,
-      "loss": 1.405,
+      "epoch": 0.45,
+      "grad_norm": 12.409043312072754,
+      "learning_rate": 1.7016274107852574e-05,
+      "loss": 1.8563,
       "step": 3567
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 16.532649993896484,
-      "learning_rate": 1.2849553974140525e-05,
-      "loss": 1.5799,
+      "epoch": 0.45,
+      "grad_norm": 8.90954303741455,
+      "learning_rate": 1.7015437392795884e-05,
+      "loss": 1.0523,
       "step": 3568
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 14.281306266784668,
-      "learning_rate": 1.2847549363536134e-05,
-      "loss": 1.1582,
+      "epoch": 0.45,
+      "grad_norm": 25.932857513427734,
+      "learning_rate": 1.7014600677739198e-05,
+      "loss": 1.5185,
       "step": 3569
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 9.972299575805664,
-      "learning_rate": 1.2845544752931745e-05,
-      "loss": 1.2031,
+      "epoch": 0.45,
+      "grad_norm": 21.852935791015625,
+      "learning_rate": 1.701376396268251e-05,
+      "loss": 2.45,
       "step": 3570
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 14.451814651489258,
-      "learning_rate": 1.2843540142327354e-05,
-      "loss": 1.5313,
+      "epoch": 0.45,
+      "grad_norm": 18.34440803527832,
+      "learning_rate": 1.7012927247625825e-05,
+      "loss": 2.2476,
       "step": 3571
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 15.139348030090332,
-      "learning_rate": 1.2841535531722964e-05,
-      "loss": 1.9977,
+      "epoch": 0.45,
+      "grad_norm": 8.53489875793457,
+      "learning_rate": 1.7012090532569135e-05,
+      "loss": 0.7178,
       "step": 3572
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 25.56646156311035,
-      "learning_rate": 1.2839530921118576e-05,
-      "loss": 2.2934,
+      "epoch": 0.45,
+      "grad_norm": 14.003607749938965,
+      "learning_rate": 1.701125381751245e-05,
+      "loss": 1.3264,
       "step": 3573
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 11.112027168273926,
-      "learning_rate": 1.2837526310514184e-05,
-      "loss": 1.5678,
+      "epoch": 0.45,
+      "grad_norm": 22.757652282714844,
+      "learning_rate": 1.7010417102455762e-05,
+      "loss": 2.5163,
       "step": 3574
     },
     {
-      "epoch": 1.07,
-      "grad_norm": 14.32687759399414,
-      "learning_rate": 1.2835521699909794e-05,
-      "loss": 1.1919,
+      "epoch": 0.45,
+      "grad_norm": 9.419745445251465,
+      "learning_rate": 1.7009580387399073e-05,
+      "loss": 0.6182,
       "step": 3575
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 11.444255828857422,
-      "learning_rate": 1.2833517089305402e-05,
-      "loss": 2.2952,
+      "epoch": 0.45,
+      "grad_norm": 12.34254264831543,
+      "learning_rate": 1.7008743672342386e-05,
+      "loss": 2.8273,
       "step": 3576
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 34.31683349609375,
-      "learning_rate": 1.2831512478701014e-05,
-      "loss": 1.8054,
+      "epoch": 0.45,
+      "grad_norm": 6.758653163909912,
+      "learning_rate": 1.7007906957285697e-05,
+      "loss": 0.4665,
       "step": 3577
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 16.4061336517334,
-      "learning_rate": 1.2829507868096623e-05,
-      "loss": 1.1196,
+      "epoch": 0.45,
+      "grad_norm": 11.435264587402344,
+      "learning_rate": 1.700707024222901e-05,
+      "loss": 1.9646,
       "step": 3578
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 12.305190086364746,
-      "learning_rate": 1.2827503257492233e-05,
-      "loss": 2.2021,
+      "epoch": 0.45,
+      "grad_norm": 8.873312950134277,
+      "learning_rate": 1.7006233527172324e-05,
+      "loss": 2.4812,
       "step": 3579
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 16.25429916381836,
-      "learning_rate": 1.2825498646887845e-05,
-      "loss": 1.629,
+      "epoch": 0.45,
+      "grad_norm": 15.715291976928711,
+      "learning_rate": 1.7005396812115634e-05,
+      "loss": 1.6792,
       "step": 3580
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 25.61264419555664,
-      "learning_rate": 1.2823494036283453e-05,
-      "loss": 2.7876,
+      "epoch": 0.45,
+      "grad_norm": 23.683542251586914,
+      "learning_rate": 1.7004560097058948e-05,
+      "loss": 1.7782,
       "step": 3581
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 20.26491928100586,
-      "learning_rate": 1.2821489425679063e-05,
-      "loss": 1.1916,
+      "epoch": 0.45,
+      "grad_norm": 36.01570129394531,
+      "learning_rate": 1.7003723382002258e-05,
+      "loss": 1.6319,
       "step": 3582
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 10.56248950958252,
-      "learning_rate": 1.2819484815074671e-05,
-      "loss": 1.302,
+      "epoch": 0.45,
+      "grad_norm": 13.715401649475098,
+      "learning_rate": 1.700288666694557e-05,
+      "loss": 3.0127,
       "step": 3583
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 15.701934814453125,
-      "learning_rate": 1.2817480204470283e-05,
-      "loss": 1.5957,
+      "epoch": 0.45,
+      "grad_norm": 87.07559204101562,
+      "learning_rate": 1.7002049951888885e-05,
+      "loss": 3.2245,
       "step": 3584
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 17.989017486572266,
-      "learning_rate": 1.2815475593865893e-05,
-      "loss": 1.2969,
+      "epoch": 0.45,
+      "grad_norm": 18.235490798950195,
+      "learning_rate": 1.70012132368322e-05,
+      "loss": 3.6731,
       "step": 3585
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 22.148818969726562,
-      "learning_rate": 1.2813470983261502e-05,
-      "loss": 2.5261,
+      "epoch": 0.45,
+      "grad_norm": 16.872243881225586,
+      "learning_rate": 1.700037652177551e-05,
+      "loss": 3.3169,
       "step": 3586
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 26.945255279541016,
-      "learning_rate": 1.2811466372657113e-05,
-      "loss": 1.8008,
+      "epoch": 0.45,
+      "grad_norm": 21.76947784423828,
+      "learning_rate": 1.6999539806718823e-05,
+      "loss": 1.0939,
       "step": 3587
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 13.003181457519531,
-      "learning_rate": 1.2809461762052722e-05,
-      "loss": 1.4448,
+      "epoch": 0.45,
+      "grad_norm": 9.35506820678711,
+      "learning_rate": 1.6998703091662137e-05,
+      "loss": 1.6434,
       "step": 3588
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 13.033709526062012,
-      "learning_rate": 1.2807457151448332e-05,
-      "loss": 2.1353,
+      "epoch": 0.45,
+      "grad_norm": 8.141793251037598,
+      "learning_rate": 1.6997866376605447e-05,
+      "loss": 2.0123,
       "step": 3589
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 16.16973876953125,
-      "learning_rate": 1.2805452540843942e-05,
-      "loss": 1.847,
+      "epoch": 0.45,
+      "grad_norm": 14.612222671508789,
+      "learning_rate": 1.699702966154876e-05,
+      "loss": 2.2924,
       "step": 3590
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 14.109622955322266,
-      "learning_rate": 1.2803447930239552e-05,
-      "loss": 2.0083,
+      "epoch": 0.45,
+      "grad_norm": 6.75926399230957,
+      "learning_rate": 1.6996192946492074e-05,
+      "loss": 1.8664,
       "step": 3591
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 13.897767066955566,
-      "learning_rate": 1.2801443319635162e-05,
-      "loss": 1.2688,
+      "epoch": 0.45,
+      "grad_norm": 30.943069458007812,
+      "learning_rate": 1.6995356231435388e-05,
+      "loss": 2.8354,
       "step": 3592
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 18.122596740722656,
-      "learning_rate": 1.2799438709030772e-05,
-      "loss": 1.9293,
+      "epoch": 0.45,
+      "grad_norm": 13.30562686920166,
+      "learning_rate": 1.6994519516378698e-05,
+      "loss": 1.5587,
       "step": 3593
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 41.87525939941406,
-      "learning_rate": 1.2797434098426382e-05,
-      "loss": 3.1812,
+      "epoch": 0.45,
+      "grad_norm": 17.608930587768555,
+      "learning_rate": 1.699368280132201e-05,
+      "loss": 1.6954,
       "step": 3594
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 22.537267684936523,
-      "learning_rate": 1.279542948782199e-05,
-      "loss": 1.9805,
+      "epoch": 0.45,
+      "grad_norm": 13.332588195800781,
+      "learning_rate": 1.6992846086265325e-05,
+      "loss": 2.9673,
       "step": 3595
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 16.393918991088867,
-      "learning_rate": 1.2793424877217602e-05,
-      "loss": 1.6751,
+      "epoch": 0.45,
+      "grad_norm": 18.777090072631836,
+      "learning_rate": 1.6992009371208636e-05,
+      "loss": 2.5276,
       "step": 3596
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 15.888949394226074,
-      "learning_rate": 1.2791420266613211e-05,
-      "loss": 2.2047,
+      "epoch": 0.45,
+      "grad_norm": 8.829140663146973,
+      "learning_rate": 1.699117265615195e-05,
+      "loss": 2.4567,
       "step": 3597
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 13.120323181152344,
-      "learning_rate": 1.2789415656008821e-05,
-      "loss": 1.527,
+      "epoch": 0.45,
+      "grad_norm": 11.514774322509766,
+      "learning_rate": 1.6990335941095263e-05,
+      "loss": 0.7589,
       "step": 3598
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 14.314374923706055,
-      "learning_rate": 1.2787411045404433e-05,
-      "loss": 2.3968,
+      "epoch": 0.45,
+      "grad_norm": 17.70444679260254,
+      "learning_rate": 1.6989499226038577e-05,
+      "loss": 1.0188,
       "step": 3599
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 20.696285247802734,
-      "learning_rate": 1.2785406434800041e-05,
-      "loss": 1.8987,
+      "epoch": 0.45,
+      "grad_norm": 3.07735276222229,
+      "learning_rate": 1.6988662510981887e-05,
+      "loss": 0.0933,
       "step": 3600
     },
     {
-      "epoch": 1.08,
-      "eval_loss": 0.23791223764419556,
-      "eval_runtime": 43.6058,
-      "eval_samples_per_second": 33.917,
-      "eval_steps_per_second": 33.917,
+      "epoch": 0.45,
+      "eval_loss": 0.15236322581768036,
+      "eval_runtime": 93.9952,
+      "eval_samples_per_second": 37.683,
+      "eval_steps_per_second": 37.683,
       "step": 3600
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 24.901500701904297,
-      "learning_rate": 1.2783401824195651e-05,
-      "loss": 1.9859,
+      "epoch": 0.45,
+      "grad_norm": 12.646539688110352,
+      "learning_rate": 1.69878257959252e-05,
+      "loss": 2.7868,
       "step": 3601
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 15.03045654296875,
-      "learning_rate": 1.278139721359126e-05,
-      "loss": 1.6967,
+      "epoch": 0.45,
+      "grad_norm": 16.227340698242188,
+      "learning_rate": 1.698698908086851e-05,
+      "loss": 2.7049,
       "step": 3602
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 16.57175064086914,
-      "learning_rate": 1.2779392602986871e-05,
-      "loss": 1.8149,
+      "epoch": 0.45,
+      "grad_norm": 12.179350852966309,
+      "learning_rate": 1.6986152365811824e-05,
+      "loss": 1.3154,
       "step": 3603
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 15.594796180725098,
-      "learning_rate": 1.2777387992382481e-05,
-      "loss": 1.4157,
+      "epoch": 0.45,
+      "grad_norm": 28.343910217285156,
+      "learning_rate": 1.6985315650755138e-05,
+      "loss": 3.3812,
       "step": 3604
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 14.430771827697754,
-      "learning_rate": 1.277538338177809e-05,
-      "loss": 2.3501,
+      "epoch": 0.45,
+      "grad_norm": 11.430498123168945,
+      "learning_rate": 1.6984478935698448e-05,
+      "loss": 4.2946,
       "step": 3605
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 38.617862701416016,
-      "learning_rate": 1.2773378771173702e-05,
-      "loss": 1.98,
+      "epoch": 0.45,
+      "grad_norm": 23.60525131225586,
+      "learning_rate": 1.6983642220641762e-05,
+      "loss": 3.3857,
       "step": 3606
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 19.550251007080078,
-      "learning_rate": 1.277137416056931e-05,
-      "loss": 1.2882,
+      "epoch": 0.45,
+      "grad_norm": 18.278339385986328,
+      "learning_rate": 1.6982805505585072e-05,
+      "loss": 1.5645,
       "step": 3607
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 19.823959350585938,
-      "learning_rate": 1.276936954996492e-05,
-      "loss": 2.7407,
+      "epoch": 0.45,
+      "grad_norm": 26.03818130493164,
+      "learning_rate": 1.6981968790528386e-05,
+      "loss": 3.5829,
       "step": 3608
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 46.88768005371094,
-      "learning_rate": 1.2767364939360528e-05,
-      "loss": 1.9532,
+      "epoch": 0.45,
+      "grad_norm": 24.34980583190918,
+      "learning_rate": 1.69811320754717e-05,
+      "loss": 2.2011,
       "step": 3609
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 15.034628868103027,
-      "learning_rate": 1.276536032875614e-05,
-      "loss": 1.2125,
+      "epoch": 0.45,
+      "grad_norm": 30.7168025970459,
+      "learning_rate": 1.698029536041501e-05,
+      "loss": 0.6755,
       "step": 3610
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 24.965927124023438,
-      "learning_rate": 1.276335571815175e-05,
-      "loss": 2.4233,
+      "epoch": 0.45,
+      "grad_norm": 12.209492683410645,
+      "learning_rate": 1.6979458645358323e-05,
+      "loss": 2.1438,
       "step": 3611
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 26.463823318481445,
-      "learning_rate": 1.2761351107547359e-05,
-      "loss": 2.3175,
+      "epoch": 0.45,
+      "grad_norm": 8.702494621276855,
+      "learning_rate": 1.6978621930301637e-05,
+      "loss": 1.2021,
       "step": 3612
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 15.110565185546875,
-      "learning_rate": 1.275934649694297e-05,
-      "loss": 2.4443,
+      "epoch": 0.45,
+      "grad_norm": 9.897283554077148,
+      "learning_rate": 1.697778521524495e-05,
+      "loss": 1.2587,
       "step": 3613
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 18.711685180664062,
-      "learning_rate": 1.2757341886338579e-05,
-      "loss": 1.5244,
+      "epoch": 0.45,
+      "grad_norm": 22.805618286132812,
+      "learning_rate": 1.697694850018826e-05,
+      "loss": 0.8335,
       "step": 3614
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 21.489322662353516,
-      "learning_rate": 1.2755337275734189e-05,
-      "loss": 1.9058,
+      "epoch": 0.45,
+      "grad_norm": 49.473880767822266,
+      "learning_rate": 1.6976111785131575e-05,
+      "loss": 1.9427,
       "step": 3615
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 15.04526424407959,
-      "learning_rate": 1.27533326651298e-05,
-      "loss": 1.4516,
+      "epoch": 0.45,
+      "grad_norm": 55.079551696777344,
+      "learning_rate": 1.6975275070074888e-05,
+      "loss": 1.6532,
       "step": 3616
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 12.51637077331543,
-      "learning_rate": 1.2751328054525409e-05,
-      "loss": 1.7847,
+      "epoch": 0.45,
+      "grad_norm": 12.680688858032227,
+      "learning_rate": 1.69744383550182e-05,
+      "loss": 2.6464,
       "step": 3617
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 29.387882232666016,
-      "learning_rate": 1.274932344392102e-05,
-      "loss": 2.5068,
+      "epoch": 0.45,
+      "grad_norm": 27.3753662109375,
+      "learning_rate": 1.6973601639961512e-05,
+      "loss": 3.3583,
       "step": 3618
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 18.071352005004883,
-      "learning_rate": 1.274731883331663e-05,
-      "loss": 2.6933,
+      "epoch": 0.45,
+      "grad_norm": 4.351609706878662,
+      "learning_rate": 1.6972764924904826e-05,
+      "loss": 0.6152,
       "step": 3619
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 20.221973419189453,
-      "learning_rate": 1.274531422271224e-05,
-      "loss": 1.3898,
+      "epoch": 0.45,
+      "grad_norm": 14.723485946655273,
+      "learning_rate": 1.697192820984814e-05,
+      "loss": 2.1768,
       "step": 3620
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 8.028789520263672,
-      "learning_rate": 1.2743309612107848e-05,
-      "loss": 1.2193,
+      "epoch": 0.45,
+      "grad_norm": 43.82529830932617,
+      "learning_rate": 1.697109149479145e-05,
+      "loss": 1.5029,
       "step": 3621
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 14.530386924743652,
-      "learning_rate": 1.274130500150346e-05,
-      "loss": 2.0765,
+      "epoch": 0.45,
+      "grad_norm": 14.625968933105469,
+      "learning_rate": 1.6970254779734763e-05,
+      "loss": 2.3635,
       "step": 3622
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 16.220090866088867,
-      "learning_rate": 1.273930039089907e-05,
-      "loss": 1.7681,
+      "epoch": 0.45,
+      "grad_norm": 18.74462890625,
+      "learning_rate": 1.6969418064678077e-05,
+      "loss": 2.8262,
       "step": 3623
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 21.331134796142578,
-      "learning_rate": 1.2737295780294678e-05,
-      "loss": 2.7775,
+      "epoch": 0.45,
+      "grad_norm": 12.347823143005371,
+      "learning_rate": 1.6968581349621387e-05,
+      "loss": 2.0164,
       "step": 3624
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 11.270216941833496,
-      "learning_rate": 1.273529116969029e-05,
-      "loss": 2.3103,
+      "epoch": 0.45,
+      "grad_norm": 12.963733673095703,
+      "learning_rate": 1.69677446345647e-05,
+      "loss": 1.8139,
       "step": 3625
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 16.944839477539062,
-      "learning_rate": 1.2733286559085898e-05,
-      "loss": 1.7782,
+      "epoch": 0.46,
+      "grad_norm": 20.923049926757812,
+      "learning_rate": 1.6966907919508015e-05,
+      "loss": 2.7124,
       "step": 3626
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 8.570146560668945,
-      "learning_rate": 1.2731281948481508e-05,
-      "loss": 1.6687,
+      "epoch": 0.46,
+      "grad_norm": 9.197908401489258,
+      "learning_rate": 1.6966071204451328e-05,
+      "loss": 0.5815,
       "step": 3627
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 53.58586502075195,
-      "learning_rate": 1.272927733787712e-05,
-      "loss": 3.6285,
+      "epoch": 0.46,
+      "grad_norm": 25.95100975036621,
+      "learning_rate": 1.696523448939464e-05,
+      "loss": 2.5109,
       "step": 3628
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 24.34233856201172,
-      "learning_rate": 1.2727272727272728e-05,
-      "loss": 2.7752,
+      "epoch": 0.46,
+      "grad_norm": 12.633675575256348,
+      "learning_rate": 1.6964397774337952e-05,
+      "loss": 0.919,
       "step": 3629
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 14.564997673034668,
-      "learning_rate": 1.2725268116668339e-05,
-      "loss": 2.5383,
+      "epoch": 0.46,
+      "grad_norm": 12.344071388244629,
+      "learning_rate": 1.6963561059281262e-05,
+      "loss": 1.5926,
       "step": 3630
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 6.960912227630615,
-      "learning_rate": 1.2723263506063947e-05,
-      "loss": 0.6939,
+      "epoch": 0.46,
+      "grad_norm": 7.051496505737305,
+      "learning_rate": 1.6962724344224576e-05,
+      "loss": 0.7164,
       "step": 3631
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 17.01753044128418,
-      "learning_rate": 1.2721258895459559e-05,
-      "loss": 1.6664,
+      "epoch": 0.46,
+      "grad_norm": 11.98715877532959,
+      "learning_rate": 1.696188762916789e-05,
+      "loss": 2.2112,
       "step": 3632
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 11.095540046691895,
-      "learning_rate": 1.2719254284855167e-05,
-      "loss": 1.6678,
+      "epoch": 0.46,
+      "grad_norm": 21.208057403564453,
+      "learning_rate": 1.69610509141112e-05,
+      "loss": 3.4837,
       "step": 3633
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 11.197050094604492,
-      "learning_rate": 1.2717249674250777e-05,
-      "loss": 1.9901,
+      "epoch": 0.46,
+      "grad_norm": 8.355364799499512,
+      "learning_rate": 1.6960214199054514e-05,
+      "loss": 1.5355,
       "step": 3634
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 11.85187816619873,
-      "learning_rate": 1.2715245063646389e-05,
-      "loss": 0.7637,
+      "epoch": 0.46,
+      "grad_norm": 20.413406372070312,
+      "learning_rate": 1.6959377483997824e-05,
+      "loss": 2.0411,
       "step": 3635
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 46.356109619140625,
-      "learning_rate": 1.2713240453041997e-05,
-      "loss": 1.686,
+      "epoch": 0.46,
+      "grad_norm": 17.831262588500977,
+      "learning_rate": 1.6958540768941137e-05,
+      "loss": 3.666,
       "step": 3636
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 72.00715637207031,
-      "learning_rate": 1.2711235842437607e-05,
-      "loss": 2.1146,
+      "epoch": 0.46,
+      "grad_norm": 28.424673080444336,
+      "learning_rate": 1.695770405388445e-05,
+      "loss": 3.6548,
       "step": 3637
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 23.22759437561035,
-      "learning_rate": 1.2709231231833218e-05,
-      "loss": 1.1876,
+      "epoch": 0.46,
+      "grad_norm": 15.798102378845215,
+      "learning_rate": 1.695686733882776e-05,
+      "loss": 2.4799,
       "step": 3638
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 16.499618530273438,
-      "learning_rate": 1.2707226621228828e-05,
-      "loss": 1.4633,
+      "epoch": 0.46,
+      "grad_norm": 8.27879810333252,
+      "learning_rate": 1.6956030623771075e-05,
+      "loss": 0.5847,
       "step": 3639
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 35.80891418457031,
-      "learning_rate": 1.2705222010624436e-05,
-      "loss": 1.7426,
+      "epoch": 0.46,
+      "grad_norm": 9.31849479675293,
+      "learning_rate": 1.695519390871439e-05,
+      "loss": 2.0426,
       "step": 3640
     },
     {
-      "epoch": 1.09,
-      "grad_norm": 18.7283992767334,
-      "learning_rate": 1.2703217400020048e-05,
-      "loss": 1.8939,
+      "epoch": 0.46,
+      "grad_norm": 13.492223739624023,
+      "learning_rate": 1.6954357193657702e-05,
+      "loss": 1.4945,
       "step": 3641
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 14.170825958251953,
-      "learning_rate": 1.2701212789415658e-05,
-      "loss": 1.992,
+      "epoch": 0.46,
+      "grad_norm": 12.793575286865234,
+      "learning_rate": 1.6953520478601013e-05,
+      "loss": 1.4826,
       "step": 3642
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 11.59808349609375,
-      "learning_rate": 1.2699208178811266e-05,
-      "loss": 1.9479,
+      "epoch": 0.46,
+      "grad_norm": 17.997392654418945,
+      "learning_rate": 1.6952683763544326e-05,
+      "loss": 1.0301,
       "step": 3643
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 27.232282638549805,
-      "learning_rate": 1.2697203568206878e-05,
-      "loss": 1.9665,
+      "epoch": 0.46,
+      "grad_norm": 15.711397171020508,
+      "learning_rate": 1.695184704848764e-05,
+      "loss": 1.4051,
       "step": 3644
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 13.292805671691895,
-      "learning_rate": 1.2695198957602486e-05,
-      "loss": 1.4353,
+      "epoch": 0.46,
+      "grad_norm": 29.958099365234375,
+      "learning_rate": 1.695101033343095e-05,
+      "loss": 1.2986,
       "step": 3645
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 8.74724292755127,
-      "learning_rate": 1.2693194346998097e-05,
-      "loss": 1.3016,
+      "epoch": 0.46,
+      "grad_norm": 11.872001647949219,
+      "learning_rate": 1.6950173618374264e-05,
+      "loss": 1.717,
       "step": 3646
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 19.11276626586914,
-      "learning_rate": 1.2691189736393708e-05,
-      "loss": 2.3302,
+      "epoch": 0.46,
+      "grad_norm": 12.301345825195312,
+      "learning_rate": 1.6949336903317577e-05,
+      "loss": 1.7587,
       "step": 3647
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 46.45100021362305,
-      "learning_rate": 1.2689185125789317e-05,
-      "loss": 2.1202,
+      "epoch": 0.46,
+      "grad_norm": 12.868897438049316,
+      "learning_rate": 1.694850018826089e-05,
+      "loss": 1.4267,
       "step": 3648
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 15.36929702758789,
-      "learning_rate": 1.2687180515184927e-05,
-      "loss": 1.7896,
+      "epoch": 0.46,
+      "grad_norm": 16.243345260620117,
+      "learning_rate": 1.69476634732042e-05,
+      "loss": 2.4037,
       "step": 3649
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 20.07041358947754,
-      "learning_rate": 1.2685175904580535e-05,
-      "loss": 1.826,
+      "epoch": 0.46,
+      "grad_norm": 14.367687225341797,
+      "learning_rate": 1.6946826758147515e-05,
+      "loss": 1.7668,
       "step": 3650
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 20.192039489746094,
-      "learning_rate": 1.2683171293976147e-05,
-      "loss": 1.8944,
+      "epoch": 0.46,
+      "grad_norm": 16.331218719482422,
+      "learning_rate": 1.694599004309083e-05,
+      "loss": 2.1331,
       "step": 3651
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 15.358658790588379,
-      "learning_rate": 1.2681166683371755e-05,
-      "loss": 2.1073,
+      "epoch": 0.46,
+      "grad_norm": 12.379274368286133,
+      "learning_rate": 1.694515332803414e-05,
+      "loss": 2.9811,
       "step": 3652
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 11.35305404663086,
-      "learning_rate": 1.2679162072767365e-05,
-      "loss": 1.167,
+      "epoch": 0.46,
+      "grad_norm": 11.794244766235352,
+      "learning_rate": 1.6944316612977453e-05,
+      "loss": 3.0356,
       "step": 3653
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 14.477778434753418,
-      "learning_rate": 1.2677157462162977e-05,
-      "loss": 2.1406,
+      "epoch": 0.46,
+      "grad_norm": 5.233652591705322,
+      "learning_rate": 1.6943479897920766e-05,
+      "loss": 1.1373,
       "step": 3654
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 25.864349365234375,
-      "learning_rate": 1.2675152851558586e-05,
-      "loss": 1.9479,
+      "epoch": 0.46,
+      "grad_norm": 9.71679973602295,
+      "learning_rate": 1.6942643182864076e-05,
+      "loss": 1.6338,
       "step": 3655
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 14.707035064697266,
-      "learning_rate": 1.2673148240954196e-05,
-      "loss": 2.2159,
+      "epoch": 0.46,
+      "grad_norm": 5.850435256958008,
+      "learning_rate": 1.694180646780739e-05,
+      "loss": 2.8447,
       "step": 3656
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 9.004791259765625,
-      "learning_rate": 1.2671143630349804e-05,
-      "loss": 0.9894,
+      "epoch": 0.46,
+      "grad_norm": 9.849655151367188,
+      "learning_rate": 1.6940969752750704e-05,
+      "loss": 1.5747,
       "step": 3657
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 16.240209579467773,
-      "learning_rate": 1.2669139019745416e-05,
-      "loss": 1.6444,
-      "step": 3658
+      "epoch": 0.46,
+      "grad_norm": 15.464215278625488,
+      "learning_rate": 1.6940133037694014e-05,
+      "loss": 1.9339,
+      "step": 3658
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 15.315890312194824,
-      "learning_rate": 1.2667134409141026e-05,
-      "loss": 2.046,
+      "epoch": 0.46,
+      "grad_norm": 6.046443939208984,
+      "learning_rate": 1.6939296322637328e-05,
+      "loss": 1.5449,
       "step": 3659
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 14.05276870727539,
-      "learning_rate": 1.2665129798536634e-05,
-      "loss": 2.0308,
+      "epoch": 0.46,
+      "grad_norm": 22.741769790649414,
+      "learning_rate": 1.6938459607580638e-05,
+      "loss": 0.8067,
       "step": 3660
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 12.463837623596191,
-      "learning_rate": 1.2663125187932246e-05,
-      "loss": 1.3266,
+      "epoch": 0.46,
+      "grad_norm": 15.346259117126465,
+      "learning_rate": 1.693762289252395e-05,
+      "loss": 2.7109,
       "step": 3661
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 10.216002464294434,
-      "learning_rate": 1.2661120577327854e-05,
-      "loss": 1.4467,
+      "epoch": 0.46,
+      "grad_norm": 11.457945823669434,
+      "learning_rate": 1.6936786177467265e-05,
+      "loss": 2.1529,
       "step": 3662
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 52.47303771972656,
-      "learning_rate": 1.2659115966723465e-05,
-      "loss": 3.601,
+      "epoch": 0.46,
+      "grad_norm": 20.453115463256836,
+      "learning_rate": 1.6935949462410575e-05,
+      "loss": 3.8016,
       "step": 3663
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 19.366987228393555,
-      "learning_rate": 1.2657111356119075e-05,
-      "loss": 1.9766,
+      "epoch": 0.46,
+      "grad_norm": 24.550575256347656,
+      "learning_rate": 1.693511274735389e-05,
+      "loss": 3.59,
       "step": 3664
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 11.65963077545166,
-      "learning_rate": 1.2655106745514685e-05,
-      "loss": 1.6192,
+      "epoch": 0.46,
+      "grad_norm": 23.053932189941406,
+      "learning_rate": 1.6934276032297203e-05,
+      "loss": 2.6321,
       "step": 3665
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 14.95705509185791,
-      "learning_rate": 1.2653102134910295e-05,
-      "loss": 1.8748,
+      "epoch": 0.46,
+      "grad_norm": 9.720477104187012,
+      "learning_rate": 1.6933439317240513e-05,
+      "loss": 2.066,
       "step": 3666
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 21.516704559326172,
-      "learning_rate": 1.2651097524305905e-05,
-      "loss": 2.2989,
+      "epoch": 0.46,
+      "grad_norm": 8.113615036010742,
+      "learning_rate": 1.6932602602183827e-05,
+      "loss": 1.7078,
       "step": 3667
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 12.6599760055542,
-      "learning_rate": 1.2649092913701515e-05,
-      "loss": 1.9577,
+      "epoch": 0.46,
+      "grad_norm": 11.102977752685547,
+      "learning_rate": 1.693176588712714e-05,
+      "loss": 2.9131,
       "step": 3668
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 20.421907424926758,
-      "learning_rate": 1.2647088303097123e-05,
-      "loss": 1.8006,
+      "epoch": 0.46,
+      "grad_norm": 9.469147682189941,
+      "learning_rate": 1.6930929172070454e-05,
+      "loss": 1.2561,
       "step": 3669
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 21.49991226196289,
-      "learning_rate": 1.2645083692492735e-05,
-      "loss": 1.5208,
+      "epoch": 0.46,
+      "grad_norm": 39.29673767089844,
+      "learning_rate": 1.6930092457013764e-05,
+      "loss": 2.2055,
       "step": 3670
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 16.71851348876953,
-      "learning_rate": 1.2643079081888345e-05,
-      "loss": 1.3844,
+      "epoch": 0.46,
+      "grad_norm": 22.410480499267578,
+      "learning_rate": 1.6929255741957078e-05,
+      "loss": 2.6519,
       "step": 3671
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 12.691519737243652,
-      "learning_rate": 1.2641074471283954e-05,
-      "loss": 1.6805,
+      "epoch": 0.46,
+      "grad_norm": 6.620682716369629,
+      "learning_rate": 1.692841902690039e-05,
+      "loss": 0.4279,
       "step": 3672
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 13.60329532623291,
-      "learning_rate": 1.2639069860679565e-05,
-      "loss": 1.3261,
+      "epoch": 0.46,
+      "grad_norm": 20.994232177734375,
+      "learning_rate": 1.6927582311843702e-05,
+      "loss": 1.4075,
       "step": 3673
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 12.163605690002441,
-      "learning_rate": 1.2637065250075174e-05,
-      "loss": 1.2984,
+      "epoch": 0.46,
+      "grad_norm": 14.912308692932129,
+      "learning_rate": 1.6926745596787015e-05,
+      "loss": 2.9889,
       "step": 3674
     },
     {
-      "epoch": 1.1,
-      "grad_norm": 11.735462188720703,
-      "learning_rate": 1.2635060639470784e-05,
-      "loss": 1.3691,
+      "epoch": 0.46,
+      "grad_norm": 23.213647842407227,
+      "learning_rate": 1.692590888173033e-05,
+      "loss": 2.9207,
       "step": 3675
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 10.237204551696777,
-      "learning_rate": 1.2633056028866392e-05,
-      "loss": 1.4796,
+      "epoch": 0.46,
+      "grad_norm": 16.951650619506836,
+      "learning_rate": 1.6925072166673643e-05,
+      "loss": 1.9713,
       "step": 3676
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 16.992961883544922,
-      "learning_rate": 1.2631051418262004e-05,
-      "loss": 1.5819,
+      "epoch": 0.46,
+      "grad_norm": 11.632166862487793,
+      "learning_rate": 1.6924235451616953e-05,
+      "loss": 1.0241,
       "step": 3677
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 25.052305221557617,
-      "learning_rate": 1.2629046807657614e-05,
-      "loss": 2.6452,
+      "epoch": 0.46,
+      "grad_norm": 17.42290687561035,
+      "learning_rate": 1.6923398736560267e-05,
+      "loss": 2.1315,
       "step": 3678
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 34.612735748291016,
-      "learning_rate": 1.2627042197053223e-05,
-      "loss": 1.888,
+      "epoch": 0.46,
+      "grad_norm": 3.987685203552246,
+      "learning_rate": 1.692256202150358e-05,
+      "loss": 0.2983,
       "step": 3679
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 12.312444686889648,
-      "learning_rate": 1.2625037586448834e-05,
-      "loss": 1.6545,
+      "epoch": 0.46,
+      "grad_norm": 12.946403503417969,
+      "learning_rate": 1.692172530644689e-05,
+      "loss": 1.0987,
       "step": 3680
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 17.487689971923828,
-      "learning_rate": 1.2623032975844443e-05,
-      "loss": 2.225,
+      "epoch": 0.46,
+      "grad_norm": 13.340381622314453,
+      "learning_rate": 1.6920888591390204e-05,
+      "loss": 2.515,
       "step": 3681
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 12.337897300720215,
-      "learning_rate": 1.2621028365240053e-05,
-      "loss": 1.285,
+      "epoch": 0.46,
+      "grad_norm": 9.504284858703613,
+      "learning_rate": 1.6920051876333518e-05,
+      "loss": 1.8395,
       "step": 3682
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 12.9970121383667,
-      "learning_rate": 1.2619023754635661e-05,
-      "loss": 1.5476,
+      "epoch": 0.46,
+      "grad_norm": 15.068203926086426,
+      "learning_rate": 1.6919215161276828e-05,
+      "loss": 2.3602,
       "step": 3683
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 36.430641174316406,
-      "learning_rate": 1.2617019144031273e-05,
-      "loss": 2.409,
+      "epoch": 0.46,
+      "grad_norm": 14.113088607788086,
+      "learning_rate": 1.6918378446220142e-05,
+      "loss": 2.6057,
       "step": 3684
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 11.384819030761719,
-      "learning_rate": 1.2615014533426883e-05,
-      "loss": 1.2839,
+      "epoch": 0.46,
+      "grad_norm": 37.74652862548828,
+      "learning_rate": 1.6917541731163455e-05,
+      "loss": 1.9992,
       "step": 3685
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 14.997862815856934,
-      "learning_rate": 1.2613009922822491e-05,
-      "loss": 2.1861,
+      "epoch": 0.46,
+      "grad_norm": 20.58168601989746,
+      "learning_rate": 1.6916705016106766e-05,
+      "loss": 2.5807,
       "step": 3686
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 9.791463851928711,
-      "learning_rate": 1.2611005312218103e-05,
-      "loss": 1.6862,
+      "epoch": 0.46,
+      "grad_norm": 24.944747924804688,
+      "learning_rate": 1.691586830105008e-05,
+      "loss": 1.9458,
       "step": 3687
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 20.38786506652832,
-      "learning_rate": 1.2609000701613712e-05,
-      "loss": 1.8915,
+      "epoch": 0.46,
+      "grad_norm": 8.984786033630371,
+      "learning_rate": 1.691503158599339e-05,
+      "loss": 1.8054,
       "step": 3688
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 56.49119567871094,
-      "learning_rate": 1.2606996091009323e-05,
-      "loss": 1.4701,
+      "epoch": 0.46,
+      "grad_norm": 12.315727233886719,
+      "learning_rate": 1.6914194870936703e-05,
+      "loss": 4.6691,
       "step": 3689
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 9.169239044189453,
-      "learning_rate": 1.2604991480404933e-05,
-      "loss": 1.0245,
+      "epoch": 0.46,
+      "grad_norm": 17.678552627563477,
+      "learning_rate": 1.6913358155880017e-05,
+      "loss": 1.7385,
       "step": 3690
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 18.84854507446289,
-      "learning_rate": 1.2602986869800542e-05,
-      "loss": 1.6371,
+      "epoch": 0.46,
+      "grad_norm": 14.618388175964355,
+      "learning_rate": 1.6912521440823327e-05,
+      "loss": 2.2065,
       "step": 3691
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 15.238136291503906,
-      "learning_rate": 1.2600982259196154e-05,
-      "loss": 1.4109,
+      "epoch": 0.46,
+      "grad_norm": 14.206609725952148,
+      "learning_rate": 1.691168472576664e-05,
+      "loss": 2.552,
       "step": 3692
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 24.6544189453125,
-      "learning_rate": 1.2598977648591762e-05,
-      "loss": 1.8289,
+      "epoch": 0.46,
+      "grad_norm": 13.141923904418945,
+      "learning_rate": 1.6910848010709954e-05,
+      "loss": 2.2018,
       "step": 3693
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 15.802862167358398,
-      "learning_rate": 1.2596973037987372e-05,
-      "loss": 2.074,
+      "epoch": 0.46,
+      "grad_norm": 10.221967697143555,
+      "learning_rate": 1.6910011295653265e-05,
+      "loss": 0.7906,
       "step": 3694
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 21.196699142456055,
-      "learning_rate": 1.259496842738298e-05,
-      "loss": 2.4331,
+      "epoch": 0.46,
+      "grad_norm": 7.5393147468566895,
+      "learning_rate": 1.690917458059658e-05,
+      "loss": 1.7885,
       "step": 3695
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 14.442291259765625,
-      "learning_rate": 1.2592963816778592e-05,
-      "loss": 1.1208,
+      "epoch": 0.46,
+      "grad_norm": 10.141867637634277,
+      "learning_rate": 1.6908337865539892e-05,
+      "loss": 1.7606,
       "step": 3696
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 16.67438316345215,
-      "learning_rate": 1.2590959206174202e-05,
-      "loss": 1.9479,
+      "epoch": 0.46,
+      "grad_norm": 30.746566772460938,
+      "learning_rate": 1.6907501150483206e-05,
+      "loss": 3.0198,
       "step": 3697
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 30.792251586914062,
-      "learning_rate": 1.258895459556981e-05,
-      "loss": 1.0711,
+      "epoch": 0.46,
+      "grad_norm": 198.8036346435547,
+      "learning_rate": 1.6906664435426516e-05,
+      "loss": 1.4779,
       "step": 3698
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 21.835054397583008,
-      "learning_rate": 1.2586949984965423e-05,
-      "loss": 1.4139,
+      "epoch": 0.46,
+      "grad_norm": 20.587520599365234,
+      "learning_rate": 1.690582772036983e-05,
+      "loss": 1.9991,
       "step": 3699
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 23.054439544677734,
-      "learning_rate": 1.2584945374361031e-05,
-      "loss": 2.3756,
+      "epoch": 0.46,
+      "grad_norm": 20.063173294067383,
+      "learning_rate": 1.6904991005313143e-05,
+      "loss": 2.9646,
       "step": 3700
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 15.562328338623047,
-      "learning_rate": 1.2582940763756641e-05,
-      "loss": 1.6298,
+      "epoch": 0.46,
+      "grad_norm": 12.660489082336426,
+      "learning_rate": 1.6904154290256453e-05,
+      "loss": 1.038,
       "step": 3701
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 19.600475311279297,
-      "learning_rate": 1.2580936153152253e-05,
-      "loss": 2.0049,
+      "epoch": 0.46,
+      "grad_norm": 14.9254150390625,
+      "learning_rate": 1.6903317575199767e-05,
+      "loss": 1.8169,
       "step": 3702
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 15.934944152832031,
-      "learning_rate": 1.2578931542547861e-05,
-      "loss": 2.3699,
+      "epoch": 0.46,
+      "grad_norm": 9.880672454833984,
+      "learning_rate": 1.690248086014308e-05,
+      "loss": 1.564,
       "step": 3703
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 13.748173713684082,
-      "learning_rate": 1.2576926931943471e-05,
-      "loss": 2.802,
+      "epoch": 0.46,
+      "grad_norm": 13.512040138244629,
+      "learning_rate": 1.6901644145086394e-05,
+      "loss": 3.2549,
       "step": 3704
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 10.191062927246094,
-      "learning_rate": 1.257492232133908e-05,
-      "loss": 0.9145,
+      "epoch": 0.46,
+      "grad_norm": 8.003437042236328,
+      "learning_rate": 1.6900807430029705e-05,
+      "loss": 0.9277,
       "step": 3705
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 14.744169235229492,
-      "learning_rate": 1.2572917710734691e-05,
-      "loss": 1.7588,
+      "epoch": 0.47,
+      "grad_norm": 15.903909683227539,
+      "learning_rate": 1.689997071497302e-05,
+      "loss": 3.1238,
       "step": 3706
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 17.905502319335938,
-      "learning_rate": 1.25709131001303e-05,
-      "loss": 1.4409,
+      "epoch": 0.47,
+      "grad_norm": 9.854524612426758,
+      "learning_rate": 1.6899133999916332e-05,
+      "loss": 0.7731,
       "step": 3707
     },
     {
-      "epoch": 1.11,
-      "grad_norm": 22.17319679260254,
-      "learning_rate": 1.256890848952591e-05,
-      "loss": 1.6653,
+      "epoch": 0.47,
+      "grad_norm": 21.493879318237305,
+      "learning_rate": 1.6898297284859642e-05,
+      "loss": 2.0419,
       "step": 3708
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 14.994551658630371,
-      "learning_rate": 1.2566903878921522e-05,
-      "loss": 2.6141,
+      "epoch": 0.47,
+      "grad_norm": 15.608744621276855,
+      "learning_rate": 1.6897460569802956e-05,
+      "loss": 2.7725,
       "step": 3709
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 27.965063095092773,
-      "learning_rate": 1.256489926831713e-05,
-      "loss": 1.4174,
+      "epoch": 0.47,
+      "grad_norm": 15.87423324584961,
+      "learning_rate": 1.689662385474627e-05,
+      "loss": 2.4522,
       "step": 3710
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 12.438477516174316,
-      "learning_rate": 1.256289465771274e-05,
-      "loss": 2.034,
+      "epoch": 0.47,
+      "grad_norm": 12.656645774841309,
+      "learning_rate": 1.689578713968958e-05,
+      "loss": 2.3673,
       "step": 3711
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 18.60542106628418,
-      "learning_rate": 1.256089004710835e-05,
-      "loss": 2.1583,
+      "epoch": 0.47,
+      "grad_norm": 8.62553882598877,
+      "learning_rate": 1.6894950424632893e-05,
+      "loss": 0.6039,
       "step": 3712
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 11.205132484436035,
-      "learning_rate": 1.255888543650396e-05,
-      "loss": 2.2042,
+      "epoch": 0.47,
+      "grad_norm": 17.59114646911621,
+      "learning_rate": 1.6894113709576204e-05,
+      "loss": 1.466,
       "step": 3713
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 21.031993865966797,
-      "learning_rate": 1.255688082589957e-05,
-      "loss": 1.3351,
+      "epoch": 0.47,
+      "grad_norm": 14.852526664733887,
+      "learning_rate": 1.6893276994519517e-05,
+      "loss": 2.2656,
       "step": 3714
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 22.70185661315918,
-      "learning_rate": 1.255487621529518e-05,
-      "loss": 2.8246,
+      "epoch": 0.47,
+      "grad_norm": 10.289872169494629,
+      "learning_rate": 1.689244027946283e-05,
+      "loss": 2.5505,
       "step": 3715
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 19.754451751708984,
-      "learning_rate": 1.255287160469079e-05,
-      "loss": 1.9262,
+      "epoch": 0.47,
+      "grad_norm": 8.868825912475586,
+      "learning_rate": 1.689160356440614e-05,
+      "loss": 1.4586,
       "step": 3716
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 14.8783597946167,
-      "learning_rate": 1.2550866994086399e-05,
-      "loss": 1.958,
+      "epoch": 0.47,
+      "grad_norm": 15.438570976257324,
+      "learning_rate": 1.6890766849349455e-05,
+      "loss": 1.4362,
       "step": 3717
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 13.51906681060791,
-      "learning_rate": 1.254886238348201e-05,
-      "loss": 2.0377,
+      "epoch": 0.47,
+      "grad_norm": 12.702199935913086,
+      "learning_rate": 1.688993013429277e-05,
+      "loss": 2.9112,
       "step": 3718
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 13.766884803771973,
-      "learning_rate": 1.2546857772877619e-05,
-      "loss": 1.8277,
+      "epoch": 0.47,
+      "grad_norm": 13.10950756072998,
+      "learning_rate": 1.688909341923608e-05,
+      "loss": 0.9233,
       "step": 3719
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 11.940971374511719,
-      "learning_rate": 1.254485316227323e-05,
-      "loss": 2.0253,
-      "step": 3720
-    },
-    {
-      "epoch": 1.12,
-      "eval_loss": 0.280333936214447,
-      "eval_runtime": 43.5973,
-      "eval_samples_per_second": 33.924,
-      "eval_steps_per_second": 33.924,
+      "epoch": 0.47,
+      "grad_norm": 9.561256408691406,
+      "learning_rate": 1.6888256704179392e-05,
+      "loss": 0.7627,
       "step": 3720
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 13.200506210327148,
-      "learning_rate": 1.2542848551668841e-05,
-      "loss": 1.8569,
+      "epoch": 0.47,
+      "grad_norm": 10.910165786743164,
+      "learning_rate": 1.6887419989122706e-05,
+      "loss": 2.432,
       "step": 3721
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 11.850561141967773,
-      "learning_rate": 1.254084394106445e-05,
-      "loss": 1.2431,
+      "epoch": 0.47,
+      "grad_norm": 25.648988723754883,
+      "learning_rate": 1.6886583274066016e-05,
+      "loss": 1.7628,
       "step": 3722
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 14.307988166809082,
-      "learning_rate": 1.253883933046006e-05,
-      "loss": 1.2535,
+      "epoch": 0.47,
+      "grad_norm": 14.003104209899902,
+      "learning_rate": 1.688574655900933e-05,
+      "loss": 3.3926,
       "step": 3723
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 33.0361328125,
-      "learning_rate": 1.2536834719855668e-05,
-      "loss": 2.3736,
+      "epoch": 0.47,
+      "grad_norm": 11.824846267700195,
+      "learning_rate": 1.6884909843952644e-05,
+      "loss": 2.1877,
       "step": 3724
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 14.496783256530762,
-      "learning_rate": 1.253483010925128e-05,
-      "loss": 1.5165,
+      "epoch": 0.47,
+      "grad_norm": 8.314187049865723,
+      "learning_rate": 1.6884073128895957e-05,
+      "loss": 0.9777,
       "step": 3725
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 20.50676155090332,
-      "learning_rate": 1.2532825498646888e-05,
-      "loss": 1.7852,
+      "epoch": 0.47,
+      "grad_norm": 10.272077560424805,
+      "learning_rate": 1.6883236413839268e-05,
+      "loss": 1.4694,
       "step": 3726
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 16.414701461791992,
-      "learning_rate": 1.2530820888042498e-05,
-      "loss": 1.5958,
+      "epoch": 0.47,
+      "grad_norm": 12.01590633392334,
+      "learning_rate": 1.688239969878258e-05,
+      "loss": 0.8113,
       "step": 3727
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 9.845596313476562,
-      "learning_rate": 1.252881627743811e-05,
-      "loss": 1.4953,
+      "epoch": 0.47,
+      "grad_norm": 14.180240631103516,
+      "learning_rate": 1.6881562983725895e-05,
+      "loss": 3.053,
       "step": 3728
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 34.321861267089844,
-      "learning_rate": 1.2526811666833718e-05,
-      "loss": 1.9955,
+      "epoch": 0.47,
+      "grad_norm": 6.562134742736816,
+      "learning_rate": 1.6880726268669205e-05,
+      "loss": 0.6537,
       "step": 3729
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 11.001937866210938,
-      "learning_rate": 1.2524807056229328e-05,
-      "loss": 1.7248,
+      "epoch": 0.47,
+      "grad_norm": 8.070725440979004,
+      "learning_rate": 1.687988955361252e-05,
+      "loss": 0.89,
       "step": 3730
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 14.16842269897461,
-      "learning_rate": 1.2522802445624937e-05,
-      "loss": 0.9204,
+      "epoch": 0.47,
+      "grad_norm": 26.202648162841797,
+      "learning_rate": 1.6879052838555832e-05,
+      "loss": 2.0353,
       "step": 3731
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 44.32683181762695,
-      "learning_rate": 1.2520797835020549e-05,
-      "loss": 1.5575,
+      "epoch": 0.47,
+      "grad_norm": 11.653743743896484,
+      "learning_rate": 1.6878216123499146e-05,
+      "loss": 2.4323,
       "step": 3732
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 21.35757064819336,
-      "learning_rate": 1.2518793224416159e-05,
-      "loss": 2.7646,
+      "epoch": 0.47,
+      "grad_norm": 10.062137603759766,
+      "learning_rate": 1.6877379408442456e-05,
+      "loss": 3.2083,
       "step": 3733
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 10.033564567565918,
-      "learning_rate": 1.2516788613811767e-05,
-      "loss": 1.054,
+      "epoch": 0.47,
+      "grad_norm": 13.19880485534668,
+      "learning_rate": 1.687654269338577e-05,
+      "loss": 1.9633,
       "step": 3734
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 11.14671802520752,
-      "learning_rate": 1.2514784003207379e-05,
-      "loss": 1.4729,
+      "epoch": 0.47,
+      "grad_norm": 10.756131172180176,
+      "learning_rate": 1.6875705978329084e-05,
+      "loss": 1.5122,
       "step": 3735
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 57.583152770996094,
-      "learning_rate": 1.2512779392602987e-05,
-      "loss": 4.8937,
+      "epoch": 0.47,
+      "grad_norm": 7.860194683074951,
+      "learning_rate": 1.6874869263272394e-05,
+      "loss": 1.4545,
       "step": 3736
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 10.513443946838379,
-      "learning_rate": 1.2510774781998597e-05,
-      "loss": 1.1766,
+      "epoch": 0.47,
+      "grad_norm": 11.343194007873535,
+      "learning_rate": 1.6874032548215708e-05,
+      "loss": 1.8502,
       "step": 3737
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 20.811086654663086,
-      "learning_rate": 1.2508770171394207e-05,
-      "loss": 1.86,
+      "epoch": 0.47,
+      "grad_norm": 9.31357192993164,
+      "learning_rate": 1.687319583315902e-05,
+      "loss": 2.8584,
       "step": 3738
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 15.33969497680664,
-      "learning_rate": 1.2506765560789817e-05,
-      "loss": 1.9234,
+      "epoch": 0.47,
+      "grad_norm": 43.624847412109375,
+      "learning_rate": 1.687235911810233e-05,
+      "loss": 2.5072,
       "step": 3739
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 16.466957092285156,
-      "learning_rate": 1.250476095018543e-05,
-      "loss": 1.0944,
+      "epoch": 0.47,
+      "grad_norm": 8.754593849182129,
+      "learning_rate": 1.6871522403045645e-05,
+      "loss": 2.2842,
       "step": 3740
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 11.51232624053955,
-      "learning_rate": 1.2502756339581038e-05,
-      "loss": 1.9455,
+      "epoch": 0.47,
+      "grad_norm": 8.47415542602539,
+      "learning_rate": 1.6870685687988955e-05,
+      "loss": 0.5375,
       "step": 3741
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 16.893362045288086,
-      "learning_rate": 1.2500751728976648e-05,
-      "loss": 2.0934,
+      "epoch": 0.47,
+      "grad_norm": 10.735517501831055,
+      "learning_rate": 1.686984897293227e-05,
+      "loss": 1.5768,
       "step": 3742
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 10.207427024841309,
-      "learning_rate": 1.2498747118372256e-05,
-      "loss": 1.5444,
+      "epoch": 0.47,
+      "grad_norm": 15.74870491027832,
+      "learning_rate": 1.6869012257875583e-05,
+      "loss": 2.2971,
       "step": 3743
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 14.303349494934082,
-      "learning_rate": 1.2496742507767868e-05,
-      "loss": 2.0237,
+      "epoch": 0.47,
+      "grad_norm": 21.359169006347656,
+      "learning_rate": 1.6868175542818893e-05,
+      "loss": 1.9346,
       "step": 3744
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 15.821151733398438,
-      "learning_rate": 1.2494737897163478e-05,
-      "loss": 1.2218,
+      "epoch": 0.47,
+      "grad_norm": 38.049407958984375,
+      "learning_rate": 1.6867338827762207e-05,
+      "loss": 2.1763,
       "step": 3745
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 19.704866409301758,
-      "learning_rate": 1.2492733286559086e-05,
-      "loss": 1.8249,
+      "epoch": 0.47,
+      "grad_norm": 15.668584823608398,
+      "learning_rate": 1.686650211270552e-05,
+      "loss": 1.9172,
       "step": 3746
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 15.844225883483887,
-      "learning_rate": 1.2490728675954698e-05,
-      "loss": 1.4068,
+      "epoch": 0.47,
+      "grad_norm": 18.220617294311523,
+      "learning_rate": 1.686566539764883e-05,
+      "loss": 1.6622,
       "step": 3747
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 26.448633193969727,
-      "learning_rate": 1.2488724065350306e-05,
-      "loss": 2.5801,
+      "epoch": 0.47,
+      "grad_norm": 26.26845359802246,
+      "learning_rate": 1.6864828682592144e-05,
+      "loss": 1.8791,
       "step": 3748
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 31.50472640991211,
-      "learning_rate": 1.2486719454745917e-05,
-      "loss": 2.1224,
+      "epoch": 0.47,
+      "grad_norm": 12.238680839538574,
+      "learning_rate": 1.6863991967535458e-05,
+      "loss": 2.2067,
       "step": 3749
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 11.211418151855469,
-      "learning_rate": 1.2484714844141525e-05,
-      "loss": 1.3734,
+      "epoch": 0.47,
+      "grad_norm": 10.083091735839844,
+      "learning_rate": 1.6863155252478768e-05,
+      "loss": 1.3996,
       "step": 3750
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 16.562360763549805,
-      "learning_rate": 1.2482710233537137e-05,
-      "loss": 2.8084,
+      "epoch": 0.47,
+      "grad_norm": 6.8707194328308105,
+      "learning_rate": 1.6862318537422082e-05,
+      "loss": 0.3253,
       "step": 3751
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 14.236701965332031,
-      "learning_rate": 1.2480705622932747e-05,
-      "loss": 1.3424,
+      "epoch": 0.47,
+      "grad_norm": 50.154117584228516,
+      "learning_rate": 1.6861481822365395e-05,
+      "loss": 3.0243,
       "step": 3752
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 17.27580451965332,
-      "learning_rate": 1.2478701012328355e-05,
-      "loss": 2.5758,
+      "epoch": 0.47,
+      "grad_norm": 9.788782119750977,
+      "learning_rate": 1.6860645107308706e-05,
+      "loss": 1.2306,
       "step": 3753
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 12.679486274719238,
-      "learning_rate": 1.2476696401723967e-05,
-      "loss": 1.6972,
+      "epoch": 0.47,
+      "grad_norm": 10.830899238586426,
+      "learning_rate": 1.685980839225202e-05,
+      "loss": 1.7717,
       "step": 3754
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 17.295305252075195,
-      "learning_rate": 1.2474691791119575e-05,
-      "loss": 1.4085,
+      "epoch": 0.47,
+      "grad_norm": 14.998096466064453,
+      "learning_rate": 1.6858971677195333e-05,
+      "loss": 3.7802,
       "step": 3755
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 14.256155014038086,
-      "learning_rate": 1.2472687180515185e-05,
-      "loss": 1.5598,
+      "epoch": 0.47,
+      "grad_norm": 7.246328830718994,
+      "learning_rate": 1.6858134962138647e-05,
+      "loss": 0.4949,
       "step": 3756
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 14.317242622375488,
-      "learning_rate": 1.2470682569910797e-05,
-      "loss": 1.98,
+      "epoch": 0.47,
+      "grad_norm": 7.857738494873047,
+      "learning_rate": 1.6857298247081957e-05,
+      "loss": 2.1915,
       "step": 3757
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 14.427403450012207,
-      "learning_rate": 1.2468677959306406e-05,
-      "loss": 1.4377,
+      "epoch": 0.47,
+      "grad_norm": 12.347123146057129,
+      "learning_rate": 1.685646153202527e-05,
+      "loss": 2.5454,
       "step": 3758
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 19.83202362060547,
-      "learning_rate": 1.2466673348702016e-05,
-      "loss": 1.7858,
+      "epoch": 0.47,
+      "grad_norm": 13.190940856933594,
+      "learning_rate": 1.6855624816968584e-05,
+      "loss": 0.8198,
       "step": 3759
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 10.979972839355469,
-      "learning_rate": 1.2464668738097626e-05,
-      "loss": 0.8914,
+      "epoch": 0.47,
+      "grad_norm": 10.892491340637207,
+      "learning_rate": 1.6854788101911894e-05,
+      "loss": 1.7259,
       "step": 3760
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 25.953853607177734,
-      "learning_rate": 1.2462664127493236e-05,
-      "loss": 1.768,
+      "epoch": 0.47,
+      "grad_norm": 14.431472778320312,
+      "learning_rate": 1.6853951386855208e-05,
+      "loss": 2.6695,
       "step": 3761
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 17.417144775390625,
-      "learning_rate": 1.2460659516888844e-05,
-      "loss": 1.9067,
+      "epoch": 0.47,
+      "grad_norm": 5.824015140533447,
+      "learning_rate": 1.685311467179852e-05,
+      "loss": 1.4626,
       "step": 3762
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 18.727876663208008,
-      "learning_rate": 1.2458654906284456e-05,
-      "loss": 1.8154,
+      "epoch": 0.47,
+      "grad_norm": 10.591657638549805,
+      "learning_rate": 1.6852277956741835e-05,
+      "loss": 1.7026,
       "step": 3763
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 28.132505416870117,
-      "learning_rate": 1.2456650295680066e-05,
-      "loss": 2.6384,
+      "epoch": 0.47,
+      "grad_norm": 19.413419723510742,
+      "learning_rate": 1.6851441241685146e-05,
+      "loss": 1.3262,
       "step": 3764
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 11.409829139709473,
-      "learning_rate": 1.2454645685075675e-05,
-      "loss": 1.3703,
+      "epoch": 0.47,
+      "grad_norm": 9.156866073608398,
+      "learning_rate": 1.685060452662846e-05,
+      "loss": 1.5683,
       "step": 3765
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 12.39699935913086,
-      "learning_rate": 1.2452641074471286e-05,
-      "loss": 1.56,
+      "epoch": 0.47,
+      "grad_norm": 15.880305290222168,
+      "learning_rate": 1.684976781157177e-05,
+      "loss": 2.2642,
       "step": 3766
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 12.608708381652832,
-      "learning_rate": 1.2450636463866895e-05,
-      "loss": 1.0913,
+      "epoch": 0.47,
+      "grad_norm": 13.627202033996582,
+      "learning_rate": 1.6848931096515083e-05,
+      "loss": 0.9258,
       "step": 3767
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 9.895720481872559,
-      "learning_rate": 1.2448631853262505e-05,
-      "loss": 1.6166,
+      "epoch": 0.47,
+      "grad_norm": 201.3279571533203,
+      "learning_rate": 1.6848094381458397e-05,
+      "loss": 1.3532,
       "step": 3768
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 15.616667747497559,
-      "learning_rate": 1.2446627242658113e-05,
-      "loss": 2.4802,
+      "epoch": 0.47,
+      "grad_norm": 7.930774688720703,
+      "learning_rate": 1.6847257666401707e-05,
+      "loss": 2.4595,
       "step": 3769
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 28.047149658203125,
-      "learning_rate": 1.2444622632053725e-05,
-      "loss": 2.0519,
+      "epoch": 0.47,
+      "grad_norm": 22.180068969726562,
+      "learning_rate": 1.684642095134502e-05,
+      "loss": 2.2954,
       "step": 3770
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 15.117725372314453,
-      "learning_rate": 1.2442618021449335e-05,
-      "loss": 2.0924,
+      "epoch": 0.47,
+      "grad_norm": 6.394443035125732,
+      "learning_rate": 1.684558423628833e-05,
+      "loss": 0.7102,
       "step": 3771
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 11.850608825683594,
-      "learning_rate": 1.2440613410844943e-05,
-      "loss": 1.4544,
+      "epoch": 0.47,
+      "grad_norm": 37.96142578125,
+      "learning_rate": 1.6844747521231645e-05,
+      "loss": 2.581,
       "step": 3772
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 13.68972396850586,
-      "learning_rate": 1.2438608800240555e-05,
-      "loss": 2.0332,
+      "epoch": 0.47,
+      "grad_norm": 19.41549301147461,
+      "learning_rate": 1.6843910806174958e-05,
+      "loss": 1.3952,
       "step": 3773
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 12.880284309387207,
-      "learning_rate": 1.2436604189636164e-05,
-      "loss": 2.0553,
+      "epoch": 0.47,
+      "grad_norm": 9.572205543518066,
+      "learning_rate": 1.684307409111827e-05,
+      "loss": 2.2393,
       "step": 3774
     },
     {
-      "epoch": 1.13,
-      "grad_norm": 14.321048736572266,
-      "learning_rate": 1.2434599579031774e-05,
-      "loss": 1.2826,
+      "epoch": 0.47,
+      "grad_norm": 17.878549575805664,
+      "learning_rate": 1.6842237376061582e-05,
+      "loss": 1.6974,
       "step": 3775
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 22.247779846191406,
-      "learning_rate": 1.2432594968427385e-05,
-      "loss": 1.7716,
+      "epoch": 0.47,
+      "grad_norm": 11.891609191894531,
+      "learning_rate": 1.6841400661004896e-05,
+      "loss": 2.4472,
       "step": 3776
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 10.209996223449707,
-      "learning_rate": 1.2430590357822994e-05,
-      "loss": 1.5767,
+      "epoch": 0.47,
+      "grad_norm": 15.71055793762207,
+      "learning_rate": 1.684056394594821e-05,
+      "loss": 1.1206,
       "step": 3777
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 12.013636589050293,
-      "learning_rate": 1.2428585747218604e-05,
-      "loss": 1.114,
+      "epoch": 0.47,
+      "grad_norm": 18.522485733032227,
+      "learning_rate": 1.683972723089152e-05,
+      "loss": 0.7849,
       "step": 3778
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 51.70674514770508,
-      "learning_rate": 1.2426581136614212e-05,
-      "loss": 2.013,
+      "epoch": 0.47,
+      "grad_norm": 23.750144958496094,
+      "learning_rate": 1.6838890515834833e-05,
+      "loss": 2.3989,
       "step": 3779
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 18.054279327392578,
-      "learning_rate": 1.2424576526009824e-05,
-      "loss": 1.4477,
+      "epoch": 0.47,
+      "grad_norm": 7.889320373535156,
+      "learning_rate": 1.6838053800778147e-05,
+      "loss": 0.5604,
       "step": 3780
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 19.923124313354492,
-      "learning_rate": 1.2422571915405432e-05,
-      "loss": 1.3768,
+      "epoch": 0.47,
+      "grad_norm": 20.605247497558594,
+      "learning_rate": 1.6837217085721457e-05,
+      "loss": 2.7653,
       "step": 3781
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 23.104625701904297,
-      "learning_rate": 1.2420567304801043e-05,
-      "loss": 2.1138,
+      "epoch": 0.47,
+      "grad_norm": 10.587319374084473,
+      "learning_rate": 1.683638037066477e-05,
+      "loss": 1.3875,
       "step": 3782
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 8.69080638885498,
-      "learning_rate": 1.2418562694196654e-05,
-      "loss": 1.1123,
+      "epoch": 0.47,
+      "grad_norm": 7.350720405578613,
+      "learning_rate": 1.6835543655608085e-05,
+      "loss": 2.1851,
       "step": 3783
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 15.912267684936523,
-      "learning_rate": 1.2416558083592263e-05,
-      "loss": 2.4516,
+      "epoch": 0.47,
+      "grad_norm": 16.164628982543945,
+      "learning_rate": 1.6834706940551398e-05,
+      "loss": 1.4781,
       "step": 3784
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 12.12271785736084,
-      "learning_rate": 1.2414553472987873e-05,
-      "loss": 1.4647,
+      "epoch": 0.48,
+      "grad_norm": 11.476985931396484,
+      "learning_rate": 1.683387022549471e-05,
+      "loss": 0.9459,
       "step": 3785
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 9.104388236999512,
-      "learning_rate": 1.2412548862383483e-05,
-      "loss": 1.8779,
+      "epoch": 0.48,
+      "grad_norm": 7.0805792808532715,
+      "learning_rate": 1.6833033510438022e-05,
+      "loss": 2.0153,
       "step": 3786
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 17.152158737182617,
-      "learning_rate": 1.2410544251779093e-05,
-      "loss": 1.4711,
+      "epoch": 0.48,
+      "grad_norm": 19.46834373474121,
+      "learning_rate": 1.6832196795381336e-05,
+      "loss": 0.8265,
       "step": 3787
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 9.907875061035156,
-      "learning_rate": 1.2408539641174703e-05,
-      "loss": 1.1329,
+      "epoch": 0.48,
+      "grad_norm": 7.584367752075195,
+      "learning_rate": 1.6831360080324646e-05,
+      "loss": 1.703,
       "step": 3788
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 20.072593688964844,
-      "learning_rate": 1.2406535030570313e-05,
-      "loss": 1.9876,
+      "epoch": 0.48,
+      "grad_norm": 33.80003356933594,
+      "learning_rate": 1.683052336526796e-05,
+      "loss": 3.119,
       "step": 3789
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 14.026973724365234,
-      "learning_rate": 1.2404530419965923e-05,
-      "loss": 0.7741,
+      "epoch": 0.48,
+      "grad_norm": 20.491531372070312,
+      "learning_rate": 1.6829686650211273e-05,
+      "loss": 1.7394,
       "step": 3790
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 15.969365119934082,
-      "learning_rate": 1.2402525809361532e-05,
-      "loss": 1.6111,
+      "epoch": 0.48,
+      "grad_norm": 12.311060905456543,
+      "learning_rate": 1.6828849935154584e-05,
+      "loss": 1.6401,
       "step": 3791
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 15.787557601928711,
-      "learning_rate": 1.2400521198757143e-05,
-      "loss": 1.5455,
+      "epoch": 0.48,
+      "grad_norm": 10.289591789245605,
+      "learning_rate": 1.6828013220097897e-05,
+      "loss": 1.0717,
       "step": 3792
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 14.39417839050293,
-      "learning_rate": 1.2398516588152752e-05,
-      "loss": 1.0369,
+      "epoch": 0.48,
+      "grad_norm": 5.99196195602417,
+      "learning_rate": 1.682717650504121e-05,
+      "loss": 0.359,
       "step": 3793
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 13.072067260742188,
-      "learning_rate": 1.2396511977548362e-05,
-      "loss": 2.3197,
+      "epoch": 0.48,
+      "grad_norm": 9.215882301330566,
+      "learning_rate": 1.682633978998452e-05,
+      "loss": 1.2833,
       "step": 3794
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 14.899435997009277,
-      "learning_rate": 1.2394507366943974e-05,
-      "loss": 2.7447,
+      "epoch": 0.48,
+      "grad_norm": 13.91613483428955,
+      "learning_rate": 1.6825503074927835e-05,
+      "loss": 2.0734,
       "step": 3795
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 15.155303001403809,
-      "learning_rate": 1.2392502756339582e-05,
-      "loss": 1.4796,
+      "epoch": 0.48,
+      "grad_norm": 25.02107810974121,
+      "learning_rate": 1.682466635987115e-05,
+      "loss": 2.1832,
       "step": 3796
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 28.191987991333008,
-      "learning_rate": 1.2390498145735192e-05,
-      "loss": 1.7832,
+      "epoch": 0.48,
+      "grad_norm": 15.298924446105957,
+      "learning_rate": 1.682382964481446e-05,
+      "loss": 1.4136,
       "step": 3797
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 14.950464248657227,
-      "learning_rate": 1.23884935351308e-05,
-      "loss": 2.1336,
+      "epoch": 0.48,
+      "grad_norm": 18.208202362060547,
+      "learning_rate": 1.6822992929757772e-05,
+      "loss": 1.7215,
       "step": 3798
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 21.477602005004883,
-      "learning_rate": 1.2386488924526412e-05,
-      "loss": 1.7525,
+      "epoch": 0.48,
+      "grad_norm": 16.074575424194336,
+      "learning_rate": 1.6822156214701083e-05,
+      "loss": 1.9483,
       "step": 3799
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 25.496963500976562,
-      "learning_rate": 1.238448431392202e-05,
-      "loss": 1.6689,
+      "epoch": 0.48,
+      "grad_norm": 14.745115280151367,
+      "learning_rate": 1.6821319499644396e-05,
+      "loss": 1.6011,
       "step": 3800
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 12.373186111450195,
-      "learning_rate": 1.238247970331763e-05,
-      "loss": 2.1899,
+      "epoch": 0.48,
+      "grad_norm": 9.882591247558594,
+      "learning_rate": 1.682048278458771e-05,
+      "loss": 2.5467,
       "step": 3801
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 20.684795379638672,
-      "learning_rate": 1.2380475092713243e-05,
-      "loss": 2.6932,
+      "epoch": 0.48,
+      "grad_norm": 10.847468376159668,
+      "learning_rate": 1.681964606953102e-05,
+      "loss": 2.0858,
       "step": 3802
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 9.77327823638916,
-      "learning_rate": 1.2378470482108851e-05,
-      "loss": 1.5063,
+      "epoch": 0.48,
+      "grad_norm": 23.348876953125,
+      "learning_rate": 1.6818809354474334e-05,
+      "loss": 3.2153,
       "step": 3803
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 28.525667190551758,
-      "learning_rate": 1.2376465871504461e-05,
-      "loss": 1.532,
+      "epoch": 0.48,
+      "grad_norm": 21.69083023071289,
+      "learning_rate": 1.6817972639417648e-05,
+      "loss": 2.6129,
       "step": 3804
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 14.671927452087402,
-      "learning_rate": 1.2374461260900071e-05,
-      "loss": 2.0253,
+      "epoch": 0.48,
+      "grad_norm": 28.29841423034668,
+      "learning_rate": 1.681713592436096e-05,
+      "loss": 1.6809,
       "step": 3805
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 30.834918975830078,
-      "learning_rate": 1.2372456650295681e-05,
-      "loss": 2.39,
+      "epoch": 0.48,
+      "grad_norm": 11.64278507232666,
+      "learning_rate": 1.681629920930427e-05,
+      "loss": 0.8851,
       "step": 3806
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 16.612564086914062,
-      "learning_rate": 1.2370452039691291e-05,
-      "loss": 1.8819,
+      "epoch": 0.48,
+      "grad_norm": 10.83991813659668,
+      "learning_rate": 1.6815462494247585e-05,
+      "loss": 0.7002,
       "step": 3807
     },
     {
-      "epoch": 1.14,
-      "grad_norm": 13.142746925354004,
-      "learning_rate": 1.2368447429086901e-05,
-      "loss": 1.6061,
+      "epoch": 0.48,
+      "grad_norm": 20.13542938232422,
+      "learning_rate": 1.68146257791909e-05,
+      "loss": 2.0806,
       "step": 3808
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 18.517765045166016,
-      "learning_rate": 1.2366442818482511e-05,
-      "loss": 1.6218,
+      "epoch": 0.48,
+      "grad_norm": 3.9735586643218994,
+      "learning_rate": 1.681378906413421e-05,
+      "loss": 0.3442,
       "step": 3809
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 19.177627563476562,
-      "learning_rate": 1.236443820787812e-05,
-      "loss": 2.3172,
+      "epoch": 0.48,
+      "grad_norm": 10.199394226074219,
+      "learning_rate": 1.6812952349077523e-05,
+      "loss": 1.8834,
       "step": 3810
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 21.176328659057617,
-      "learning_rate": 1.2362433597273732e-05,
-      "loss": 1.8578,
+      "epoch": 0.48,
+      "grad_norm": 16.84869384765625,
+      "learning_rate": 1.6812115634020836e-05,
+      "loss": 2.7995,
       "step": 3811
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 21.381567001342773,
-      "learning_rate": 1.236042898666934e-05,
-      "loss": 1.5194,
+      "epoch": 0.48,
+      "grad_norm": 19.027019500732422,
+      "learning_rate": 1.681127891896415e-05,
+      "loss": 2.2333,
       "step": 3812
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 11.9020414352417,
-      "learning_rate": 1.235842437606495e-05,
-      "loss": 1.6908,
+      "epoch": 0.48,
+      "grad_norm": 13.716145515441895,
+      "learning_rate": 1.681044220390746e-05,
+      "loss": 2.0771,
       "step": 3813
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 39.47920608520508,
-      "learning_rate": 1.2356419765460562e-05,
-      "loss": 1.8726,
+      "epoch": 0.48,
+      "grad_norm": 7.44636344909668,
+      "learning_rate": 1.6809605488850774e-05,
+      "loss": 2.2718,
       "step": 3814
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 10.872967720031738,
-      "learning_rate": 1.235441515485617e-05,
-      "loss": 1.5569,
+      "epoch": 0.48,
+      "grad_norm": 28.244157791137695,
+      "learning_rate": 1.6808768773794087e-05,
+      "loss": 2.5986,
       "step": 3815
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 14.20240592956543,
-      "learning_rate": 1.235241054425178e-05,
-      "loss": 1.9525,
+      "epoch": 0.48,
+      "grad_norm": 11.873055458068848,
+      "learning_rate": 1.6807932058737398e-05,
+      "loss": 2.1554,
       "step": 3816
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 16.04691505432129,
-      "learning_rate": 1.2350405933647389e-05,
-      "loss": 1.4398,
+      "epoch": 0.48,
+      "grad_norm": 7.707242488861084,
+      "learning_rate": 1.680709534368071e-05,
+      "loss": 2.8684,
       "step": 3817
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 19.002132415771484,
-      "learning_rate": 1.2348401323043e-05,
-      "loss": 1.9253,
+      "epoch": 0.48,
+      "grad_norm": 9.765034675598145,
+      "learning_rate": 1.6806258628624025e-05,
+      "loss": 1.7397,
       "step": 3818
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 9.962849617004395,
-      "learning_rate": 1.234639671243861e-05,
-      "loss": 2.0496,
+      "epoch": 0.48,
+      "grad_norm": 11.618304252624512,
+      "learning_rate": 1.6805421913567335e-05,
+      "loss": 2.0023,
       "step": 3819
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 10.417901039123535,
-      "learning_rate": 1.2344392101834219e-05,
-      "loss": 1.9296,
+      "epoch": 0.48,
+      "grad_norm": 26.45114517211914,
+      "learning_rate": 1.680458519851065e-05,
+      "loss": 2.2092,
       "step": 3820
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 16.502777099609375,
-      "learning_rate": 1.234238749122983e-05,
-      "loss": 1.6108,
+      "epoch": 0.48,
+      "grad_norm": 133.9401092529297,
+      "learning_rate": 1.6803748483453963e-05,
+      "loss": 1.8951,
       "step": 3821
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 18.25497817993164,
-      "learning_rate": 1.234038288062544e-05,
-      "loss": 1.7453,
+      "epoch": 0.48,
+      "grad_norm": 15.46374797821045,
+      "learning_rate": 1.6802911768397273e-05,
+      "loss": 1.0543,
       "step": 3822
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 20.428895950317383,
-      "learning_rate": 1.233837827002105e-05,
-      "loss": 2.3424,
+      "epoch": 0.48,
+      "grad_norm": 12.88305377960205,
+      "learning_rate": 1.6802075053340587e-05,
+      "loss": 3.2489,
       "step": 3823
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 13.153669357299805,
-      "learning_rate": 1.2336373659416658e-05,
-      "loss": 1.3317,
+      "epoch": 0.48,
+      "grad_norm": 14.470881462097168,
+      "learning_rate": 1.6801238338283897e-05,
+      "loss": 1.6371,
       "step": 3824
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 13.739114761352539,
-      "learning_rate": 1.233436904881227e-05,
-      "loss": 1.4802,
+      "epoch": 0.48,
+      "grad_norm": 17.17293357849121,
+      "learning_rate": 1.680040162322721e-05,
+      "loss": 2.6381,
       "step": 3825
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 17.079265594482422,
-      "learning_rate": 1.233236443820788e-05,
-      "loss": 1.6825,
+      "epoch": 0.48,
+      "grad_norm": 26.487239837646484,
+      "learning_rate": 1.6799564908170524e-05,
+      "loss": 2.1349,
       "step": 3826
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 15.86989974975586,
-      "learning_rate": 1.2330359827603488e-05,
-      "loss": 1.7154,
+      "epoch": 0.48,
+      "grad_norm": 15.372435569763184,
+      "learning_rate": 1.6798728193113834e-05,
+      "loss": 1.4839,
       "step": 3827
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 14.82314682006836,
-      "learning_rate": 1.23283552169991e-05,
-      "loss": 1.3117,
+      "epoch": 0.48,
+      "grad_norm": 32.98752975463867,
+      "learning_rate": 1.6797891478057148e-05,
+      "loss": 2.9483,
       "step": 3828
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 21.10222625732422,
-      "learning_rate": 1.2326350606394708e-05,
-      "loss": 1.4309,
+      "epoch": 0.48,
+      "grad_norm": 14.529047012329102,
+      "learning_rate": 1.679705476300046e-05,
+      "loss": 2.0955,
       "step": 3829
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 15.547961235046387,
-      "learning_rate": 1.2324345995790318e-05,
-      "loss": 1.3606,
+      "epoch": 0.48,
+      "grad_norm": 9.023208618164062,
+      "learning_rate": 1.6796218047943772e-05,
+      "loss": 1.2004,
       "step": 3830
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 13.123003005981445,
-      "learning_rate": 1.232234138518593e-05,
-      "loss": 2.1664,
+      "epoch": 0.48,
+      "grad_norm": 12.670555114746094,
+      "learning_rate": 1.6795381332887086e-05,
+      "loss": 3.1739,
       "step": 3831
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 14.233585357666016,
-      "learning_rate": 1.2320336774581538e-05,
-      "loss": 1.9827,
+      "epoch": 0.48,
+      "grad_norm": 25.908464431762695,
+      "learning_rate": 1.67945446178304e-05,
+      "loss": 2.7281,
       "step": 3832
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 8.499350547790527,
-      "learning_rate": 1.2318332163977148e-05,
-      "loss": 0.8647,
+      "epoch": 0.48,
+      "grad_norm": 5.578554153442383,
+      "learning_rate": 1.6793707902773713e-05,
+      "loss": 0.2401,
       "step": 3833
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 24.76041603088379,
-      "learning_rate": 1.2316327553372758e-05,
-      "loss": 2.5669,
+      "epoch": 0.48,
+      "grad_norm": 14.278937339782715,
+      "learning_rate": 1.6792871187717023e-05,
+      "loss": 1.7784,
       "step": 3834
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 11.308053970336914,
-      "learning_rate": 1.2314322942768369e-05,
-      "loss": 1.5768,
+      "epoch": 0.48,
+      "grad_norm": 15.300288200378418,
+      "learning_rate": 1.6792034472660337e-05,
+      "loss": 1.2436,
       "step": 3835
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 25.03925132751465,
-      "learning_rate": 1.2312318332163977e-05,
-      "loss": 1.7572,
+      "epoch": 0.48,
+      "grad_norm": 21.30611228942871,
+      "learning_rate": 1.679119775760365e-05,
+      "loss": 1.885,
       "step": 3836
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 19.462387084960938,
-      "learning_rate": 1.2310313721559589e-05,
-      "loss": 2.0888,
+      "epoch": 0.48,
+      "grad_norm": 16.563570022583008,
+      "learning_rate": 1.679036104254696e-05,
+      "loss": 1.578,
       "step": 3837
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 12.561948776245117,
-      "learning_rate": 1.2308309110955199e-05,
-      "loss": 1.1124,
+      "epoch": 0.48,
+      "grad_norm": 19.17254066467285,
+      "learning_rate": 1.6789524327490274e-05,
+      "loss": 2.2793,
       "step": 3838
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 34.78009796142578,
-      "learning_rate": 1.2306304500350807e-05,
-      "loss": 1.5277,
+      "epoch": 0.48,
+      "grad_norm": 9.509984016418457,
+      "learning_rate": 1.6788687612433588e-05,
+      "loss": 1.495,
       "step": 3839
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 26.024934768676758,
-      "learning_rate": 1.2304299889746419e-05,
-      "loss": 2.0164,
-      "step": 3840
-    },
-    {
-      "epoch": 1.15,
-      "eval_loss": 0.21982242166996002,
-      "eval_runtime": 43.5261,
-      "eval_samples_per_second": 33.98,
-      "eval_steps_per_second": 33.98,
+      "epoch": 0.48,
+      "grad_norm": 33.51737594604492,
+      "learning_rate": 1.67878508973769e-05,
+      "loss": 2.7438,
       "step": 3840
     },
     {
-      "epoch": 1.15,
-      "grad_norm": 6.806481838226318,
-      "learning_rate": 1.2302295279142027e-05,
-      "loss": 0.7476,
+      "epoch": 0.48,
+      "grad_norm": 20.427261352539062,
+      "learning_rate": 1.6787014182320212e-05,
+      "loss": 3.2905,
       "step": 3841
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 18.72743797302246,
-      "learning_rate": 1.2300290668537637e-05,
-      "loss": 1.4762,
+      "epoch": 0.48,
+      "grad_norm": 11.561785697937012,
+      "learning_rate": 1.6786177467263526e-05,
+      "loss": 3.0269,
       "step": 3842
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 55.82676315307617,
-      "learning_rate": 1.2298286057933246e-05,
-      "loss": 1.809,
+      "epoch": 0.48,
+      "grad_norm": 19.586536407470703,
+      "learning_rate": 1.678534075220684e-05,
+      "loss": 2.9902,
       "step": 3843
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 13.187909126281738,
-      "learning_rate": 1.2296281447328858e-05,
-      "loss": 1.9105,
+      "epoch": 0.48,
+      "grad_norm": 10.160581588745117,
+      "learning_rate": 1.678450403715015e-05,
+      "loss": 1.4402,
       "step": 3844
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 11.694255828857422,
-      "learning_rate": 1.2294276836724468e-05,
-      "loss": 1.6571,
+      "epoch": 0.48,
+      "grad_norm": 21.57250213623047,
+      "learning_rate": 1.6783667322093463e-05,
+      "loss": 1.9489,
       "step": 3845
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 20.324060440063477,
-      "learning_rate": 1.2292272226120076e-05,
-      "loss": 1.3941,
+      "epoch": 0.48,
+      "grad_norm": 14.982203483581543,
+      "learning_rate": 1.6782830607036777e-05,
+      "loss": 1.0441,
       "step": 3846
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 13.865283012390137,
-      "learning_rate": 1.2290267615515688e-05,
-      "loss": 1.9631,
+      "epoch": 0.48,
+      "grad_norm": 14.22341251373291,
+      "learning_rate": 1.6781993891980087e-05,
+      "loss": 1.8044,
       "step": 3847
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 18.18462562561035,
-      "learning_rate": 1.2288263004911296e-05,
-      "loss": 2.1097,
+      "epoch": 0.48,
+      "grad_norm": 6.044835567474365,
+      "learning_rate": 1.67811571769234e-05,
+      "loss": 0.6106,
       "step": 3848
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 12.211462020874023,
-      "learning_rate": 1.2286258394306906e-05,
-      "loss": 1.8528,
+      "epoch": 0.48,
+      "grad_norm": 28.15514373779297,
+      "learning_rate": 1.6780320461866714e-05,
+      "loss": 1.8341,
       "step": 3849
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 12.799270629882812,
-      "learning_rate": 1.2284253783702518e-05,
-      "loss": 1.9839,
+      "epoch": 0.48,
+      "grad_norm": 13.517406463623047,
+      "learning_rate": 1.6779483746810025e-05,
+      "loss": 1.1603,
       "step": 3850
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 36.99559020996094,
-      "learning_rate": 1.2282249173098127e-05,
-      "loss": 1.7886,
+      "epoch": 0.48,
+      "grad_norm": 12.102564811706543,
+      "learning_rate": 1.6778647031753338e-05,
+      "loss": 1.1951,
       "step": 3851
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 15.477313995361328,
-      "learning_rate": 1.2280244562493737e-05,
-      "loss": 1.6013,
+      "epoch": 0.48,
+      "grad_norm": 22.357250213623047,
+      "learning_rate": 1.677781031669665e-05,
+      "loss": 1.5787,
       "step": 3852
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 12.97021198272705,
-      "learning_rate": 1.2278239951889345e-05,
-      "loss": 2.1934,
+      "epoch": 0.48,
+      "grad_norm": 8.78569507598877,
+      "learning_rate": 1.6776973601639962e-05,
+      "loss": 0.8656,
       "step": 3853
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 9.071000099182129,
-      "learning_rate": 1.2276235341284957e-05,
-      "loss": 1.2502,
+      "epoch": 0.48,
+      "grad_norm": 14.004049301147461,
+      "learning_rate": 1.6776136886583276e-05,
+      "loss": 2.3198,
       "step": 3854
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 13.090424537658691,
-      "learning_rate": 1.2274230730680565e-05,
-      "loss": 1.4692,
+      "epoch": 0.48,
+      "grad_norm": 11.207050323486328,
+      "learning_rate": 1.6775300171526586e-05,
+      "loss": 2.6709,
       "step": 3855
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 15.393147468566895,
-      "learning_rate": 1.2272226120076175e-05,
-      "loss": 1.1286,
+      "epoch": 0.48,
+      "grad_norm": 17.077152252197266,
+      "learning_rate": 1.67744634564699e-05,
+      "loss": 1.5795,
       "step": 3856
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 38.615726470947266,
-      "learning_rate": 1.2270221509471787e-05,
-      "loss": 2.5387,
+      "epoch": 0.48,
+      "grad_norm": 25.284257888793945,
+      "learning_rate": 1.6773626741413213e-05,
+      "loss": 3.127,
       "step": 3857
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 28.15321922302246,
-      "learning_rate": 1.2268216898867395e-05,
-      "loss": 2.187,
+      "epoch": 0.48,
+      "grad_norm": 12.849029541015625,
+      "learning_rate": 1.6772790026356524e-05,
+      "loss": 1.9975,
       "step": 3858
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 12.39357852935791,
-      "learning_rate": 1.2266212288263007e-05,
-      "loss": 2.2882,
+      "epoch": 0.48,
+      "grad_norm": 12.927366256713867,
+      "learning_rate": 1.6771953311299837e-05,
+      "loss": 3.0397,
       "step": 3859
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 30.955167770385742,
-      "learning_rate": 1.2264207677658616e-05,
-      "loss": 1.8665,
+      "epoch": 0.48,
+      "grad_norm": 7.013041973114014,
+      "learning_rate": 1.677111659624315e-05,
+      "loss": 0.2361,
       "step": 3860
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 16.133106231689453,
-      "learning_rate": 1.2262203067054226e-05,
-      "loss": 1.8293,
+      "epoch": 0.48,
+      "grad_norm": 10.52235221862793,
+      "learning_rate": 1.6770279881186465e-05,
+      "loss": 1.1664,
       "step": 3861
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 7.856118202209473,
-      "learning_rate": 1.2260198456449837e-05,
-      "loss": 0.9778,
+      "epoch": 0.48,
+      "grad_norm": 5.555588722229004,
+      "learning_rate": 1.6769443166129775e-05,
+      "loss": 0.4259,
       "step": 3862
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 12.470169067382812,
-      "learning_rate": 1.2258193845845446e-05,
-      "loss": 1.4214,
+      "epoch": 0.48,
+      "grad_norm": 6.372686862945557,
+      "learning_rate": 1.676860645107309e-05,
+      "loss": 1.5433,
       "step": 3863
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 19.56519317626953,
-      "learning_rate": 1.2256189235241056e-05,
-      "loss": 2.2586,
+      "epoch": 0.48,
+      "grad_norm": 12.421819686889648,
+      "learning_rate": 1.6767769736016402e-05,
+      "loss": 1.1898,
       "step": 3864
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 23.616268157958984,
-      "learning_rate": 1.2254184624636664e-05,
-      "loss": 2.7421,
+      "epoch": 0.49,
+      "grad_norm": 13.235730171203613,
+      "learning_rate": 1.6766933020959712e-05,
+      "loss": 1.0993,
       "step": 3865
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 15.930326461791992,
-      "learning_rate": 1.2252180014032276e-05,
-      "loss": 2.1892,
+      "epoch": 0.49,
+      "grad_norm": 12.569911003112793,
+      "learning_rate": 1.6766096305903026e-05,
+      "loss": 2.0045,
       "step": 3866
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 24.9573974609375,
-      "learning_rate": 1.2250175403427884e-05,
-      "loss": 2.7173,
+      "epoch": 0.49,
+      "grad_norm": 19.18654441833496,
+      "learning_rate": 1.676525959084634e-05,
+      "loss": 1.0166,
       "step": 3867
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 8.64990234375,
-      "learning_rate": 1.2248170792823495e-05,
-      "loss": 1.8587,
+      "epoch": 0.49,
+      "grad_norm": 12.57631778717041,
+      "learning_rate": 1.6764422875789653e-05,
+      "loss": 1.365,
       "step": 3868
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 29.846378326416016,
-      "learning_rate": 1.2246166182219106e-05,
-      "loss": 1.8677,
+      "epoch": 0.49,
+      "grad_norm": 12.228401184082031,
+      "learning_rate": 1.6763586160732964e-05,
+      "loss": 1.8901,
       "step": 3869
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 11.002440452575684,
-      "learning_rate": 1.2244161571614715e-05,
-      "loss": 1.6396,
+      "epoch": 0.49,
+      "grad_norm": 18.046661376953125,
+      "learning_rate": 1.6762749445676277e-05,
+      "loss": 2.1157,
       "step": 3870
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 16.57766342163086,
-      "learning_rate": 1.2242156961010325e-05,
-      "loss": 1.381,
+      "epoch": 0.49,
+      "grad_norm": 17.057580947875977,
+      "learning_rate": 1.676191273061959e-05,
+      "loss": 1.4784,
       "step": 3871
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 16.003095626831055,
-      "learning_rate": 1.2240152350405933e-05,
-      "loss": 1.7541,
+      "epoch": 0.49,
+      "grad_norm": 2.8317675590515137,
+      "learning_rate": 1.67610760155629e-05,
+      "loss": 0.1153,
       "step": 3872
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 16.007123947143555,
-      "learning_rate": 1.2238147739801545e-05,
-      "loss": 1.9818,
+      "epoch": 0.49,
+      "grad_norm": 8.640043258666992,
+      "learning_rate": 1.6760239300506215e-05,
+      "loss": 1.1723,
       "step": 3873
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 16.965221405029297,
-      "learning_rate": 1.2236143129197155e-05,
-      "loss": 2.3105,
+      "epoch": 0.49,
+      "grad_norm": 9.828316688537598,
+      "learning_rate": 1.675940258544953e-05,
+      "loss": 2.7181,
       "step": 3874
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 27.36821174621582,
-      "learning_rate": 1.2234138518592763e-05,
-      "loss": 2.4866,
+      "epoch": 0.49,
+      "grad_norm": 5.815865993499756,
+      "learning_rate": 1.675856587039284e-05,
+      "loss": 0.4049,
       "step": 3875
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 13.598553657531738,
-      "learning_rate": 1.2232133907988375e-05,
-      "loss": 1.654,
+      "epoch": 0.49,
+      "grad_norm": 25.681018829345703,
+      "learning_rate": 1.6757729155336152e-05,
+      "loss": 3.4645,
       "step": 3876
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 15.706884384155273,
-      "learning_rate": 1.2230129297383984e-05,
-      "loss": 2.687,
+      "epoch": 0.49,
+      "grad_norm": 9.771852493286133,
+      "learning_rate": 1.6756892440279463e-05,
+      "loss": 2.0329,
       "step": 3877
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 16.47588348388672,
-      "learning_rate": 1.2228124686779594e-05,
-      "loss": 1.6308,
+      "epoch": 0.49,
+      "grad_norm": 24.728055953979492,
+      "learning_rate": 1.6756055725222776e-05,
+      "loss": 3.5055,
       "step": 3878
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 12.876527786254883,
-      "learning_rate": 1.2226120076175204e-05,
-      "loss": 1.4511,
+      "epoch": 0.49,
+      "grad_norm": 8.333413124084473,
+      "learning_rate": 1.675521901016609e-05,
+      "loss": 2.9655,
       "step": 3879
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 20.67323112487793,
-      "learning_rate": 1.2224115465570814e-05,
-      "loss": 1.4801,
+      "epoch": 0.49,
+      "grad_norm": 15.079514503479004,
+      "learning_rate": 1.67543822951094e-05,
+      "loss": 2.9962,
       "step": 3880
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 29.967750549316406,
-      "learning_rate": 1.2222110854966424e-05,
-      "loss": 2.2102,
+      "epoch": 0.49,
+      "grad_norm": 16.224035263061523,
+      "learning_rate": 1.6753545580052714e-05,
+      "loss": 1.4763,
       "step": 3881
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 12.949405670166016,
-      "learning_rate": 1.2220106244362034e-05,
-      "loss": 1.8167,
+      "epoch": 0.49,
+      "grad_norm": 6.735492706298828,
+      "learning_rate": 1.6752708864996027e-05,
+      "loss": 0.4744,
       "step": 3882
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 26.90522575378418,
-      "learning_rate": 1.2218101633757644e-05,
-      "loss": 1.7836,
+      "epoch": 0.49,
+      "grad_norm": 13.388711929321289,
+      "learning_rate": 1.6751872149939338e-05,
+      "loss": 0.7654,
       "step": 3883
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 14.43256950378418,
-      "learning_rate": 1.2216097023153253e-05,
-      "loss": 1.5144,
+      "epoch": 0.49,
+      "grad_norm": 25.84760093688965,
+      "learning_rate": 1.675103543488265e-05,
+      "loss": 3.6068,
       "step": 3884
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 10.095649719238281,
-      "learning_rate": 1.2214092412548864e-05,
-      "loss": 0.8749,
+      "epoch": 0.49,
+      "grad_norm": 11.907512664794922,
+      "learning_rate": 1.6750198719825965e-05,
+      "loss": 1.642,
       "step": 3885
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 7.504431247711182,
-      "learning_rate": 1.2212087801944473e-05,
-      "loss": 1.5074,
+      "epoch": 0.49,
+      "grad_norm": 12.980208396911621,
+      "learning_rate": 1.6749362004769275e-05,
+      "loss": 1.0291,
       "step": 3886
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 19.273103713989258,
-      "learning_rate": 1.2210083191340083e-05,
-      "loss": 2.6758,
+      "epoch": 0.49,
+      "grad_norm": 12.23293685913086,
+      "learning_rate": 1.674852528971259e-05,
+      "loss": 2.3221,
       "step": 3887
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 24.052509307861328,
-      "learning_rate": 1.2208078580735695e-05,
-      "loss": 1.6407,
+      "epoch": 0.49,
+      "grad_norm": 13.18155288696289,
+      "learning_rate": 1.6747688574655903e-05,
+      "loss": 1.0233,
       "step": 3888
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 19.1132869720459,
-      "learning_rate": 1.2206073970131303e-05,
-      "loss": 1.7163,
+      "epoch": 0.49,
+      "grad_norm": 8.485978126525879,
+      "learning_rate": 1.6746851859599216e-05,
+      "loss": 1.2268,
       "step": 3889
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 51.50567626953125,
-      "learning_rate": 1.2204069359526913e-05,
-      "loss": 2.9385,
+      "epoch": 0.49,
+      "grad_norm": 9.914412498474121,
+      "learning_rate": 1.6746015144542526e-05,
+      "loss": 1.3486,
       "step": 3890
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 18.24009132385254,
-      "learning_rate": 1.2202064748922521e-05,
-      "loss": 1.7915,
+      "epoch": 0.49,
+      "grad_norm": 6.944596290588379,
+      "learning_rate": 1.674517842948584e-05,
+      "loss": 1.0267,
       "step": 3891
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 14.289188385009766,
-      "learning_rate": 1.2200060138318133e-05,
-      "loss": 1.8599,
+      "epoch": 0.49,
+      "grad_norm": 10.893056869506836,
+      "learning_rate": 1.6744341714429154e-05,
+      "loss": 0.8046,
       "step": 3892
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 35.022804260253906,
-      "learning_rate": 1.2198055527713743e-05,
-      "loss": 2.9202,
+      "epoch": 0.49,
+      "grad_norm": 20.531518936157227,
+      "learning_rate": 1.6743504999372464e-05,
+      "loss": 3.314,
       "step": 3893
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 25.463573455810547,
-      "learning_rate": 1.2196050917109352e-05,
-      "loss": 1.622,
+      "epoch": 0.49,
+      "grad_norm": 26.03514862060547,
+      "learning_rate": 1.6742668284315778e-05,
+      "loss": 1.8338,
       "step": 3894
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 10.089998245239258,
-      "learning_rate": 1.2194046306504963e-05,
-      "loss": 1.1327,
+      "epoch": 0.49,
+      "grad_norm": 14.674633026123047,
+      "learning_rate": 1.674183156925909e-05,
+      "loss": 2.1986,
       "step": 3895
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 23.384336471557617,
-      "learning_rate": 1.2192041695900572e-05,
-      "loss": 1.2661,
+      "epoch": 0.49,
+      "grad_norm": 22.857954025268555,
+      "learning_rate": 1.6740994854202405e-05,
+      "loss": 3.8863,
       "step": 3896
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 12.663224220275879,
-      "learning_rate": 1.2190037085296182e-05,
-      "loss": 2.6986,
+      "epoch": 0.49,
+      "grad_norm": 16.738014221191406,
+      "learning_rate": 1.6740158139145715e-05,
+      "loss": 2.1353,
       "step": 3897
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 16.6176700592041,
-      "learning_rate": 1.218803247469179e-05,
-      "loss": 1.894,
+      "epoch": 0.49,
+      "grad_norm": 14.383560180664062,
+      "learning_rate": 1.673932142408903e-05,
+      "loss": 3.1573,
       "step": 3898
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 12.314738273620605,
-      "learning_rate": 1.2186027864087402e-05,
-      "loss": 1.2068,
+      "epoch": 0.49,
+      "grad_norm": 31.759464263916016,
+      "learning_rate": 1.6738484709032343e-05,
+      "loss": 2.1836,
       "step": 3899
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 23.991886138916016,
-      "learning_rate": 1.2184023253483012e-05,
-      "loss": 1.6332,
+      "epoch": 0.49,
+      "grad_norm": 9.206013679504395,
+      "learning_rate": 1.6737647993975653e-05,
+      "loss": 1.3652,
       "step": 3900
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 23.119836807250977,
-      "learning_rate": 1.218201864287862e-05,
-      "loss": 2.1423,
+      "epoch": 0.49,
+      "grad_norm": 7.129506587982178,
+      "learning_rate": 1.6736811278918966e-05,
+      "loss": 1.1021,
       "step": 3901
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 29.719497680664062,
-      "learning_rate": 1.2180014032274232e-05,
-      "loss": 1.4554,
+      "epoch": 0.49,
+      "grad_norm": 21.61859703063965,
+      "learning_rate": 1.673597456386228e-05,
+      "loss": 2.0031,
       "step": 3902
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 13.610468864440918,
-      "learning_rate": 1.217800942166984e-05,
-      "loss": 1.6478,
+      "epoch": 0.49,
+      "grad_norm": 16.006519317626953,
+      "learning_rate": 1.673513784880559e-05,
+      "loss": 2.8328,
       "step": 3903
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 12.19708251953125,
-      "learning_rate": 1.217600481106545e-05,
-      "loss": 1.4697,
+      "epoch": 0.49,
+      "grad_norm": 14.28494930267334,
+      "learning_rate": 1.6734301133748904e-05,
+      "loss": 1.6195,
       "step": 3904
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 14.37850570678711,
-      "learning_rate": 1.2174000200461063e-05,
-      "loss": 1.2426,
+      "epoch": 0.49,
+      "grad_norm": 27.926685333251953,
+      "learning_rate": 1.6733464418692214e-05,
+      "loss": 1.0091,
       "step": 3905
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 16.169696807861328,
-      "learning_rate": 1.2171995589856671e-05,
-      "loss": 1.7116,
+      "epoch": 0.49,
+      "grad_norm": 12.705221176147461,
+      "learning_rate": 1.6732627703635528e-05,
+      "loss": 1.2817,
       "step": 3906
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 14.644316673278809,
-      "learning_rate": 1.2169990979252281e-05,
-      "loss": 1.5071,
+      "epoch": 0.49,
+      "grad_norm": 13.459248542785645,
+      "learning_rate": 1.673179098857884e-05,
+      "loss": 1.7091,
       "step": 3907
     },
     {
-      "epoch": 1.17,
-      "grad_norm": 10.753741264343262,
-      "learning_rate": 1.2167986368647891e-05,
-      "loss": 1.8289,
+      "epoch": 0.49,
+      "grad_norm": 19.173797607421875,
+      "learning_rate": 1.6730954273522152e-05,
+      "loss": 3.2456,
       "step": 3908
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 17.671937942504883,
-      "learning_rate": 1.2165981758043501e-05,
-      "loss": 1.3769,
+      "epoch": 0.49,
+      "grad_norm": 16.161258697509766,
+      "learning_rate": 1.6730117558465465e-05,
+      "loss": 1.4467,
       "step": 3909
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 10.696758270263672,
-      "learning_rate": 1.216397714743911e-05,
-      "loss": 1.7544,
+      "epoch": 0.49,
+      "grad_norm": 24.011537551879883,
+      "learning_rate": 1.672928084340878e-05,
+      "loss": 2.3174,
       "step": 3910
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 10.894145011901855,
-      "learning_rate": 1.2161972536834721e-05,
-      "loss": 1.3638,
+      "epoch": 0.49,
+      "grad_norm": 15.91162395477295,
+      "learning_rate": 1.672844412835209e-05,
+      "loss": 4.2072,
       "step": 3911
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 14.590895652770996,
-      "learning_rate": 1.2159967926230332e-05,
-      "loss": 1.4625,
+      "epoch": 0.49,
+      "grad_norm": 24.110483169555664,
+      "learning_rate": 1.6727607413295403e-05,
+      "loss": 1.7958,
       "step": 3912
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 20.419675827026367,
-      "learning_rate": 1.215796331562594e-05,
-      "loss": 2.223,
+      "epoch": 0.49,
+      "grad_norm": 8.102417945861816,
+      "learning_rate": 1.6726770698238717e-05,
+      "loss": 1.6236,
       "step": 3913
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 29.404043197631836,
-      "learning_rate": 1.2155958705021552e-05,
-      "loss": 1.6396,
+      "epoch": 0.49,
+      "grad_norm": 8.216350555419922,
+      "learning_rate": 1.6725933983182027e-05,
+      "loss": 3.1026,
       "step": 3914
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 24.126564025878906,
-      "learning_rate": 1.215395409441716e-05,
-      "loss": 1.6138,
+      "epoch": 0.49,
+      "grad_norm": 9.680403709411621,
+      "learning_rate": 1.672509726812534e-05,
+      "loss": 2.7506,
       "step": 3915
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 13.458683967590332,
-      "learning_rate": 1.215194948381277e-05,
-      "loss": 2.0191,
+      "epoch": 0.49,
+      "grad_norm": 11.488420486450195,
+      "learning_rate": 1.6724260553068654e-05,
+      "loss": 2.0227,
       "step": 3916
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 9.993988037109375,
-      "learning_rate": 1.2149944873208382e-05,
-      "loss": 1.3779,
+      "epoch": 0.49,
+      "grad_norm": 32.506500244140625,
+      "learning_rate": 1.6723423838011968e-05,
+      "loss": 1.5272,
       "step": 3917
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 13.048202514648438,
-      "learning_rate": 1.214794026260399e-05,
-      "loss": 1.8113,
+      "epoch": 0.49,
+      "grad_norm": 5.9329352378845215,
+      "learning_rate": 1.6722587122955278e-05,
+      "loss": 0.4988,
       "step": 3918
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 11.234829902648926,
-      "learning_rate": 1.21459356519996e-05,
-      "loss": 1.4256,
+      "epoch": 0.49,
+      "grad_norm": 7.876817226409912,
+      "learning_rate": 1.6721750407898592e-05,
+      "loss": 1.166,
       "step": 3919
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 25.56989860534668,
-      "learning_rate": 1.2143931041395209e-05,
-      "loss": 2.0883,
+      "epoch": 0.49,
+      "grad_norm": 11.694708824157715,
+      "learning_rate": 1.6720913692841905e-05,
+      "loss": 1.7381,
       "step": 3920
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 12.80074405670166,
-      "learning_rate": 1.214192643079082e-05,
-      "loss": 1.7903,
+      "epoch": 0.49,
+      "grad_norm": 17.446664810180664,
+      "learning_rate": 1.6720076977785216e-05,
+      "loss": 1.4075,
       "step": 3921
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 26.22979736328125,
-      "learning_rate": 1.2139921820186429e-05,
-      "loss": 2.3038,
+      "epoch": 0.49,
+      "grad_norm": 15.675682067871094,
+      "learning_rate": 1.671924026272853e-05,
+      "loss": 2.8233,
       "step": 3922
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 15.948308944702148,
-      "learning_rate": 1.2137917209582039e-05,
-      "loss": 1.9847,
+      "epoch": 0.49,
+      "grad_norm": 22.835962295532227,
+      "learning_rate": 1.6718403547671843e-05,
+      "loss": 1.0312,
       "step": 3923
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 20.231281280517578,
-      "learning_rate": 1.213591259897765e-05,
-      "loss": 2.6807,
+      "epoch": 0.49,
+      "grad_norm": 9.870447158813477,
+      "learning_rate": 1.6717566832615157e-05,
+      "loss": 1.2746,
       "step": 3924
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 26.88908576965332,
-      "learning_rate": 1.213390798837326e-05,
-      "loss": 2.2021,
+      "epoch": 0.49,
+      "grad_norm": 43.03246307373047,
+      "learning_rate": 1.6716730117558467e-05,
+      "loss": 2.252,
       "step": 3925
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 9.647951126098633,
-      "learning_rate": 1.213190337776887e-05,
-      "loss": 1.7718,
+      "epoch": 0.49,
+      "grad_norm": 10.218485832214355,
+      "learning_rate": 1.671589340250178e-05,
+      "loss": 0.9047,
       "step": 3926
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 25.476966857910156,
-      "learning_rate": 1.212989876716448e-05,
-      "loss": 1.7512,
+      "epoch": 0.49,
+      "grad_norm": 9.962735176086426,
+      "learning_rate": 1.6715056687445094e-05,
+      "loss": 1.1066,
       "step": 3927
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 17.493688583374023,
-      "learning_rate": 1.212789415656009e-05,
-      "loss": 2.0739,
+      "epoch": 0.49,
+      "grad_norm": 11.211706161499023,
+      "learning_rate": 1.6714219972388404e-05,
+      "loss": 1.5452,
       "step": 3928
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 18.608076095581055,
-      "learning_rate": 1.2125889545955698e-05,
-      "loss": 1.8991,
+      "epoch": 0.49,
+      "grad_norm": 11.349990844726562,
+      "learning_rate": 1.6713383257331718e-05,
+      "loss": 0.8992,
       "step": 3929
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 26.32684326171875,
-      "learning_rate": 1.212388493535131e-05,
-      "loss": 1.3218,
+      "epoch": 0.49,
+      "grad_norm": 14.319396018981934,
+      "learning_rate": 1.671254654227503e-05,
+      "loss": 2.6085,
       "step": 3930
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 14.290658950805664,
-      "learning_rate": 1.212188032474692e-05,
-      "loss": 1.6064,
+      "epoch": 0.49,
+      "grad_norm": 34.960479736328125,
+      "learning_rate": 1.6711709827218342e-05,
+      "loss": 2.4196,
       "step": 3931
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 8.176145553588867,
-      "learning_rate": 1.2119875714142528e-05,
-      "loss": 1.101,
+      "epoch": 0.49,
+      "grad_norm": 18.694778442382812,
+      "learning_rate": 1.6710873112161656e-05,
+      "loss": 1.6912,
       "step": 3932
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 77.52426147460938,
-      "learning_rate": 1.211787110353814e-05,
-      "loss": 1.8667,
+      "epoch": 0.49,
+      "grad_norm": 33.8570442199707,
+      "learning_rate": 1.6710036397104966e-05,
+      "loss": 2.4436,
       "step": 3933
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 13.218364715576172,
-      "learning_rate": 1.2115866492933748e-05,
-      "loss": 2.0845,
+      "epoch": 0.49,
+      "grad_norm": 6.813931465148926,
+      "learning_rate": 1.670919968204828e-05,
+      "loss": 4.083,
       "step": 3934
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 13.180416107177734,
-      "learning_rate": 1.2113861882329358e-05,
-      "loss": 1.7175,
+      "epoch": 0.49,
+      "grad_norm": 4.025628089904785,
+      "learning_rate": 1.670836296699159e-05,
+      "loss": 0.09,
       "step": 3935
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 27.077486038208008,
-      "learning_rate": 1.211185727172497e-05,
-      "loss": 1.5133,
+      "epoch": 0.49,
+      "grad_norm": 7.611042499542236,
+      "learning_rate": 1.6707526251934903e-05,
+      "loss": 0.9509,
       "step": 3936
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 14.540903091430664,
-      "learning_rate": 1.2109852661120579e-05,
-      "loss": 2.1343,
+      "epoch": 0.49,
+      "grad_norm": 56.0538330078125,
+      "learning_rate": 1.6706689536878217e-05,
+      "loss": 1.2226,
       "step": 3937
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 18.087007522583008,
-      "learning_rate": 1.2107848050516189e-05,
-      "loss": 1.6534,
+      "epoch": 0.49,
+      "grad_norm": 12.239697456359863,
+      "learning_rate": 1.670585282182153e-05,
+      "loss": 2.6821,
       "step": 3938
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 18.4873046875,
-      "learning_rate": 1.2105843439911797e-05,
-      "loss": 1.8601,
+      "epoch": 0.49,
+      "grad_norm": 21.050329208374023,
+      "learning_rate": 1.670501610676484e-05,
+      "loss": 1.2584,
       "step": 3939
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 11.12307071685791,
-      "learning_rate": 1.2103838829307409e-05,
-      "loss": 1.4969,
+      "epoch": 0.49,
+      "grad_norm": 46.02468490600586,
+      "learning_rate": 1.6704179391708155e-05,
+      "loss": 2.3161,
       "step": 3940
     },
     {
-      "epoch": 1.18,
-      "grad_norm": 22.688817977905273,
-      "learning_rate": 1.2101834218703017e-05,
-      "loss": 2.0138,
+      "epoch": 0.49,
+      "grad_norm": 31.32345199584961,
+      "learning_rate": 1.670334267665147e-05,
+      "loss": 2.216,
       "step": 3941
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 32.38698196411133,
-      "learning_rate": 1.2099829608098627e-05,
-      "loss": 3.5612,
+      "epoch": 0.49,
+      "grad_norm": 13.726557731628418,
+      "learning_rate": 1.670250596159478e-05,
+      "loss": 2.2354,
       "step": 3942
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 40.410884857177734,
-      "learning_rate": 1.2097824997494239e-05,
-      "loss": 2.6222,
+      "epoch": 0.49,
+      "grad_norm": 12.190067291259766,
+      "learning_rate": 1.6701669246538092e-05,
+      "loss": 1.2092,
       "step": 3943
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 11.373454093933105,
-      "learning_rate": 1.2095820386889847e-05,
-      "loss": 1.8325,
+      "epoch": 0.49,
+      "grad_norm": 19.066221237182617,
+      "learning_rate": 1.6700832531481406e-05,
+      "loss": 2.4095,
       "step": 3944
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 57.750003814697266,
-      "learning_rate": 1.2093815776285458e-05,
-      "loss": 2.7973,
+      "epoch": 0.5,
+      "grad_norm": 48.04729080200195,
+      "learning_rate": 1.669999581642472e-05,
+      "loss": 1.9395,
       "step": 3945
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 18.441944122314453,
-      "learning_rate": 1.2091811165681066e-05,
-      "loss": 2.9912,
+      "epoch": 0.5,
+      "grad_norm": 15.870589256286621,
+      "learning_rate": 1.669915910136803e-05,
+      "loss": 2.7708,
       "step": 3946
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 32.177513122558594,
-      "learning_rate": 1.2089806555076678e-05,
-      "loss": 1.9056,
+      "epoch": 0.5,
+      "grad_norm": 8.784354209899902,
+      "learning_rate": 1.6698322386311343e-05,
+      "loss": 2.7359,
       "step": 3947
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 22.477294921875,
-      "learning_rate": 1.2087801944472288e-05,
-      "loss": 2.2548,
+      "epoch": 0.5,
+      "grad_norm": 19.37180519104004,
+      "learning_rate": 1.6697485671254657e-05,
+      "loss": 1.7255,
       "step": 3948
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 20.485315322875977,
-      "learning_rate": 1.2085797333867896e-05,
-      "loss": 1.4721,
+      "epoch": 0.5,
+      "grad_norm": 16.548009872436523,
+      "learning_rate": 1.6696648956197967e-05,
+      "loss": 1.3249,
       "step": 3949
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 14.799079895019531,
-      "learning_rate": 1.2083792723263508e-05,
-      "loss": 1.8698,
+      "epoch": 0.5,
+      "grad_norm": 8.36740493774414,
+      "learning_rate": 1.669581224114128e-05,
+      "loss": 0.97,
       "step": 3950
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 13.152766227722168,
-      "learning_rate": 1.2081788112659116e-05,
-      "loss": 2.6729,
+      "epoch": 0.5,
+      "grad_norm": 14.328258514404297,
+      "learning_rate": 1.6694975526084595e-05,
+      "loss": 1.6861,
       "step": 3951
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 10.631035804748535,
-      "learning_rate": 1.2079783502054726e-05,
-      "loss": 1.7254,
+      "epoch": 0.5,
+      "grad_norm": 10.262837409973145,
+      "learning_rate": 1.6694138811027908e-05,
+      "loss": 1.8354,
       "step": 3952
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 15.786643981933594,
-      "learning_rate": 1.2077778891450336e-05,
-      "loss": 1.1798,
+      "epoch": 0.5,
+      "grad_norm": 10.428544044494629,
+      "learning_rate": 1.669330209597122e-05,
+      "loss": 1.3457,
       "step": 3953
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 12.110862731933594,
-      "learning_rate": 1.2075774280845947e-05,
-      "loss": 1.4401,
+      "epoch": 0.5,
+      "grad_norm": 13.362865447998047,
+      "learning_rate": 1.6692465380914532e-05,
+      "loss": 2.1365,
       "step": 3954
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 15.332069396972656,
-      "learning_rate": 1.2073769670241557e-05,
-      "loss": 2.4265,
+      "epoch": 0.5,
+      "grad_norm": 11.466238975524902,
+      "learning_rate": 1.6691628665857842e-05,
+      "loss": 1.1351,
       "step": 3955
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 26.413911819458008,
-      "learning_rate": 1.2071765059637167e-05,
-      "loss": 1.7556,
+      "epoch": 0.5,
+      "grad_norm": 17.091264724731445,
+      "learning_rate": 1.6690791950801156e-05,
+      "loss": 2.4696,
       "step": 3956
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 10.954570770263672,
-      "learning_rate": 1.2069760449032777e-05,
-      "loss": 0.8213,
+      "epoch": 0.5,
+      "grad_norm": 18.040620803833008,
+      "learning_rate": 1.668995523574447e-05,
+      "loss": 2.4856,
       "step": 3957
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 20.662534713745117,
-      "learning_rate": 1.2067755838428385e-05,
-      "loss": 1.6746,
+      "epoch": 0.5,
+      "grad_norm": 10.580212593078613,
+      "learning_rate": 1.668911852068778e-05,
+      "loss": 1.9885,
       "step": 3958
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 9.891891479492188,
-      "learning_rate": 1.2065751227823997e-05,
-      "loss": 1.3557,
+      "epoch": 0.5,
+      "grad_norm": 7.360715389251709,
+      "learning_rate": 1.6688281805631094e-05,
+      "loss": 0.5205,
       "step": 3959
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 13.939960479736328,
-      "learning_rate": 1.2063746617219607e-05,
-      "loss": 1.6665,
-      "step": 3960
-    },
-    {
-      "epoch": 1.19,
-      "eval_loss": 0.22422289848327637,
-      "eval_runtime": 43.495,
-      "eval_samples_per_second": 34.004,
-      "eval_steps_per_second": 34.004,
+      "epoch": 0.5,
+      "grad_norm": 17.456523895263672,
+      "learning_rate": 1.6687445090574407e-05,
+      "loss": 1.7968,
       "step": 3960
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 50.82015609741211,
-      "learning_rate": 1.2061742006615215e-05,
-      "loss": 2.8102,
+      "epoch": 0.5,
+      "grad_norm": 10.726287841796875,
+      "learning_rate": 1.6686608375517718e-05,
+      "loss": 1.5141,
       "step": 3961
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 10.213366508483887,
-      "learning_rate": 1.2059737396010827e-05,
-      "loss": 1.7045,
+      "epoch": 0.5,
+      "grad_norm": 8.44047737121582,
+      "learning_rate": 1.668577166046103e-05,
+      "loss": 1.1776,
       "step": 3962
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 9.090950965881348,
-      "learning_rate": 1.2057732785406436e-05,
-      "loss": 1.0682,
+      "epoch": 0.5,
+      "grad_norm": 16.860645294189453,
+      "learning_rate": 1.668493494540434e-05,
+      "loss": 5.1519,
       "step": 3963
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 14.20166015625,
-      "learning_rate": 1.2055728174802046e-05,
-      "loss": 1.6642,
+      "epoch": 0.5,
+      "grad_norm": 11.764198303222656,
+      "learning_rate": 1.6684098230347655e-05,
+      "loss": 1.5988,
       "step": 3964
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 14.564234733581543,
-      "learning_rate": 1.2053723564197654e-05,
-      "loss": 1.6248,
+      "epoch": 0.5,
+      "grad_norm": 5.626613616943359,
+      "learning_rate": 1.668326151529097e-05,
+      "loss": 1.0498,
       "step": 3965
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 16.231090545654297,
-      "learning_rate": 1.2051718953593266e-05,
-      "loss": 1.2251,
+      "epoch": 0.5,
+      "grad_norm": 53.96699523925781,
+      "learning_rate": 1.6682424800234282e-05,
+      "loss": 1.1779,
       "step": 3966
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 15.789569854736328,
-      "learning_rate": 1.2049714342988876e-05,
-      "loss": 2.4118,
+      "epoch": 0.5,
+      "grad_norm": 4.14115571975708,
+      "learning_rate": 1.6681588085177593e-05,
+      "loss": 0.1223,
       "step": 3967
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 18.06235122680664,
-      "learning_rate": 1.2047709732384484e-05,
-      "loss": 1.4758,
+      "epoch": 0.5,
+      "grad_norm": 35.95234680175781,
+      "learning_rate": 1.6680751370120906e-05,
+      "loss": 1.6813,
       "step": 3968
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 18.72103500366211,
-      "learning_rate": 1.2045705121780096e-05,
-      "loss": 1.416,
+      "epoch": 0.5,
+      "grad_norm": 11.989574432373047,
+      "learning_rate": 1.667991465506422e-05,
+      "loss": 1.5719,
       "step": 3969
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 67.29039001464844,
-      "learning_rate": 1.2043700511175705e-05,
-      "loss": 3.1432,
+      "epoch": 0.5,
+      "grad_norm": 9.279258728027344,
+      "learning_rate": 1.667907794000753e-05,
+      "loss": 1.3138,
       "step": 3970
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 9.234892845153809,
-      "learning_rate": 1.2041695900571315e-05,
-      "loss": 1.0072,
+      "epoch": 0.5,
+      "grad_norm": 7.835389614105225,
+      "learning_rate": 1.6678241224950844e-05,
+      "loss": 1.4085,
       "step": 3971
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 16.655811309814453,
-      "learning_rate": 1.2039691289966923e-05,
-      "loss": 1.4894,
+      "epoch": 0.5,
+      "grad_norm": 10.986465454101562,
+      "learning_rate": 1.6677404509894158e-05,
+      "loss": 1.6206,
       "step": 3972
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 24.546369552612305,
-      "learning_rate": 1.2037686679362535e-05,
-      "loss": 1.6944,
+      "epoch": 0.5,
+      "grad_norm": 21.686466217041016,
+      "learning_rate": 1.667656779483747e-05,
+      "loss": 1.826,
       "step": 3973
     },
     {
-      "epoch": 1.19,
-      "grad_norm": 22.591285705566406,
-      "learning_rate": 1.2035682068758145e-05,
-      "loss": 1.4835,
+      "epoch": 0.5,
+      "grad_norm": 10.920929908752441,
+      "learning_rate": 1.667573107978078e-05,
+      "loss": 1.2232,
       "step": 3974
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 19.460481643676758,
-      "learning_rate": 1.2033677458153753e-05,
-      "loss": 2.3772,
+      "epoch": 0.5,
+      "grad_norm": 14.287152290344238,
+      "learning_rate": 1.6674894364724095e-05,
+      "loss": 0.3653,
       "step": 3975
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 10.695039749145508,
-      "learning_rate": 1.2031672847549365e-05,
-      "loss": 1.3966,
+      "epoch": 0.5,
+      "grad_norm": 40.73248291015625,
+      "learning_rate": 1.667405764966741e-05,
+      "loss": 3.1087,
       "step": 3976
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 16.96231460571289,
-      "learning_rate": 1.2029668236944973e-05,
-      "loss": 1.5844,
+      "epoch": 0.5,
+      "grad_norm": 10.77466106414795,
+      "learning_rate": 1.667322093461072e-05,
+      "loss": 1.8408,
       "step": 3977
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 13.368218421936035,
-      "learning_rate": 1.2027663626340585e-05,
-      "loss": 2.2137,
+      "epoch": 0.5,
+      "grad_norm": 16.43342399597168,
+      "learning_rate": 1.6672384219554033e-05,
+      "loss": 1.9936,
       "step": 3978
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 16.881895065307617,
-      "learning_rate": 1.2025659015736195e-05,
-      "loss": 1.809,
+      "epoch": 0.5,
+      "grad_norm": 12.749366760253906,
+      "learning_rate": 1.6671547504497346e-05,
+      "loss": 1.4893,
       "step": 3979
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 16.86216926574707,
-      "learning_rate": 1.2023654405131804e-05,
-      "loss": 1.7264,
+      "epoch": 0.5,
+      "grad_norm": 22.858173370361328,
+      "learning_rate": 1.667071078944066e-05,
+      "loss": 2.2102,
       "step": 3980
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 19.68351173400879,
-      "learning_rate": 1.2021649794527415e-05,
-      "loss": 1.4559,
+      "epoch": 0.5,
+      "grad_norm": 10.709026336669922,
+      "learning_rate": 1.666987407438397e-05,
+      "loss": 0.931,
       "step": 3981
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 27.27216339111328,
-      "learning_rate": 1.2019645183923024e-05,
-      "loss": 1.6525,
+      "epoch": 0.5,
+      "grad_norm": 12.102494239807129,
+      "learning_rate": 1.6669037359327284e-05,
+      "loss": 1.8199,
       "step": 3982
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 48.12042999267578,
-      "learning_rate": 1.2017640573318634e-05,
-      "loss": 2.7933,
+      "epoch": 0.5,
+      "grad_norm": 29.741167068481445,
+      "learning_rate": 1.6668200644270594e-05,
+      "loss": 1.8224,
       "step": 3983
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 17.5889892578125,
-      "learning_rate": 1.2015635962714242e-05,
-      "loss": 1.2232,
+      "epoch": 0.5,
+      "grad_norm": 10.0697021484375,
+      "learning_rate": 1.6667363929213908e-05,
+      "loss": 2.046,
       "step": 3984
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 15.102045059204102,
-      "learning_rate": 1.2013631352109854e-05,
-      "loss": 2.1175,
+      "epoch": 0.5,
+      "grad_norm": 29.015119552612305,
+      "learning_rate": 1.666652721415722e-05,
+      "loss": 1.8735,
       "step": 3985
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 15.093314170837402,
-      "learning_rate": 1.2011626741505464e-05,
-      "loss": 2.0942,
+      "epoch": 0.5,
+      "grad_norm": 7.788076400756836,
+      "learning_rate": 1.6665690499100532e-05,
+      "loss": 2.3206,
       "step": 3986
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 14.149928092956543,
-      "learning_rate": 1.2009622130901073e-05,
-      "loss": 1.4488,
+      "epoch": 0.5,
+      "grad_norm": 39.918521881103516,
+      "learning_rate": 1.6664853784043845e-05,
+      "loss": 1.9093,
       "step": 3987
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 23.17877197265625,
-      "learning_rate": 1.2007617520296684e-05,
-      "loss": 2.1352,
+      "epoch": 0.5,
+      "grad_norm": 14.87916374206543,
+      "learning_rate": 1.6664017068987156e-05,
+      "loss": 1.5678,
       "step": 3988
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 11.108948707580566,
-      "learning_rate": 1.2005612909692293e-05,
-      "loss": 1.4723,
+      "epoch": 0.5,
+      "grad_norm": 11.87180233001709,
+      "learning_rate": 1.666318035393047e-05,
+      "loss": 1.7662,
       "step": 3989
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 41.934715270996094,
-      "learning_rate": 1.2003608299087903e-05,
-      "loss": 1.3572,
+      "epoch": 0.5,
+      "grad_norm": 26.39105224609375,
+      "learning_rate": 1.6662343638873783e-05,
+      "loss": 1.7568,
       "step": 3990
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 30.752838134765625,
-      "learning_rate": 1.2001603688483515e-05,
-      "loss": 1.7268,
+      "epoch": 0.5,
+      "grad_norm": 8.435544967651367,
+      "learning_rate": 1.6661506923817093e-05,
+      "loss": 3.2938,
       "step": 3991
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 19.63275718688965,
-      "learning_rate": 1.1999599077879123e-05,
-      "loss": 1.8781,
+      "epoch": 0.5,
+      "grad_norm": 13.673137664794922,
+      "learning_rate": 1.6660670208760407e-05,
+      "loss": 1.3841,
       "step": 3992
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 16.867294311523438,
-      "learning_rate": 1.1997594467274733e-05,
-      "loss": 1.2249,
+      "epoch": 0.5,
+      "grad_norm": 10.718171119689941,
+      "learning_rate": 1.665983349370372e-05,
+      "loss": 0.8371,
       "step": 3993
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 12.371650695800781,
-      "learning_rate": 1.1995589856670341e-05,
-      "loss": 2.7085,
+      "epoch": 0.5,
+      "grad_norm": 18.63213539123535,
+      "learning_rate": 1.6658996778647034e-05,
+      "loss": 1.8301,
       "step": 3994
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 14.14538860321045,
-      "learning_rate": 1.1993585246065953e-05,
-      "loss": 1.7304,
+      "epoch": 0.5,
+      "grad_norm": 11.364863395690918,
+      "learning_rate": 1.6658160063590344e-05,
+      "loss": 2.5042,
       "step": 3995
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 28.63239860534668,
-      "learning_rate": 1.1991580635461562e-05,
-      "loss": 1.6735,
+      "epoch": 0.5,
+      "grad_norm": 77.68419647216797,
+      "learning_rate": 1.6657323348533658e-05,
+      "loss": 0.9009,
       "step": 3996
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 10.542112350463867,
-      "learning_rate": 1.1989576024857172e-05,
-      "loss": 2.628,
+      "epoch": 0.5,
+      "grad_norm": 17.809261322021484,
+      "learning_rate": 1.665648663347697e-05,
+      "loss": 1.5477,
       "step": 3997
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 15.562253952026367,
-      "learning_rate": 1.1987571414252784e-05,
-      "loss": 1.8914,
+      "epoch": 0.5,
+      "grad_norm": 34.99085998535156,
+      "learning_rate": 1.6655649918420282e-05,
+      "loss": 3.4252,
       "step": 3998
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 14.955403327941895,
-      "learning_rate": 1.1985566803648392e-05,
-      "loss": 1.4507,
+      "epoch": 0.5,
+      "grad_norm": 18.85693359375,
+      "learning_rate": 1.6654813203363596e-05,
+      "loss": 2.2447,
       "step": 3999
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 8.927510261535645,
-      "learning_rate": 1.1983562193044002e-05,
-      "loss": 1.8247,
+      "epoch": 0.5,
+      "grad_norm": 10.359712600708008,
+      "learning_rate": 1.665397648830691e-05,
+      "loss": 3.4757,
       "step": 4000
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 40.698150634765625,
-      "learning_rate": 1.1981557582439612e-05,
-      "loss": 2.7692,
+      "epoch": 0.5,
+      "eval_loss": 0.14231130480766296,
+      "eval_runtime": 94.7497,
+      "eval_samples_per_second": 37.383,
+      "eval_steps_per_second": 37.383,
+      "step": 4000
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.207204818725586,
+      "learning_rate": 1.6653139773250223e-05,
+      "loss": 1.625,
       "step": 4001
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 8.478541374206543,
-      "learning_rate": 1.1979552971835222e-05,
-      "loss": 1.0191,
+      "epoch": 0.5,
+      "grad_norm": 12.07330322265625,
+      "learning_rate": 1.6652303058193533e-05,
+      "loss": 2.6738,
       "step": 4002
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 29.216922760009766,
-      "learning_rate": 1.1977548361230832e-05,
-      "loss": 1.7092,
+      "epoch": 0.5,
+      "grad_norm": 16.51776123046875,
+      "learning_rate": 1.6651466343136847e-05,
+      "loss": 1.036,
       "step": 4003
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 11.635241508483887,
-      "learning_rate": 1.1975543750626442e-05,
-      "loss": 1.7223,
+      "epoch": 0.5,
+      "grad_norm": 19.95757293701172,
+      "learning_rate": 1.665062962808016e-05,
+      "loss": 2.9243,
       "step": 4004
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 23.52641487121582,
-      "learning_rate": 1.1973539140022052e-05,
-      "loss": 1.6989,
+      "epoch": 0.5,
+      "grad_norm": 71.45176696777344,
+      "learning_rate": 1.664979291302347e-05,
+      "loss": 1.9166,
       "step": 4005
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 18.005016326904297,
-      "learning_rate": 1.197153452941766e-05,
-      "loss": 2.5197,
+      "epoch": 0.5,
+      "grad_norm": 44.509273529052734,
+      "learning_rate": 1.6648956197966784e-05,
+      "loss": 2.8575,
       "step": 4006
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 59.00607681274414,
-      "learning_rate": 1.1969529918813273e-05,
-      "loss": 3.0074,
+      "epoch": 0.5,
+      "grad_norm": 17.11343765258789,
+      "learning_rate": 1.6648119482910098e-05,
+      "loss": 3.3717,
       "step": 4007
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 23.461963653564453,
-      "learning_rate": 1.1967525308208881e-05,
-      "loss": 2.5943,
+      "epoch": 0.5,
+      "grad_norm": 19.97622299194336,
+      "learning_rate": 1.6647282767853408e-05,
+      "loss": 1.7877,
       "step": 4008
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 30.815887451171875,
-      "learning_rate": 1.1965520697604491e-05,
-      "loss": 1.9424,
+      "epoch": 0.5,
+      "grad_norm": 15.097450256347656,
+      "learning_rate": 1.6646446052796722e-05,
+      "loss": 2.227,
       "step": 4009
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 29.489526748657227,
-      "learning_rate": 1.1963516087000103e-05,
-      "loss": 2.0126,
+      "epoch": 0.5,
+      "grad_norm": 23.042789459228516,
+      "learning_rate": 1.6645609337740036e-05,
+      "loss": 2.6331,
       "step": 4010
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 7.134322166442871,
-      "learning_rate": 1.1961511476395711e-05,
-      "loss": 1.2801,
+      "epoch": 0.5,
+      "grad_norm": 15.941010475158691,
+      "learning_rate": 1.6644772622683346e-05,
+      "loss": 1.7747,
       "step": 4011
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 12.04995059967041,
-      "learning_rate": 1.1959506865791321e-05,
-      "loss": 1.0723,
+      "epoch": 0.5,
+      "grad_norm": 33.62506103515625,
+      "learning_rate": 1.664393590762666e-05,
+      "loss": 1.3494,
       "step": 4012
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 13.849966049194336,
-      "learning_rate": 1.195750225518693e-05,
-      "loss": 1.0122,
+      "epoch": 0.5,
+      "grad_norm": 19.71503257751465,
+      "learning_rate": 1.6643099192569973e-05,
+      "loss": 1.6179,
       "step": 4013
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 7.143357753753662,
-      "learning_rate": 1.1955497644582541e-05,
-      "loss": 1.961,
+      "epoch": 0.5,
+      "grad_norm": 16.13291358947754,
+      "learning_rate": 1.6642262477513283e-05,
+      "loss": 2.6505,
       "step": 4014
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 16.170339584350586,
-      "learning_rate": 1.195349303397815e-05,
-      "loss": 1.9324,
+      "epoch": 0.5,
+      "grad_norm": 16.614009857177734,
+      "learning_rate": 1.6641425762456597e-05,
+      "loss": 2.215,
       "step": 4015
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 22.960071563720703,
-      "learning_rate": 1.195148842337376e-05,
-      "loss": 2.5535,
+      "epoch": 0.5,
+      "grad_norm": 20.98224449157715,
+      "learning_rate": 1.6640589047399907e-05,
+      "loss": 1.7982,
       "step": 4016
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 15.098960876464844,
-      "learning_rate": 1.1949483812769372e-05,
-      "loss": 1.142,
+      "epoch": 0.5,
+      "grad_norm": 18.615270614624023,
+      "learning_rate": 1.663975233234322e-05,
+      "loss": 4.0383,
       "step": 4017
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 19.63709831237793,
-      "learning_rate": 1.194747920216498e-05,
-      "loss": 2.072,
+      "epoch": 0.5,
+      "grad_norm": 21.422555923461914,
+      "learning_rate": 1.6638915617286535e-05,
+      "loss": 0.7228,
       "step": 4018
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 14.952527046203613,
-      "learning_rate": 1.194547459156059e-05,
-      "loss": 1.2427,
+      "epoch": 0.5,
+      "grad_norm": 18.596290588378906,
+      "learning_rate": 1.6638078902229845e-05,
+      "loss": 1.7599,
       "step": 4019
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 13.70450210571289,
-      "learning_rate": 1.1943469980956199e-05,
-      "loss": 1.8315,
+      "epoch": 0.5,
+      "grad_norm": 12.224632263183594,
+      "learning_rate": 1.663724218717316e-05,
+      "loss": 1.4104,
       "step": 4020
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 13.074018478393555,
-      "learning_rate": 1.194146537035181e-05,
-      "loss": 0.9992,
+      "epoch": 0.5,
+      "grad_norm": 13.065860748291016,
+      "learning_rate": 1.6636405472116472e-05,
+      "loss": 2.3884,
       "step": 4021
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 12.277508735656738,
-      "learning_rate": 1.193946075974742e-05,
-      "loss": 1.8067,
+      "epoch": 0.5,
+      "grad_norm": 22.903993606567383,
+      "learning_rate": 1.6635568757059786e-05,
+      "loss": 3.8494,
       "step": 4022
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 27.505800247192383,
-      "learning_rate": 1.1937456149143029e-05,
-      "loss": 2.1141,
+      "epoch": 0.5,
+      "grad_norm": 12.7423734664917,
+      "learning_rate": 1.6634732042003096e-05,
+      "loss": 1.5295,
       "step": 4023
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 18.097551345825195,
-      "learning_rate": 1.193545153853864e-05,
-      "loss": 1.631,
+      "epoch": 0.51,
+      "grad_norm": 16.80319595336914,
+      "learning_rate": 1.663389532694641e-05,
+      "loss": 1.9948,
       "step": 4024
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 15.072179794311523,
-      "learning_rate": 1.1933446927934249e-05,
-      "loss": 1.5262,
+      "epoch": 0.51,
+      "grad_norm": 8.247112274169922,
+      "learning_rate": 1.6633058611889723e-05,
+      "loss": 1.9971,
       "step": 4025
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 13.110054016113281,
-      "learning_rate": 1.1931442317329859e-05,
-      "loss": 1.4835,
+      "epoch": 0.51,
+      "grad_norm": 16.039772033691406,
+      "learning_rate": 1.6632221896833034e-05,
+      "loss": 1.5694,
       "step": 4026
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 11.2367525100708,
-      "learning_rate": 1.192943770672547e-05,
-      "loss": 1.5453,
+      "epoch": 0.51,
+      "grad_norm": 13.967763900756836,
+      "learning_rate": 1.6631385181776347e-05,
+      "loss": 1.4187,
       "step": 4027
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 19.51766014099121,
-      "learning_rate": 1.192743309612108e-05,
-      "loss": 1.4438,
+      "epoch": 0.51,
+      "grad_norm": 17.23777198791504,
+      "learning_rate": 1.663054846671966e-05,
+      "loss": 1.1976,
       "step": 4028
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 58.76639938354492,
-      "learning_rate": 1.1925428485516691e-05,
-      "loss": 2.2727,
+      "epoch": 0.51,
+      "grad_norm": 25.99625587463379,
+      "learning_rate": 1.6629711751662975e-05,
+      "loss": 4.8848,
       "step": 4029
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 16.14882469177246,
-      "learning_rate": 1.19234238749123e-05,
-      "loss": 1.4155,
+      "epoch": 0.51,
+      "grad_norm": 17.495437622070312,
+      "learning_rate": 1.6628875036606285e-05,
+      "loss": 3.902,
       "step": 4030
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 9.466064453125,
-      "learning_rate": 1.192141926430791e-05,
-      "loss": 1.3346,
+      "epoch": 0.51,
+      "grad_norm": 26.159387588500977,
+      "learning_rate": 1.66280383215496e-05,
+      "loss": 3.2425,
       "step": 4031
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 16.83454704284668,
-      "learning_rate": 1.1919414653703518e-05,
-      "loss": 1.3689,
+      "epoch": 0.51,
+      "grad_norm": 32.6840934753418,
+      "learning_rate": 1.6627201606492912e-05,
+      "loss": 1.4698,
       "step": 4032
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 11.721967697143555,
-      "learning_rate": 1.191741004309913e-05,
-      "loss": 1.5468,
+      "epoch": 0.51,
+      "grad_norm": 12.527408599853516,
+      "learning_rate": 1.6626364891436222e-05,
+      "loss": 1.9287,
       "step": 4033
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 9.986775398254395,
-      "learning_rate": 1.191540543249474e-05,
-      "loss": 2.5393,
+      "epoch": 0.51,
+      "grad_norm": 25.900415420532227,
+      "learning_rate": 1.6625528176379536e-05,
+      "loss": 2.039,
       "step": 4034
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 11.999430656433105,
-      "learning_rate": 1.1913400821890348e-05,
-      "loss": 1.7203,
+      "epoch": 0.51,
+      "grad_norm": 10.852729797363281,
+      "learning_rate": 1.662469146132285e-05,
+      "loss": 1.3046,
       "step": 4035
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 15.086469650268555,
-      "learning_rate": 1.191139621128596e-05,
-      "loss": 1.8978,
+      "epoch": 0.51,
+      "grad_norm": 14.895544052124023,
+      "learning_rate": 1.662385474626616e-05,
+      "loss": 1.8385,
       "step": 4036
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 20.604961395263672,
-      "learning_rate": 1.1909391600681568e-05,
-      "loss": 1.0557,
+      "epoch": 0.51,
+      "grad_norm": 16.182043075561523,
+      "learning_rate": 1.6623018031209474e-05,
+      "loss": 2.2905,
       "step": 4037
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 11.977767944335938,
-      "learning_rate": 1.1907386990077178e-05,
-      "loss": 1.9318,
+      "epoch": 0.51,
+      "grad_norm": 11.084127426147461,
+      "learning_rate": 1.6622181316152787e-05,
+      "loss": 4.2573,
       "step": 4038
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 17.97518539428711,
-      "learning_rate": 1.1905382379472787e-05,
-      "loss": 0.948,
+      "epoch": 0.51,
+      "grad_norm": 18.71702766418457,
+      "learning_rate": 1.6621344601096097e-05,
+      "loss": 2.1855,
       "step": 4039
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 9.713574409484863,
-      "learning_rate": 1.1903377768868399e-05,
-      "loss": 1.0446,
+      "epoch": 0.51,
+      "grad_norm": 10.081755638122559,
+      "learning_rate": 1.662050788603941e-05,
+      "loss": 1.9548,
       "step": 4040
     },
     {
-      "epoch": 1.21,
-      "grad_norm": 18.080841064453125,
-      "learning_rate": 1.1901373158264009e-05,
-      "loss": 1.703,
+      "epoch": 0.51,
+      "grad_norm": 17.882465362548828,
+      "learning_rate": 1.661967117098272e-05,
+      "loss": 3.3668,
       "step": 4041
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 35.3074951171875,
-      "learning_rate": 1.1899368547659617e-05,
-      "loss": 2.8692,
+      "epoch": 0.51,
+      "grad_norm": 28.202253341674805,
+      "learning_rate": 1.6618834455926035e-05,
+      "loss": 1.74,
       "step": 4042
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 12.872424125671387,
-      "learning_rate": 1.1897363937055229e-05,
-      "loss": 1.2558,
+      "epoch": 0.51,
+      "grad_norm": 13.790504455566406,
+      "learning_rate": 1.661799774086935e-05,
+      "loss": 2.8872,
       "step": 4043
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 20.24639129638672,
-      "learning_rate": 1.1895359326450837e-05,
-      "loss": 2.013,
+      "epoch": 0.51,
+      "grad_norm": 19.462251663208008,
+      "learning_rate": 1.661716102581266e-05,
+      "loss": 0.7643,
       "step": 4044
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 19.4423885345459,
-      "learning_rate": 1.1893354715846447e-05,
-      "loss": 1.2018,
+      "epoch": 0.51,
+      "grad_norm": 13.855001449584961,
+      "learning_rate": 1.6616324310755973e-05,
+      "loss": 1.5681,
       "step": 4045
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 28.46637725830078,
-      "learning_rate": 1.1891350105242059e-05,
-      "loss": 2.3941,
+      "epoch": 0.51,
+      "grad_norm": 8.267029762268066,
+      "learning_rate": 1.6615487595699286e-05,
+      "loss": 0.942,
       "step": 4046
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 15.011366844177246,
-      "learning_rate": 1.1889345494637667e-05,
-      "loss": 1.5718,
+      "epoch": 0.51,
+      "grad_norm": 14.253353118896484,
+      "learning_rate": 1.6614650880642597e-05,
+      "loss": 3.2943,
       "step": 4047
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 30.27291488647461,
-      "learning_rate": 1.1887340884033278e-05,
-      "loss": 2.3006,
+      "epoch": 0.51,
+      "grad_norm": 4.903005599975586,
+      "learning_rate": 1.661381416558591e-05,
+      "loss": 1.309,
       "step": 4048
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 18.10991859436035,
-      "learning_rate": 1.1885336273428888e-05,
-      "loss": 1.7672,
+      "epoch": 0.51,
+      "grad_norm": 6.407052040100098,
+      "learning_rate": 1.6612977450529224e-05,
+      "loss": 0.5181,
       "step": 4049
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 34.54472732543945,
-      "learning_rate": 1.1883331662824498e-05,
-      "loss": 2.484,
+      "epoch": 0.51,
+      "grad_norm": 17.52642250061035,
+      "learning_rate": 1.6612140735472537e-05,
+      "loss": 1.4306,
       "step": 4050
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 12.643610954284668,
-      "learning_rate": 1.1881327052220106e-05,
-      "loss": 2.0557,
+      "epoch": 0.51,
+      "grad_norm": 14.540874481201172,
+      "learning_rate": 1.6611304020415848e-05,
+      "loss": 1.8284,
       "step": 4051
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 12.419498443603516,
-      "learning_rate": 1.1879322441615718e-05,
-      "loss": 1.8953,
+      "epoch": 0.51,
+      "grad_norm": 8.986979484558105,
+      "learning_rate": 1.661046730535916e-05,
+      "loss": 0.7586,
       "step": 4052
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 9.909022331237793,
-      "learning_rate": 1.1877317831011328e-05,
-      "loss": 0.9697,
+      "epoch": 0.51,
+      "grad_norm": 8.945680618286133,
+      "learning_rate": 1.6609630590302475e-05,
+      "loss": 1.2829,
       "step": 4053
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 213.60519409179688,
-      "learning_rate": 1.1875313220406936e-05,
-      "loss": 2.7859,
+      "epoch": 0.51,
+      "grad_norm": 19.894546508789062,
+      "learning_rate": 1.6608793875245785e-05,
+      "loss": 2.1166,
       "step": 4054
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 43.98239517211914,
-      "learning_rate": 1.1873308609802548e-05,
-      "loss": 2.0812,
+      "epoch": 0.51,
+      "grad_norm": 13.819477081298828,
+      "learning_rate": 1.66079571601891e-05,
+      "loss": 1.7849,
       "step": 4055
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 22.185237884521484,
-      "learning_rate": 1.1871303999198157e-05,
-      "loss": 1.778,
+      "epoch": 0.51,
+      "grad_norm": 13.106219291687012,
+      "learning_rate": 1.6607120445132413e-05,
+      "loss": 1.3608,
       "step": 4056
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 20.283220291137695,
-      "learning_rate": 1.1869299388593767e-05,
-      "loss": 1.6555,
+      "epoch": 0.51,
+      "grad_norm": 18.80193328857422,
+      "learning_rate": 1.6606283730075726e-05,
+      "loss": 2.3694,
       "step": 4057
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 52.456600189208984,
-      "learning_rate": 1.1867294777989375e-05,
-      "loss": 2.7659,
+      "epoch": 0.51,
+      "grad_norm": 11.726581573486328,
+      "learning_rate": 1.6605447015019036e-05,
+      "loss": 2.1741,
       "step": 4058
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 27.247201919555664,
-      "learning_rate": 1.1865290167384987e-05,
-      "loss": 1.9769,
+      "epoch": 0.51,
+      "grad_norm": 8.649958610534668,
+      "learning_rate": 1.660461029996235e-05,
+      "loss": 0.8855,
       "step": 4059
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 19.406150817871094,
-      "learning_rate": 1.1863285556780597e-05,
-      "loss": 2.0779,
+      "epoch": 0.51,
+      "grad_norm": 13.534453392028809,
+      "learning_rate": 1.6603773584905664e-05,
+      "loss": 2.5613,
       "step": 4060
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 7.287651062011719,
-      "learning_rate": 1.1861280946176205e-05,
-      "loss": 1.3887,
+      "epoch": 0.51,
+      "grad_norm": 21.837444305419922,
+      "learning_rate": 1.6602936869848974e-05,
+      "loss": 3.0084,
       "step": 4061
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 12.292468070983887,
-      "learning_rate": 1.1859276335571817e-05,
-      "loss": 1.7488,
+      "epoch": 0.51,
+      "grad_norm": 9.196706771850586,
+      "learning_rate": 1.6602100154792288e-05,
+      "loss": 2.6228,
       "step": 4062
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 9.17116928100586,
-      "learning_rate": 1.1857271724967425e-05,
-      "loss": 1.2331,
+      "epoch": 0.51,
+      "grad_norm": 13.78110408782959,
+      "learning_rate": 1.66012634397356e-05,
+      "loss": 1.6917,
       "step": 4063
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 12.708617210388184,
-      "learning_rate": 1.1855267114363036e-05,
-      "loss": 1.5956,
+      "epoch": 0.51,
+      "grad_norm": 6.981208801269531,
+      "learning_rate": 1.660042672467891e-05,
+      "loss": 0.9407,
       "step": 4064
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 19.217905044555664,
-      "learning_rate": 1.1853262503758647e-05,
-      "loss": 3.3333,
+      "epoch": 0.51,
+      "grad_norm": 14.598764419555664,
+      "learning_rate": 1.6599590009622225e-05,
+      "loss": 0.7875,
       "step": 4065
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 9.486623764038086,
-      "learning_rate": 1.1851257893154256e-05,
-      "loss": 1.1654,
+      "epoch": 0.51,
+      "grad_norm": 22.41578483581543,
+      "learning_rate": 1.6598753294565536e-05,
+      "loss": 2.7584,
       "step": 4066
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 15.519346237182617,
-      "learning_rate": 1.1849253282549866e-05,
-      "loss": 1.3905,
+      "epoch": 0.51,
+      "grad_norm": 15.423556327819824,
+      "learning_rate": 1.659791657950885e-05,
+      "loss": 1.6977,
       "step": 4067
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 27.3670597076416,
-      "learning_rate": 1.1847248671945474e-05,
-      "loss": 1.8644,
+      "epoch": 0.51,
+      "grad_norm": 26.116514205932617,
+      "learning_rate": 1.6597079864452163e-05,
+      "loss": 3.1091,
       "step": 4068
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 12.652219772338867,
-      "learning_rate": 1.1845244061341086e-05,
-      "loss": 1.7432,
+      "epoch": 0.51,
+      "grad_norm": 12.65701675415039,
+      "learning_rate": 1.6596243149395473e-05,
+      "loss": 0.8299,
       "step": 4069
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 7.705514430999756,
-      "learning_rate": 1.1843239450736694e-05,
-      "loss": 0.8268,
+      "epoch": 0.51,
+      "grad_norm": 6.301980018615723,
+      "learning_rate": 1.6595406434338787e-05,
+      "loss": 1.3544,
       "step": 4070
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 13.351577758789062,
-      "learning_rate": 1.1841234840132304e-05,
-      "loss": 2.1508,
+      "epoch": 0.51,
+      "grad_norm": 11.021586418151855,
+      "learning_rate": 1.65945697192821e-05,
+      "loss": 0.9748,
       "step": 4071
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 44.65596008300781,
-      "learning_rate": 1.1839230229527916e-05,
-      "loss": 2.0205,
+      "epoch": 0.51,
+      "grad_norm": 15.616560935974121,
+      "learning_rate": 1.659373300422541e-05,
+      "loss": 3.0321,
       "step": 4072
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 24.321109771728516,
-      "learning_rate": 1.1837225618923525e-05,
-      "loss": 1.7858,
+      "epoch": 0.51,
+      "grad_norm": 32.11990737915039,
+      "learning_rate": 1.6592896289168724e-05,
+      "loss": 2.4086,
       "step": 4073
     },
     {
-      "epoch": 1.22,
-      "grad_norm": 48.96630096435547,
-      "learning_rate": 1.1835221008319135e-05,
-      "loss": 1.9235,
+      "epoch": 0.51,
+      "grad_norm": 16.586008071899414,
+      "learning_rate": 1.6592059574112038e-05,
+      "loss": 3.0542,
       "step": 4074
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 16.92742156982422,
-      "learning_rate": 1.1833216397714745e-05,
-      "loss": 1.7837,
+      "epoch": 0.51,
+      "grad_norm": 11.527971267700195,
+      "learning_rate": 1.6591222859055348e-05,
+      "loss": 2.2782,
       "step": 4075
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 42.10576629638672,
-      "learning_rate": 1.1831211787110355e-05,
-      "loss": 3.2812,
+      "epoch": 0.51,
+      "grad_norm": 22.980730056762695,
+      "learning_rate": 1.6590386143998662e-05,
+      "loss": 0.79,
       "step": 4076
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 13.819901466369629,
-      "learning_rate": 1.1829207176505965e-05,
-      "loss": 1.4294,
+      "epoch": 0.51,
+      "grad_norm": 13.224310874938965,
+      "learning_rate": 1.6589549428941975e-05,
+      "loss": 1.4557,
       "step": 4077
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 11.949419021606445,
-      "learning_rate": 1.1827202565901575e-05,
-      "loss": 1.3175,
+      "epoch": 0.51,
+      "grad_norm": 21.20461082458496,
+      "learning_rate": 1.658871271388529e-05,
+      "loss": 1.6549,
       "step": 4078
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 23.273836135864258,
-      "learning_rate": 1.1825197955297185e-05,
-      "loss": 2.5785,
+      "epoch": 0.51,
+      "grad_norm": 21.756563186645508,
+      "learning_rate": 1.65878759988286e-05,
+      "loss": 2.6471,
       "step": 4079
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 15.00057601928711,
-      "learning_rate": 1.1823193344692793e-05,
-      "loss": 1.4673,
-      "step": 4080
-    },
-    {
-      "epoch": 1.23,
-      "eval_loss": 0.22429148852825165,
-      "eval_runtime": 43.6464,
-      "eval_samples_per_second": 33.886,
-      "eval_steps_per_second": 33.886,
+      "epoch": 0.51,
+      "grad_norm": 26.183300018310547,
+      "learning_rate": 1.6587039283771913e-05,
+      "loss": 0.6442,
       "step": 4080
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 16.4818172454834,
-      "learning_rate": 1.1821188734088405e-05,
-      "loss": 1.8456,
+      "epoch": 0.51,
+      "grad_norm": 10.016181945800781,
+      "learning_rate": 1.6586202568715227e-05,
+      "loss": 1.7346,
       "step": 4081
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 9.311663627624512,
-      "learning_rate": 1.1819184123484014e-05,
-      "loss": 1.5563,
+      "epoch": 0.51,
+      "grad_norm": 87.37271881103516,
+      "learning_rate": 1.6585365853658537e-05,
+      "loss": 0.7923,
       "step": 4082
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 19.122093200683594,
-      "learning_rate": 1.1817179512879624e-05,
-      "loss": 2.2633,
+      "epoch": 0.51,
+      "grad_norm": 65.72235107421875,
+      "learning_rate": 1.658452913860185e-05,
+      "loss": 1.9297,
       "step": 4083
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 33.5509033203125,
-      "learning_rate": 1.1815174902275236e-05,
-      "loss": 1.9949,
+      "epoch": 0.51,
+      "grad_norm": 9.46326732635498,
+      "learning_rate": 1.6583692423545164e-05,
+      "loss": 0.6906,
       "step": 4084
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 14.471306800842285,
-      "learning_rate": 1.1813170291670844e-05,
-      "loss": 1.8224,
+      "epoch": 0.51,
+      "grad_norm": 8.229296684265137,
+      "learning_rate": 1.6582855708488478e-05,
+      "loss": 0.719,
       "step": 4085
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 11.821321487426758,
-      "learning_rate": 1.1811165681066454e-05,
-      "loss": 1.8433,
+      "epoch": 0.51,
+      "grad_norm": 9.068437576293945,
+      "learning_rate": 1.6582018993431788e-05,
+      "loss": 1.4107,
       "step": 4086
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 26.970781326293945,
-      "learning_rate": 1.1809161070462062e-05,
-      "loss": 2.336,
+      "epoch": 0.51,
+      "grad_norm": 16.31252670288086,
+      "learning_rate": 1.6581182278375102e-05,
+      "loss": 3.5396,
       "step": 4087
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 18.373268127441406,
-      "learning_rate": 1.1807156459857674e-05,
-      "loss": 1.4686,
+      "epoch": 0.51,
+      "grad_norm": 7.02438497543335,
+      "learning_rate": 1.6580345563318415e-05,
+      "loss": 1.0429,
       "step": 4088
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 17.756778717041016,
-      "learning_rate": 1.1805151849253284e-05,
-      "loss": 1.4351,
+      "epoch": 0.51,
+      "grad_norm": 29.739078521728516,
+      "learning_rate": 1.6579508848261726e-05,
+      "loss": 2.1456,
       "step": 4089
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 21.345794677734375,
-      "learning_rate": 1.1803147238648893e-05,
-      "loss": 1.7058,
+      "epoch": 0.51,
+      "grad_norm": 54.58208084106445,
+      "learning_rate": 1.657867213320504e-05,
+      "loss": 2.1292,
       "step": 4090
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 99.40449523925781,
-      "learning_rate": 1.1801142628044504e-05,
-      "loss": 2.2699,
+      "epoch": 0.51,
+      "grad_norm": 20.13058090209961,
+      "learning_rate": 1.6577835418148353e-05,
+      "loss": 3.2348,
       "step": 4091
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 29.920364379882812,
-      "learning_rate": 1.1799138017440113e-05,
-      "loss": 1.7795,
+      "epoch": 0.51,
+      "grad_norm": 14.049749374389648,
+      "learning_rate": 1.6576998703091663e-05,
+      "loss": 3.1451,
       "step": 4092
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 16.23194122314453,
-      "learning_rate": 1.1797133406835723e-05,
-      "loss": 1.8951,
+      "epoch": 0.51,
+      "grad_norm": 21.84672737121582,
+      "learning_rate": 1.6576161988034977e-05,
+      "loss": 3.6311,
       "step": 4093
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 16.518890380859375,
-      "learning_rate": 1.1795128796231331e-05,
-      "loss": 2.0618,
+      "epoch": 0.51,
+      "grad_norm": 17.8717041015625,
+      "learning_rate": 1.6575325272978287e-05,
+      "loss": 1.4837,
       "step": 4094
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 15.834227561950684,
-      "learning_rate": 1.1793124185626943e-05,
-      "loss": 1.3648,
+      "epoch": 0.51,
+      "grad_norm": 35.71775817871094,
+      "learning_rate": 1.65744885579216e-05,
+      "loss": 1.683,
       "step": 4095
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 21.73334503173828,
-      "learning_rate": 1.1791119575022553e-05,
-      "loss": 1.6835,
+      "epoch": 0.51,
+      "grad_norm": 11.681072235107422,
+      "learning_rate": 1.6573651842864914e-05,
+      "loss": 1.1817,
       "step": 4096
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 16.20547103881836,
-      "learning_rate": 1.1789114964418163e-05,
-      "loss": 2.8748,
+      "epoch": 0.51,
+      "grad_norm": 14.100946426391602,
+      "learning_rate": 1.6572815127808225e-05,
+      "loss": 2.141,
       "step": 4097
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 13.940154075622559,
-      "learning_rate": 1.1787110353813773e-05,
-      "loss": 1.7865,
+      "epoch": 0.51,
+      "grad_norm": 10.091500282287598,
+      "learning_rate": 1.657197841275154e-05,
+      "loss": 0.9144,
       "step": 4098
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 10.902242660522461,
-      "learning_rate": 1.1785105743209382e-05,
-      "loss": 1.4294,
+      "epoch": 0.51,
+      "grad_norm": 7.164851665496826,
+      "learning_rate": 1.6571141697694852e-05,
+      "loss": 0.4938,
       "step": 4099
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 20.06895637512207,
-      "learning_rate": 1.1783101132604993e-05,
-      "loss": 2.0791,
+      "epoch": 0.51,
+      "grad_norm": 30.006725311279297,
+      "learning_rate": 1.6570304982638162e-05,
+      "loss": 3.7963,
       "step": 4100
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 21.698528289794922,
-      "learning_rate": 1.1781096522000602e-05,
-      "loss": 1.8301,
+      "epoch": 0.51,
+      "grad_norm": 12.350128173828125,
+      "learning_rate": 1.6569468267581476e-05,
+      "loss": 2.2554,
       "step": 4101
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 50.30092239379883,
-      "learning_rate": 1.1779091911396212e-05,
-      "loss": 2.0178,
+      "epoch": 0.51,
+      "grad_norm": 13.280160903930664,
+      "learning_rate": 1.656863155252479e-05,
+      "loss": 2.8461,
       "step": 4102
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 29.254417419433594,
-      "learning_rate": 1.1777087300791824e-05,
-      "loss": 1.7604,
+      "epoch": 0.51,
+      "grad_norm": 36.530006408691406,
+      "learning_rate": 1.65677948374681e-05,
+      "loss": 3.6085,
       "step": 4103
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 13.801785469055176,
-      "learning_rate": 1.1775082690187432e-05,
-      "loss": 1.9442,
+      "epoch": 0.52,
+      "grad_norm": 14.720293045043945,
+      "learning_rate": 1.6566958122411414e-05,
+      "loss": 1.2472,
       "step": 4104
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 14.859130859375,
-      "learning_rate": 1.1773078079583042e-05,
-      "loss": 2.0129,
+      "epoch": 0.52,
+      "grad_norm": 11.89719009399414,
+      "learning_rate": 1.6566121407354727e-05,
+      "loss": 1.7108,
       "step": 4105
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 14.485882759094238,
-      "learning_rate": 1.177107346897865e-05,
-      "loss": 1.7607,
+      "epoch": 0.52,
+      "grad_norm": 15.50600528717041,
+      "learning_rate": 1.6565284692298037e-05,
+      "loss": 1.8948,
       "step": 4106
     },
     {
-      "epoch": 1.23,
-      "grad_norm": 16.90434455871582,
-      "learning_rate": 1.1769068858374262e-05,
-      "loss": 1.5387,
+      "epoch": 0.52,
+      "grad_norm": 19.96320152282715,
+      "learning_rate": 1.656444797724135e-05,
+      "loss": 2.2534,
       "step": 4107
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 13.56885051727295,
-      "learning_rate": 1.1767064247769872e-05,
-      "loss": 1.2367,
+      "epoch": 0.52,
+      "grad_norm": 11.358124732971191,
+      "learning_rate": 1.6563611262184665e-05,
+      "loss": 1.7402,
       "step": 4108
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 14.099568367004395,
-      "learning_rate": 1.176505963716548e-05,
-      "loss": 0.9839,
+      "epoch": 0.52,
+      "grad_norm": 11.969573974609375,
+      "learning_rate": 1.656277454712798e-05,
+      "loss": 1.468,
       "step": 4109
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 22.063432693481445,
-      "learning_rate": 1.1763055026561093e-05,
-      "loss": 1.3803,
+      "epoch": 0.52,
+      "grad_norm": 15.62781810760498,
+      "learning_rate": 1.656193783207129e-05,
+      "loss": 1.5013,
       "step": 4110
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 16.2562255859375,
-      "learning_rate": 1.1761050415956701e-05,
-      "loss": 1.5013,
+      "epoch": 0.52,
+      "grad_norm": 17.239166259765625,
+      "learning_rate": 1.6561101117014602e-05,
+      "loss": 3.8252,
       "step": 4111
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 26.141794204711914,
-      "learning_rate": 1.1759045805352311e-05,
-      "loss": 2.49,
+      "epoch": 0.52,
+      "grad_norm": 7.174729824066162,
+      "learning_rate": 1.6560264401957916e-05,
+      "loss": 0.6201,
       "step": 4112
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 16.62359619140625,
-      "learning_rate": 1.175704119474792e-05,
-      "loss": 0.9543,
+      "epoch": 0.52,
+      "grad_norm": 23.233652114868164,
+      "learning_rate": 1.6559427686901226e-05,
+      "loss": 1.8877,
       "step": 4113
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 21.430021286010742,
-      "learning_rate": 1.1755036584143531e-05,
-      "loss": 1.2435,
+      "epoch": 0.52,
+      "grad_norm": 7.3985066413879395,
+      "learning_rate": 1.655859097184454e-05,
+      "loss": 2.5075,
       "step": 4114
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 15.07040786743164,
-      "learning_rate": 1.1753031973539141e-05,
-      "loss": 0.9931,
+      "epoch": 0.52,
+      "grad_norm": 31.08193588256836,
+      "learning_rate": 1.6557754256787853e-05,
+      "loss": 3.7956,
       "step": 4115
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 9.915048599243164,
-      "learning_rate": 1.175102736293475e-05,
-      "loss": 1.498,
+      "epoch": 0.52,
+      "grad_norm": 5.321239471435547,
+      "learning_rate": 1.6556917541731167e-05,
+      "loss": 0.406,
       "step": 4116
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 16.46501922607422,
-      "learning_rate": 1.1749022752330362e-05,
-      "loss": 2.0349,
+      "epoch": 0.52,
+      "grad_norm": 14.6619234085083,
+      "learning_rate": 1.6556080826674477e-05,
+      "loss": 1.5106,
       "step": 4117
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 16.73762321472168,
-      "learning_rate": 1.174701814172597e-05,
-      "loss": 1.9466,
+      "epoch": 0.52,
+      "grad_norm": 9.001350402832031,
+      "learning_rate": 1.655524411161779e-05,
+      "loss": 1.8031,
       "step": 4118
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 19.883028030395508,
-      "learning_rate": 1.174501353112158e-05,
-      "loss": 1.8937,
+      "epoch": 0.52,
+      "grad_norm": 11.533523559570312,
+      "learning_rate": 1.65544073965611e-05,
+      "loss": 0.8387,
       "step": 4119
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 47.894859313964844,
-      "learning_rate": 1.1743008920517192e-05,
-      "loss": 1.9303,
+      "epoch": 0.52,
+      "grad_norm": 8.868361473083496,
+      "learning_rate": 1.6553570681504415e-05,
+      "loss": 0.2606,
       "step": 4120
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 14.77435302734375,
-      "learning_rate": 1.17410043099128e-05,
-      "loss": 1.8005,
+      "epoch": 0.52,
+      "grad_norm": 34.47975158691406,
+      "learning_rate": 1.655273396644773e-05,
+      "loss": 2.8862,
       "step": 4121
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 9.72576904296875,
-      "learning_rate": 1.173899969930841e-05,
-      "loss": 1.5267,
+      "epoch": 0.52,
+      "grad_norm": 20.224817276000977,
+      "learning_rate": 1.655189725139104e-05,
+      "loss": 2.4562,
       "step": 4122
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 27.231468200683594,
-      "learning_rate": 1.173699508870402e-05,
-      "loss": 1.5438,
+      "epoch": 0.52,
+      "grad_norm": 12.108320236206055,
+      "learning_rate": 1.6551060536334353e-05,
+      "loss": 1.7544,
       "step": 4123
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 14.572410583496094,
-      "learning_rate": 1.173499047809963e-05,
-      "loss": 1.7702,
+      "epoch": 0.52,
+      "grad_norm": 12.343804359436035,
+      "learning_rate": 1.6550223821277666e-05,
+      "loss": 2.9308,
       "step": 4124
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 14.047436714172363,
-      "learning_rate": 1.1732985867495239e-05,
-      "loss": 1.8277,
+      "epoch": 0.52,
+      "grad_norm": 24.018522262573242,
+      "learning_rate": 1.6549387106220976e-05,
+      "loss": 3.1213,
       "step": 4125
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 20.151954650878906,
-      "learning_rate": 1.173098125689085e-05,
-      "loss": 1.7252,
+      "epoch": 0.52,
+      "grad_norm": 42.93147659301758,
+      "learning_rate": 1.654855039116429e-05,
+      "loss": 1.9207,
       "step": 4126
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 13.929177284240723,
-      "learning_rate": 1.172897664628646e-05,
-      "loss": 2.2859,
+      "epoch": 0.52,
+      "grad_norm": 12.159256935119629,
+      "learning_rate": 1.65477136761076e-05,
+      "loss": 0.8922,
       "step": 4127
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 16.374547958374023,
-      "learning_rate": 1.1726972035682069e-05,
-      "loss": 1.4863,
+      "epoch": 0.52,
+      "grad_norm": 12.611270904541016,
+      "learning_rate": 1.6546876961050914e-05,
+      "loss": 1.4998,
       "step": 4128
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 20.026716232299805,
-      "learning_rate": 1.172496742507768e-05,
-      "loss": 1.4962,
+      "epoch": 0.52,
+      "grad_norm": 11.294777870178223,
+      "learning_rate": 1.6546040245994228e-05,
+      "loss": 0.698,
       "step": 4129
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 14.383737564086914,
-      "learning_rate": 1.172296281447329e-05,
-      "loss": 1.8605,
+      "epoch": 0.52,
+      "grad_norm": 15.277084350585938,
+      "learning_rate": 1.654520353093754e-05,
+      "loss": 2.7802,
       "step": 4130
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 8.625479698181152,
-      "learning_rate": 1.17209582038689e-05,
-      "loss": 1.3414,
+      "epoch": 0.52,
+      "grad_norm": 10.900196075439453,
+      "learning_rate": 1.654436681588085e-05,
+      "loss": 0.6085,
       "step": 4131
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 9.765339851379395,
-      "learning_rate": 1.1718953593264511e-05,
-      "loss": 2.2172,
+      "epoch": 0.52,
+      "grad_norm": 18.502803802490234,
+      "learning_rate": 1.6543530100824165e-05,
+      "loss": 1.4703,
       "step": 4132
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 17.315092086791992,
-      "learning_rate": 1.171694898266012e-05,
-      "loss": 1.8639,
+      "epoch": 0.52,
+      "grad_norm": 34.35626220703125,
+      "learning_rate": 1.654269338576748e-05,
+      "loss": 2.4679,
       "step": 4133
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 28.305118560791016,
-      "learning_rate": 1.171494437205573e-05,
-      "loss": 1.2452,
+      "epoch": 0.52,
+      "grad_norm": 21.993507385253906,
+      "learning_rate": 1.654185667071079e-05,
+      "loss": 3.4082,
       "step": 4134
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 23.19142723083496,
-      "learning_rate": 1.1712939761451338e-05,
-      "loss": 1.5799,
+      "epoch": 0.52,
+      "grad_norm": 23.336885452270508,
+      "learning_rate": 1.6541019955654103e-05,
+      "loss": 3.8204,
       "step": 4135
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 38.168418884277344,
-      "learning_rate": 1.171093515084695e-05,
-      "loss": 2.4729,
+      "epoch": 0.52,
+      "grad_norm": 14.885393142700195,
+      "learning_rate": 1.6540183240597416e-05,
+      "loss": 1.2218,
       "step": 4136
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 12.799455642700195,
-      "learning_rate": 1.1708930540242558e-05,
-      "loss": 0.9976,
+      "epoch": 0.52,
+      "grad_norm": 12.772022247314453,
+      "learning_rate": 1.653934652554073e-05,
+      "loss": 1.872,
       "step": 4137
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 11.151683807373047,
-      "learning_rate": 1.1706925929638168e-05,
-      "loss": 2.1855,
+      "epoch": 0.52,
+      "grad_norm": 7.038129806518555,
+      "learning_rate": 1.653850981048404e-05,
+      "loss": 1.8513,
       "step": 4138
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 55.58821487426758,
-      "learning_rate": 1.170492131903378e-05,
-      "loss": 1.4985,
+      "epoch": 0.52,
+      "grad_norm": 4.405773639678955,
+      "learning_rate": 1.6537673095427354e-05,
+      "loss": 1.6504,
       "step": 4139
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 12.804335594177246,
-      "learning_rate": 1.1702916708429388e-05,
-      "loss": 2.2843,
+      "epoch": 0.52,
+      "grad_norm": 12.794690132141113,
+      "learning_rate": 1.6536836380370668e-05,
+      "loss": 1.3373,
       "step": 4140
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 18.690269470214844,
-      "learning_rate": 1.1700912097824998e-05,
-      "loss": 1.8465,
+      "epoch": 0.52,
+      "grad_norm": 69.78988647460938,
+      "learning_rate": 1.6535999665313978e-05,
+      "loss": 1.8227,
       "step": 4141
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 13.506994247436523,
-      "learning_rate": 1.1698907487220607e-05,
-      "loss": 1.2933,
+      "epoch": 0.52,
+      "grad_norm": 13.622889518737793,
+      "learning_rate": 1.653516295025729e-05,
+      "loss": 1.7115,
       "step": 4142
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 12.685111999511719,
-      "learning_rate": 1.1696902876616219e-05,
-      "loss": 1.1366,
+      "epoch": 0.52,
+      "grad_norm": 10.512645721435547,
+      "learning_rate": 1.6534326235200605e-05,
+      "loss": 0.8185,
       "step": 4143
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 51.746055603027344,
-      "learning_rate": 1.1694898266011827e-05,
-      "loss": 2.0853,
+      "epoch": 0.52,
+      "grad_norm": 9.174022674560547,
+      "learning_rate": 1.653348952014392e-05,
+      "loss": 0.811,
       "step": 4144
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 16.30849838256836,
-      "learning_rate": 1.1692893655407437e-05,
-      "loss": 1.8595,
+      "epoch": 0.52,
+      "grad_norm": 14.15351676940918,
+      "learning_rate": 1.653265280508723e-05,
+      "loss": 2.5374,
       "step": 4145
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 24.40019989013672,
-      "learning_rate": 1.1690889044803049e-05,
-      "loss": 2.4656,
+      "epoch": 0.52,
+      "grad_norm": 14.107945442199707,
+      "learning_rate": 1.6531816090030543e-05,
+      "loss": 1.8294,
       "step": 4146
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 30.043943405151367,
-      "learning_rate": 1.1688884434198657e-05,
-      "loss": 1.5323,
+      "epoch": 0.52,
+      "grad_norm": 46.87839126586914,
+      "learning_rate": 1.6530979374973853e-05,
+      "loss": 3.7714,
       "step": 4147
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 16.770814895629883,
-      "learning_rate": 1.1686879823594269e-05,
-      "loss": 2.9425,
+      "epoch": 0.52,
+      "grad_norm": 27.021465301513672,
+      "learning_rate": 1.6530142659917167e-05,
+      "loss": 1.1433,
       "step": 4148
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 11.892877578735352,
-      "learning_rate": 1.1684875212989877e-05,
-      "loss": 2.216,
+      "epoch": 0.52,
+      "grad_norm": 57.1532096862793,
+      "learning_rate": 1.652930594486048e-05,
+      "loss": 1.6439,
       "step": 4149
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 20.787343978881836,
-      "learning_rate": 1.1682870602385488e-05,
-      "loss": 2.4695,
+      "epoch": 0.52,
+      "grad_norm": 25.887805938720703,
+      "learning_rate": 1.652846922980379e-05,
+      "loss": 2.2957,
       "step": 4150
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 23.898902893066406,
-      "learning_rate": 1.16808659917811e-05,
-      "loss": 1.5012,
+      "epoch": 0.52,
+      "grad_norm": 15.23217487335205,
+      "learning_rate": 1.6527632514747104e-05,
+      "loss": 2.052,
       "step": 4151
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 24.14310073852539,
-      "learning_rate": 1.1678861381176708e-05,
-      "loss": 1.8126,
+      "epoch": 0.52,
+      "grad_norm": 10.493826866149902,
+      "learning_rate": 1.6526795799690414e-05,
+      "loss": 1.3312,
       "step": 4152
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 16.704151153564453,
-      "learning_rate": 1.1676856770572318e-05,
-      "loss": 2.1569,
+      "epoch": 0.52,
+      "grad_norm": 10.333160400390625,
+      "learning_rate": 1.6525959084633728e-05,
+      "loss": 1.345,
       "step": 4153
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 11.908331871032715,
-      "learning_rate": 1.1674852159967926e-05,
-      "loss": 1.5541,
+      "epoch": 0.52,
+      "grad_norm": 32.34117889404297,
+      "learning_rate": 1.6525122369577042e-05,
+      "loss": 2.4136,
       "step": 4154
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 14.777883529663086,
-      "learning_rate": 1.1672847549363538e-05,
-      "loss": 1.7314,
+      "epoch": 0.52,
+      "grad_norm": 18.288616180419922,
+      "learning_rate": 1.6524285654520352e-05,
+      "loss": 0.7964,
       "step": 4155
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 13.375772476196289,
-      "learning_rate": 1.1670842938759146e-05,
-      "loss": 1.1609,
+      "epoch": 0.52,
+      "grad_norm": 11.544230461120605,
+      "learning_rate": 1.6523448939463666e-05,
+      "loss": 2.2476,
       "step": 4156
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 9.32558536529541,
-      "learning_rate": 1.1668838328154756e-05,
-      "loss": 1.6317,
+      "epoch": 0.52,
+      "grad_norm": 13.22861099243164,
+      "learning_rate": 1.652261222440698e-05,
+      "loss": 1.4952,
       "step": 4157
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 14.819250106811523,
-      "learning_rate": 1.1666833717550368e-05,
-      "loss": 1.5169,
+      "epoch": 0.52,
+      "grad_norm": 48.550697326660156,
+      "learning_rate": 1.6521775509350293e-05,
+      "loss": 1.0027,
       "step": 4158
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 7.862631797790527,
-      "learning_rate": 1.1664829106945977e-05,
-      "loss": 1.187,
+      "epoch": 0.52,
+      "grad_norm": 8.484000205993652,
+      "learning_rate": 1.6520938794293603e-05,
+      "loss": 0.4071,
       "step": 4159
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 26.258054733276367,
-      "learning_rate": 1.1662824496341587e-05,
-      "loss": 2.8375,
+      "epoch": 0.52,
+      "grad_norm": 13.84500503540039,
+      "learning_rate": 1.6520102079236917e-05,
+      "loss": 1.1636,
       "step": 4160
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 16.67561912536621,
-      "learning_rate": 1.1660819885737195e-05,
-      "loss": 1.5601,
+      "epoch": 0.52,
+      "grad_norm": 15.75895881652832,
+      "learning_rate": 1.651926536418023e-05,
+      "loss": 1.3895,
       "step": 4161
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 22.342544555664062,
-      "learning_rate": 1.1658815275132807e-05,
-      "loss": 1.5789,
+      "epoch": 0.52,
+      "grad_norm": 23.719133377075195,
+      "learning_rate": 1.651842864912354e-05,
+      "loss": 2.8539,
       "step": 4162
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 13.550854682922363,
-      "learning_rate": 1.1656810664528417e-05,
-      "loss": 2.4655,
+      "epoch": 0.52,
+      "grad_norm": 30.87020492553711,
+      "learning_rate": 1.6517591934066854e-05,
+      "loss": 1.6645,
       "step": 4163
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 33.92632293701172,
-      "learning_rate": 1.1654806053924025e-05,
-      "loss": 1.5478,
+      "epoch": 0.52,
+      "grad_norm": 9.798362731933594,
+      "learning_rate": 1.6516755219010168e-05,
+      "loss": 1.4841,
       "step": 4164
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 10.819427490234375,
-      "learning_rate": 1.1652801443319637e-05,
-      "loss": 1.5479,
+      "epoch": 0.52,
+      "grad_norm": 8.731616973876953,
+      "learning_rate": 1.6515918503953482e-05,
+      "loss": 0.8007,
       "step": 4165
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 16.504961013793945,
-      "learning_rate": 1.1650796832715245e-05,
-      "loss": 1.4122,
+      "epoch": 0.52,
+      "grad_norm": 8.03121566772461,
+      "learning_rate": 1.6515081788896792e-05,
+      "loss": 0.4783,
       "step": 4166
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 22.91649055480957,
-      "learning_rate": 1.1648792222110856e-05,
-      "loss": 1.9459,
+      "epoch": 0.52,
+      "grad_norm": 8.199400901794434,
+      "learning_rate": 1.6514245073840106e-05,
+      "loss": 2.1517,
       "step": 4167
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 9.941856384277344,
-      "learning_rate": 1.1646787611506466e-05,
-      "loss": 1.7419,
+      "epoch": 0.52,
+      "grad_norm": 17.17656898498535,
+      "learning_rate": 1.651340835878342e-05,
+      "loss": 1.3263,
       "step": 4168
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 12.914164543151855,
-      "learning_rate": 1.1644783000902076e-05,
-      "loss": 1.2351,
+      "epoch": 0.52,
+      "grad_norm": 12.302600860595703,
+      "learning_rate": 1.651257164372673e-05,
+      "loss": 0.6993,
       "step": 4169
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 10.681536674499512,
-      "learning_rate": 1.1642778390297686e-05,
-      "loss": 1.7278,
+      "epoch": 0.52,
+      "grad_norm": 34.4046745300293,
+      "learning_rate": 1.6511734928670043e-05,
+      "loss": 1.5034,
       "step": 4170
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 17.06873321533203,
-      "learning_rate": 1.1640773779693296e-05,
-      "loss": 2.0348,
+      "epoch": 0.52,
+      "grad_norm": 15.348669052124023,
+      "learning_rate": 1.6510898213613357e-05,
+      "loss": 1.8687,
       "step": 4171
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 20.173357009887695,
-      "learning_rate": 1.1638769169088906e-05,
-      "loss": 1.6212,
+      "epoch": 0.52,
+      "grad_norm": 8.436819076538086,
+      "learning_rate": 1.6510061498556667e-05,
+      "loss": 1.0849,
       "step": 4172
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 6.546261310577393,
-      "learning_rate": 1.1636764558484514e-05,
-      "loss": 0.4214,
+      "epoch": 0.52,
+      "grad_norm": 12.729174613952637,
+      "learning_rate": 1.650922478349998e-05,
+      "loss": 1.7756,
       "step": 4173
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 27.385225296020508,
-      "learning_rate": 1.1634759947880126e-05,
-      "loss": 1.6822,
+      "epoch": 0.52,
+      "grad_norm": 9.80980110168457,
+      "learning_rate": 1.6508388068443294e-05,
+      "loss": 1.7873,
       "step": 4174
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 37.10486602783203,
-      "learning_rate": 1.1632755337275736e-05,
-      "loss": 1.5469,
+      "epoch": 0.52,
+      "grad_norm": 16.981359481811523,
+      "learning_rate": 1.6507551353386605e-05,
+      "loss": 2.1876,
       "step": 4175
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 13.431403160095215,
-      "learning_rate": 1.1630750726671345e-05,
-      "loss": 2.7734,
+      "epoch": 0.52,
+      "grad_norm": 7.826014041900635,
+      "learning_rate": 1.6506714638329918e-05,
+      "loss": 2.1796,
       "step": 4176
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 17.27809715270996,
-      "learning_rate": 1.1628746116066956e-05,
-      "loss": 1.8377,
+      "epoch": 0.52,
+      "grad_norm": 28.039058685302734,
+      "learning_rate": 1.650587792327323e-05,
+      "loss": 1.1134,
       "step": 4177
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 6.829921245574951,
-      "learning_rate": 1.1626741505462565e-05,
-      "loss": 1.0531,
+      "epoch": 0.52,
+      "grad_norm": 8.88306999206543,
+      "learning_rate": 1.6505041208216542e-05,
+      "loss": 1.0974,
       "step": 4178
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 15.239294052124023,
-      "learning_rate": 1.1624736894858175e-05,
-      "loss": 1.8001,
+      "epoch": 0.52,
+      "grad_norm": 6.60862398147583,
+      "learning_rate": 1.6504204493159856e-05,
+      "loss": 0.5625,
       "step": 4179
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 12.837756156921387,
-      "learning_rate": 1.1622732284253783e-05,
-      "loss": 1.1665,
+      "epoch": 0.52,
+      "grad_norm": 10.781895637512207,
+      "learning_rate": 1.6503367778103166e-05,
+      "loss": 1.1298,
       "step": 4180
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 16.143875122070312,
-      "learning_rate": 1.1620727673649395e-05,
-      "loss": 0.9885,
+      "epoch": 0.52,
+      "grad_norm": 18.55240249633789,
+      "learning_rate": 1.650253106304648e-05,
+      "loss": 2.801,
       "step": 4181
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 12.66282844543457,
-      "learning_rate": 1.1618723063045005e-05,
-      "loss": 1.219,
+      "epoch": 0.52,
+      "grad_norm": 16.69265365600586,
+      "learning_rate": 1.6501694347989793e-05,
+      "loss": 1.61,
       "step": 4182
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 27.858238220214844,
-      "learning_rate": 1.1616718452440614e-05,
-      "loss": 2.2014,
+      "epoch": 0.52,
+      "grad_norm": 17.6204891204834,
+      "learning_rate": 1.6500857632933104e-05,
+      "loss": 3.775,
       "step": 4183
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 21.537065505981445,
-      "learning_rate": 1.1614713841836225e-05,
-      "loss": 2.1228,
+      "epoch": 0.53,
+      "grad_norm": 9.196520805358887,
+      "learning_rate": 1.6500020917876417e-05,
+      "loss": 0.6398,
       "step": 4184
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 43.390296936035156,
-      "learning_rate": 1.1612709231231834e-05,
-      "loss": 2.343,
+      "epoch": 0.53,
+      "grad_norm": 15.22008228302002,
+      "learning_rate": 1.649918420281973e-05,
+      "loss": 2.4216,
       "step": 4185
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 24.188064575195312,
-      "learning_rate": 1.1610704620627444e-05,
-      "loss": 1.1875,
+      "epoch": 0.53,
+      "grad_norm": 41.116641998291016,
+      "learning_rate": 1.6498347487763045e-05,
+      "loss": 2.4155,
       "step": 4186
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 13.248528480529785,
-      "learning_rate": 1.1608700010023052e-05,
-      "loss": 1.3365,
+      "epoch": 0.53,
+      "grad_norm": 33.585166931152344,
+      "learning_rate": 1.6497510772706355e-05,
+      "loss": 2.9035,
       "step": 4187
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 26.479053497314453,
-      "learning_rate": 1.1606695399418664e-05,
-      "loss": 2.2343,
+      "epoch": 0.53,
+      "grad_norm": 7.945860385894775,
+      "learning_rate": 1.649667405764967e-05,
+      "loss": 0.5704,
       "step": 4188
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 20.167842864990234,
-      "learning_rate": 1.1604690788814274e-05,
-      "loss": 1.9832,
+      "epoch": 0.53,
+      "grad_norm": 13.593266487121582,
+      "learning_rate": 1.6495837342592982e-05,
+      "loss": 2.459,
       "step": 4189
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 27.049171447753906,
-      "learning_rate": 1.1602686178209882e-05,
-      "loss": 1.6917,
+      "epoch": 0.53,
+      "grad_norm": 99.65393829345703,
+      "learning_rate": 1.6495000627536292e-05,
+      "loss": 2.3606,
       "step": 4190
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 7.952903747558594,
-      "learning_rate": 1.1600681567605494e-05,
-      "loss": 1.1251,
+      "epoch": 0.53,
+      "grad_norm": 15.27071762084961,
+      "learning_rate": 1.6494163912479606e-05,
+      "loss": 2.4064,
       "step": 4191
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 9.99161148071289,
-      "learning_rate": 1.1598676957001103e-05,
-      "loss": 1.3309,
+      "epoch": 0.53,
+      "grad_norm": 13.85755443572998,
+      "learning_rate": 1.649332719742292e-05,
+      "loss": 0.7231,
       "step": 4192
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 23.682477951049805,
-      "learning_rate": 1.1596672346396713e-05,
-      "loss": 1.7625,
+      "epoch": 0.53,
+      "grad_norm": 17.547557830810547,
+      "learning_rate": 1.6492490482366233e-05,
+      "loss": 1.8349,
       "step": 4193
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 14.787646293640137,
-      "learning_rate": 1.1594667735792324e-05,
-      "loss": 0.8436,
+      "epoch": 0.53,
+      "grad_norm": 13.032142639160156,
+      "learning_rate": 1.6491653767309544e-05,
+      "loss": 1.0245,
       "step": 4194
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 16.3641414642334,
-      "learning_rate": 1.1592663125187933e-05,
-      "loss": 1.5916,
+      "epoch": 0.53,
+      "grad_norm": 37.35247039794922,
+      "learning_rate": 1.6490817052252857e-05,
+      "loss": 3.246,
       "step": 4195
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 15.075904846191406,
-      "learning_rate": 1.1590658514583543e-05,
-      "loss": 1.5184,
+      "epoch": 0.53,
+      "grad_norm": 27.185203552246094,
+      "learning_rate": 1.648998033719617e-05,
+      "loss": 3.0728,
       "step": 4196
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 13.050864219665527,
-      "learning_rate": 1.1588653903979153e-05,
-      "loss": 1.7963,
+      "epoch": 0.53,
+      "grad_norm": 10.853189468383789,
+      "learning_rate": 1.648914362213948e-05,
+      "loss": 2.0268,
       "step": 4197
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 10.601302146911621,
-      "learning_rate": 1.1586649293374763e-05,
-      "loss": 1.7489,
+      "epoch": 0.53,
+      "grad_norm": 16.356616973876953,
+      "learning_rate": 1.6488306907082795e-05,
+      "loss": 1.4985,
       "step": 4198
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 21.2014217376709,
-      "learning_rate": 1.1584644682770371e-05,
-      "loss": 1.6054,
+      "epoch": 0.53,
+      "grad_norm": 16.33892822265625,
+      "learning_rate": 1.648747019202611e-05,
+      "loss": 3.2044,
       "step": 4199
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 17.15383529663086,
-      "learning_rate": 1.1582640072165983e-05,
-      "loss": 1.2979,
+      "epoch": 0.53,
+      "grad_norm": 16.992094039916992,
+      "learning_rate": 1.648663347696942e-05,
+      "loss": 2.4081,
       "step": 4200
     },
     {
-      "epoch": 1.26,
-      "eval_loss": 0.2202432006597519,
-      "eval_runtime": 43.4456,
-      "eval_samples_per_second": 34.043,
-      "eval_steps_per_second": 34.043,
-      "step": 4200
+      "epoch": 0.53,
+      "grad_norm": 19.227170944213867,
+      "learning_rate": 1.6485796761912732e-05,
+      "loss": 0.8864,
+      "step": 4201
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 14.757326126098633,
-      "learning_rate": 1.1580635461561593e-05,
-      "loss": 1.03,
-      "step": 4201
-    },
-    {
-      "epoch": 1.26,
-      "grad_norm": 12.233736038208008,
-      "learning_rate": 1.1578630850957202e-05,
-      "loss": 1.4463,
+      "epoch": 0.53,
+      "grad_norm": 154.24375915527344,
+      "learning_rate": 1.6484960046856046e-05,
+      "loss": 0.8654,
       "step": 4202
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 17.687862396240234,
-      "learning_rate": 1.1576626240352814e-05,
-      "loss": 1.9516,
+      "epoch": 0.53,
+      "grad_norm": 8.544039726257324,
+      "learning_rate": 1.6484123331799356e-05,
+      "loss": 0.8526,
       "step": 4203
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 18.68252944946289,
-      "learning_rate": 1.1574621629748422e-05,
-      "loss": 2.6357,
+      "epoch": 0.53,
+      "grad_norm": 30.925003051757812,
+      "learning_rate": 1.648328661674267e-05,
+      "loss": 2.7016,
       "step": 4204
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 24.503774642944336,
-      "learning_rate": 1.1572617019144032e-05,
-      "loss": 1.4327,
+      "epoch": 0.53,
+      "grad_norm": 10.509055137634277,
+      "learning_rate": 1.648244990168598e-05,
+      "loss": 0.7868,
       "step": 4205
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 14.029136657714844,
-      "learning_rate": 1.1570612408539644e-05,
-      "loss": 2.0086,
+      "epoch": 0.53,
+      "grad_norm": 10.42081069946289,
+      "learning_rate": 1.6481613186629294e-05,
+      "loss": 1.9535,
       "step": 4206
     },
     {
-      "epoch": 1.26,
-      "grad_norm": 13.784452438354492,
-      "learning_rate": 1.1568607797935252e-05,
-      "loss": 2.5528,
+      "epoch": 0.53,
+      "grad_norm": 4.504967212677002,
+      "learning_rate": 1.6480776471572608e-05,
+      "loss": 0.3315,
       "step": 4207
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 29.253156661987305,
-      "learning_rate": 1.1566603187330862e-05,
-      "loss": 0.5661,
+      "epoch": 0.53,
+      "grad_norm": 79.56721496582031,
+      "learning_rate": 1.6479939756515918e-05,
+      "loss": 3.8614,
       "step": 4208
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 26.327823638916016,
-      "learning_rate": 1.156459857672647e-05,
-      "loss": 1.8774,
+      "epoch": 0.53,
+      "grad_norm": 15.625540733337402,
+      "learning_rate": 1.647910304145923e-05,
+      "loss": 1.8213,
       "step": 4209
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 50.058536529541016,
-      "learning_rate": 1.1562593966122082e-05,
-      "loss": 2.2888,
+      "epoch": 0.53,
+      "grad_norm": 9.894253730773926,
+      "learning_rate": 1.6478266326402545e-05,
+      "loss": 1.1941,
       "step": 4210
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 8.850481033325195,
-      "learning_rate": 1.156058935551769e-05,
-      "loss": 1.476,
+      "epoch": 0.53,
+      "grad_norm": 17.677467346191406,
+      "learning_rate": 1.6477429611345855e-05,
+      "loss": 3.2213,
       "step": 4211
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 12.578634262084961,
-      "learning_rate": 1.1558584744913301e-05,
-      "loss": 1.5757,
+      "epoch": 0.53,
+      "grad_norm": 4.6228766441345215,
+      "learning_rate": 1.647659289628917e-05,
+      "loss": 1.6235,
       "step": 4212
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 42.25316619873047,
-      "learning_rate": 1.1556580134308913e-05,
-      "loss": 1.574,
+      "epoch": 0.53,
+      "grad_norm": 30.595361709594727,
+      "learning_rate": 1.6475756181232483e-05,
+      "loss": 2.2286,
       "step": 4213
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 15.227287292480469,
-      "learning_rate": 1.1554575523704521e-05,
-      "loss": 2.4708,
+      "epoch": 0.53,
+      "grad_norm": 22.9339542388916,
+      "learning_rate": 1.6474919466175796e-05,
+      "loss": 3.3339,
       "step": 4214
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 10.333723068237305,
-      "learning_rate": 1.1552570913100131e-05,
-      "loss": 1.6319,
+      "epoch": 0.53,
+      "grad_norm": 11.161800384521484,
+      "learning_rate": 1.6474082751119107e-05,
+      "loss": 2.5971,
       "step": 4215
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 9.808292388916016,
-      "learning_rate": 1.1550566302495741e-05,
-      "loss": 2.1226,
+      "epoch": 0.53,
+      "grad_norm": 8.695982933044434,
+      "learning_rate": 1.647324603606242e-05,
+      "loss": 1.6401,
       "step": 4216
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 39.865562438964844,
-      "learning_rate": 1.1548561691891351e-05,
-      "loss": 2.9798,
+      "epoch": 0.53,
+      "grad_norm": 31.54930877685547,
+      "learning_rate": 1.6472409321005734e-05,
+      "loss": 2.0755,
       "step": 4217
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 47.3433723449707,
-      "learning_rate": 1.154655708128696e-05,
-      "loss": 1.7633,
+      "epoch": 0.53,
+      "grad_norm": 19.136375427246094,
+      "learning_rate": 1.6471572605949044e-05,
+      "loss": 1.2303,
       "step": 4218
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 18.767004013061523,
-      "learning_rate": 1.1544552470682571e-05,
-      "loss": 1.0271,
+      "epoch": 0.53,
+      "grad_norm": 25.4252986907959,
+      "learning_rate": 1.6470735890892358e-05,
+      "loss": 2.7373,
       "step": 4219
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 5.930094242095947,
-      "learning_rate": 1.1542547860078182e-05,
-      "loss": 0.7385,
+      "epoch": 0.53,
+      "grad_norm": 14.933919906616211,
+      "learning_rate": 1.646989917583567e-05,
+      "loss": 2.7044,
       "step": 4220
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 17.10114097595215,
-      "learning_rate": 1.154054324947379e-05,
-      "loss": 1.4075,
+      "epoch": 0.53,
+      "grad_norm": 25.30756187438965,
+      "learning_rate": 1.6469062460778985e-05,
+      "loss": 2.2519,
       "step": 4221
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 22.982498168945312,
-      "learning_rate": 1.1538538638869402e-05,
-      "loss": 2.3147,
+      "epoch": 0.53,
+      "grad_norm": 3.7429444789886475,
+      "learning_rate": 1.6468225745722295e-05,
+      "loss": 0.5936,
       "step": 4222
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 29.882640838623047,
-      "learning_rate": 1.153653402826501e-05,
-      "loss": 1.3968,
+      "epoch": 0.53,
+      "grad_norm": 12.389649391174316,
+      "learning_rate": 1.646738903066561e-05,
+      "loss": 1.0459,
       "step": 4223
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 17.916343688964844,
-      "learning_rate": 1.153452941766062e-05,
-      "loss": 1.5234,
+      "epoch": 0.53,
+      "grad_norm": 27.49970054626465,
+      "learning_rate": 1.6466552315608923e-05,
+      "loss": 2.9375,
       "step": 4224
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 12.272050857543945,
-      "learning_rate": 1.1532524807056232e-05,
-      "loss": 0.895,
+      "epoch": 0.53,
+      "grad_norm": 4.344343185424805,
+      "learning_rate": 1.6465715600552233e-05,
+      "loss": 0.0751,
       "step": 4225
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 14.076766967773438,
-      "learning_rate": 1.153052019645184e-05,
-      "loss": 1.1994,
+      "epoch": 0.53,
+      "grad_norm": 14.881327629089355,
+      "learning_rate": 1.6464878885495547e-05,
+      "loss": 2.584,
       "step": 4226
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 16.016265869140625,
-      "learning_rate": 1.152851558584745e-05,
-      "loss": 1.0423,
+      "epoch": 0.53,
+      "grad_norm": 14.25925064086914,
+      "learning_rate": 1.646404217043886e-05,
+      "loss": 1.7455,
       "step": 4227
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 10.317485809326172,
-      "learning_rate": 1.1526510975243059e-05,
-      "loss": 1.3147,
+      "epoch": 0.53,
+      "grad_norm": 13.312737464904785,
+      "learning_rate": 1.646320545538217e-05,
+      "loss": 2.3142,
       "step": 4228
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 12.768622398376465,
-      "learning_rate": 1.152450636463867e-05,
-      "loss": 2.2527,
+      "epoch": 0.53,
+      "grad_norm": 22.862720489501953,
+      "learning_rate": 1.6462368740325484e-05,
+      "loss": 2.6122,
       "step": 4229
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 15.519412994384766,
-      "learning_rate": 1.1522501754034279e-05,
-      "loss": 1.4902,
+      "epoch": 0.53,
+      "grad_norm": 19.689655303955078,
+      "learning_rate": 1.6461532025268794e-05,
+      "loss": 3.9217,
       "step": 4230
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 18.289718627929688,
-      "learning_rate": 1.1520497143429889e-05,
-      "loss": 2.5218,
+      "epoch": 0.53,
+      "grad_norm": 22.79598045349121,
+      "learning_rate": 1.6460695310212108e-05,
+      "loss": 0.636,
       "step": 4231
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 17.28578758239746,
-      "learning_rate": 1.1518492532825501e-05,
-      "loss": 0.8004,
+      "epoch": 0.53,
+      "grad_norm": 16.043413162231445,
+      "learning_rate": 1.645985859515542e-05,
+      "loss": 2.4597,
       "step": 4232
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 23.51228141784668,
-      "learning_rate": 1.151648792222111e-05,
-      "loss": 1.8972,
+      "epoch": 0.53,
+      "grad_norm": 19.496339797973633,
+      "learning_rate": 1.6459021880098732e-05,
+      "loss": 2.9435,
       "step": 4233
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 12.242810249328613,
-      "learning_rate": 1.151448331161672e-05,
-      "loss": 1.0982,
+      "epoch": 0.53,
+      "grad_norm": 9.446452140808105,
+      "learning_rate": 1.6458185165042046e-05,
+      "loss": 1.7816,
       "step": 4234
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 16.621652603149414,
-      "learning_rate": 1.1512478701012328e-05,
-      "loss": 1.8477,
+      "epoch": 0.53,
+      "grad_norm": 12.899818420410156,
+      "learning_rate": 1.645734844998536e-05,
+      "loss": 2.1706,
       "step": 4235
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 9.91673755645752,
-      "learning_rate": 1.151047409040794e-05,
-      "loss": 1.4897,
+      "epoch": 0.53,
+      "grad_norm": 12.458577156066895,
+      "learning_rate": 1.645651173492867e-05,
+      "loss": 1.667,
       "step": 4236
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 17.222322463989258,
-      "learning_rate": 1.150846947980355e-05,
-      "loss": 1.2186,
+      "epoch": 0.53,
+      "grad_norm": 27.420490264892578,
+      "learning_rate": 1.6455675019871983e-05,
+      "loss": 2.0797,
       "step": 4237
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 11.030045509338379,
-      "learning_rate": 1.1506464869199158e-05,
-      "loss": 1.0489,
+      "epoch": 0.53,
+      "grad_norm": 9.165604591369629,
+      "learning_rate": 1.6454838304815297e-05,
+      "loss": 1.4257,
       "step": 4238
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 13.282185554504395,
-      "learning_rate": 1.150446025859477e-05,
-      "loss": 2.1427,
+      "epoch": 0.53,
+      "grad_norm": 11.653155326843262,
+      "learning_rate": 1.6454001589758607e-05,
+      "loss": 1.6472,
       "step": 4239
     },
     {
-      "epoch": 1.27,
-      "grad_norm": 18.355653762817383,
-      "learning_rate": 1.1502455647990378e-05,
-      "loss": 1.5696,
+      "epoch": 0.53,
+      "grad_norm": 52.71607208251953,
+      "learning_rate": 1.645316487470192e-05,
+      "loss": 3.3165,
       "step": 4240
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 15.735917091369629,
-      "learning_rate": 1.1500451037385988e-05,
-      "loss": 1.789,
+      "epoch": 0.53,
+      "grad_norm": 60.9556999206543,
+      "learning_rate": 1.6452328159645234e-05,
+      "loss": 2.1788,
       "step": 4241
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 28.514938354492188,
-      "learning_rate": 1.1498446426781598e-05,
-      "loss": 2.2112,
+      "epoch": 0.53,
+      "grad_norm": 8.82602310180664,
+      "learning_rate": 1.6451491444588548e-05,
+      "loss": 0.8412,
       "step": 4242
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 16.573684692382812,
-      "learning_rate": 1.1496441816177208e-05,
-      "loss": 1.6606,
+      "epoch": 0.53,
+      "grad_norm": 11.567743301391602,
+      "learning_rate": 1.6450654729531858e-05,
+      "loss": 2.6234,
       "step": 4243
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 18.012393951416016,
-      "learning_rate": 1.1494437205572819e-05,
-      "loss": 1.3518,
+      "epoch": 0.53,
+      "grad_norm": 15.222989082336426,
+      "learning_rate": 1.6449818014475172e-05,
+      "loss": 1.9354,
       "step": 4244
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 21.622119903564453,
-      "learning_rate": 1.1492432594968429e-05,
-      "loss": 1.9794,
+      "epoch": 0.53,
+      "grad_norm": 20.00623321533203,
+      "learning_rate": 1.6448981299418486e-05,
+      "loss": 2.37,
       "step": 4245
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 11.884101867675781,
-      "learning_rate": 1.1490427984364039e-05,
-      "loss": 1.4142,
+      "epoch": 0.53,
+      "grad_norm": 19.969297409057617,
+      "learning_rate": 1.6448144584361796e-05,
+      "loss": 2.7817,
       "step": 4246
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 19.552974700927734,
-      "learning_rate": 1.1488423373759647e-05,
-      "loss": 1.0571,
+      "epoch": 0.53,
+      "grad_norm": 10.317906379699707,
+      "learning_rate": 1.644730786930511e-05,
+      "loss": 1.6364,
       "step": 4247
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 17.979076385498047,
-      "learning_rate": 1.1486418763155259e-05,
-      "loss": 1.6985,
+      "epoch": 0.53,
+      "grad_norm": 16.51494789123535,
+      "learning_rate": 1.6446471154248423e-05,
+      "loss": 2.9893,
       "step": 4248
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 19.238636016845703,
-      "learning_rate": 1.1484414152550869e-05,
-      "loss": 1.396,
+      "epoch": 0.53,
+      "grad_norm": 12.647660255432129,
+      "learning_rate": 1.6445634439191737e-05,
+      "loss": 2.065,
       "step": 4249
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 46.26264953613281,
-      "learning_rate": 1.1482409541946477e-05,
-      "loss": 2.4083,
+      "epoch": 0.53,
+      "grad_norm": 11.133133888244629,
+      "learning_rate": 1.6444797724135047e-05,
+      "loss": 1.5552,
       "step": 4250
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 13.219773292541504,
-      "learning_rate": 1.1480404931342089e-05,
-      "loss": 0.8623,
+      "epoch": 0.53,
+      "grad_norm": 10.225388526916504,
+      "learning_rate": 1.644396100907836e-05,
+      "loss": 0.9422,
       "step": 4251
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 9.744564056396484,
-      "learning_rate": 1.1478400320737697e-05,
-      "loss": 2.33,
+      "epoch": 0.53,
+      "grad_norm": 13.449052810668945,
+      "learning_rate": 1.6443124294021674e-05,
+      "loss": 0.5701,
       "step": 4252
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 17.062484741210938,
-      "learning_rate": 1.1476395710133308e-05,
-      "loss": 1.4111,
+      "epoch": 0.53,
+      "grad_norm": 16.118934631347656,
+      "learning_rate": 1.6442287578964985e-05,
+      "loss": 1.8372,
       "step": 4253
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 11.308719635009766,
-      "learning_rate": 1.1474391099528916e-05,
-      "loss": 1.0592,
+      "epoch": 0.53,
+      "grad_norm": 12.920221328735352,
+      "learning_rate": 1.6441450863908298e-05,
+      "loss": 2.2682,
       "step": 4254
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 12.195525169372559,
-      "learning_rate": 1.1472386488924528e-05,
-      "loss": 1.8724,
+      "epoch": 0.53,
+      "grad_norm": 17.492290496826172,
+      "learning_rate": 1.6440614148851612e-05,
+      "loss": 1.5939,
       "step": 4255
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 13.376428604125977,
-      "learning_rate": 1.1470381878320138e-05,
-      "loss": 1.5216,
+      "epoch": 0.53,
+      "grad_norm": 23.524873733520508,
+      "learning_rate": 1.6439777433794922e-05,
+      "loss": 1.9309,
       "step": 4256
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 7.64074182510376,
-      "learning_rate": 1.1468377267715746e-05,
-      "loss": 0.7377,
+      "epoch": 0.53,
+      "grad_norm": 14.729205131530762,
+      "learning_rate": 1.6438940718738236e-05,
+      "loss": 1.0595,
       "step": 4257
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 19.55640983581543,
-      "learning_rate": 1.1466372657111358e-05,
-      "loss": 2.3189,
+      "epoch": 0.53,
+      "grad_norm": 11.834205627441406,
+      "learning_rate": 1.6438104003681546e-05,
+      "loss": 1.9227,
       "step": 4258
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 202.1298370361328,
-      "learning_rate": 1.1464368046506966e-05,
-      "loss": 1.7217,
+      "epoch": 0.53,
+      "grad_norm": 18.87928009033203,
+      "learning_rate": 1.643726728862486e-05,
+      "loss": 2.661,
       "step": 4259
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 44.232112884521484,
-      "learning_rate": 1.1462363435902576e-05,
-      "loss": 1.1155,
+      "epoch": 0.53,
+      "grad_norm": 9.27440357208252,
+      "learning_rate": 1.6436430573568173e-05,
+      "loss": 0.9423,
       "step": 4260
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 20.715179443359375,
-      "learning_rate": 1.1460358825298185e-05,
-      "loss": 1.5698,
+      "epoch": 0.53,
+      "grad_norm": 49.268157958984375,
+      "learning_rate": 1.6435593858511484e-05,
+      "loss": 2.913,
       "step": 4261
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 16.672870635986328,
-      "learning_rate": 1.1458354214693797e-05,
-      "loss": 2.0041,
+      "epoch": 0.53,
+      "grad_norm": 13.71524715423584,
+      "learning_rate": 1.6434757143454797e-05,
+      "loss": 1.9741,
       "step": 4262
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 15.357919692993164,
-      "learning_rate": 1.1456349604089407e-05,
-      "loss": 1.9595,
+      "epoch": 0.53,
+      "grad_norm": 15.70528507232666,
+      "learning_rate": 1.643392042839811e-05,
+      "loss": 1.6815,
       "step": 4263
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 25.542970657348633,
-      "learning_rate": 1.1454344993485015e-05,
-      "loss": 3.4864,
+      "epoch": 0.54,
+      "grad_norm": 7.827919006347656,
+      "learning_rate": 1.643308371334142e-05,
+      "loss": 1.6473,
       "step": 4264
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 20.45473861694336,
-      "learning_rate": 1.1452340382880627e-05,
-      "loss": 1.8289,
+      "epoch": 0.54,
+      "grad_norm": 9.871508598327637,
+      "learning_rate": 1.6432246998284735e-05,
+      "loss": 1.4339,
       "step": 4265
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 18.233806610107422,
-      "learning_rate": 1.1450335772276235e-05,
-      "loss": 1.9064,
+      "epoch": 0.54,
+      "grad_norm": 16.842849731445312,
+      "learning_rate": 1.643141028322805e-05,
+      "loss": 3.094,
       "step": 4266
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 8.359683990478516,
-      "learning_rate": 1.1448331161671847e-05,
-      "loss": 1.201,
+      "epoch": 0.54,
+      "grad_norm": 15.442558288574219,
+      "learning_rate": 1.643057356817136e-05,
+      "loss": 1.0509,
       "step": 4267
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 9.61418628692627,
-      "learning_rate": 1.1446326551067457e-05,
-      "loss": 2.0508,
+      "epoch": 0.54,
+      "grad_norm": 7.148504734039307,
+      "learning_rate": 1.6429736853114672e-05,
+      "loss": 1.6355,
       "step": 4268
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 12.493593215942383,
-      "learning_rate": 1.1444321940463066e-05,
-      "loss": 1.8421,
+      "epoch": 0.54,
+      "grad_norm": 12.161744117736816,
+      "learning_rate": 1.6428900138057986e-05,
+      "loss": 1.2154,
       "step": 4269
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 19.759090423583984,
-      "learning_rate": 1.1442317329858677e-05,
-      "loss": 2.7833,
+      "epoch": 0.54,
+      "grad_norm": 50.82536697387695,
+      "learning_rate": 1.64280634230013e-05,
+      "loss": 3.1307,
       "step": 4270
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 12.588993072509766,
-      "learning_rate": 1.1440312719254286e-05,
-      "loss": 1.1135,
+      "epoch": 0.54,
+      "grad_norm": 12.391793251037598,
+      "learning_rate": 1.642722670794461e-05,
+      "loss": 2.1355,
       "step": 4271
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 22.17607879638672,
-      "learning_rate": 1.1438308108649896e-05,
-      "loss": 2.3915,
+      "epoch": 0.54,
+      "grad_norm": 9.488656997680664,
+      "learning_rate": 1.6426389992887924e-05,
+      "loss": 1.2715,
       "step": 4272
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 11.478544235229492,
-      "learning_rate": 1.1436303498045504e-05,
-      "loss": 0.9938,
+      "epoch": 0.54,
+      "grad_norm": 12.979984283447266,
+      "learning_rate": 1.6425553277831237e-05,
+      "loss": 3.0628,
       "step": 4273
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 10.766488075256348,
-      "learning_rate": 1.1434298887441116e-05,
-      "loss": 1.0582,
+      "epoch": 0.54,
+      "grad_norm": 11.4072265625,
+      "learning_rate": 1.6424716562774547e-05,
+      "loss": 1.2632,
       "step": 4274
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 20.56522560119629,
-      "learning_rate": 1.1432294276836726e-05,
-      "loss": 1.6918,
+      "epoch": 0.54,
+      "grad_norm": 32.04029083251953,
+      "learning_rate": 1.642387984771786e-05,
+      "loss": 1.761,
       "step": 4275
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 9.447360038757324,
-      "learning_rate": 1.1430289666232334e-05,
-      "loss": 1.839,
+      "epoch": 0.54,
+      "grad_norm": 15.586896896362305,
+      "learning_rate": 1.6423043132661175e-05,
+      "loss": 4.1178,
       "step": 4276
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 9.256118774414062,
-      "learning_rate": 1.1428285055627946e-05,
-      "loss": 1.307,
+      "epoch": 0.54,
+      "grad_norm": 29.479475021362305,
+      "learning_rate": 1.642220641760449e-05,
+      "loss": 0.9489,
       "step": 4277
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 13.01025676727295,
-      "learning_rate": 1.1426280445023555e-05,
-      "loss": 1.6272,
+      "epoch": 0.54,
+      "grad_norm": 16.579774856567383,
+      "learning_rate": 1.64213697025478e-05,
+      "loss": 2.2051,
       "step": 4278
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 11.885642051696777,
-      "learning_rate": 1.1424275834419165e-05,
-      "loss": 0.5883,
+      "epoch": 0.54,
+      "grad_norm": 17.332616806030273,
+      "learning_rate": 1.6420532987491112e-05,
+      "loss": 2.4483,
       "step": 4279
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 13.292377471923828,
-      "learning_rate": 1.1422271223814776e-05,
-      "loss": 1.4237,
+      "epoch": 0.54,
+      "grad_norm": 36.76666259765625,
+      "learning_rate": 1.6419696272434426e-05,
+      "loss": 2.0568,
       "step": 4280
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 7.443901538848877,
-      "learning_rate": 1.1420266613210385e-05,
-      "loss": 0.9989,
+      "epoch": 0.54,
+      "grad_norm": 31.958080291748047,
+      "learning_rate": 1.6418859557377736e-05,
+      "loss": 4.2887,
       "step": 4281
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 83.39949035644531,
-      "learning_rate": 1.1418262002605995e-05,
-      "loss": 1.8078,
+      "epoch": 0.54,
+      "grad_norm": 19.1790771484375,
+      "learning_rate": 1.641802284232105e-05,
+      "loss": 2.8956,
       "step": 4282
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 12.701087951660156,
-      "learning_rate": 1.1416257392001603e-05,
-      "loss": 1.5752,
+      "epoch": 0.54,
+      "grad_norm": 4.67531681060791,
+      "learning_rate": 1.641718612726436e-05,
+      "loss": 0.6016,
       "step": 4283
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 58.16669464111328,
-      "learning_rate": 1.1414252781397215e-05,
-      "loss": 2.8532,
+      "epoch": 0.54,
+      "grad_norm": 29.13949203491211,
+      "learning_rate": 1.6416349412207674e-05,
+      "loss": 1.4015,
       "step": 4284
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 21.291078567504883,
-      "learning_rate": 1.1412248170792823e-05,
-      "loss": 1.2318,
+      "epoch": 0.54,
+      "grad_norm": 19.934898376464844,
+      "learning_rate": 1.6415512697150987e-05,
+      "loss": 0.4292,
       "step": 4285
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 18.41106605529785,
-      "learning_rate": 1.1410243560188434e-05,
-      "loss": 2.264,
+      "epoch": 0.54,
+      "grad_norm": 3.7015771865844727,
+      "learning_rate": 1.6414675982094298e-05,
+      "loss": 0.1326,
       "step": 4286
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 20.537443161010742,
-      "learning_rate": 1.1408238949584045e-05,
-      "loss": 2.6269,
+      "epoch": 0.54,
+      "grad_norm": 5.187589645385742,
+      "learning_rate": 1.641383926703761e-05,
+      "loss": 2.0533,
       "step": 4287
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 15.127706527709961,
-      "learning_rate": 1.1406234338979654e-05,
-      "loss": 1.9341,
+      "epoch": 0.54,
+      "grad_norm": 12.003097534179688,
+      "learning_rate": 1.641300255198092e-05,
+      "loss": 0.6824,
       "step": 4288
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 20.711387634277344,
-      "learning_rate": 1.1404229728375264e-05,
-      "loss": 1.7335,
+      "epoch": 0.54,
+      "grad_norm": 9.616144180297852,
+      "learning_rate": 1.6412165836924235e-05,
+      "loss": 0.643,
       "step": 4289
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 50.79159164428711,
-      "learning_rate": 1.1402225117770874e-05,
-      "loss": 2.1342,
+      "epoch": 0.54,
+      "grad_norm": 27.701128005981445,
+      "learning_rate": 1.641132912186755e-05,
+      "loss": 1.868,
       "step": 4290
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 29.805246353149414,
-      "learning_rate": 1.1400220507166484e-05,
-      "loss": 2.0907,
+      "epoch": 0.54,
+      "grad_norm": 17.69377326965332,
+      "learning_rate": 1.6410492406810863e-05,
+      "loss": 1.8682,
       "step": 4291
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 12.824877738952637,
-      "learning_rate": 1.1398215896562094e-05,
-      "loss": 1.1383,
+      "epoch": 0.54,
+      "grad_norm": 46.678592681884766,
+      "learning_rate": 1.6409655691754173e-05,
+      "loss": 3.8657,
       "step": 4292
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 7.294422149658203,
-      "learning_rate": 1.1396211285957704e-05,
-      "loss": 1.3885,
+      "epoch": 0.54,
+      "grad_norm": 14.063295364379883,
+      "learning_rate": 1.6408818976697486e-05,
+      "loss": 1.2708,
       "step": 4293
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 31.934511184692383,
-      "learning_rate": 1.1394206675353314e-05,
-      "loss": 2.1207,
+      "epoch": 0.54,
+      "grad_norm": 18.941165924072266,
+      "learning_rate": 1.64079822616408e-05,
+      "loss": 1.2093,
       "step": 4294
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 18.717044830322266,
-      "learning_rate": 1.1392202064748923e-05,
-      "loss": 1.4102,
+      "epoch": 0.54,
+      "grad_norm": 16.60546112060547,
+      "learning_rate": 1.640714554658411e-05,
+      "loss": 1.9364,
       "step": 4295
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 14.38543701171875,
-      "learning_rate": 1.1390197454144534e-05,
-      "loss": 1.9233,
+      "epoch": 0.54,
+      "grad_norm": 5.379274368286133,
+      "learning_rate": 1.6406308831527424e-05,
+      "loss": 0.3526,
       "step": 4296
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 20.979101181030273,
-      "learning_rate": 1.1388192843540143e-05,
-      "loss": 1.509,
+      "epoch": 0.54,
+      "grad_norm": 16.46821403503418,
+      "learning_rate": 1.6405472116470738e-05,
+      "loss": 1.3559,
       "step": 4297
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 16.605932235717773,
-      "learning_rate": 1.1386188232935753e-05,
-      "loss": 1.4354,
+      "epoch": 0.54,
+      "grad_norm": 8.183782577514648,
+      "learning_rate": 1.640463540141405e-05,
+      "loss": 2.0754,
       "step": 4298
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 9.505638122558594,
-      "learning_rate": 1.1384183622331365e-05,
-      "loss": 1.1145,
+      "epoch": 0.54,
+      "grad_norm": 11.519583702087402,
+      "learning_rate": 1.640379868635736e-05,
+      "loss": 1.7285,
       "step": 4299
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 15.347785949707031,
-      "learning_rate": 1.1382179011726973e-05,
-      "loss": 1.6882,
+      "epoch": 0.54,
+      "grad_norm": 6.20827579498291,
+      "learning_rate": 1.6402961971300675e-05,
+      "loss": 0.2051,
       "step": 4300
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 20.905508041381836,
-      "learning_rate": 1.1380174401122583e-05,
-      "loss": 1.5622,
+      "epoch": 0.54,
+      "grad_norm": 17.898536682128906,
+      "learning_rate": 1.640212525624399e-05,
+      "loss": 1.0117,
       "step": 4301
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 41.201194763183594,
-      "learning_rate": 1.1378169790518192e-05,
-      "loss": 3.4766,
+      "epoch": 0.54,
+      "grad_norm": 14.119236946105957,
+      "learning_rate": 1.64012885411873e-05,
+      "loss": 2.6476,
       "step": 4302
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 15.540892601013184,
-      "learning_rate": 1.1376165179913803e-05,
-      "loss": 1.4535,
+      "epoch": 0.54,
+      "grad_norm": 36.42744064331055,
+      "learning_rate": 1.6400451826130613e-05,
+      "loss": 3.5639,
       "step": 4303
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 24.336341857910156,
-      "learning_rate": 1.1374160569309412e-05,
-      "loss": 1.7003,
+      "epoch": 0.54,
+      "grad_norm": 36.01893615722656,
+      "learning_rate": 1.6399615111073926e-05,
+      "loss": 3.8818,
       "step": 4304
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 28.202116012573242,
-      "learning_rate": 1.1372155958705022e-05,
-      "loss": 1.2576,
+      "epoch": 0.54,
+      "grad_norm": 10.309459686279297,
+      "learning_rate": 1.639877839601724e-05,
+      "loss": 4.1736,
       "step": 4305
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 12.418293952941895,
-      "learning_rate": 1.1370151348100634e-05,
-      "loss": 1.8355,
+      "epoch": 0.54,
+      "grad_norm": 9.84698486328125,
+      "learning_rate": 1.639794168096055e-05,
+      "loss": 4.2613,
       "step": 4306
     },
     {
-      "epoch": 1.29,
-      "grad_norm": 35.69652557373047,
-      "learning_rate": 1.1368146737496242e-05,
-      "loss": 2.1658,
+      "epoch": 0.54,
+      "grad_norm": 20.667268753051758,
+      "learning_rate": 1.6397104965903864e-05,
+      "loss": 2.7518,
       "step": 4307
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 17.411029815673828,
-      "learning_rate": 1.1366142126891852e-05,
-      "loss": 2.4983,
+      "epoch": 0.54,
+      "grad_norm": 14.868730545043945,
+      "learning_rate": 1.6396268250847178e-05,
+      "loss": 1.341,
       "step": 4308
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 10.031928062438965,
-      "learning_rate": 1.136413751628746e-05,
-      "loss": 1.7212,
+      "epoch": 0.54,
+      "grad_norm": 13.767081260681152,
+      "learning_rate": 1.6395431535790488e-05,
+      "loss": 1.7816,
       "step": 4309
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 11.622937202453613,
-      "learning_rate": 1.1362132905683072e-05,
-      "loss": 1.4201,
+      "epoch": 0.54,
+      "grad_norm": 6.044618606567383,
+      "learning_rate": 1.63945948207338e-05,
+      "loss": 0.4184,
       "step": 4310
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 19.42033576965332,
-      "learning_rate": 1.1360128295078682e-05,
-      "loss": 1.8876,
+      "epoch": 0.54,
+      "grad_norm": 15.502923011779785,
+      "learning_rate": 1.6393758105677112e-05,
+      "loss": 1.5554,
       "step": 4311
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 24.995628356933594,
-      "learning_rate": 1.135812368447429e-05,
-      "loss": 1.682,
+      "epoch": 0.54,
+      "grad_norm": 24.838972091674805,
+      "learning_rate": 1.6392921390620425e-05,
+      "loss": 3.4248,
       "step": 4312
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 11.356565475463867,
-      "learning_rate": 1.1356119073869902e-05,
-      "loss": 1.3858,
+      "epoch": 0.54,
+      "grad_norm": 13.437346458435059,
+      "learning_rate": 1.639208467556374e-05,
+      "loss": 1.5421,
       "step": 4313
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 18.608964920043945,
-      "learning_rate": 1.1354114463265511e-05,
-      "loss": 1.2763,
+      "epoch": 0.54,
+      "grad_norm": 19.534582138061523,
+      "learning_rate": 1.639124796050705e-05,
+      "loss": 2.9816,
       "step": 4314
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 16.34697723388672,
-      "learning_rate": 1.1352109852661121e-05,
-      "loss": 1.2931,
+      "epoch": 0.54,
+      "grad_norm": 15.092426300048828,
+      "learning_rate": 1.6390411245450363e-05,
+      "loss": 1.3546,
       "step": 4315
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 17.903200149536133,
-      "learning_rate": 1.1350105242056731e-05,
-      "loss": 1.8466,
+      "epoch": 0.54,
+      "grad_norm": 21.328441619873047,
+      "learning_rate": 1.6389574530393673e-05,
+      "loss": 3.1882,
       "step": 4316
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 7.10983419418335,
-      "learning_rate": 1.1348100631452341e-05,
-      "loss": 0.967,
+      "epoch": 0.54,
+      "grad_norm": 13.988789558410645,
+      "learning_rate": 1.6388737815336987e-05,
+      "loss": 0.9068,
       "step": 4317
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 13.738561630249023,
-      "learning_rate": 1.1346096020847951e-05,
-      "loss": 1.4251,
+      "epoch": 0.54,
+      "grad_norm": 24.274625778198242,
+      "learning_rate": 1.63879011002803e-05,
+      "loss": 1.8792,
       "step": 4318
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 14.061918258666992,
-      "learning_rate": 1.1344091410243561e-05,
-      "loss": 1.8931,
+      "epoch": 0.54,
+      "grad_norm": 27.35269546508789,
+      "learning_rate": 1.6387064385223614e-05,
+      "loss": 4.0709,
       "step": 4319
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 28.247364044189453,
-      "learning_rate": 1.1342086799639171e-05,
-      "loss": 2.5285,
-      "step": 4320
-    },
-    {
-      "epoch": 1.3,
-      "eval_loss": 0.21285276114940643,
-      "eval_runtime": 43.3572,
-      "eval_samples_per_second": 34.112,
-      "eval_steps_per_second": 34.112,
+      "epoch": 0.54,
+      "grad_norm": 18.570083618164062,
+      "learning_rate": 1.6386227670166924e-05,
+      "loss": 2.2045,
       "step": 4320
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 11.178879737854004,
-      "learning_rate": 1.134008218903478e-05,
-      "loss": 1.5172,
+      "epoch": 0.54,
+      "grad_norm": 14.518431663513184,
+      "learning_rate": 1.6385390955110238e-05,
+      "loss": 1.7629,
       "step": 4321
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 16.7074031829834,
-      "learning_rate": 1.1338077578430392e-05,
-      "loss": 1.3458,
+      "epoch": 0.54,
+      "grad_norm": 150.36700439453125,
+      "learning_rate": 1.6384554240053552e-05,
+      "loss": 0.7615,
       "step": 4322
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 21.143476486206055,
-      "learning_rate": 1.1336072967826002e-05,
-      "loss": 2.4362,
+      "epoch": 0.54,
+      "grad_norm": 8.598651885986328,
+      "learning_rate": 1.6383717524996862e-05,
+      "loss": 0.6577,
       "step": 4323
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 22.83983039855957,
-      "learning_rate": 1.133406835722161e-05,
-      "loss": 2.1729,
+      "epoch": 0.54,
+      "grad_norm": 6.166103363037109,
+      "learning_rate": 1.6382880809940176e-05,
+      "loss": 0.7035,
       "step": 4324
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 16.511369705200195,
-      "learning_rate": 1.1332063746617222e-05,
-      "loss": 1.3165,
+      "epoch": 0.54,
+      "grad_norm": 15.817057609558105,
+      "learning_rate": 1.638204409488349e-05,
+      "loss": 1.1081,
       "step": 4325
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 16.411027908325195,
-      "learning_rate": 1.133005913601283e-05,
-      "loss": 2.1342,
+      "epoch": 0.54,
+      "grad_norm": 14.033500671386719,
+      "learning_rate": 1.6381207379826803e-05,
+      "loss": 0.422,
       "step": 4326
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 16.996347427368164,
-      "learning_rate": 1.132805452540844e-05,
-      "loss": 1.9439,
+      "epoch": 0.54,
+      "grad_norm": 20.49435806274414,
+      "learning_rate": 1.6380370664770113e-05,
+      "loss": 1.372,
       "step": 4327
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 27.14801788330078,
-      "learning_rate": 1.1326049914804049e-05,
-      "loss": 2.3698,
+      "epoch": 0.54,
+      "grad_norm": 3.773137331008911,
+      "learning_rate": 1.6379533949713427e-05,
+      "loss": 1.6234,
       "step": 4328
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 10.485852241516113,
-      "learning_rate": 1.132404530419966e-05,
-      "loss": 1.1316,
+      "epoch": 0.54,
+      "grad_norm": 8.239606857299805,
+      "learning_rate": 1.637869723465674e-05,
+      "loss": 1.6802,
       "step": 4329
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 48.32770538330078,
-      "learning_rate": 1.132204069359527e-05,
-      "loss": 1.8271,
+      "epoch": 0.54,
+      "grad_norm": 12.494515419006348,
+      "learning_rate": 1.637786051960005e-05,
+      "loss": 1.7986,
       "step": 4330
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 68.1049575805664,
-      "learning_rate": 1.1320036082990879e-05,
-      "loss": 2.5344,
+      "epoch": 0.54,
+      "grad_norm": 7.956020832061768,
+      "learning_rate": 1.6377023804543364e-05,
+      "loss": 0.1914,
       "step": 4331
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 22.41739273071289,
-      "learning_rate": 1.131803147238649e-05,
-      "loss": 2.1376,
+      "epoch": 0.54,
+      "grad_norm": 12.799524307250977,
+      "learning_rate": 1.6376187089486678e-05,
+      "loss": 1.3183,
       "step": 4332
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 61.55850601196289,
-      "learning_rate": 1.1316026861782099e-05,
-      "loss": 3.21,
+      "epoch": 0.54,
+      "grad_norm": 11.716842651367188,
+      "learning_rate": 1.6375350374429992e-05,
+      "loss": 2.1588,
       "step": 4333
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 13.802007675170898,
-      "learning_rate": 1.1314022251177709e-05,
-      "loss": 1.7577,
+      "epoch": 0.54,
+      "grad_norm": 18.58441162109375,
+      "learning_rate": 1.6374513659373302e-05,
+      "loss": 3.564,
       "step": 4334
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 16.4052734375,
-      "learning_rate": 1.1312017640573321e-05,
-      "loss": 1.2041,
+      "epoch": 0.54,
+      "grad_norm": 12.418187141418457,
+      "learning_rate": 1.6373676944316616e-05,
+      "loss": 0.9168,
       "step": 4335
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 26.579936981201172,
-      "learning_rate": 1.131001302996893e-05,
-      "loss": 1.6785,
+      "epoch": 0.54,
+      "grad_norm": 19.849315643310547,
+      "learning_rate": 1.6372840229259926e-05,
+      "loss": 3.2139,
       "step": 4336
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 24.084991455078125,
-      "learning_rate": 1.130800841936454e-05,
-      "loss": 1.5441,
+      "epoch": 0.54,
+      "grad_norm": 11.36044979095459,
+      "learning_rate": 1.637200351420324e-05,
+      "loss": 0.7428,
       "step": 4337
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 21.47726058959961,
-      "learning_rate": 1.130600380876015e-05,
-      "loss": 1.9604,
+      "epoch": 0.54,
+      "grad_norm": 6.952395439147949,
+      "learning_rate": 1.6371166799146553e-05,
+      "loss": 2.1326,
       "step": 4338
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 38.121421813964844,
-      "learning_rate": 1.130399919815576e-05,
-      "loss": 2.2498,
+      "epoch": 0.54,
+      "grad_norm": 17.800748825073242,
+      "learning_rate": 1.6370330084089863e-05,
+      "loss": 1.45,
       "step": 4339
     },
     {
-      "epoch": 1.3,
-      "grad_norm": 23.063236236572266,
-      "learning_rate": 1.1301994587551368e-05,
-      "loss": 1.8673,
+      "epoch": 0.54,
+      "grad_norm": 8.255887985229492,
+      "learning_rate": 1.6369493369033177e-05,
+      "loss": 1.0562,
       "step": 4340
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 11.506025314331055,
-      "learning_rate": 1.129998997694698e-05,
-      "loss": 1.3873,
+      "epoch": 0.54,
+      "grad_norm": 9.223470687866211,
+      "learning_rate": 1.6368656653976487e-05,
+      "loss": 0.4263,
       "step": 4341
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 11.516486167907715,
-      "learning_rate": 1.129798536634259e-05,
-      "loss": 1.2646,
+      "epoch": 0.54,
+      "grad_norm": 46.6847038269043,
+      "learning_rate": 1.63678199389198e-05,
+      "loss": 1.5089,
       "step": 4342
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 15.611605644226074,
-      "learning_rate": 1.1295980755738198e-05,
-      "loss": 1.2307,
+      "epoch": 0.55,
+      "grad_norm": 12.898388862609863,
+      "learning_rate": 1.6366983223863115e-05,
+      "loss": 2.1125,
       "step": 4343
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 17.53089141845703,
-      "learning_rate": 1.129397614513381e-05,
-      "loss": 1.2414,
+      "epoch": 0.55,
+      "grad_norm": 4.78568172454834,
+      "learning_rate": 1.6366146508806425e-05,
+      "loss": 0.3338,
       "step": 4344
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 14.397602081298828,
-      "learning_rate": 1.1291971534529418e-05,
-      "loss": 1.3744,
+      "epoch": 0.55,
+      "grad_norm": 16.7759952545166,
+      "learning_rate": 1.636530979374974e-05,
+      "loss": 2.0696,
       "step": 4345
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 23.23992156982422,
-      "learning_rate": 1.1289966923925028e-05,
-      "loss": 1.8092,
+      "epoch": 0.55,
+      "grad_norm": 59.63092041015625,
+      "learning_rate": 1.6364473078693052e-05,
+      "loss": 2.7622,
       "step": 4346
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 15.50413990020752,
-      "learning_rate": 1.1287962313320637e-05,
-      "loss": 1.819,
+      "epoch": 0.55,
+      "grad_norm": 26.8160400390625,
+      "learning_rate": 1.6363636363636366e-05,
+      "loss": 1.609,
       "step": 4347
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 12.4719877243042,
-      "learning_rate": 1.1285957702716249e-05,
-      "loss": 0.959,
+      "epoch": 0.55,
+      "grad_norm": 9.036898612976074,
+      "learning_rate": 1.6362799648579676e-05,
+      "loss": 1.0178,
       "step": 4348
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 8.75976276397705,
-      "learning_rate": 1.1283953092111859e-05,
-      "loss": 1.0429,
+      "epoch": 0.55,
+      "grad_norm": 18.46747398376465,
+      "learning_rate": 1.636196293352299e-05,
+      "loss": 2.2952,
       "step": 4349
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 34.667945861816406,
-      "learning_rate": 1.1281948481507467e-05,
-      "loss": 2.2953,
+      "epoch": 0.55,
+      "grad_norm": 7.6316447257995605,
+      "learning_rate": 1.6361126218466303e-05,
+      "loss": 1.4112,
       "step": 4350
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 15.772024154663086,
-      "learning_rate": 1.1279943870903079e-05,
-      "loss": 1.67,
+      "epoch": 0.55,
+      "grad_norm": 22.83441734313965,
+      "learning_rate": 1.6360289503409614e-05,
+      "loss": 2.9932,
       "step": 4351
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 19.514604568481445,
-      "learning_rate": 1.1277939260298687e-05,
-      "loss": 2.0143,
+      "epoch": 0.55,
+      "grad_norm": 11.766851425170898,
+      "learning_rate": 1.6359452788352927e-05,
+      "loss": 0.7483,
       "step": 4352
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 22.16571807861328,
-      "learning_rate": 1.1275934649694297e-05,
-      "loss": 1.2593,
+      "epoch": 0.55,
+      "grad_norm": 11.485483169555664,
+      "learning_rate": 1.635861607329624e-05,
+      "loss": 2.5164,
       "step": 4353
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 17.157798767089844,
-      "learning_rate": 1.1273930039089909e-05,
-      "loss": 1.6187,
+      "epoch": 0.55,
+      "grad_norm": 16.853513717651367,
+      "learning_rate": 1.6357779358239555e-05,
+      "loss": 2.0076,
       "step": 4354
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 33.16340637207031,
-      "learning_rate": 1.1271925428485518e-05,
-      "loss": 2.0045,
+      "epoch": 0.55,
+      "grad_norm": 15.183171272277832,
+      "learning_rate": 1.6356942643182865e-05,
+      "loss": 1.3851,
       "step": 4355
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 14.194746971130371,
-      "learning_rate": 1.1269920817881128e-05,
-      "loss": 1.671,
+      "epoch": 0.55,
+      "grad_norm": 9.89082145690918,
+      "learning_rate": 1.635610592812618e-05,
+      "loss": 0.6442,
       "step": 4356
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 15.4343900680542,
-      "learning_rate": 1.1267916207276736e-05,
-      "loss": 1.2274,
+      "epoch": 0.55,
+      "grad_norm": 24.128459930419922,
+      "learning_rate": 1.6355269213069492e-05,
+      "loss": 1.1141,
       "step": 4357
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 21.623807907104492,
-      "learning_rate": 1.1265911596672348e-05,
-      "loss": 2.5802,
+      "epoch": 0.55,
+      "grad_norm": 22.751218795776367,
+      "learning_rate": 1.6354432498012802e-05,
+      "loss": 1.3387,
       "step": 4358
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 12.904182434082031,
-      "learning_rate": 1.1263906986067956e-05,
-      "loss": 1.4189,
+      "epoch": 0.55,
+      "grad_norm": 20.522287368774414,
+      "learning_rate": 1.6353595782956116e-05,
+      "loss": 1.8312,
       "step": 4359
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 16.601791381835938,
-      "learning_rate": 1.1261902375463566e-05,
-      "loss": 2.787,
+      "epoch": 0.55,
+      "grad_norm": 39.677337646484375,
+      "learning_rate": 1.635275906789943e-05,
+      "loss": 2.4149,
       "step": 4360
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 12.915867805480957,
-      "learning_rate": 1.1259897764859178e-05,
-      "loss": 1.5663,
+      "epoch": 0.55,
+      "grad_norm": 17.512928009033203,
+      "learning_rate": 1.6351922352842743e-05,
+      "loss": 0.6399,
       "step": 4361
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 17.556509017944336,
-      "learning_rate": 1.1257893154254786e-05,
-      "loss": 1.8852,
+      "epoch": 0.55,
+      "grad_norm": 13.060235977172852,
+      "learning_rate": 1.6351085637786054e-05,
+      "loss": 2.7061,
       "step": 4362
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 23.893291473388672,
-      "learning_rate": 1.1255888543650397e-05,
-      "loss": 1.4539,
+      "epoch": 0.55,
+      "grad_norm": 53.84843063354492,
+      "learning_rate": 1.6350248922729367e-05,
+      "loss": 2.6682,
       "step": 4363
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 11.98607063293457,
-      "learning_rate": 1.1253883933046007e-05,
-      "loss": 1.7452,
+      "epoch": 0.55,
+      "grad_norm": 10.398431777954102,
+      "learning_rate": 1.6349412207672678e-05,
+      "loss": 1.0722,
       "step": 4364
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 13.887165069580078,
-      "learning_rate": 1.1251879322441617e-05,
-      "loss": 1.6717,
+      "epoch": 0.55,
+      "grad_norm": 17.36014175415039,
+      "learning_rate": 1.634857549261599e-05,
+      "loss": 2.3584,
       "step": 4365
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 7.588750839233398,
-      "learning_rate": 1.1249874711837227e-05,
-      "loss": 0.8518,
+      "epoch": 0.55,
+      "grad_norm": 8.186704635620117,
+      "learning_rate": 1.6347738777559305e-05,
+      "loss": 0.9661,
       "step": 4366
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 27.59679412841797,
-      "learning_rate": 1.1247870101232837e-05,
-      "loss": 2.483,
+      "epoch": 0.55,
+      "grad_norm": 100.78430938720703,
+      "learning_rate": 1.6346902062502615e-05,
+      "loss": 2.5819,
       "step": 4367
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 19.16243553161621,
-      "learning_rate": 1.1245865490628447e-05,
-      "loss": 1.4322,
+      "epoch": 0.55,
+      "grad_norm": 117.7176513671875,
+      "learning_rate": 1.634606534744593e-05,
+      "loss": 1.7862,
       "step": 4368
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 16.66512107849121,
-      "learning_rate": 1.1243860880024055e-05,
-      "loss": 1.9586,
+      "epoch": 0.55,
+      "grad_norm": 9.360919952392578,
+      "learning_rate": 1.634522863238924e-05,
+      "loss": 1.8567,
       "step": 4369
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 17.893173217773438,
-      "learning_rate": 1.1241856269419667e-05,
-      "loss": 2.8739,
+      "epoch": 0.55,
+      "grad_norm": 12.167970657348633,
+      "learning_rate": 1.6344391917332553e-05,
+      "loss": 1.3236,
       "step": 4370
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 16.915111541748047,
-      "learning_rate": 1.1239851658815275e-05,
-      "loss": 1.3351,
+      "epoch": 0.55,
+      "grad_norm": 5.873944282531738,
+      "learning_rate": 1.6343555202275866e-05,
+      "loss": 0.6095,
       "step": 4371
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 20.749107360839844,
-      "learning_rate": 1.1237847048210886e-05,
-      "loss": 2.3157,
+      "epoch": 0.55,
+      "grad_norm": 21.591901779174805,
+      "learning_rate": 1.6342718487219177e-05,
+      "loss": 1.4569,
       "step": 4372
     },
     {
-      "epoch": 1.31,
-      "grad_norm": 33.895973205566406,
-      "learning_rate": 1.1235842437606497e-05,
-      "loss": 3.2345,
+      "epoch": 0.55,
+      "grad_norm": 6.27478551864624,
+      "learning_rate": 1.634188177216249e-05,
+      "loss": 0.9596,
       "step": 4373
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 16.17300033569336,
-      "learning_rate": 1.1233837827002106e-05,
-      "loss": 1.0639,
+      "epoch": 0.55,
+      "grad_norm": 28.359424591064453,
+      "learning_rate": 1.6341045057105804e-05,
+      "loss": 1.5756,
       "step": 4374
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 11.130797386169434,
-      "learning_rate": 1.1231833216397716e-05,
-      "loss": 2.0059,
+      "epoch": 0.55,
+      "grad_norm": 17.1900577545166,
+      "learning_rate": 1.6340208342049118e-05,
+      "loss": 1.7516,
       "step": 4375
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 12.532829284667969,
-      "learning_rate": 1.1229828605793324e-05,
-      "loss": 1.1101,
+      "epoch": 0.55,
+      "grad_norm": 12.922842979431152,
+      "learning_rate": 1.6339371626992428e-05,
+      "loss": 2.1068,
       "step": 4376
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 10.261481285095215,
-      "learning_rate": 1.1227823995188936e-05,
-      "loss": 1.7796,
+      "epoch": 0.55,
+      "grad_norm": 24.80573844909668,
+      "learning_rate": 1.633853491193574e-05,
+      "loss": 0.9404,
       "step": 4377
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 18.463781356811523,
-      "learning_rate": 1.1225819384584546e-05,
-      "loss": 1.9479,
+      "epoch": 0.55,
+      "grad_norm": 42.843414306640625,
+      "learning_rate": 1.6337698196879055e-05,
+      "loss": 2.0427,
       "step": 4378
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 12.4391450881958,
-      "learning_rate": 1.1223814773980154e-05,
-      "loss": 2.0404,
+      "epoch": 0.55,
+      "grad_norm": 9.03561782836914,
+      "learning_rate": 1.6336861481822365e-05,
+      "loss": 1.7024,
       "step": 4379
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 15.823043823242188,
-      "learning_rate": 1.1221810163375766e-05,
-      "loss": 1.4798,
+      "epoch": 0.55,
+      "grad_norm": 25.199548721313477,
+      "learning_rate": 1.633602476676568e-05,
+      "loss": 1.7186,
       "step": 4380
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 19.075504302978516,
-      "learning_rate": 1.1219805552771375e-05,
-      "loss": 1.712,
+      "epoch": 0.55,
+      "grad_norm": 26.4344482421875,
+      "learning_rate": 1.6335188051708993e-05,
+      "loss": 3.121,
       "step": 4381
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 13.849617958068848,
-      "learning_rate": 1.1217800942166985e-05,
-      "loss": 1.8711,
+      "epoch": 0.55,
+      "grad_norm": 11.396100044250488,
+      "learning_rate": 1.6334351336652306e-05,
+      "loss": 1.4493,
       "step": 4382
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 17.942249298095703,
-      "learning_rate": 1.1215796331562593e-05,
-      "loss": 2.6901,
+      "epoch": 0.55,
+      "grad_norm": 13.396303176879883,
+      "learning_rate": 1.6333514621595617e-05,
+      "loss": 1.7732,
       "step": 4383
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 9.185188293457031,
-      "learning_rate": 1.1213791720958205e-05,
-      "loss": 1.0339,
+      "epoch": 0.55,
+      "grad_norm": 48.93893814086914,
+      "learning_rate": 1.633267790653893e-05,
+      "loss": 1.8056,
       "step": 4384
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 12.996100425720215,
-      "learning_rate": 1.1211787110353815e-05,
-      "loss": 1.7259,
+      "epoch": 0.55,
+      "grad_norm": 11.110664367675781,
+      "learning_rate": 1.6331841191482244e-05,
+      "loss": 0.7474,
       "step": 4385
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 31.83272933959961,
-      "learning_rate": 1.1209782499749425e-05,
-      "loss": 2.1627,
+      "epoch": 0.55,
+      "grad_norm": 11.114632606506348,
+      "learning_rate": 1.6331004476425554e-05,
+      "loss": 3.6277,
       "step": 4386
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 11.160262107849121,
-      "learning_rate": 1.1207777889145035e-05,
-      "loss": 1.7325,
+      "epoch": 0.55,
+      "grad_norm": 12.204710006713867,
+      "learning_rate": 1.6330167761368868e-05,
+      "loss": 2.2624,
       "step": 4387
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 19.457317352294922,
-      "learning_rate": 1.1205773278540644e-05,
-      "loss": 1.6398,
+      "epoch": 0.55,
+      "grad_norm": 5.68466329574585,
+      "learning_rate": 1.632933104631218e-05,
+      "loss": 0.5796,
       "step": 4388
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 15.514799118041992,
-      "learning_rate": 1.1203768667936255e-05,
-      "loss": 1.5179,
+      "epoch": 0.55,
+      "grad_norm": 8.876830101013184,
+      "learning_rate": 1.6328494331255492e-05,
+      "loss": 2.9876,
       "step": 4389
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 55.15019607543945,
-      "learning_rate": 1.1201764057331864e-05,
-      "loss": 4.2946,
+      "epoch": 0.55,
+      "grad_norm": 13.933097839355469,
+      "learning_rate": 1.6327657616198805e-05,
+      "loss": 2.1435,
       "step": 4390
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 17.135194778442383,
-      "learning_rate": 1.1199759446727474e-05,
-      "loss": 2.0419,
+      "epoch": 0.55,
+      "grad_norm": 20.075345993041992,
+      "learning_rate": 1.632682090114212e-05,
+      "loss": 2.5147,
       "step": 4391
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 11.672941207885742,
-      "learning_rate": 1.1197754836123086e-05,
-      "loss": 1.4943,
+      "epoch": 0.55,
+      "grad_norm": 20.869224548339844,
+      "learning_rate": 1.632598418608543e-05,
+      "loss": 2.4408,
       "step": 4392
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 27.473955154418945,
-      "learning_rate": 1.1195750225518694e-05,
-      "loss": 2.4105,
+      "epoch": 0.55,
+      "grad_norm": 17.259469985961914,
+      "learning_rate": 1.6325147471028743e-05,
+      "loss": 2.3515,
       "step": 4393
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 12.319040298461914,
-      "learning_rate": 1.1193745614914304e-05,
-      "loss": 1.1325,
+      "epoch": 0.55,
+      "grad_norm": 7.999358654022217,
+      "learning_rate": 1.6324310755972053e-05,
+      "loss": 0.331,
       "step": 4394
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 19.93411636352539,
-      "learning_rate": 1.1191741004309912e-05,
-      "loss": 2.4709,
+      "epoch": 0.55,
+      "grad_norm": 18.779560089111328,
+      "learning_rate": 1.6323474040915367e-05,
+      "loss": 1.2501,
       "step": 4395
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 25.42889976501465,
-      "learning_rate": 1.1189736393705524e-05,
-      "loss": 2.7937,
+      "epoch": 0.55,
+      "grad_norm": 6.969943046569824,
+      "learning_rate": 1.632263732585868e-05,
+      "loss": 0.4008,
       "step": 4396
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 31.769067764282227,
-      "learning_rate": 1.1187731783101134e-05,
-      "loss": 2.7676,
+      "epoch": 0.55,
+      "grad_norm": 17.99399757385254,
+      "learning_rate": 1.632180061080199e-05,
+      "loss": 1.9365,
       "step": 4397
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 21.329421997070312,
-      "learning_rate": 1.1185727172496743e-05,
-      "loss": 2.0521,
+      "epoch": 0.55,
+      "grad_norm": 32.63505554199219,
+      "learning_rate": 1.6320963895745304e-05,
+      "loss": 3.0524,
       "step": 4398
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 17.158628463745117,
-      "learning_rate": 1.1183722561892354e-05,
-      "loss": 1.5289,
+      "epoch": 0.55,
+      "grad_norm": 28.14476203918457,
+      "learning_rate": 1.6320127180688618e-05,
+      "loss": 2.3284,
       "step": 4399
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 13.490479469299316,
-      "learning_rate": 1.1181717951287963e-05,
-      "loss": 1.4183,
+      "epoch": 0.55,
+      "grad_norm": 12.068603515625,
+      "learning_rate": 1.6319290465631928e-05,
+      "loss": 3.1424,
       "step": 4400
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 8.754962921142578,
-      "learning_rate": 1.1179713340683573e-05,
-      "loss": 1.2418,
+      "epoch": 0.55,
+      "eval_loss": 0.1459905505180359,
+      "eval_runtime": 94.1606,
+      "eval_samples_per_second": 37.617,
+      "eval_steps_per_second": 37.617,
+      "step": 4400
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.126096725463867,
+      "learning_rate": 1.6318453750575242e-05,
+      "loss": 2.3201,
       "step": 4401
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 11.73357105255127,
-      "learning_rate": 1.1177708730079181e-05,
-      "loss": 1.3139,
+      "epoch": 0.55,
+      "grad_norm": 13.402315139770508,
+      "learning_rate": 1.6317617035518556e-05,
+      "loss": 1.3034,
       "step": 4402
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 13.230740547180176,
-      "learning_rate": 1.1175704119474793e-05,
-      "loss": 1.6772,
+      "epoch": 0.55,
+      "grad_norm": 11.152032852172852,
+      "learning_rate": 1.631678032046187e-05,
+      "loss": 0.9552,
       "step": 4403
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 17.465307235717773,
-      "learning_rate": 1.1173699508870403e-05,
-      "loss": 1.6562,
+      "epoch": 0.55,
+      "grad_norm": 20.30156898498535,
+      "learning_rate": 1.631594360540518e-05,
+      "loss": 1.2899,
       "step": 4404
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 32.38056945800781,
-      "learning_rate": 1.1171694898266012e-05,
-      "loss": 2.3572,
+      "epoch": 0.55,
+      "grad_norm": 28.718151092529297,
+      "learning_rate": 1.6315106890348493e-05,
+      "loss": 3.3766,
       "step": 4405
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 14.467717170715332,
-      "learning_rate": 1.1169690287661623e-05,
-      "loss": 1.2874,
+      "epoch": 0.55,
+      "grad_norm": 17.650142669677734,
+      "learning_rate": 1.6314270175291807e-05,
+      "loss": 2.6818,
       "step": 4406
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 17.778472900390625,
-      "learning_rate": 1.1167685677057232e-05,
-      "loss": 1.9954,
+      "epoch": 0.55,
+      "grad_norm": 15.46431827545166,
+      "learning_rate": 1.6313433460235117e-05,
+      "loss": 2.8227,
       "step": 4407
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 8.146272659301758,
-      "learning_rate": 1.1165681066452842e-05,
-      "loss": 0.6304,
+      "epoch": 0.55,
+      "grad_norm": 31.419979095458984,
+      "learning_rate": 1.631259674517843e-05,
+      "loss": 2.973,
       "step": 4408
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 18.66156005859375,
-      "learning_rate": 1.1163676455848454e-05,
-      "loss": 2.0366,
+      "epoch": 0.55,
+      "grad_norm": 12.229551315307617,
+      "learning_rate": 1.6311760030121744e-05,
+      "loss": 2.2177,
       "step": 4409
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 13.149731636047363,
-      "learning_rate": 1.1161671845244062e-05,
-      "loss": 1.5874,
+      "epoch": 0.55,
+      "grad_norm": 16.591360092163086,
+      "learning_rate": 1.6310923315065058e-05,
+      "loss": 2.3748,
       "step": 4410
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 15.526867866516113,
-      "learning_rate": 1.1159667234639672e-05,
-      "loss": 1.4642,
+      "epoch": 0.55,
+      "grad_norm": 12.206480979919434,
+      "learning_rate": 1.6310086600008368e-05,
+      "loss": 2.0689,
       "step": 4411
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 27.83363151550293,
-      "learning_rate": 1.1157662624035282e-05,
-      "loss": 2.0596,
+      "epoch": 0.55,
+      "grad_norm": 19.842300415039062,
+      "learning_rate": 1.6309249884951682e-05,
+      "loss": 2.4016,
       "step": 4412
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 82.84793090820312,
-      "learning_rate": 1.1155658013430892e-05,
-      "loss": 1.8469,
+      "epoch": 0.55,
+      "grad_norm": 33.59919357299805,
+      "learning_rate": 1.6308413169894996e-05,
+      "loss": 2.0409,
       "step": 4413
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 16.59908676147461,
-      "learning_rate": 1.11536534028265e-05,
-      "loss": 1.9362,
+      "epoch": 0.55,
+      "grad_norm": 14.440958023071289,
+      "learning_rate": 1.6307576454838306e-05,
+      "loss": 2.1244,
       "step": 4414
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 8.759167671203613,
-      "learning_rate": 1.1151648792222112e-05,
-      "loss": 2.0726,
+      "epoch": 0.55,
+      "grad_norm": 17.500349044799805,
+      "learning_rate": 1.630673973978162e-05,
+      "loss": 1.1437,
       "step": 4415
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 32.26197814941406,
-      "learning_rate": 1.1149644181617723e-05,
-      "loss": 2.0811,
+      "epoch": 0.55,
+      "grad_norm": 9.321260452270508,
+      "learning_rate": 1.6305903024724933e-05,
+      "loss": 1.6508,
       "step": 4416
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 8.401446342468262,
-      "learning_rate": 1.1147639571013331e-05,
-      "loss": 1.3336,
+      "epoch": 0.55,
+      "grad_norm": 15.433318138122559,
+      "learning_rate": 1.6305066309668243e-05,
+      "loss": 1.8476,
       "step": 4417
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 10.307430267333984,
-      "learning_rate": 1.1145634960408943e-05,
-      "loss": 1.1188,
+      "epoch": 0.55,
+      "grad_norm": 10.021492958068848,
+      "learning_rate": 1.6304229594611557e-05,
+      "loss": 1.6934,
       "step": 4418
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 16.372766494750977,
-      "learning_rate": 1.1143630349804551e-05,
-      "loss": 1.8886,
+      "epoch": 0.55,
+      "grad_norm": 17.209562301635742,
+      "learning_rate": 1.630339287955487e-05,
+      "loss": 1.5899,
       "step": 4419
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 39.85802459716797,
-      "learning_rate": 1.1141625739200161e-05,
-      "loss": 2.3996,
+      "epoch": 0.55,
+      "grad_norm": 4.5861005783081055,
+      "learning_rate": 1.630255616449818e-05,
+      "loss": 0.5776,
       "step": 4420
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 14.044843673706055,
-      "learning_rate": 1.1139621128595773e-05,
-      "loss": 1.4335,
+      "epoch": 0.55,
+      "grad_norm": 15.866487503051758,
+      "learning_rate": 1.6301719449441495e-05,
+      "loss": 2.145,
       "step": 4421
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 23.040878295898438,
-      "learning_rate": 1.1137616517991381e-05,
-      "loss": 1.8451,
+      "epoch": 0.55,
+      "grad_norm": 12.779694557189941,
+      "learning_rate": 1.6300882734384805e-05,
+      "loss": 1.1097,
       "step": 4422
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 17.075387954711914,
-      "learning_rate": 1.1135611907386991e-05,
-      "loss": 2.2248,
+      "epoch": 0.56,
+      "grad_norm": 9.33056926727295,
+      "learning_rate": 1.630004601932812e-05,
+      "loss": 2.8792,
       "step": 4423
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 110.57589721679688,
-      "learning_rate": 1.11336072967826e-05,
-      "loss": 1.5452,
+      "epoch": 0.56,
+      "grad_norm": 15.881376266479492,
+      "learning_rate": 1.6299209304271432e-05,
+      "loss": 1.7005,
       "step": 4424
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 15.424068450927734,
-      "learning_rate": 1.1131602686178212e-05,
-      "loss": 1.5217,
+      "epoch": 0.56,
+      "grad_norm": 12.460254669189453,
+      "learning_rate": 1.6298372589214742e-05,
+      "loss": 1.3359,
       "step": 4425
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 23.823379516601562,
-      "learning_rate": 1.112959807557382e-05,
-      "loss": 1.2601,
+      "epoch": 0.56,
+      "grad_norm": 10.446908950805664,
+      "learning_rate": 1.6297535874158056e-05,
+      "loss": 2.2929,
       "step": 4426
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 10.928523063659668,
-      "learning_rate": 1.112759346496943e-05,
-      "loss": 1.4933,
+      "epoch": 0.56,
+      "grad_norm": 14.90433406829834,
+      "learning_rate": 1.629669915910137e-05,
+      "loss": 1.2793,
       "step": 4427
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 13.319860458374023,
-      "learning_rate": 1.1125588854365042e-05,
-      "loss": 1.4411,
+      "epoch": 0.56,
+      "grad_norm": 21.37445068359375,
+      "learning_rate": 1.629586244404468e-05,
+      "loss": 3.7241,
       "step": 4428
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 11.283712387084961,
-      "learning_rate": 1.112358424376065e-05,
-      "loss": 2.1932,
+      "epoch": 0.56,
+      "grad_norm": 3.737511396408081,
+      "learning_rate": 1.6295025728987994e-05,
+      "loss": 0.4612,
       "step": 4429
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 7.913817882537842,
-      "learning_rate": 1.112157963315626e-05,
-      "loss": 1.3824,
+      "epoch": 0.56,
+      "grad_norm": 15.169925689697266,
+      "learning_rate": 1.6294189013931307e-05,
+      "loss": 3.6944,
       "step": 4430
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 10.032218933105469,
-      "learning_rate": 1.1119575022551869e-05,
-      "loss": 1.2246,
+      "epoch": 0.56,
+      "grad_norm": 35.48377990722656,
+      "learning_rate": 1.629335229887462e-05,
+      "loss": 2.125,
       "step": 4431
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 25.357025146484375,
-      "learning_rate": 1.111757041194748e-05,
-      "loss": 1.7075,
+      "epoch": 0.56,
+      "grad_norm": 11.017925262451172,
+      "learning_rate": 1.629251558381793e-05,
+      "loss": 1.7788,
       "step": 4432
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 16.777484893798828,
-      "learning_rate": 1.1115565801343089e-05,
-      "loss": 1.6018,
+      "epoch": 0.56,
+      "grad_norm": 18.080007553100586,
+      "learning_rate": 1.6291678868761245e-05,
+      "loss": 3.4252,
       "step": 4433
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 27.090877532958984,
-      "learning_rate": 1.1113561190738699e-05,
-      "loss": 2.4302,
+      "epoch": 0.56,
+      "grad_norm": 13.970507621765137,
+      "learning_rate": 1.629084215370456e-05,
+      "loss": 1.3873,
       "step": 4434
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 24.513057708740234,
-      "learning_rate": 1.111155658013431e-05,
-      "loss": 2.2252,
+      "epoch": 0.56,
+      "grad_norm": 17.634294509887695,
+      "learning_rate": 1.629000543864787e-05,
+      "loss": 2.8027,
       "step": 4435
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 20.253862380981445,
-      "learning_rate": 1.1109551969529919e-05,
-      "loss": 1.4782,
+      "epoch": 0.56,
+      "grad_norm": 37.78720474243164,
+      "learning_rate": 1.6289168723591182e-05,
+      "loss": 1.2082,
       "step": 4436
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 10.198271751403809,
-      "learning_rate": 1.1107547358925531e-05,
-      "loss": 2.0537,
+      "epoch": 0.56,
+      "grad_norm": 13.17506217956543,
+      "learning_rate": 1.6288332008534496e-05,
+      "loss": 0.7018,
       "step": 4437
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 16.26251792907715,
-      "learning_rate": 1.110554274832114e-05,
-      "loss": 1.6344,
+      "epoch": 0.56,
+      "grad_norm": 21.82538414001465,
+      "learning_rate": 1.628749529347781e-05,
+      "loss": 1.4864,
       "step": 4438
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 17.53883934020996,
-      "learning_rate": 1.110353813771675e-05,
-      "loss": 1.4399,
+      "epoch": 0.56,
+      "grad_norm": 18.28236961364746,
+      "learning_rate": 1.628665857842112e-05,
+      "loss": 1.2984,
       "step": 4439
     },
     {
-      "epoch": 1.33,
-      "grad_norm": 9.351251602172852,
-      "learning_rate": 1.1101533527112361e-05,
-      "loss": 1.2029,
-      "step": 4440
-    },
-    {
-      "epoch": 1.33,
-      "eval_loss": 0.22466596961021423,
-      "eval_runtime": 43.3151,
-      "eval_samples_per_second": 34.145,
-      "eval_steps_per_second": 34.145,
+      "epoch": 0.56,
+      "grad_norm": 10.602968215942383,
+      "learning_rate": 1.6285821863364434e-05,
+      "loss": 1.3774,
       "step": 4440
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 18.78108787536621,
-      "learning_rate": 1.109952891650797e-05,
-      "loss": 1.8896,
+      "epoch": 0.56,
+      "grad_norm": 14.304097175598145,
+      "learning_rate": 1.6284985148307747e-05,
+      "loss": 1.4304,
       "step": 4441
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 14.223852157592773,
-      "learning_rate": 1.109752430590358e-05,
-      "loss": 1.8473,
+      "epoch": 0.56,
+      "grad_norm": 13.172176361083984,
+      "learning_rate": 1.6284148433251058e-05,
+      "loss": 1.0238,
       "step": 4442
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 29.34437370300293,
-      "learning_rate": 1.1095519695299188e-05,
-      "loss": 1.8663,
+      "epoch": 0.56,
+      "grad_norm": 16.124061584472656,
+      "learning_rate": 1.628331171819437e-05,
+      "loss": 0.9346,
       "step": 4443
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 11.203985214233398,
-      "learning_rate": 1.10935150846948e-05,
-      "loss": 1.4461,
+      "epoch": 0.56,
+      "grad_norm": 7.68695592880249,
+      "learning_rate": 1.6282475003137685e-05,
+      "loss": 1.2211,
       "step": 4444
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 13.125014305114746,
-      "learning_rate": 1.1091510474090408e-05,
-      "loss": 1.3763,
+      "epoch": 0.56,
+      "grad_norm": 18.085830688476562,
+      "learning_rate": 1.6281638288080995e-05,
+      "loss": 1.9951,
       "step": 4445
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 17.99578094482422,
-      "learning_rate": 1.1089505863486018e-05,
-      "loss": 2.1525,
+      "epoch": 0.56,
+      "grad_norm": 12.31620979309082,
+      "learning_rate": 1.628080157302431e-05,
+      "loss": 1.1265,
       "step": 4446
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 36.31547927856445,
-      "learning_rate": 1.108750125288163e-05,
-      "loss": 2.2111,
+      "epoch": 0.56,
+      "grad_norm": 9.333024024963379,
+      "learning_rate": 1.627996485796762e-05,
+      "loss": 2.1302,
       "step": 4447
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 13.750075340270996,
-      "learning_rate": 1.1085496642277238e-05,
-      "loss": 1.3206,
+      "epoch": 0.56,
+      "grad_norm": 10.498042106628418,
+      "learning_rate": 1.6279128142910933e-05,
+      "loss": 0.7364,
       "step": 4448
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 15.520805358886719,
-      "learning_rate": 1.1083492031672849e-05,
-      "loss": 3.3127,
+      "epoch": 0.56,
+      "grad_norm": 20.063735961914062,
+      "learning_rate": 1.6278291427854246e-05,
+      "loss": 3.5866,
       "step": 4449
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 16.05813217163086,
-      "learning_rate": 1.1081487421068457e-05,
-      "loss": 1.5862,
+      "epoch": 0.56,
+      "grad_norm": 13.666686058044434,
+      "learning_rate": 1.6277454712797557e-05,
+      "loss": 2.8051,
       "step": 4450
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 18.96892738342285,
-      "learning_rate": 1.1079482810464069e-05,
-      "loss": 1.4905,
+      "epoch": 0.56,
+      "grad_norm": 13.316627502441406,
+      "learning_rate": 1.627661799774087e-05,
+      "loss": 1.6376,
       "step": 4451
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 11.585920333862305,
-      "learning_rate": 1.1077478199859679e-05,
-      "loss": 2.2658,
+      "epoch": 0.56,
+      "grad_norm": 11.108987808227539,
+      "learning_rate": 1.6275781282684184e-05,
+      "loss": 2.7904,
       "step": 4452
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 21.121341705322266,
-      "learning_rate": 1.1075473589255287e-05,
-      "loss": 0.8119,
+      "epoch": 0.56,
+      "grad_norm": 49.902618408203125,
+      "learning_rate": 1.6274944567627494e-05,
+      "loss": 2.4332,
       "step": 4453
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 39.30922317504883,
-      "learning_rate": 1.1073468978650899e-05,
-      "loss": 2.4467,
+      "epoch": 0.56,
+      "grad_norm": 18.142854690551758,
+      "learning_rate": 1.6274107852570808e-05,
+      "loss": 3.3683,
       "step": 4454
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 15.98094654083252,
-      "learning_rate": 1.1071464368046507e-05,
-      "loss": 2.3305,
+      "epoch": 0.56,
+      "grad_norm": 31.451210021972656,
+      "learning_rate": 1.627327113751412e-05,
+      "loss": 1.4711,
       "step": 4455
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 9.627312660217285,
-      "learning_rate": 1.1069459757442117e-05,
-      "loss": 1.2344,
+      "epoch": 0.56,
+      "grad_norm": 12.135986328125,
+      "learning_rate": 1.627243442245743e-05,
+      "loss": 2.2472,
       "step": 4456
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 14.245975494384766,
-      "learning_rate": 1.1067455146837727e-05,
-      "loss": 0.9343,
+      "epoch": 0.56,
+      "grad_norm": 12.158326148986816,
+      "learning_rate": 1.6271597707400745e-05,
+      "loss": 1.5944,
       "step": 4457
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 29.139923095703125,
-      "learning_rate": 1.1065450536233338e-05,
-      "loss": 2.0948,
+      "epoch": 0.56,
+      "grad_norm": 39.65235137939453,
+      "learning_rate": 1.627076099234406e-05,
+      "loss": 2.0681,
       "step": 4458
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 23.93746566772461,
-      "learning_rate": 1.1063445925628948e-05,
-      "loss": 1.8753,
+      "epoch": 0.56,
+      "grad_norm": 19.438718795776367,
+      "learning_rate": 1.6269924277287373e-05,
+      "loss": 2.1166,
       "step": 4459
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 8.223407745361328,
-      "learning_rate": 1.1061441315024558e-05,
-      "loss": 1.0049,
+      "epoch": 0.56,
+      "grad_norm": 7.061102390289307,
+      "learning_rate": 1.6269087562230683e-05,
+      "loss": 0.2029,
       "step": 4460
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 28.680850982666016,
-      "learning_rate": 1.1059436704420168e-05,
-      "loss": 1.8643,
+      "epoch": 0.56,
+      "grad_norm": 7.9185919761657715,
+      "learning_rate": 1.6268250847173997e-05,
+      "loss": 0.6045,
       "step": 4461
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 27.831514358520508,
-      "learning_rate": 1.1057432093815776e-05,
-      "loss": 1.8748,
+      "epoch": 0.56,
+      "grad_norm": 24.92377281188965,
+      "learning_rate": 1.626741413211731e-05,
+      "loss": 2.6631,
       "step": 4462
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 17.656972885131836,
-      "learning_rate": 1.1055427483211388e-05,
-      "loss": 1.8081,
+      "epoch": 0.56,
+      "grad_norm": 14.203178405761719,
+      "learning_rate": 1.626657741706062e-05,
+      "loss": 0.8789,
       "step": 4463
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 12.41053581237793,
-      "learning_rate": 1.1053422872606998e-05,
-      "loss": 1.6312,
+      "epoch": 0.56,
+      "grad_norm": 14.638683319091797,
+      "learning_rate": 1.6265740702003934e-05,
+      "loss": 2.8277,
       "step": 4464
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 14.747664451599121,
-      "learning_rate": 1.1051418262002606e-05,
-      "loss": 1.8963,
+      "epoch": 0.56,
+      "grad_norm": 8.197480201721191,
+      "learning_rate": 1.6264903986947248e-05,
+      "loss": 2.014,
       "step": 4465
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 20.91558074951172,
-      "learning_rate": 1.1049413651398218e-05,
-      "loss": 1.0479,
+      "epoch": 0.56,
+      "grad_norm": 55.8455924987793,
+      "learning_rate": 1.6264067271890558e-05,
+      "loss": 3.2082,
       "step": 4466
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 18.06478500366211,
-      "learning_rate": 1.1047409040793827e-05,
-      "loss": 1.513,
+      "epoch": 0.56,
+      "grad_norm": 45.411041259765625,
+      "learning_rate": 1.626323055683387e-05,
+      "loss": 2.6712,
       "step": 4467
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 25.323034286499023,
-      "learning_rate": 1.1045404430189437e-05,
-      "loss": 2.1898,
+      "epoch": 0.56,
+      "grad_norm": 9.196112632751465,
+      "learning_rate": 1.6262393841777185e-05,
+      "loss": 2.336,
       "step": 4468
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 10.666196823120117,
-      "learning_rate": 1.1043399819585045e-05,
-      "loss": 1.6848,
+      "epoch": 0.56,
+      "grad_norm": 7.315990924835205,
+      "learning_rate": 1.62615571267205e-05,
+      "loss": 1.1812,
       "step": 4469
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 13.26004695892334,
-      "learning_rate": 1.1041395208980657e-05,
-      "loss": 1.3236,
+      "epoch": 0.56,
+      "grad_norm": 17.87982749938965,
+      "learning_rate": 1.626072041166381e-05,
+      "loss": 1.2415,
       "step": 4470
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 11.206985473632812,
-      "learning_rate": 1.1039390598376267e-05,
-      "loss": 1.2008,
+      "epoch": 0.56,
+      "grad_norm": 9.346860885620117,
+      "learning_rate": 1.6259883696607123e-05,
+      "loss": 1.159,
       "step": 4471
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 22.27272605895996,
-      "learning_rate": 1.1037385987771875e-05,
-      "loss": 1.5483,
+      "epoch": 0.56,
+      "grad_norm": 9.384839057922363,
+      "learning_rate": 1.6259046981550436e-05,
+      "loss": 2.3701,
       "step": 4472
     },
     {
-      "epoch": 1.34,
-      "grad_norm": 13.044824600219727,
-      "learning_rate": 1.1035381377167487e-05,
-      "loss": 1.4811,
+      "epoch": 0.56,
+      "grad_norm": 16.067934036254883,
+      "learning_rate": 1.6258210266493747e-05,
+      "loss": 2.8654,
       "step": 4473
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 15.387162208557129,
-      "learning_rate": 1.1033376766563096e-05,
-      "loss": 1.7151,
+      "epoch": 0.56,
+      "grad_norm": 6.0859880447387695,
+      "learning_rate": 1.625737355143706e-05,
+      "loss": 2.1036,
       "step": 4474
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 20.35818862915039,
-      "learning_rate": 1.1031372155958706e-05,
-      "loss": 1.6009,
+      "epoch": 0.56,
+      "grad_norm": 6.0310750007629395,
+      "learning_rate": 1.625653683638037e-05,
+      "loss": 0.6741,
       "step": 4475
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 12.326359748840332,
-      "learning_rate": 1.1029367545354314e-05,
-      "loss": 1.7638,
+      "epoch": 0.56,
+      "grad_norm": 42.43614959716797,
+      "learning_rate": 1.6255700121323684e-05,
+      "loss": 3.6709,
       "step": 4476
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 29.4976863861084,
-      "learning_rate": 1.1027362934749926e-05,
-      "loss": 1.8569,
+      "epoch": 0.56,
+      "grad_norm": 7.449599266052246,
+      "learning_rate": 1.6254863406266998e-05,
+      "loss": 2.0419,
       "step": 4477
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 11.03281307220459,
-      "learning_rate": 1.1025358324145536e-05,
-      "loss": 1.0993,
+      "epoch": 0.56,
+      "grad_norm": 18.67700958251953,
+      "learning_rate": 1.6254026691210308e-05,
+      "loss": 1.4978,
       "step": 4478
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 20.460525512695312,
-      "learning_rate": 1.1023353713541144e-05,
-      "loss": 1.4282,
+      "epoch": 0.56,
+      "grad_norm": 14.114124298095703,
+      "learning_rate": 1.6253189976153622e-05,
+      "loss": 1.6081,
       "step": 4479
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 11.63111400604248,
-      "learning_rate": 1.1021349102936756e-05,
-      "loss": 2.2856,
+      "epoch": 0.56,
+      "grad_norm": 17.91291046142578,
+      "learning_rate": 1.6252353261096932e-05,
+      "loss": 2.4313,
       "step": 4480
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 17.048723220825195,
-      "learning_rate": 1.1019344492332364e-05,
-      "loss": 2.037,
+      "epoch": 0.56,
+      "grad_norm": 19.789363861083984,
+      "learning_rate": 1.6251516546040246e-05,
+      "loss": 2.3497,
       "step": 4481
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 7.7821502685546875,
-      "learning_rate": 1.1017339881727975e-05,
-      "loss": 1.2083,
+      "epoch": 0.56,
+      "grad_norm": 5.772512435913086,
+      "learning_rate": 1.625067983098356e-05,
+      "loss": 0.6537,
       "step": 4482
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 16.613143920898438,
-      "learning_rate": 1.1015335271123586e-05,
-      "loss": 1.3636,
+      "epoch": 0.56,
+      "grad_norm": 11.66680908203125,
+      "learning_rate": 1.6249843115926873e-05,
+      "loss": 1.1615,
       "step": 4483
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 15.978687286376953,
-      "learning_rate": 1.1013330660519195e-05,
-      "loss": 1.5447,
+      "epoch": 0.56,
+      "grad_norm": 22.826932907104492,
+      "learning_rate": 1.6249006400870183e-05,
+      "loss": 1.9935,
       "step": 4484
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 12.686722755432129,
-      "learning_rate": 1.1011326049914805e-05,
-      "loss": 1.308,
+      "epoch": 0.56,
+      "grad_norm": 20.87784194946289,
+      "learning_rate": 1.6248169685813497e-05,
+      "loss": 0.682,
       "step": 4485
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 15.303074836730957,
-      "learning_rate": 1.1009321439310415e-05,
-      "loss": 2.4245,
+      "epoch": 0.56,
+      "grad_norm": 22.138505935668945,
+      "learning_rate": 1.624733297075681e-05,
+      "loss": 0.5427,
       "step": 4486
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 12.841423988342285,
-      "learning_rate": 1.1007316828706025e-05,
-      "loss": 2.0611,
+      "epoch": 0.56,
+      "grad_norm": 25.351593017578125,
+      "learning_rate": 1.624649625570012e-05,
+      "loss": 1.4485,
       "step": 4487
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 14.324323654174805,
-      "learning_rate": 1.1005312218101633e-05,
-      "loss": 1.9151,
+      "epoch": 0.56,
+      "grad_norm": 18.775243759155273,
+      "learning_rate": 1.6245659540643435e-05,
+      "loss": 1.9783,
       "step": 4488
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 10.316316604614258,
-      "learning_rate": 1.1003307607497245e-05,
-      "loss": 1.8989,
+      "epoch": 0.56,
+      "grad_norm": 13.692991256713867,
+      "learning_rate": 1.6244822825586748e-05,
+      "loss": 2.1927,
       "step": 4489
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 11.872175216674805,
-      "learning_rate": 1.1001302996892855e-05,
-      "loss": 1.4971,
+      "epoch": 0.56,
+      "grad_norm": 45.6162109375,
+      "learning_rate": 1.6243986110530062e-05,
+      "loss": 2.8494,
       "step": 4490
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 13.584701538085938,
-      "learning_rate": 1.0999298386288464e-05,
-      "loss": 1.6093,
+      "epoch": 0.56,
+      "grad_norm": 18.776546478271484,
+      "learning_rate": 1.6243149395473372e-05,
+      "loss": 1.6298,
       "step": 4491
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 23.555721282958984,
-      "learning_rate": 1.0997293775684075e-05,
-      "loss": 1.4708,
+      "epoch": 0.56,
+      "grad_norm": 8.89008903503418,
+      "learning_rate": 1.6242312680416686e-05,
+      "loss": 0.6049,
       "step": 4492
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 13.753824234008789,
-      "learning_rate": 1.0995289165079684e-05,
-      "loss": 1.4468,
+      "epoch": 0.56,
+      "grad_norm": 14.663054466247559,
+      "learning_rate": 1.624147596536e-05,
+      "loss": 2.9699,
       "step": 4493
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 23.030887603759766,
-      "learning_rate": 1.0993284554475294e-05,
-      "loss": 1.9327,
+      "epoch": 0.56,
+      "grad_norm": 14.046426773071289,
+      "learning_rate": 1.624063925030331e-05,
+      "loss": 2.0183,
       "step": 4494
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 43.81611251831055,
-      "learning_rate": 1.0991279943870906e-05,
-      "loss": 2.3426,
+      "epoch": 0.56,
+      "grad_norm": 26.707117080688477,
+      "learning_rate": 1.6239802535246623e-05,
+      "loss": 1.9285,
       "step": 4495
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 22.91530418395996,
-      "learning_rate": 1.0989275333266514e-05,
-      "loss": 2.1606,
+      "epoch": 0.56,
+      "grad_norm": 6.997550964355469,
+      "learning_rate": 1.6238965820189937e-05,
+      "loss": 1.8572,
       "step": 4496
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 30.3094425201416,
-      "learning_rate": 1.0987270722662124e-05,
-      "loss": 2.3948,
+      "epoch": 0.56,
+      "grad_norm": 7.119721412658691,
+      "learning_rate": 1.623812910513325e-05,
+      "loss": 1.4935,
       "step": 4497
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 90.9547348022461,
-      "learning_rate": 1.0985266112057732e-05,
-      "loss": 2.5067,
+      "epoch": 0.56,
+      "grad_norm": 27.102155685424805,
+      "learning_rate": 1.623729239007656e-05,
+      "loss": 2.3526,
       "step": 4498
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 11.258960723876953,
-      "learning_rate": 1.0983261501453344e-05,
-      "loss": 2.0156,
+      "epoch": 0.56,
+      "grad_norm": 11.541977882385254,
+      "learning_rate": 1.6236455675019875e-05,
+      "loss": 1.8764,
       "step": 4499
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 66.39472961425781,
-      "learning_rate": 1.0981256890848953e-05,
-      "loss": 2.6879,
+      "epoch": 0.56,
+      "grad_norm": 12.703827857971191,
+      "learning_rate": 1.6235618959963185e-05,
+      "loss": 3.0699,
       "step": 4500
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 20.91155242919922,
-      "learning_rate": 1.0979252280244563e-05,
-      "loss": 1.7934,
+      "epoch": 0.56,
+      "grad_norm": 11.643512725830078,
+      "learning_rate": 1.62347822449065e-05,
+      "loss": 1.1002,
       "step": 4501
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 18.402379989624023,
-      "learning_rate": 1.0977247669640175e-05,
-      "loss": 1.3445,
+      "epoch": 0.56,
+      "grad_norm": 9.464075088500977,
+      "learning_rate": 1.6233945529849812e-05,
+      "loss": 1.2133,
       "step": 4502
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 15.267261505126953,
-      "learning_rate": 1.0975243059035783e-05,
-      "loss": 1.642,
+      "epoch": 0.57,
+      "grad_norm": 17.377323150634766,
+      "learning_rate": 1.6233108814793122e-05,
+      "loss": 1.7252,
       "step": 4503
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 50.35695266723633,
-      "learning_rate": 1.0973238448431393e-05,
-      "loss": 2.0173,
+      "epoch": 0.57,
+      "grad_norm": 14.55684757232666,
+      "learning_rate": 1.6232272099736436e-05,
+      "loss": 2.0068,
       "step": 4504
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 10.630228042602539,
-      "learning_rate": 1.0971233837827003e-05,
-      "loss": 1.9372,
+      "epoch": 0.57,
+      "grad_norm": 14.520007133483887,
+      "learning_rate": 1.6231435384679746e-05,
+      "loss": 1.9361,
       "step": 4505
     },
     {
-      "epoch": 1.35,
-      "grad_norm": 32.916053771972656,
-      "learning_rate": 1.0969229227222613e-05,
-      "loss": 2.2776,
+      "epoch": 0.57,
+      "grad_norm": 12.38199520111084,
+      "learning_rate": 1.623059866962306e-05,
+      "loss": 1.497,
       "step": 4506
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 22.993738174438477,
-      "learning_rate": 1.0967224616618223e-05,
-      "loss": 2.2662,
+      "epoch": 0.57,
+      "grad_norm": 11.052742958068848,
+      "learning_rate": 1.6229761954566374e-05,
+      "loss": 1.7926,
       "step": 4507
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 13.480685234069824,
-      "learning_rate": 1.0965220006013833e-05,
-      "loss": 1.4548,
+      "epoch": 0.57,
+      "grad_norm": 13.624369621276855,
+      "learning_rate": 1.6228925239509684e-05,
+      "loss": 2.5462,
       "step": 4508
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 22.65932273864746,
-      "learning_rate": 1.0963215395409443e-05,
-      "loss": 2.1417,
+      "epoch": 0.57,
+      "grad_norm": 12.981353759765625,
+      "learning_rate": 1.6228088524452997e-05,
+      "loss": 0.9325,
       "step": 4509
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 9.409554481506348,
-      "learning_rate": 1.0961210784805052e-05,
-      "loss": 1.1043,
+      "epoch": 0.57,
+      "grad_norm": 16.13641357421875,
+      "learning_rate": 1.622725180939631e-05,
+      "loss": 1.8671,
       "step": 4510
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 51.40016555786133,
-      "learning_rate": 1.0959206174200664e-05,
-      "loss": 2.5073,
+      "epoch": 0.57,
+      "grad_norm": 14.619405746459961,
+      "learning_rate": 1.6226415094339625e-05,
+      "loss": 1.7138,
       "step": 4511
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 32.164634704589844,
-      "learning_rate": 1.0957201563596272e-05,
-      "loss": 1.7867,
+      "epoch": 0.57,
+      "grad_norm": 17.65199089050293,
+      "learning_rate": 1.6225578379282935e-05,
+      "loss": 4.2354,
       "step": 4512
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 29.420494079589844,
-      "learning_rate": 1.0955196952991882e-05,
-      "loss": 1.7466,
+      "epoch": 0.57,
+      "grad_norm": 26.615278244018555,
+      "learning_rate": 1.622474166422625e-05,
+      "loss": 3.0798,
       "step": 4513
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 10.119473457336426,
-      "learning_rate": 1.0953192342387494e-05,
-      "loss": 1.3604,
+      "epoch": 0.57,
+      "grad_norm": 9.01013469696045,
+      "learning_rate": 1.6223904949169562e-05,
+      "loss": 1.7211,
       "step": 4514
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 10.170442581176758,
-      "learning_rate": 1.0951187731783102e-05,
-      "loss": 1.2242,
+      "epoch": 0.57,
+      "grad_norm": 21.100177764892578,
+      "learning_rate": 1.6223068234112873e-05,
+      "loss": 4.2199,
       "step": 4515
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 11.852310180664062,
-      "learning_rate": 1.0949183121178712e-05,
-      "loss": 1.2755,
+      "epoch": 0.57,
+      "grad_norm": 13.066444396972656,
+      "learning_rate": 1.6222231519056186e-05,
+      "loss": 0.5538,
       "step": 4516
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 14.793402671813965,
-      "learning_rate": 1.094717851057432e-05,
-      "loss": 1.752,
+      "epoch": 0.57,
+      "grad_norm": 11.702042579650879,
+      "learning_rate": 1.62213948039995e-05,
+      "loss": 1.5223,
       "step": 4517
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 42.632144927978516,
-      "learning_rate": 1.0945173899969932e-05,
-      "loss": 1.7613,
+      "epoch": 0.57,
+      "grad_norm": 17.563871383666992,
+      "learning_rate": 1.6220558088942814e-05,
+      "loss": 1.169,
       "step": 4518
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 19.947559356689453,
-      "learning_rate": 1.0943169289365541e-05,
-      "loss": 1.9094,
+      "epoch": 0.57,
+      "grad_norm": 15.5070161819458,
+      "learning_rate": 1.6219721373886124e-05,
+      "loss": 1.4809,
       "step": 4519
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 15.529866218566895,
-      "learning_rate": 1.0941164678761151e-05,
-      "loss": 1.3777,
+      "epoch": 0.57,
+      "grad_norm": 13.678018569946289,
+      "learning_rate": 1.6218884658829437e-05,
+      "loss": 2.9076,
       "step": 4520
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 21.871538162231445,
-      "learning_rate": 1.0939160068156763e-05,
-      "loss": 1.91,
+      "epoch": 0.57,
+      "grad_norm": 8.950277328491211,
+      "learning_rate": 1.621804794377275e-05,
+      "loss": 2.0884,
       "step": 4521
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 18.05958366394043,
-      "learning_rate": 1.0937155457552371e-05,
-      "loss": 1.6152,
+      "epoch": 0.57,
+      "grad_norm": 11.462152481079102,
+      "learning_rate": 1.621721122871606e-05,
+      "loss": 0.8893,
       "step": 4522
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 13.304652214050293,
-      "learning_rate": 1.0935150846947981e-05,
-      "loss": 1.4861,
+      "epoch": 0.57,
+      "grad_norm": 47.0164680480957,
+      "learning_rate": 1.6216374513659375e-05,
+      "loss": 2.5958,
       "step": 4523
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 72.59398651123047,
-      "learning_rate": 1.093314623634359e-05,
-      "loss": 2.0853,
+      "epoch": 0.57,
+      "grad_norm": 13.051041603088379,
+      "learning_rate": 1.621553779860269e-05,
+      "loss": 2.353,
       "step": 4524
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 11.113508224487305,
-      "learning_rate": 1.0931141625739201e-05,
-      "loss": 1.5427,
+      "epoch": 0.57,
+      "grad_norm": 6.730961322784424,
+      "learning_rate": 1.6214701083546002e-05,
+      "loss": 0.7855,
       "step": 4525
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 11.754325866699219,
-      "learning_rate": 1.0929137015134811e-05,
-      "loss": 1.8286,
+      "epoch": 0.57,
+      "grad_norm": 14.580228805541992,
+      "learning_rate": 1.6213864368489313e-05,
+      "loss": 3.6427,
       "step": 4526
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 16.077978134155273,
-      "learning_rate": 1.092713240453042e-05,
-      "loss": 1.3027,
+      "epoch": 0.57,
+      "grad_norm": 12.631464004516602,
+      "learning_rate": 1.6213027653432626e-05,
+      "loss": 3.1712,
       "step": 4527
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 14.799029350280762,
-      "learning_rate": 1.0925127793926032e-05,
-      "loss": 1.222,
+      "epoch": 0.57,
+      "grad_norm": 9.510821342468262,
+      "learning_rate": 1.6212190938375936e-05,
+      "loss": 1.8789,
       "step": 4528
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 12.316436767578125,
-      "learning_rate": 1.092312318332164e-05,
-      "loss": 1.5382,
+      "epoch": 0.57,
+      "grad_norm": 18.531097412109375,
+      "learning_rate": 1.621135422331925e-05,
+      "loss": 2.4732,
       "step": 4529
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 11.659561157226562,
-      "learning_rate": 1.092111857271725e-05,
-      "loss": 1.0118,
+      "epoch": 0.57,
+      "grad_norm": 25.47800636291504,
+      "learning_rate": 1.6210517508262564e-05,
+      "loss": 2.4834,
       "step": 4530
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 10.741880416870117,
-      "learning_rate": 1.091911396211286e-05,
-      "loss": 1.1201,
+      "epoch": 0.57,
+      "grad_norm": 22.9942626953125,
+      "learning_rate": 1.6209680793205874e-05,
+      "loss": 1.6036,
       "step": 4531
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 17.27454376220703,
-      "learning_rate": 1.091710935150847e-05,
-      "loss": 1.8521,
+      "epoch": 0.57,
+      "grad_norm": 24.319677352905273,
+      "learning_rate": 1.6208844078149188e-05,
+      "loss": 1.4908,
       "step": 4532
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 19.821794509887695,
-      "learning_rate": 1.091510474090408e-05,
-      "loss": 1.1864,
+      "epoch": 0.57,
+      "grad_norm": 25.627683639526367,
+      "learning_rate": 1.6208007363092498e-05,
+      "loss": 1.5463,
       "step": 4533
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 42.41746520996094,
-      "learning_rate": 1.091310013029969e-05,
-      "loss": 1.7798,
+      "epoch": 0.57,
+      "grad_norm": 11.917311668395996,
+      "learning_rate": 1.620717064803581e-05,
+      "loss": 2.1978,
       "step": 4534
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 17.059722900390625,
-      "learning_rate": 1.09110955196953e-05,
-      "loss": 2.3114,
+      "epoch": 0.57,
+      "grad_norm": 11.899550437927246,
+      "learning_rate": 1.6206333932979125e-05,
+      "loss": 1.598,
       "step": 4535
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 38.7147331237793,
-      "learning_rate": 1.0909090909090909e-05,
-      "loss": 2.4257,
+      "epoch": 0.57,
+      "grad_norm": 15.763212203979492,
+      "learning_rate": 1.6205497217922435e-05,
+      "loss": 1.3893,
       "step": 4536
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 17.778711318969727,
-      "learning_rate": 1.090708629848652e-05,
-      "loss": 1.598,
+      "epoch": 0.57,
+      "grad_norm": 23.417814254760742,
+      "learning_rate": 1.620466050286575e-05,
+      "loss": 1.5707,
       "step": 4537
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 55.522010803222656,
-      "learning_rate": 1.090508168788213e-05,
-      "loss": 1.7511,
+      "epoch": 0.57,
+      "grad_norm": 18.423015594482422,
+      "learning_rate": 1.6203823787809063e-05,
+      "loss": 2.6211,
       "step": 4538
     },
     {
-      "epoch": 1.36,
-      "grad_norm": 31.59695053100586,
-      "learning_rate": 1.090307707727774e-05,
-      "loss": 1.6838,
+      "epoch": 0.57,
+      "grad_norm": 10.28055477142334,
+      "learning_rate": 1.6202987072752376e-05,
+      "loss": 1.2414,
       "step": 4539
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 14.828858375549316,
-      "learning_rate": 1.0901072466673351e-05,
-      "loss": 1.3515,
+      "epoch": 0.57,
+      "grad_norm": 7.889482498168945,
+      "learning_rate": 1.6202150357695687e-05,
+      "loss": 2.6767,
       "step": 4540
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 18.156906127929688,
-      "learning_rate": 1.089906785606896e-05,
-      "loss": 2.1229,
+      "epoch": 0.57,
+      "grad_norm": 65.32218933105469,
+      "learning_rate": 1.6201313642639e-05,
+      "loss": 2.8384,
       "step": 4541
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 14.6245698928833,
-      "learning_rate": 1.089706324546457e-05,
-      "loss": 1.0952,
+      "epoch": 0.57,
+      "grad_norm": 13.456228256225586,
+      "learning_rate": 1.6200476927582314e-05,
+      "loss": 1.3431,
       "step": 4542
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 14.46635913848877,
-      "learning_rate": 1.0895058634860178e-05,
-      "loss": 1.6335,
+      "epoch": 0.57,
+      "grad_norm": 11.8510160446167,
+      "learning_rate": 1.6199640212525624e-05,
+      "loss": 1.0315,
       "step": 4543
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 11.85185432434082,
-      "learning_rate": 1.089305402425579e-05,
-      "loss": 1.6408,
+      "epoch": 0.57,
+      "grad_norm": 20.21044921875,
+      "learning_rate": 1.6198803497468938e-05,
+      "loss": 2.6707,
       "step": 4544
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 9.316162109375,
-      "learning_rate": 1.08910494136514e-05,
-      "loss": 1.0231,
+      "epoch": 0.57,
+      "grad_norm": 14.6653470993042,
+      "learning_rate": 1.619796678241225e-05,
+      "loss": 2.2289,
       "step": 4545
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 16.85337257385254,
-      "learning_rate": 1.0889044803047008e-05,
-      "loss": 1.486,
+      "epoch": 0.57,
+      "grad_norm": 8.660666465759277,
+      "learning_rate": 1.6197130067355565e-05,
+      "loss": 1.8441,
       "step": 4546
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 13.948755264282227,
-      "learning_rate": 1.088704019244262e-05,
-      "loss": 2.3643,
+      "epoch": 0.57,
+      "grad_norm": 7.917065143585205,
+      "learning_rate": 1.6196293352298875e-05,
+      "loss": 1.0702,
       "step": 4547
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 22.807424545288086,
-      "learning_rate": 1.0885035581838228e-05,
-      "loss": 1.6381,
+      "epoch": 0.57,
+      "grad_norm": 33.09801483154297,
+      "learning_rate": 1.619545663724219e-05,
+      "loss": 1.7716,
       "step": 4548
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 17.668928146362305,
-      "learning_rate": 1.0883030971233838e-05,
-      "loss": 1.8671,
+      "epoch": 0.57,
+      "grad_norm": 9.391871452331543,
+      "learning_rate": 1.6194619922185503e-05,
+      "loss": 1.15,
       "step": 4549
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 61.22587966918945,
-      "learning_rate": 1.088102636062945e-05,
-      "loss": 2.4726,
+      "epoch": 0.57,
+      "grad_norm": 33.01217269897461,
+      "learning_rate": 1.6193783207128813e-05,
+      "loss": 1.6423,
       "step": 4550
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 15.125883102416992,
-      "learning_rate": 1.0879021750025058e-05,
-      "loss": 1.6064,
+      "epoch": 0.57,
+      "grad_norm": 12.896798133850098,
+      "learning_rate": 1.6192946492072127e-05,
+      "loss": 1.6256,
       "step": 4551
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 12.281584739685059,
-      "learning_rate": 1.0877017139420669e-05,
-      "loss": 1.9036,
+      "epoch": 0.57,
+      "grad_norm": 33.08998107910156,
+      "learning_rate": 1.619210977701544e-05,
+      "loss": 2.8372,
       "step": 4552
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 20.28942108154297,
-      "learning_rate": 1.0875012528816277e-05,
-      "loss": 2.022,
+      "epoch": 0.57,
+      "grad_norm": 6.515081882476807,
+      "learning_rate": 1.619127306195875e-05,
+      "loss": 0.5338,
       "step": 4553
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 12.765894889831543,
-      "learning_rate": 1.0873007918211889e-05,
-      "loss": 1.2068,
+      "epoch": 0.57,
+      "grad_norm": 6.269851207733154,
+      "learning_rate": 1.6190436346902064e-05,
+      "loss": 0.8314,
       "step": 4554
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 14.457947731018066,
-      "learning_rate": 1.0871003307607497e-05,
-      "loss": 1.2275,
+      "epoch": 0.57,
+      "grad_norm": 6.227377891540527,
+      "learning_rate": 1.6189599631845378e-05,
+      "loss": 2.0353,
       "step": 4555
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 26.644542694091797,
-      "learning_rate": 1.0868998697003109e-05,
-      "loss": 3.3596,
+      "epoch": 0.57,
+      "grad_norm": 17.164512634277344,
+      "learning_rate": 1.6188762916788688e-05,
+      "loss": 0.7052,
       "step": 4556
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 17.826154708862305,
-      "learning_rate": 1.0866994086398719e-05,
-      "loss": 2.4254,
+      "epoch": 0.57,
+      "grad_norm": 24.479812622070312,
+      "learning_rate": 1.6187926201732002e-05,
+      "loss": 1.5668,
       "step": 4557
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 26.188709259033203,
-      "learning_rate": 1.0864989475794327e-05,
-      "loss": 1.8766,
+      "epoch": 0.57,
+      "grad_norm": 6.707143783569336,
+      "learning_rate": 1.6187089486675312e-05,
+      "loss": 0.2343,
       "step": 4558
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 15.313033103942871,
-      "learning_rate": 1.0862984865189939e-05,
-      "loss": 1.5112,
+      "epoch": 0.57,
+      "grad_norm": 12.978679656982422,
+      "learning_rate": 1.6186252771618626e-05,
+      "loss": 0.7059,
       "step": 4559
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 15.900823593139648,
-      "learning_rate": 1.0860980254585548e-05,
-      "loss": 2.0939,
+      "epoch": 0.57,
+      "grad_norm": 14.660187721252441,
+      "learning_rate": 1.618541605656194e-05,
+      "loss": 2.3506,
       "step": 4560
     },
     {
-      "epoch": 1.37,
-      "eval_loss": 0.21247684955596924,
-      "eval_runtime": 43.2712,
-      "eval_samples_per_second": 34.18,
-      "eval_steps_per_second": 34.18,
-      "step": 4560
+      "epoch": 0.57,
+      "grad_norm": 8.758925437927246,
+      "learning_rate": 1.618457934150525e-05,
+      "loss": 1.3554,
+      "step": 4561
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 9.430264472961426,
-      "learning_rate": 1.0858975643981158e-05,
-      "loss": 1.7693,
-      "step": 4561
-    },
-    {
-      "epoch": 1.37,
-      "grad_norm": 9.332898139953613,
-      "learning_rate": 1.0856971033376766e-05,
-      "loss": 1.048,
+      "epoch": 0.57,
+      "grad_norm": 9.19585132598877,
+      "learning_rate": 1.6183742626448563e-05,
+      "loss": 2.1781,
       "step": 4562
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 15.586241722106934,
-      "learning_rate": 1.0854966422772378e-05,
-      "loss": 1.3119,
+      "epoch": 0.57,
+      "grad_norm": 15.945395469665527,
+      "learning_rate": 1.6182905911391877e-05,
+      "loss": 0.7306,
       "step": 4563
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 23.395587921142578,
-      "learning_rate": 1.0852961812167988e-05,
-      "loss": 2.1297,
+      "epoch": 0.57,
+      "grad_norm": 10.57514762878418,
+      "learning_rate": 1.6182069196335187e-05,
+      "loss": 1.2656,
       "step": 4564
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 11.959819793701172,
-      "learning_rate": 1.0850957201563596e-05,
-      "loss": 1.2409,
+      "epoch": 0.57,
+      "grad_norm": 18.79050064086914,
+      "learning_rate": 1.61812324812785e-05,
+      "loss": 1.7017,
       "step": 4565
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 9.007245063781738,
-      "learning_rate": 1.0848952590959208e-05,
-      "loss": 1.714,
+      "epoch": 0.57,
+      "grad_norm": 15.656498908996582,
+      "learning_rate": 1.6180395766221814e-05,
+      "loss": 2.3983,
       "step": 4566
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 56.485538482666016,
-      "learning_rate": 1.0846947980354816e-05,
-      "loss": 2.3396,
+      "epoch": 0.57,
+      "grad_norm": 12.713788986206055,
+      "learning_rate": 1.6179559051165128e-05,
+      "loss": 3.1377,
       "step": 4567
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 12.010300636291504,
-      "learning_rate": 1.0844943369750427e-05,
-      "loss": 2.0262,
+      "epoch": 0.57,
+      "grad_norm": 11.3866548538208,
+      "learning_rate": 1.617872233610844e-05,
+      "loss": 2.89,
       "step": 4568
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 37.940059661865234,
-      "learning_rate": 1.0842938759146038e-05,
-      "loss": 1.7123,
+      "epoch": 0.57,
+      "grad_norm": 15.300727844238281,
+      "learning_rate": 1.6177885621051752e-05,
+      "loss": 2.2796,
       "step": 4569
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 13.00949764251709,
-      "learning_rate": 1.0840934148541647e-05,
-      "loss": 1.3661,
+      "epoch": 0.57,
+      "grad_norm": 13.088519096374512,
+      "learning_rate": 1.6177048905995066e-05,
+      "loss": 1.4001,
       "step": 4570
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 15.618173599243164,
-      "learning_rate": 1.0838929537937257e-05,
-      "loss": 1.9199,
+      "epoch": 0.57,
+      "grad_norm": 12.312095642089844,
+      "learning_rate": 1.6176212190938376e-05,
+      "loss": 1.031,
       "step": 4571
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 22.606422424316406,
-      "learning_rate": 1.0836924927332865e-05,
-      "loss": 1.8386,
+      "epoch": 0.57,
+      "grad_norm": 25.807453155517578,
+      "learning_rate": 1.617537547588169e-05,
+      "loss": 1.5651,
       "step": 4572
     },
     {
-      "epoch": 1.37,
-      "grad_norm": 13.844141960144043,
-      "learning_rate": 1.0834920316728477e-05,
-      "loss": 1.3548,
+      "epoch": 0.57,
+      "grad_norm": 7.2591400146484375,
+      "learning_rate": 1.6174538760825003e-05,
+      "loss": 0.7109,
       "step": 4573
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 11.279787063598633,
-      "learning_rate": 1.0832915706124085e-05,
-      "loss": 1.5721,
+      "epoch": 0.57,
+      "grad_norm": 13.769720077514648,
+      "learning_rate": 1.6173702045768317e-05,
+      "loss": 1.0747,
       "step": 4574
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 13.838190078735352,
-      "learning_rate": 1.0830911095519695e-05,
-      "loss": 1.6001,
+      "epoch": 0.57,
+      "grad_norm": 11.24695110321045,
+      "learning_rate": 1.6172865330711627e-05,
+      "loss": 1.2303,
       "step": 4575
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 18.21175765991211,
-      "learning_rate": 1.0828906484915307e-05,
-      "loss": 1.9498,
+      "epoch": 0.57,
+      "grad_norm": 31.83026123046875,
+      "learning_rate": 1.617202861565494e-05,
+      "loss": 2.0739,
       "step": 4576
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 11.330121994018555,
-      "learning_rate": 1.0826901874310916e-05,
-      "loss": 1.3935,
+      "epoch": 0.57,
+      "grad_norm": 15.03792667388916,
+      "learning_rate": 1.6171191900598254e-05,
+      "loss": 0.9974,
       "step": 4577
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 18.719507217407227,
-      "learning_rate": 1.0824897263706526e-05,
-      "loss": 1.4156,
+      "epoch": 0.57,
+      "grad_norm": 15.796013832092285,
+      "learning_rate": 1.6170355185541565e-05,
+      "loss": 2.6044,
       "step": 4578
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 49.91222381591797,
-      "learning_rate": 1.0822892653102136e-05,
-      "loss": 2.8363,
+      "epoch": 0.57,
+      "grad_norm": 14.602738380432129,
+      "learning_rate": 1.616951847048488e-05,
+      "loss": 2.3995,
       "step": 4579
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 32.154563903808594,
-      "learning_rate": 1.0820888042497746e-05,
-      "loss": 2.5501,
+      "epoch": 0.57,
+      "grad_norm": 77.08251190185547,
+      "learning_rate": 1.6168681755428192e-05,
+      "loss": 2.5041,
       "step": 4580
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 8.45114803314209,
-      "learning_rate": 1.0818883431893356e-05,
-      "loss": 1.6866,
+      "epoch": 0.57,
+      "grad_norm": 8.562460899353027,
+      "learning_rate": 1.6167845040371502e-05,
+      "loss": 0.6393,
       "step": 4581
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 44.169677734375,
-      "learning_rate": 1.0816878821288966e-05,
-      "loss": 2.4302,
+      "epoch": 0.58,
+      "grad_norm": 7.505651473999023,
+      "learning_rate": 1.6167008325314816e-05,
+      "loss": 0.7842,
       "step": 4582
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 25.283023834228516,
-      "learning_rate": 1.0814874210684576e-05,
-      "loss": 2.2415,
+      "epoch": 0.58,
+      "grad_norm": 6.744200229644775,
+      "learning_rate": 1.616617161025813e-05,
+      "loss": 1.9606,
       "step": 4583
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 16.97034454345703,
-      "learning_rate": 1.0812869600080184e-05,
-      "loss": 1.8726,
+      "epoch": 0.58,
+      "grad_norm": 17.124282836914062,
+      "learning_rate": 1.616533489520144e-05,
+      "loss": 2.2951,
       "step": 4584
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 9.19655990600586,
-      "learning_rate": 1.0810864989475796e-05,
-      "loss": 1.1808,
+      "epoch": 0.58,
+      "grad_norm": 12.59799861907959,
+      "learning_rate": 1.6164498180144753e-05,
+      "loss": 2.3428,
       "step": 4585
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 18.277463912963867,
-      "learning_rate": 1.0808860378871405e-05,
-      "loss": 1.8823,
+      "epoch": 0.58,
+      "grad_norm": 8.041672706604004,
+      "learning_rate": 1.6163661465088064e-05,
+      "loss": 0.4739,
       "step": 4586
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 21.819673538208008,
-      "learning_rate": 1.0806855768267015e-05,
-      "loss": 2.0347,
+      "epoch": 0.58,
+      "grad_norm": 6.622467517852783,
+      "learning_rate": 1.6162824750031377e-05,
+      "loss": 0.7039,
       "step": 4587
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 9.92550277709961,
-      "learning_rate": 1.0804851157662627e-05,
-      "loss": 1.032,
+      "epoch": 0.58,
+      "grad_norm": 12.706607818603516,
+      "learning_rate": 1.616198803497469e-05,
+      "loss": 1.9807,
       "step": 4588
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 22.52004623413086,
-      "learning_rate": 1.0802846547058235e-05,
-      "loss": 2.1637,
+      "epoch": 0.58,
+      "grad_norm": 3.131267547607422,
+      "learning_rate": 1.6161151319918e-05,
+      "loss": 0.1042,
       "step": 4589
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 18.34803581237793,
-      "learning_rate": 1.0800841936453845e-05,
-      "loss": 1.9793,
+      "epoch": 0.58,
+      "grad_norm": 10.65312385559082,
+      "learning_rate": 1.6160314604861315e-05,
+      "loss": 1.5446,
       "step": 4590
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 24.83062171936035,
-      "learning_rate": 1.0798837325849453e-05,
-      "loss": 2.2503,
+      "epoch": 0.58,
+      "grad_norm": 3.5985212326049805,
+      "learning_rate": 1.615947788980463e-05,
+      "loss": 0.1319,
       "step": 4591
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 31.985370635986328,
-      "learning_rate": 1.0796832715245065e-05,
-      "loss": 1.9019,
+      "epoch": 0.58,
+      "grad_norm": 16.998870849609375,
+      "learning_rate": 1.615864117474794e-05,
+      "loss": 2.8475,
       "step": 4592
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 11.606969833374023,
-      "learning_rate": 1.0794828104640674e-05,
-      "loss": 1.5377,
+      "epoch": 0.58,
+      "grad_norm": 14.827058792114258,
+      "learning_rate": 1.6157804459691252e-05,
+      "loss": 1.9056,
       "step": 4593
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 14.753523826599121,
-      "learning_rate": 1.0792823494036284e-05,
-      "loss": 1.6545,
+      "epoch": 0.58,
+      "grad_norm": 12.889955520629883,
+      "learning_rate": 1.6156967744634566e-05,
+      "loss": 3.2301,
       "step": 4594
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 19.58027458190918,
-      "learning_rate": 1.0790818883431895e-05,
-      "loss": 1.7343,
+      "epoch": 0.58,
+      "grad_norm": 46.53408432006836,
+      "learning_rate": 1.615613102957788e-05,
+      "loss": 1.6417,
       "step": 4595
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 17.902088165283203,
-      "learning_rate": 1.0788814272827504e-05,
-      "loss": 1.763,
+      "epoch": 0.58,
+      "grad_norm": 10.040948867797852,
+      "learning_rate": 1.615529431452119e-05,
+      "loss": 1.7649,
       "step": 4596
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 10.415716171264648,
-      "learning_rate": 1.0786809662223114e-05,
-      "loss": 1.6063,
+      "epoch": 0.58,
+      "grad_norm": 6.867565631866455,
+      "learning_rate": 1.6154457599464504e-05,
+      "loss": 1.5661,
       "step": 4597
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 13.281739234924316,
-      "learning_rate": 1.0784805051618722e-05,
-      "loss": 1.8627,
+      "epoch": 0.58,
+      "grad_norm": 13.134184837341309,
+      "learning_rate": 1.6153620884407817e-05,
+      "loss": 1.8915,
       "step": 4598
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 10.292632102966309,
-      "learning_rate": 1.0782800441014334e-05,
-      "loss": 1.7306,
+      "epoch": 0.58,
+      "grad_norm": 15.050349235534668,
+      "learning_rate": 1.6152784169351128e-05,
+      "loss": 1.4449,
       "step": 4599
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 12.653688430786133,
-      "learning_rate": 1.0780795830409944e-05,
-      "loss": 1.3753,
+      "epoch": 0.58,
+      "grad_norm": 17.474966049194336,
+      "learning_rate": 1.615194745429444e-05,
+      "loss": 1.6006,
       "step": 4600
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 26.442541122436523,
-      "learning_rate": 1.0778791219805553e-05,
-      "loss": 2.2257,
+      "epoch": 0.58,
+      "grad_norm": 9.59934139251709,
+      "learning_rate": 1.6151110739237755e-05,
+      "loss": 0.7479,
       "step": 4601
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 18.668859481811523,
-      "learning_rate": 1.0776786609201164e-05,
-      "loss": 1.8788,
+      "epoch": 0.58,
+      "grad_norm": 14.621644973754883,
+      "learning_rate": 1.615027402418107e-05,
+      "loss": 2.0671,
       "step": 4602
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 19.046268463134766,
-      "learning_rate": 1.0774781998596773e-05,
-      "loss": 2.1165,
+      "epoch": 0.58,
+      "grad_norm": 14.647985458374023,
+      "learning_rate": 1.614943730912438e-05,
+      "loss": 0.7475,
       "step": 4603
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 11.989215850830078,
-      "learning_rate": 1.0772777387992383e-05,
-      "loss": 1.2069,
+      "epoch": 0.58,
+      "grad_norm": 17.761341094970703,
+      "learning_rate": 1.6148600594067692e-05,
+      "loss": 1.6466,
       "step": 4604
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 16.762723922729492,
-      "learning_rate": 1.0770772777387993e-05,
-      "loss": 1.3351,
+      "epoch": 0.58,
+      "grad_norm": 14.85982894897461,
+      "learning_rate": 1.6147763879011006e-05,
+      "loss": 1.6539,
       "step": 4605
     },
     {
-      "epoch": 1.38,
-      "grad_norm": 21.549854278564453,
-      "learning_rate": 1.0768768166783603e-05,
-      "loss": 1.8815,
+      "epoch": 0.58,
+      "grad_norm": 16.099637985229492,
+      "learning_rate": 1.6146927163954316e-05,
+      "loss": 1.3589,
       "step": 4606
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 46.833518981933594,
-      "learning_rate": 1.0766763556179213e-05,
-      "loss": 2.6215,
+      "epoch": 0.58,
+      "grad_norm": 12.23636531829834,
+      "learning_rate": 1.614609044889763e-05,
+      "loss": 1.4119,
       "step": 4607
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 13.283019065856934,
-      "learning_rate": 1.0764758945574823e-05,
-      "loss": 1.9159,
+      "epoch": 0.58,
+      "grad_norm": 12.770011901855469,
+      "learning_rate": 1.6145253733840944e-05,
+      "loss": 0.4493,
       "step": 4608
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 41.622859954833984,
-      "learning_rate": 1.0762754334970433e-05,
-      "loss": 1.4153,
+      "epoch": 0.58,
+      "grad_norm": 132.66159057617188,
+      "learning_rate": 1.6144417018784254e-05,
+      "loss": 3.0702,
       "step": 4609
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 9.129386901855469,
-      "learning_rate": 1.0760749724366042e-05,
-      "loss": 1.0581,
+      "epoch": 0.58,
+      "grad_norm": 13.396702766418457,
+      "learning_rate": 1.6143580303727568e-05,
+      "loss": 1.1574,
       "step": 4610
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 63.40047836303711,
-      "learning_rate": 1.0758745113761653e-05,
-      "loss": 1.7917,
+      "epoch": 0.58,
+      "grad_norm": 26.069164276123047,
+      "learning_rate": 1.6142743588670878e-05,
+      "loss": 2.2689,
       "step": 4611
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 119.53295135498047,
-      "learning_rate": 1.0756740503157263e-05,
-      "loss": 1.7074,
+      "epoch": 0.58,
+      "grad_norm": 11.59582233428955,
+      "learning_rate": 1.614190687361419e-05,
+      "loss": 2.3476,
       "step": 4612
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 15.246512413024902,
-      "learning_rate": 1.0754735892552872e-05,
-      "loss": 1.3584,
+      "epoch": 0.58,
+      "grad_norm": 20.005428314208984,
+      "learning_rate": 1.6141070158557505e-05,
+      "loss": 2.1947,
       "step": 4613
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 12.469449043273926,
-      "learning_rate": 1.0752731281948484e-05,
-      "loss": 1.8839,
+      "epoch": 0.58,
+      "grad_norm": 14.68831729888916,
+      "learning_rate": 1.6140233443500815e-05,
+      "loss": 1.5459,
       "step": 4614
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 38.270023345947266,
-      "learning_rate": 1.0750726671344092e-05,
-      "loss": 1.4275,
+      "epoch": 0.58,
+      "grad_norm": 5.946023464202881,
+      "learning_rate": 1.613939672844413e-05,
+      "loss": 0.2806,
       "step": 4615
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 16.700550079345703,
-      "learning_rate": 1.0748722060739702e-05,
-      "loss": 1.771,
+      "epoch": 0.58,
+      "grad_norm": 11.920265197753906,
+      "learning_rate": 1.6138560013387443e-05,
+      "loss": 0.4182,
       "step": 4616
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 23.500558853149414,
-      "learning_rate": 1.074671745013531e-05,
-      "loss": 2.2971,
+      "epoch": 0.58,
+      "grad_norm": 9.490656852722168,
+      "learning_rate": 1.6137723298330753e-05,
+      "loss": 0.8312,
       "step": 4617
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 17.58616065979004,
-      "learning_rate": 1.0744712839530922e-05,
-      "loss": 1.05,
+      "epoch": 0.58,
+      "grad_norm": 17.456974029541016,
+      "learning_rate": 1.6136886583274067e-05,
+      "loss": 1.4119,
       "step": 4618
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 12.4642972946167,
-      "learning_rate": 1.0742708228926532e-05,
-      "loss": 1.1186,
+      "epoch": 0.58,
+      "grad_norm": 29.047473907470703,
+      "learning_rate": 1.613604986821738e-05,
+      "loss": 1.9479,
       "step": 4619
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 12.4574556350708,
-      "learning_rate": 1.074070361832214e-05,
-      "loss": 1.4973,
+      "epoch": 0.58,
+      "grad_norm": 2.6419622898101807,
+      "learning_rate": 1.613521315316069e-05,
+      "loss": 0.0435,
       "step": 4620
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 10.377580642700195,
-      "learning_rate": 1.0738699007717753e-05,
-      "loss": 1.8076,
+      "epoch": 0.58,
+      "grad_norm": 14.935125350952148,
+      "learning_rate": 1.6134376438104004e-05,
+      "loss": 2.1664,
       "step": 4621
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 19.34781265258789,
-      "learning_rate": 1.0736694397113361e-05,
-      "loss": 2.2722,
+      "epoch": 0.58,
+      "grad_norm": 8.660603523254395,
+      "learning_rate": 1.6133539723047318e-05,
+      "loss": 2.3279,
       "step": 4622
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 16.955293655395508,
-      "learning_rate": 1.0734689786508971e-05,
-      "loss": 1.7219,
+      "epoch": 0.58,
+      "grad_norm": 23.47085189819336,
+      "learning_rate": 1.613270300799063e-05,
+      "loss": 2.4712,
       "step": 4623
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 18.286012649536133,
-      "learning_rate": 1.0732685175904583e-05,
-      "loss": 2.9251,
+      "epoch": 0.58,
+      "grad_norm": 15.253812789916992,
+      "learning_rate": 1.613186629293394e-05,
+      "loss": 1.809,
       "step": 4624
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 19.848609924316406,
-      "learning_rate": 1.0730680565300191e-05,
-      "loss": 2.2197,
+      "epoch": 0.58,
+      "grad_norm": 19.041500091552734,
+      "learning_rate": 1.6131029577877255e-05,
+      "loss": 1.2805,
       "step": 4625
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 10.013169288635254,
-      "learning_rate": 1.0728675954695801e-05,
-      "loss": 0.7125,
+      "epoch": 0.58,
+      "grad_norm": 17.887468338012695,
+      "learning_rate": 1.613019286282057e-05,
+      "loss": 2.9352,
       "step": 4626
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 20.401023864746094,
-      "learning_rate": 1.0726671344091411e-05,
-      "loss": 1.5097,
+      "epoch": 0.58,
+      "grad_norm": 26.64823341369629,
+      "learning_rate": 1.612935614776388e-05,
+      "loss": 1.194,
       "step": 4627
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 20.424652099609375,
-      "learning_rate": 1.0724666733487021e-05,
-      "loss": 1.6132,
+      "epoch": 0.58,
+      "grad_norm": 15.227701187133789,
+      "learning_rate": 1.6128519432707193e-05,
+      "loss": 1.1081,
       "step": 4628
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 17.268680572509766,
-      "learning_rate": 1.072266212288263e-05,
-      "loss": 1.589,
+      "epoch": 0.58,
+      "grad_norm": 9.078343391418457,
+      "learning_rate": 1.6127682717650507e-05,
+      "loss": 1.4022,
       "step": 4629
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 10.245696067810059,
-      "learning_rate": 1.0720657512278242e-05,
-      "loss": 2.4915,
+      "epoch": 0.58,
+      "grad_norm": 35.389652252197266,
+      "learning_rate": 1.612684600259382e-05,
+      "loss": 1.8525,
       "step": 4630
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 21.644811630249023,
-      "learning_rate": 1.0718652901673852e-05,
-      "loss": 2.4812,
+      "epoch": 0.58,
+      "grad_norm": 80.05748748779297,
+      "learning_rate": 1.612600928753713e-05,
+      "loss": 3.0423,
       "step": 4631
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 19.70071792602539,
-      "learning_rate": 1.071664829106946e-05,
-      "loss": 1.8817,
+      "epoch": 0.58,
+      "grad_norm": 8.841204643249512,
+      "learning_rate": 1.6125172572480444e-05,
+      "loss": 0.655,
       "step": 4632
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 22.212900161743164,
-      "learning_rate": 1.0714643680465072e-05,
-      "loss": 1.9495,
+      "epoch": 0.58,
+      "grad_norm": 14.465465545654297,
+      "learning_rate": 1.6124335857423758e-05,
+      "loss": 2.3567,
       "step": 4633
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 11.387062072753906,
-      "learning_rate": 1.071263906986068e-05,
-      "loss": 1.3702,
+      "epoch": 0.58,
+      "grad_norm": 30.510494232177734,
+      "learning_rate": 1.6123499142367068e-05,
+      "loss": 1.9619,
       "step": 4634
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 10.107064247131348,
-      "learning_rate": 1.071063445925629e-05,
-      "loss": 1.0777,
+      "epoch": 0.58,
+      "grad_norm": 14.832023620605469,
+      "learning_rate": 1.612266242731038e-05,
+      "loss": 1.2945,
       "step": 4635
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 34.227264404296875,
-      "learning_rate": 1.0708629848651899e-05,
-      "loss": 1.9404,
+      "epoch": 0.58,
+      "grad_norm": 14.375158309936523,
+      "learning_rate": 1.6121825712253695e-05,
+      "loss": 1.9831,
       "step": 4636
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 12.135993003845215,
-      "learning_rate": 1.070662523804751e-05,
-      "loss": 1.2095,
+      "epoch": 0.58,
+      "grad_norm": 12.27586555480957,
+      "learning_rate": 1.6120988997197006e-05,
+      "loss": 2.1559,
       "step": 4637
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 15.475272178649902,
-      "learning_rate": 1.070462062744312e-05,
-      "loss": 1.4376,
+      "epoch": 0.58,
+      "grad_norm": 22.290096282958984,
+      "learning_rate": 1.612015228214032e-05,
+      "loss": 2.4757,
       "step": 4638
     },
     {
-      "epoch": 1.39,
-      "grad_norm": 29.245624542236328,
-      "learning_rate": 1.0702616016838729e-05,
-      "loss": 2.0634,
+      "epoch": 0.58,
+      "grad_norm": 51.71296310424805,
+      "learning_rate": 1.611931556708363e-05,
+      "loss": 2.2293,
       "step": 4639
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 33.319969177246094,
-      "learning_rate": 1.070061140623434e-05,
-      "loss": 1.772,
+      "epoch": 0.58,
+      "grad_norm": 22.068788528442383,
+      "learning_rate": 1.6118478852026943e-05,
+      "loss": 0.9596,
       "step": 4640
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 15.358686447143555,
-      "learning_rate": 1.0698606795629949e-05,
-      "loss": 1.5593,
+      "epoch": 0.58,
+      "grad_norm": 11.041895866394043,
+      "learning_rate": 1.6117642136970257e-05,
+      "loss": 2.1294,
       "step": 4641
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 13.498563766479492,
-      "learning_rate": 1.069660218502556e-05,
-      "loss": 1.8836,
+      "epoch": 0.58,
+      "grad_norm": 9.11545467376709,
+      "learning_rate": 1.6116805421913567e-05,
+      "loss": 1.3963,
       "step": 4642
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 11.653413772583008,
-      "learning_rate": 1.0694597574421171e-05,
-      "loss": 1.3999,
+      "epoch": 0.58,
+      "grad_norm": 15.038471221923828,
+      "learning_rate": 1.611596870685688e-05,
+      "loss": 1.2528,
       "step": 4643
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 7.912506103515625,
-      "learning_rate": 1.069259296381678e-05,
-      "loss": 1.0855,
+      "epoch": 0.58,
+      "grad_norm": 12.920796394348145,
+      "learning_rate": 1.6115131991800194e-05,
+      "loss": 1.743,
       "step": 4644
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 18.94805908203125,
-      "learning_rate": 1.069058835321239e-05,
-      "loss": 2.4709,
+      "epoch": 0.58,
+      "grad_norm": 10.468306541442871,
+      "learning_rate": 1.6114295276743505e-05,
+      "loss": 1.2616,
       "step": 4645
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 14.374813079833984,
-      "learning_rate": 1.0688583742607998e-05,
-      "loss": 1.4454,
+      "epoch": 0.58,
+      "grad_norm": 15.705430030822754,
+      "learning_rate": 1.6113458561686818e-05,
+      "loss": 1.9102,
       "step": 4646
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 77.0107650756836,
-      "learning_rate": 1.068657913200361e-05,
-      "loss": 2.156,
+      "epoch": 0.58,
+      "grad_norm": 15.315054893493652,
+      "learning_rate": 1.6112621846630132e-05,
+      "loss": 4.8984,
       "step": 4647
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 14.53071403503418,
-      "learning_rate": 1.0684574521399218e-05,
-      "loss": 1.3087,
+      "epoch": 0.58,
+      "grad_norm": 27.60223388671875,
+      "learning_rate": 1.6111785131573442e-05,
+      "loss": 2.6091,
       "step": 4648
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 18.24622917175293,
-      "learning_rate": 1.0682569910794828e-05,
-      "loss": 2.3464,
+      "epoch": 0.58,
+      "grad_norm": 7.672779083251953,
+      "learning_rate": 1.6110948416516756e-05,
+      "loss": 1.8401,
       "step": 4649
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 7.772927761077881,
-      "learning_rate": 1.068056530019044e-05,
-      "loss": 1.3107,
+      "epoch": 0.58,
+      "grad_norm": 21.329416275024414,
+      "learning_rate": 1.611011170146007e-05,
+      "loss": 1.8097,
       "step": 4650
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 4.152658462524414,
-      "learning_rate": 1.0678560689586048e-05,
-      "loss": 0.467,
+      "epoch": 0.58,
+      "grad_norm": 14.985925674438477,
+      "learning_rate": 1.6109274986403383e-05,
+      "loss": 2.3634,
       "step": 4651
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 14.69251537322998,
-      "learning_rate": 1.0676556078981658e-05,
-      "loss": 1.5746,
+      "epoch": 0.58,
+      "grad_norm": 40.59479522705078,
+      "learning_rate": 1.6108438271346693e-05,
+      "loss": 2.189,
       "step": 4652
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 15.464540481567383,
-      "learning_rate": 1.0674551468377268e-05,
-      "loss": 1.5305,
+      "epoch": 0.58,
+      "grad_norm": 12.507532119750977,
+      "learning_rate": 1.6107601556290007e-05,
+      "loss": 2.6886,
       "step": 4653
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 11.303486824035645,
-      "learning_rate": 1.0672546857772879e-05,
-      "loss": 1.8223,
+      "epoch": 0.58,
+      "grad_norm": 36.30336380004883,
+      "learning_rate": 1.610676484123332e-05,
+      "loss": 0.8976,
       "step": 4654
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 23.572040557861328,
-      "learning_rate": 1.0670542247168489e-05,
-      "loss": 1.9877,
+      "epoch": 0.58,
+      "grad_norm": 25.99757194519043,
+      "learning_rate": 1.610592812617663e-05,
+      "loss": 2.6275,
       "step": 4655
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 22.017297744750977,
-      "learning_rate": 1.0668537636564099e-05,
-      "loss": 1.2985,
+      "epoch": 0.58,
+      "grad_norm": 10.68349552154541,
+      "learning_rate": 1.6105091411119945e-05,
+      "loss": 3.1259,
       "step": 4656
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 13.40971851348877,
-      "learning_rate": 1.0666533025959709e-05,
-      "loss": 1.3887,
+      "epoch": 0.58,
+      "grad_norm": 20.675979614257812,
+      "learning_rate": 1.6104254696063258e-05,
+      "loss": 1.5156,
       "step": 4657
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 17.30711555480957,
-      "learning_rate": 1.0664528415355317e-05,
-      "loss": 1.6859,
+      "epoch": 0.58,
+      "grad_norm": 13.17327880859375,
+      "learning_rate": 1.6103417981006572e-05,
+      "loss": 0.7102,
       "step": 4658
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 8.817755699157715,
-      "learning_rate": 1.0662523804750929e-05,
-      "loss": 1.308,
+      "epoch": 0.58,
+      "grad_norm": 22.457372665405273,
+      "learning_rate": 1.6102581265949882e-05,
+      "loss": 1.4038,
       "step": 4659
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 29.949804306030273,
-      "learning_rate": 1.0660519194146537e-05,
-      "loss": 1.2951,
+      "epoch": 0.58,
+      "grad_norm": 12.7407865524292,
+      "learning_rate": 1.6101744550893196e-05,
+      "loss": 2.5345,
       "step": 4660
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 83.96745300292969,
-      "learning_rate": 1.0658514583542147e-05,
-      "loss": 2.6652,
+      "epoch": 0.58,
+      "grad_norm": 35.302730560302734,
+      "learning_rate": 1.610090783583651e-05,
+      "loss": 2.2454,
       "step": 4661
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 9.81541919708252,
-      "learning_rate": 1.065650997293776e-05,
-      "loss": 0.9763,
+      "epoch": 0.59,
+      "grad_norm": 36.84288024902344,
+      "learning_rate": 1.610007112077982e-05,
+      "loss": 3.0446,
       "step": 4662
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 24.451271057128906,
-      "learning_rate": 1.0654505362333368e-05,
-      "loss": 2.2991,
+      "epoch": 0.59,
+      "grad_norm": 11.649720191955566,
+      "learning_rate": 1.6099234405723133e-05,
+      "loss": 0.5907,
       "step": 4663
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 21.38994789123535,
-      "learning_rate": 1.0652500751728978e-05,
-      "loss": 1.4465,
+      "epoch": 0.59,
+      "grad_norm": 13.00683307647705,
+      "learning_rate": 1.6098397690666444e-05,
+      "loss": 3.9629,
       "step": 4664
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 9.186866760253906,
-      "learning_rate": 1.0650496141124586e-05,
-      "loss": 0.5958,
+      "epoch": 0.59,
+      "grad_norm": 22.276634216308594,
+      "learning_rate": 1.6097560975609757e-05,
+      "loss": 2.0421,
       "step": 4665
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 18.036312103271484,
-      "learning_rate": 1.0648491530520198e-05,
-      "loss": 1.7996,
+      "epoch": 0.59,
+      "grad_norm": 3.471829891204834,
+      "learning_rate": 1.609672426055307e-05,
+      "loss": 0.071,
       "step": 4666
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 13.11405086517334,
-      "learning_rate": 1.0646486919915808e-05,
-      "loss": 0.8623,
+      "epoch": 0.59,
+      "grad_norm": 16.312870025634766,
+      "learning_rate": 1.609588754549638e-05,
+      "loss": 1.2167,
       "step": 4667
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 15.945577621459961,
-      "learning_rate": 1.0644482309311416e-05,
-      "loss": 1.5742,
+      "epoch": 0.59,
+      "grad_norm": 28.203134536743164,
+      "learning_rate": 1.6095050830439695e-05,
+      "loss": 1.3825,
       "step": 4668
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 10.415094375610352,
-      "learning_rate": 1.0642477698707028e-05,
-      "loss": 1.0011,
+      "epoch": 0.59,
+      "grad_norm": 7.57172155380249,
+      "learning_rate": 1.6094214115383005e-05,
+      "loss": 0.6187,
       "step": 4669
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 13.777579307556152,
-      "learning_rate": 1.0640473088102636e-05,
-      "loss": 1.6847,
+      "epoch": 0.59,
+      "grad_norm": 12.809304237365723,
+      "learning_rate": 1.609337740032632e-05,
+      "loss": 1.9076,
       "step": 4670
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 29.36578369140625,
-      "learning_rate": 1.0638468477498247e-05,
-      "loss": 1.2423,
+      "epoch": 0.59,
+      "grad_norm": 14.868522644042969,
+      "learning_rate": 1.6092540685269632e-05,
+      "loss": 1.2166,
       "step": 4671
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 26.05026626586914,
-      "learning_rate": 1.0636463866893855e-05,
-      "loss": 1.976,
+      "epoch": 0.59,
+      "grad_norm": 41.13257598876953,
+      "learning_rate": 1.6091703970212946e-05,
+      "loss": 2.9149,
       "step": 4672
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 26.071765899658203,
-      "learning_rate": 1.0634459256289467e-05,
-      "loss": 2.2839,
+      "epoch": 0.59,
+      "grad_norm": 14.411210060119629,
+      "learning_rate": 1.6090867255156256e-05,
+      "loss": 1.8933,
       "step": 4673
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 16.8984317779541,
-      "learning_rate": 1.0632454645685077e-05,
-      "loss": 1.3912,
+      "epoch": 0.59,
+      "grad_norm": 22.125534057617188,
+      "learning_rate": 1.609003054009957e-05,
+      "loss": 2.9775,
       "step": 4674
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 65.57967376708984,
-      "learning_rate": 1.0630450035080687e-05,
-      "loss": 2.2416,
+      "epoch": 0.59,
+      "grad_norm": 16.977636337280273,
+      "learning_rate": 1.6089193825042884e-05,
+      "loss": 2.0797,
       "step": 4675
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 12.922877311706543,
-      "learning_rate": 1.0628445424476297e-05,
-      "loss": 1.5951,
+      "epoch": 0.59,
+      "grad_norm": 15.84194564819336,
+      "learning_rate": 1.6088357109986194e-05,
+      "loss": 1.6354,
       "step": 4676
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 15.637014389038086,
-      "learning_rate": 1.0626440813871905e-05,
-      "loss": 1.9517,
+      "epoch": 0.59,
+      "grad_norm": 6.054462432861328,
+      "learning_rate": 1.6087520394929507e-05,
+      "loss": 0.5514,
       "step": 4677
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 12.687530517578125,
-      "learning_rate": 1.0624436203267517e-05,
-      "loss": 1.3298,
+      "epoch": 0.59,
+      "grad_norm": 6.776900291442871,
+      "learning_rate": 1.608668367987282e-05,
+      "loss": 0.9739,
       "step": 4678
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 47.006107330322266,
-      "learning_rate": 1.0622431592663126e-05,
-      "loss": 2.6819,
+      "epoch": 0.59,
+      "grad_norm": 13.240543365478516,
+      "learning_rate": 1.6085846964816135e-05,
+      "loss": 1.1257,
       "step": 4679
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 12.535234451293945,
-      "learning_rate": 1.0620426982058736e-05,
-      "loss": 1.4554,
-      "step": 4680
-    },
-    {
-      "epoch": 1.41,
-      "eval_loss": 0.2101414054632187,
-      "eval_runtime": 43.0905,
-      "eval_samples_per_second": 34.323,
-      "eval_steps_per_second": 34.323,
+      "epoch": 0.59,
+      "grad_norm": 16.568603515625,
+      "learning_rate": 1.6085010249759445e-05,
+      "loss": 2.5752,
       "step": 4680
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 9.197644233703613,
-      "learning_rate": 1.0618422371454347e-05,
-      "loss": 1.1242,
+      "epoch": 0.59,
+      "grad_norm": 11.308704376220703,
+      "learning_rate": 1.608417353470276e-05,
+      "loss": 1.13,
       "step": 4681
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 11.781989097595215,
-      "learning_rate": 1.0616417760849956e-05,
-      "loss": 1.9253,
+      "epoch": 0.59,
+      "grad_norm": 21.552059173583984,
+      "learning_rate": 1.6083336819646072e-05,
+      "loss": 2.0728,
       "step": 4682
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 41.59600830078125,
-      "learning_rate": 1.0614413150245566e-05,
-      "loss": 2.3303,
+      "epoch": 0.59,
+      "grad_norm": 10.485605239868164,
+      "learning_rate": 1.6082500104589383e-05,
+      "loss": 2.5726,
       "step": 4683
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 13.190827369689941,
-      "learning_rate": 1.0612408539641174e-05,
-      "loss": 0.8268,
+      "epoch": 0.59,
+      "grad_norm": 18.043210983276367,
+      "learning_rate": 1.6081663389532696e-05,
+      "loss": 1.7581,
       "step": 4684
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 57.965003967285156,
-      "learning_rate": 1.0610403929036786e-05,
-      "loss": 2.375,
+      "epoch": 0.59,
+      "grad_norm": 20.532176971435547,
+      "learning_rate": 1.608082667447601e-05,
+      "loss": 2.0524,
       "step": 4685
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 21.654773712158203,
-      "learning_rate": 1.0608399318432396e-05,
-      "loss": 2.0238,
+      "epoch": 0.59,
+      "grad_norm": 19.0378360748291,
+      "learning_rate": 1.6079989959419324e-05,
+      "loss": 1.6556,
       "step": 4686
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 21.622922897338867,
-      "learning_rate": 1.0606394707828005e-05,
-      "loss": 1.8048,
+      "epoch": 0.59,
+      "grad_norm": 14.122875213623047,
+      "learning_rate": 1.6079153244362634e-05,
+      "loss": 2.5804,
       "step": 4687
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 21.0383358001709,
-      "learning_rate": 1.0604390097223616e-05,
-      "loss": 2.2016,
+      "epoch": 0.59,
+      "grad_norm": 13.868156433105469,
+      "learning_rate": 1.6078316529305947e-05,
+      "loss": 1.3452,
       "step": 4688
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 25.455896377563477,
-      "learning_rate": 1.0602385486619225e-05,
-      "loss": 1.9829,
+      "epoch": 0.59,
+      "grad_norm": 6.920721054077148,
+      "learning_rate": 1.6077479814249258e-05,
+      "loss": 0.7653,
       "step": 4689
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 27.83086395263672,
-      "learning_rate": 1.0600380876014835e-05,
-      "loss": 1.637,
+      "epoch": 0.59,
+      "grad_norm": 46.12936782836914,
+      "learning_rate": 1.607664309919257e-05,
+      "loss": 0.5357,
       "step": 4690
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 14.252532005310059,
-      "learning_rate": 1.0598376265410443e-05,
-      "loss": 0.7723,
+      "epoch": 0.59,
+      "grad_norm": 9.634956359863281,
+      "learning_rate": 1.6075806384135885e-05,
+      "loss": 1.4596,
       "step": 4691
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 16.23126220703125,
-      "learning_rate": 1.0596371654806055e-05,
-      "loss": 1.5121,
+      "epoch": 0.59,
+      "grad_norm": 4.42297887802124,
+      "learning_rate": 1.6074969669079195e-05,
+      "loss": 0.235,
       "step": 4692
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 26.036483764648438,
-      "learning_rate": 1.0594367044201665e-05,
-      "loss": 2.4891,
+      "epoch": 0.59,
+      "grad_norm": 13.291658401489258,
+      "learning_rate": 1.607413295402251e-05,
+      "loss": 2.0819,
       "step": 4693
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 11.481901168823242,
-      "learning_rate": 1.0592362433597273e-05,
-      "loss": 1.8866,
+      "epoch": 0.59,
+      "grad_norm": 12.573540687561035,
+      "learning_rate": 1.6073296238965823e-05,
+      "loss": 2.758,
       "step": 4694
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 15.858734130859375,
-      "learning_rate": 1.0590357822992885e-05,
-      "loss": 1.6499,
+      "epoch": 0.59,
+      "grad_norm": 23.449174880981445,
+      "learning_rate": 1.6072459523909133e-05,
+      "loss": 1.1348,
       "step": 4695
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 42.61335754394531,
-      "learning_rate": 1.0588353212388494e-05,
-      "loss": 2.7187,
+      "epoch": 0.59,
+      "grad_norm": 4.471280097961426,
+      "learning_rate": 1.6071622808852446e-05,
+      "loss": 0.6692,
       "step": 4696
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 18.403488159179688,
-      "learning_rate": 1.0586348601784104e-05,
-      "loss": 1.5052,
+      "epoch": 0.59,
+      "grad_norm": 12.927807807922363,
+      "learning_rate": 1.6070786093795757e-05,
+      "loss": 1.8158,
       "step": 4697
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 12.110722541809082,
-      "learning_rate": 1.0584343991179715e-05,
-      "loss": 1.499,
+      "epoch": 0.59,
+      "grad_norm": 22.802858352661133,
+      "learning_rate": 1.606994937873907e-05,
+      "loss": 3.4538,
       "step": 4698
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 22.188568115234375,
-      "learning_rate": 1.0582339380575324e-05,
-      "loss": 2.2641,
+      "epoch": 0.59,
+      "grad_norm": 14.071301460266113,
+      "learning_rate": 1.6069112663682384e-05,
+      "loss": 1.1313,
       "step": 4699
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 22.298145294189453,
-      "learning_rate": 1.0580334769970934e-05,
-      "loss": 1.9698,
+      "epoch": 0.59,
+      "grad_norm": 32.92894744873047,
+      "learning_rate": 1.6068275948625698e-05,
+      "loss": 3.067,
       "step": 4700
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 16.0262508392334,
-      "learning_rate": 1.0578330159366544e-05,
-      "loss": 1.4256,
+      "epoch": 0.59,
+      "grad_norm": 10.579645156860352,
+      "learning_rate": 1.6067439233569008e-05,
+      "loss": 1.6233,
       "step": 4701
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 7.636526584625244,
-      "learning_rate": 1.0576325548762154e-05,
-      "loss": 0.9309,
+      "epoch": 0.59,
+      "grad_norm": 44.29985809326172,
+      "learning_rate": 1.606660251851232e-05,
+      "loss": 1.7801,
       "step": 4702
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 16.564451217651367,
-      "learning_rate": 1.0574320938157762e-05,
-      "loss": 0.9893,
+      "epoch": 0.59,
+      "grad_norm": 15.755998611450195,
+      "learning_rate": 1.6065765803455635e-05,
+      "loss": 2.3343,
       "step": 4703
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 12.611454963684082,
-      "learning_rate": 1.0572316327553374e-05,
-      "loss": 1.1581,
+      "epoch": 0.59,
+      "grad_norm": 10.83920669555664,
+      "learning_rate": 1.6064929088398946e-05,
+      "loss": 2.1506,
       "step": 4704
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 19.108123779296875,
-      "learning_rate": 1.0570311716948984e-05,
-      "loss": 2.3804,
+      "epoch": 0.59,
+      "grad_norm": 37.44891357421875,
+      "learning_rate": 1.606409237334226e-05,
+      "loss": 3.1074,
       "step": 4705
     },
     {
-      "epoch": 1.41,
-      "grad_norm": 28.391921997070312,
-      "learning_rate": 1.0568307106344593e-05,
-      "loss": 1.5563,
+      "epoch": 0.59,
+      "grad_norm": 16.268556594848633,
+      "learning_rate": 1.6063255658285573e-05,
+      "loss": 2.9378,
       "step": 4706
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 14.404617309570312,
-      "learning_rate": 1.0566302495740205e-05,
-      "loss": 1.6261,
+      "epoch": 0.59,
+      "grad_norm": 7.230185031890869,
+      "learning_rate": 1.6062418943228886e-05,
+      "loss": 2.6047,
       "step": 4707
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 16.0683650970459,
-      "learning_rate": 1.0564297885135813e-05,
-      "loss": 1.5423,
+      "epoch": 0.59,
+      "grad_norm": 42.817569732666016,
+      "learning_rate": 1.6061582228172197e-05,
+      "loss": 2.9179,
       "step": 4708
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 15.672560691833496,
-      "learning_rate": 1.0562293274531423e-05,
-      "loss": 1.2338,
+      "epoch": 0.59,
+      "grad_norm": 11.923335075378418,
+      "learning_rate": 1.606074551311551e-05,
+      "loss": 1.2606,
       "step": 4709
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 63.09310531616211,
-      "learning_rate": 1.0560288663927035e-05,
-      "loss": 3.1458,
+      "epoch": 0.59,
+      "grad_norm": 18.071735382080078,
+      "learning_rate": 1.6059908798058824e-05,
+      "loss": 1.5856,
       "step": 4710
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 31.956239700317383,
-      "learning_rate": 1.0558284053322643e-05,
-      "loss": 1.1543,
+      "epoch": 0.59,
+      "grad_norm": 24.506731033325195,
+      "learning_rate": 1.6059072083002134e-05,
+      "loss": 2.0624,
       "step": 4711
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 18.227100372314453,
-      "learning_rate": 1.0556279442718253e-05,
-      "loss": 1.5385,
+      "epoch": 0.59,
+      "grad_norm": 17.046676635742188,
+      "learning_rate": 1.6058235367945448e-05,
+      "loss": 1.8085,
       "step": 4712
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 17.809656143188477,
-      "learning_rate": 1.0554274832113862e-05,
-      "loss": 1.6656,
+      "epoch": 0.59,
+      "grad_norm": 10.442744255065918,
+      "learning_rate": 1.605739865288876e-05,
+      "loss": 2.6301,
       "step": 4713
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 14.160889625549316,
-      "learning_rate": 1.0552270221509473e-05,
-      "loss": 1.4603,
+      "epoch": 0.59,
+      "grad_norm": 11.123990058898926,
+      "learning_rate": 1.6056561937832075e-05,
+      "loss": 2.6547,
       "step": 4714
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 19.26714324951172,
-      "learning_rate": 1.0550265610905082e-05,
-      "loss": 1.4625,
+      "epoch": 0.59,
+      "grad_norm": 7.235197067260742,
+      "learning_rate": 1.6055725222775385e-05,
+      "loss": 0.9485,
       "step": 4715
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 14.98919677734375,
-      "learning_rate": 1.0548261000300692e-05,
-      "loss": 1.7723,
+      "epoch": 0.59,
+      "grad_norm": 14.619832038879395,
+      "learning_rate": 1.60548885077187e-05,
+      "loss": 1.2491,
       "step": 4716
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 15.679314613342285,
-      "learning_rate": 1.0546256389696304e-05,
-      "loss": 2.7869,
+      "epoch": 0.59,
+      "grad_norm": 22.21745491027832,
+      "learning_rate": 1.605405179266201e-05,
+      "loss": 2.5148,
       "step": 4717
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 21.745820999145508,
-      "learning_rate": 1.0544251779091912e-05,
-      "loss": 1.7639,
+      "epoch": 0.59,
+      "grad_norm": 25.608308792114258,
+      "learning_rate": 1.6053215077605323e-05,
+      "loss": 3.2723,
       "step": 4718
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 36.71981430053711,
-      "learning_rate": 1.0542247168487522e-05,
-      "loss": 1.5607,
+      "epoch": 0.59,
+      "grad_norm": 12.341256141662598,
+      "learning_rate": 1.6052378362548637e-05,
+      "loss": 2.5153,
       "step": 4719
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 37.882877349853516,
-      "learning_rate": 1.054024255788313e-05,
-      "loss": 2.3179,
+      "epoch": 0.59,
+      "grad_norm": 8.019450187683105,
+      "learning_rate": 1.6051541647491947e-05,
+      "loss": 2.0929,
       "step": 4720
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 13.184647560119629,
-      "learning_rate": 1.0538237947278742e-05,
-      "loss": 2.0276,
+      "epoch": 0.59,
+      "grad_norm": 36.901058197021484,
+      "learning_rate": 1.605070493243526e-05,
+      "loss": 3.1348,
       "step": 4721
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 7.884564399719238,
-      "learning_rate": 1.053623333667435e-05,
-      "loss": 2.1323,
+      "epoch": 0.59,
+      "grad_norm": 15.31069278717041,
+      "learning_rate": 1.604986821737857e-05,
+      "loss": 2.0974,
       "step": 4722
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 13.331478118896484,
-      "learning_rate": 1.053422872606996e-05,
-      "loss": 1.2308,
+      "epoch": 0.59,
+      "grad_norm": 12.79121208190918,
+      "learning_rate": 1.6049031502321885e-05,
+      "loss": 1.4689,
       "step": 4723
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 10.356979370117188,
-      "learning_rate": 1.0532224115465573e-05,
-      "loss": 1.2406,
+      "epoch": 0.59,
+      "grad_norm": 10.151163101196289,
+      "learning_rate": 1.6048194787265198e-05,
+      "loss": 3.6915,
       "step": 4724
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 27.083091735839844,
-      "learning_rate": 1.0530219504861181e-05,
-      "loss": 1.3356,
+      "epoch": 0.59,
+      "grad_norm": 15.297480583190918,
+      "learning_rate": 1.604735807220851e-05,
+      "loss": 2.0763,
       "step": 4725
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 14.349416732788086,
-      "learning_rate": 1.0528214894256791e-05,
-      "loss": 1.337,
+      "epoch": 0.59,
+      "grad_norm": 10.322936058044434,
+      "learning_rate": 1.6046521357151822e-05,
+      "loss": 0.5036,
       "step": 4726
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 13.042084693908691,
-      "learning_rate": 1.0526210283652401e-05,
-      "loss": 1.3127,
+      "epoch": 0.59,
+      "grad_norm": 38.53228759765625,
+      "learning_rate": 1.6045684642095136e-05,
+      "loss": 3.0067,
       "step": 4727
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 39.22003936767578,
-      "learning_rate": 1.0524205673048011e-05,
-      "loss": 1.7421,
+      "epoch": 0.59,
+      "grad_norm": 8.970709800720215,
+      "learning_rate": 1.604484792703845e-05,
+      "loss": 0.9686,
       "step": 4728
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 7.622201919555664,
-      "learning_rate": 1.0522201062443623e-05,
-      "loss": 0.9831,
+      "epoch": 0.59,
+      "grad_norm": 8.248995780944824,
+      "learning_rate": 1.604401121198176e-05,
+      "loss": 0.9467,
       "step": 4729
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 31.560455322265625,
-      "learning_rate": 1.0520196451839231e-05,
-      "loss": 2.0777,
+      "epoch": 0.59,
+      "grad_norm": 5.467845439910889,
+      "learning_rate": 1.6043174496925073e-05,
+      "loss": 0.8413,
       "step": 4730
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 13.57126235961914,
-      "learning_rate": 1.0518191841234841e-05,
-      "loss": 1.4528,
+      "epoch": 0.59,
+      "grad_norm": 12.535849571228027,
+      "learning_rate": 1.6042337781868387e-05,
+      "loss": 1.2482,
       "step": 4731
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 25.839723587036133,
-      "learning_rate": 1.051618723063045e-05,
-      "loss": 1.3498,
+      "epoch": 0.59,
+      "grad_norm": 21.349803924560547,
+      "learning_rate": 1.6041501066811697e-05,
+      "loss": 3.1314,
       "step": 4732
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 8.843565940856934,
-      "learning_rate": 1.0514182620026062e-05,
-      "loss": 2.0511,
+      "epoch": 0.59,
+      "grad_norm": 15.141033172607422,
+      "learning_rate": 1.604066435175501e-05,
+      "loss": 3.0181,
       "step": 4733
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 13.93571949005127,
-      "learning_rate": 1.051217800942167e-05,
-      "loss": 1.4334,
+      "epoch": 0.59,
+      "grad_norm": 8.8250732421875,
+      "learning_rate": 1.6039827636698324e-05,
+      "loss": 1.0692,
       "step": 4734
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 32.184539794921875,
-      "learning_rate": 1.051017339881728e-05,
-      "loss": 2.1138,
+      "epoch": 0.59,
+      "grad_norm": 9.506110191345215,
+      "learning_rate": 1.6038990921641638e-05,
+      "loss": 1.1641,
       "step": 4735
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 12.731632232666016,
-      "learning_rate": 1.0508168788212892e-05,
-      "loss": 1.506,
+      "epoch": 0.59,
+      "grad_norm": 12.257560729980469,
+      "learning_rate": 1.603815420658495e-05,
+      "loss": 1.8662,
       "step": 4736
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 9.572550773620605,
-      "learning_rate": 1.05061641776085e-05,
-      "loss": 2.4816,
+      "epoch": 0.59,
+      "grad_norm": 24.33102035522461,
+      "learning_rate": 1.6037317491528262e-05,
+      "loss": 1.547,
       "step": 4737
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 13.963497161865234,
-      "learning_rate": 1.050415956700411e-05,
-      "loss": 1.4354,
+      "epoch": 0.59,
+      "grad_norm": 15.50571346282959,
+      "learning_rate": 1.6036480776471576e-05,
+      "loss": 1.1362,
       "step": 4738
     },
     {
-      "epoch": 1.42,
-      "grad_norm": 19.97682762145996,
-      "learning_rate": 1.0502154956399719e-05,
-      "loss": 1.7492,
+      "epoch": 0.59,
+      "grad_norm": 6.502542495727539,
+      "learning_rate": 1.6035644061414886e-05,
+      "loss": 2.4382,
       "step": 4739
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 11.23585033416748,
-      "learning_rate": 1.050015034579533e-05,
-      "loss": 1.396,
+      "epoch": 0.59,
+      "grad_norm": 23.697622299194336,
+      "learning_rate": 1.60348073463582e-05,
+      "loss": 0.6613,
       "step": 4740
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 21.9536075592041,
-      "learning_rate": 1.049814573519094e-05,
-      "loss": 1.1373,
+      "epoch": 0.59,
+      "grad_norm": 25.582416534423828,
+      "learning_rate": 1.6033970631301513e-05,
+      "loss": 1.6498,
       "step": 4741
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 15.670050621032715,
-      "learning_rate": 1.0496141124586549e-05,
-      "loss": 1.9185,
+      "epoch": 0.6,
+      "grad_norm": 38.71111297607422,
+      "learning_rate": 1.6033133916244824e-05,
+      "loss": 2.4338,
       "step": 4742
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 18.927589416503906,
-      "learning_rate": 1.049413651398216e-05,
-      "loss": 1.7097,
+      "epoch": 0.6,
+      "grad_norm": 7.77809476852417,
+      "learning_rate": 1.6032297201188137e-05,
+      "loss": 0.7501,
       "step": 4743
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 14.530732154846191,
-      "learning_rate": 1.049213190337777e-05,
-      "loss": 1.1204,
+      "epoch": 0.6,
+      "grad_norm": 9.764083862304688,
+      "learning_rate": 1.603146048613145e-05,
+      "loss": 2.6929,
       "step": 4744
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 29.094755172729492,
-      "learning_rate": 1.049012729277338e-05,
-      "loss": 1.787,
+      "epoch": 0.6,
+      "grad_norm": 28.068134307861328,
+      "learning_rate": 1.603062377107476e-05,
+      "loss": 1.5478,
       "step": 4745
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 25.101224899291992,
-      "learning_rate": 1.048812268216899e-05,
-      "loss": 2.8473,
+      "epoch": 0.6,
+      "grad_norm": 12.991608619689941,
+      "learning_rate": 1.6029787056018075e-05,
+      "loss": 1.7992,
       "step": 4746
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 17.750200271606445,
-      "learning_rate": 1.04861180715646e-05,
-      "loss": 1.8474,
+      "epoch": 0.6,
+      "grad_norm": 25.247655868530273,
+      "learning_rate": 1.602895034096139e-05,
+      "loss": 3.3891,
       "step": 4747
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 22.220870971679688,
-      "learning_rate": 1.048411346096021e-05,
-      "loss": 1.8063,
+      "epoch": 0.6,
+      "grad_norm": 18.69204330444336,
+      "learning_rate": 1.60281136259047e-05,
+      "loss": 2.7039,
       "step": 4748
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 7.666170120239258,
-      "learning_rate": 1.048210885035582e-05,
-      "loss": 1.4058,
+      "epoch": 0.6,
+      "grad_norm": 17.054393768310547,
+      "learning_rate": 1.6027276910848012e-05,
+      "loss": 2.4671,
       "step": 4749
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 28.113630294799805,
-      "learning_rate": 1.048010423975143e-05,
-      "loss": 2.1896,
+      "epoch": 0.6,
+      "grad_norm": 13.61137580871582,
+      "learning_rate": 1.6026440195791323e-05,
+      "loss": 1.3915,
       "step": 4750
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 13.658616065979004,
-      "learning_rate": 1.0478099629147038e-05,
-      "loss": 1.7455,
+      "epoch": 0.6,
+      "grad_norm": 118.35962677001953,
+      "learning_rate": 1.6025603480734636e-05,
+      "loss": 2.3395,
       "step": 4751
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 11.242774963378906,
-      "learning_rate": 1.047609501854265e-05,
-      "loss": 1.9807,
+      "epoch": 0.6,
+      "grad_norm": 17.12057876586914,
+      "learning_rate": 1.602476676567795e-05,
+      "loss": 0.4615,
       "step": 4752
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 11.421576499938965,
-      "learning_rate": 1.047409040793826e-05,
-      "loss": 1.5406,
+      "epoch": 0.6,
+      "grad_norm": 8.947701454162598,
+      "learning_rate": 1.602393005062126e-05,
+      "loss": 0.7204,
       "step": 4753
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 24.668371200561523,
-      "learning_rate": 1.0472085797333868e-05,
-      "loss": 2.2752,
+      "epoch": 0.6,
+      "grad_norm": 10.49992561340332,
+      "learning_rate": 1.6023093335564574e-05,
+      "loss": 0.4165,
       "step": 4754
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 24.07050132751465,
-      "learning_rate": 1.047008118672948e-05,
-      "loss": 1.4799,
+      "epoch": 0.6,
+      "grad_norm": 23.460378646850586,
+      "learning_rate": 1.6022256620507887e-05,
+      "loss": 2.3077,
       "step": 4755
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 38.27885055541992,
-      "learning_rate": 1.0468076576125088e-05,
-      "loss": 1.5369,
+      "epoch": 0.6,
+      "grad_norm": 17.311359405517578,
+      "learning_rate": 1.60214199054512e-05,
+      "loss": 2.3508,
       "step": 4756
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 20.224210739135742,
-      "learning_rate": 1.0466071965520699e-05,
-      "loss": 1.9254,
+      "epoch": 0.6,
+      "grad_norm": 9.50903034210205,
+      "learning_rate": 1.602058319039451e-05,
+      "loss": 3.0923,
       "step": 4757
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 10.304624557495117,
-      "learning_rate": 1.0464067354916307e-05,
-      "loss": 1.5751,
+      "epoch": 0.6,
+      "grad_norm": 25.622928619384766,
+      "learning_rate": 1.6019746475337825e-05,
+      "loss": 2.3855,
       "step": 4758
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 23.387136459350586,
-      "learning_rate": 1.0462062744311919e-05,
-      "loss": 1.2816,
+      "epoch": 0.6,
+      "grad_norm": 11.12957763671875,
+      "learning_rate": 1.601890976028114e-05,
+      "loss": 0.692,
       "step": 4759
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 8.415315628051758,
-      "learning_rate": 1.0460058133707529e-05,
-      "loss": 1.2024,
+      "epoch": 0.6,
+      "grad_norm": 15.091212272644043,
+      "learning_rate": 1.601807304522445e-05,
+      "loss": 1.551,
       "step": 4760
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 15.556396484375,
-      "learning_rate": 1.0458053523103137e-05,
-      "loss": 1.7196,
+      "epoch": 0.6,
+      "grad_norm": 10.973529815673828,
+      "learning_rate": 1.6017236330167763e-05,
+      "loss": 2.5152,
       "step": 4761
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 70.80048370361328,
-      "learning_rate": 1.0456048912498749e-05,
-      "loss": 1.6606,
+      "epoch": 0.6,
+      "grad_norm": 10.444022178649902,
+      "learning_rate": 1.6016399615111076e-05,
+      "loss": 0.7957,
       "step": 4762
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 8.032504081726074,
-      "learning_rate": 1.0454044301894357e-05,
-      "loss": 1.4636,
+      "epoch": 0.6,
+      "grad_norm": 16.062118530273438,
+      "learning_rate": 1.601556290005439e-05,
+      "loss": 1.7941,
       "step": 4763
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 13.399280548095703,
-      "learning_rate": 1.0452039691289967e-05,
-      "loss": 2.0113,
+      "epoch": 0.6,
+      "grad_norm": 10.109869956970215,
+      "learning_rate": 1.60147261849977e-05,
+      "loss": 3.889,
       "step": 4764
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 10.88759708404541,
-      "learning_rate": 1.0450035080685576e-05,
-      "loss": 1.7287,
+      "epoch": 0.6,
+      "grad_norm": 21.51214027404785,
+      "learning_rate": 1.6013889469941014e-05,
+      "loss": 2.4326,
       "step": 4765
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 46.82052230834961,
-      "learning_rate": 1.0448030470081188e-05,
-      "loss": 2.7536,
+      "epoch": 0.6,
+      "grad_norm": 23.204986572265625,
+      "learning_rate": 1.6013052754884327e-05,
+      "loss": 2.5666,
       "step": 4766
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 19.954078674316406,
-      "learning_rate": 1.0446025859476798e-05,
-      "loss": 1.6187,
+      "epoch": 0.6,
+      "grad_norm": 9.954808235168457,
+      "learning_rate": 1.6012216039827638e-05,
+      "loss": 1.3047,
       "step": 4767
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 46.83697509765625,
-      "learning_rate": 1.0444021248872406e-05,
-      "loss": 2.2134,
+      "epoch": 0.6,
+      "grad_norm": 13.861639976501465,
+      "learning_rate": 1.601137932477095e-05,
+      "loss": 3.0222,
       "step": 4768
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 29.12272834777832,
-      "learning_rate": 1.0442016638268018e-05,
-      "loss": 2.3703,
+      "epoch": 0.6,
+      "grad_norm": 16.373212814331055,
+      "learning_rate": 1.6010542609714265e-05,
+      "loss": 1.7138,
       "step": 4769
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 37.321956634521484,
-      "learning_rate": 1.0440012027663626e-05,
-      "loss": 1.2239,
+      "epoch": 0.6,
+      "grad_norm": 33.26377868652344,
+      "learning_rate": 1.6009705894657575e-05,
+      "loss": 4.864,
       "step": 4770
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 18.08650779724121,
-      "learning_rate": 1.0438007417059236e-05,
-      "loss": 1.4429,
+      "epoch": 0.6,
+      "grad_norm": 21.3576602935791,
+      "learning_rate": 1.600886917960089e-05,
+      "loss": 1.6901,
       "step": 4771
     },
     {
-      "epoch": 1.43,
-      "grad_norm": 15.265766143798828,
-      "learning_rate": 1.0436002806454848e-05,
-      "loss": 1.2304,
+      "epoch": 0.6,
+      "grad_norm": 15.211508750915527,
+      "learning_rate": 1.6008032464544202e-05,
+      "loss": 2.8331,
       "step": 4772
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 16.230012893676758,
-      "learning_rate": 1.0433998195850457e-05,
-      "loss": 1.2576,
+      "epoch": 0.6,
+      "grad_norm": 12.891839981079102,
+      "learning_rate": 1.6007195749487513e-05,
+      "loss": 1.9634,
       "step": 4773
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 27.102861404418945,
-      "learning_rate": 1.0431993585246067e-05,
-      "loss": 2.3902,
+      "epoch": 0.6,
+      "grad_norm": 25.414566040039062,
+      "learning_rate": 1.6006359034430826e-05,
+      "loss": 3.7175,
       "step": 4774
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 9.05881404876709,
-      "learning_rate": 1.0429988974641677e-05,
-      "loss": 0.8901,
+      "epoch": 0.6,
+      "grad_norm": 6.307236194610596,
+      "learning_rate": 1.6005522319374137e-05,
+      "loss": 3.0202,
       "step": 4775
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 24.15192985534668,
-      "learning_rate": 1.0427984364037287e-05,
-      "loss": 1.4729,
+      "epoch": 0.6,
+      "grad_norm": 14.140206336975098,
+      "learning_rate": 1.600468560431745e-05,
+      "loss": 1.3212,
       "step": 4776
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 11.738104820251465,
-      "learning_rate": 1.0425979753432895e-05,
-      "loss": 1.5979,
+      "epoch": 0.6,
+      "grad_norm": 4.296204566955566,
+      "learning_rate": 1.6003848889260764e-05,
+      "loss": 0.3131,
       "step": 4777
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 17.82452392578125,
-      "learning_rate": 1.0423975142828507e-05,
-      "loss": 1.2219,
+      "epoch": 0.6,
+      "grad_norm": 41.11643981933594,
+      "learning_rate": 1.6003012174204074e-05,
+      "loss": 3.8694,
       "step": 4778
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 10.79785442352295,
-      "learning_rate": 1.0421970532224117e-05,
-      "loss": 1.1237,
+      "epoch": 0.6,
+      "grad_norm": 8.5755033493042,
+      "learning_rate": 1.6002175459147388e-05,
+      "loss": 0.5878,
       "step": 4779
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 66.31954193115234,
-      "learning_rate": 1.0419965921619725e-05,
-      "loss": 1.1348,
+      "epoch": 0.6,
+      "grad_norm": 15.202603340148926,
+      "learning_rate": 1.60013387440907e-05,
+      "loss": 2.3258,
       "step": 4780
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 14.166802406311035,
-      "learning_rate": 1.0417961311015337e-05,
-      "loss": 1.5118,
+      "epoch": 0.6,
+      "grad_norm": 36.033565521240234,
+      "learning_rate": 1.6000502029034012e-05,
+      "loss": 4.0675,
       "step": 4781
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 19.842845916748047,
-      "learning_rate": 1.0415956700410946e-05,
-      "loss": 1.22,
+      "epoch": 0.6,
+      "grad_norm": 14.063431739807129,
+      "learning_rate": 1.5999665313977325e-05,
+      "loss": 2.0848,
       "step": 4782
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 19.89260482788086,
-      "learning_rate": 1.0413952089806556e-05,
-      "loss": 1.4466,
+      "epoch": 0.6,
+      "grad_norm": 19.439945220947266,
+      "learning_rate": 1.599882859892064e-05,
+      "loss": 1.6212,
       "step": 4783
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 8.727401733398438,
-      "learning_rate": 1.0411947479202167e-05,
-      "loss": 1.2331,
+      "epoch": 0.6,
+      "grad_norm": 8.303755760192871,
+      "learning_rate": 1.5997991883863953e-05,
+      "loss": 1.6547,
       "step": 4784
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 16.119182586669922,
-      "learning_rate": 1.0409942868597776e-05,
-      "loss": 2.1112,
+      "epoch": 0.6,
+      "grad_norm": 8.920239448547363,
+      "learning_rate": 1.5997155168807263e-05,
+      "loss": 3.9113,
       "step": 4785
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 8.426682472229004,
-      "learning_rate": 1.0407938257993386e-05,
-      "loss": 0.7954,
+      "epoch": 0.6,
+      "grad_norm": 19.899768829345703,
+      "learning_rate": 1.5996318453750577e-05,
+      "loss": 2.0214,
       "step": 4786
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 11.057311058044434,
-      "learning_rate": 1.0405933647388994e-05,
-      "loss": 1.4438,
+      "epoch": 0.6,
+      "grad_norm": 7.666860580444336,
+      "learning_rate": 1.599548173869389e-05,
+      "loss": 1.6604,
       "step": 4787
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 11.847160339355469,
-      "learning_rate": 1.0403929036784606e-05,
-      "loss": 1.6443,
+      "epoch": 0.6,
+      "grad_norm": 23.639904022216797,
+      "learning_rate": 1.59946450236372e-05,
+      "loss": 1.0322,
       "step": 4788
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 52.34111022949219,
-      "learning_rate": 1.0401924426180214e-05,
-      "loss": 1.8,
+      "epoch": 0.6,
+      "grad_norm": 31.30626106262207,
+      "learning_rate": 1.5993808308580514e-05,
+      "loss": 0.8971,
       "step": 4789
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 12.100299835205078,
-      "learning_rate": 1.0399919815575825e-05,
-      "loss": 0.9117,
+      "epoch": 0.6,
+      "grad_norm": 25.765701293945312,
+      "learning_rate": 1.5992971593523828e-05,
+      "loss": 1.9441,
       "step": 4790
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 25.9334774017334,
-      "learning_rate": 1.0397915204971436e-05,
-      "loss": 1.6836,
+      "epoch": 0.6,
+      "grad_norm": 10.381741523742676,
+      "learning_rate": 1.599213487846714e-05,
+      "loss": 0.5661,
       "step": 4791
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 16.446218490600586,
-      "learning_rate": 1.0395910594367045e-05,
-      "loss": 1.581,
+      "epoch": 0.6,
+      "grad_norm": 13.025772094726562,
+      "learning_rate": 1.5991298163410452e-05,
+      "loss": 1.9931,
       "step": 4792
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 16.541593551635742,
-      "learning_rate": 1.0393905983762655e-05,
-      "loss": 1.6899,
+      "epoch": 0.6,
+      "grad_norm": 13.313878059387207,
+      "learning_rate": 1.5990461448353765e-05,
+      "loss": 2.152,
       "step": 4793
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 14.012060165405273,
-      "learning_rate": 1.0391901373158265e-05,
-      "loss": 1.3736,
+      "epoch": 0.6,
+      "grad_norm": 12.876245498657227,
+      "learning_rate": 1.598962473329708e-05,
+      "loss": 1.0786,
       "step": 4794
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 18.858366012573242,
-      "learning_rate": 1.0389896762553875e-05,
-      "loss": 1.2005,
+      "epoch": 0.6,
+      "grad_norm": 12.202618598937988,
+      "learning_rate": 1.598878801824039e-05,
+      "loss": 1.5297,
       "step": 4795
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 11.462150573730469,
-      "learning_rate": 1.0387892151949485e-05,
-      "loss": 1.0013,
+      "epoch": 0.6,
+      "grad_norm": 6.869583606719971,
+      "learning_rate": 1.5987951303183703e-05,
+      "loss": 0.3643,
       "step": 4796
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 13.834492683410645,
-      "learning_rate": 1.0385887541345095e-05,
-      "loss": 0.8818,
+      "epoch": 0.6,
+      "grad_norm": 17.388463973999023,
+      "learning_rate": 1.5987114588127017e-05,
+      "loss": 2.3765,
       "step": 4797
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 44.42348861694336,
-      "learning_rate": 1.0383882930740705e-05,
-      "loss": 2.2629,
+      "epoch": 0.6,
+      "grad_norm": 22.159502029418945,
+      "learning_rate": 1.5986277873070327e-05,
+      "loss": 2.6772,
       "step": 4798
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 21.56435775756836,
-      "learning_rate": 1.0381878320136314e-05,
-      "loss": 2.646,
+      "epoch": 0.6,
+      "grad_norm": 10.427502632141113,
+      "learning_rate": 1.598544115801364e-05,
+      "loss": 2.6432,
       "step": 4799
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 12.599276542663574,
-      "learning_rate": 1.0379873709531925e-05,
-      "loss": 1.2467,
+      "epoch": 0.6,
+      "grad_norm": 15.760404586791992,
+      "learning_rate": 1.5984604442956954e-05,
+      "loss": 0.9616,
       "step": 4800
     },
     {
-      "epoch": 1.44,
-      "eval_loss": 0.20445984601974487,
-      "eval_runtime": 43.0646,
-      "eval_samples_per_second": 34.344,
-      "eval_steps_per_second": 34.344,
+      "epoch": 0.6,
+      "eval_loss": 0.11780498921871185,
+      "eval_runtime": 93.8932,
+      "eval_samples_per_second": 37.724,
+      "eval_steps_per_second": 37.724,
       "step": 4800
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 21.611980438232422,
-      "learning_rate": 1.0377869098927534e-05,
-      "loss": 1.8799,
+      "epoch": 0.6,
+      "grad_norm": 13.106200218200684,
+      "learning_rate": 1.5983767727900264e-05,
+      "loss": 2.8001,
       "step": 4801
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 26.111543655395508,
-      "learning_rate": 1.0375864488323144e-05,
-      "loss": 1.8642,
+      "epoch": 0.6,
+      "grad_norm": 19.5357723236084,
+      "learning_rate": 1.5982931012843578e-05,
+      "loss": 3.5563,
       "step": 4802
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 19.41090965270996,
-      "learning_rate": 1.0373859877718756e-05,
-      "loss": 1.8098,
+      "epoch": 0.6,
+      "grad_norm": 11.600570678710938,
+      "learning_rate": 1.598209429778689e-05,
+      "loss": 1.7516,
       "step": 4803
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 23.62296485900879,
-      "learning_rate": 1.0371855267114364e-05,
-      "loss": 2.0232,
+      "epoch": 0.6,
+      "grad_norm": 63.57748031616211,
+      "learning_rate": 1.5981257582730202e-05,
+      "loss": 1.5366,
       "step": 4804
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 25.114355087280273,
-      "learning_rate": 1.0369850656509974e-05,
-      "loss": 1.348,
+      "epoch": 0.6,
+      "grad_norm": 7.862771511077881,
+      "learning_rate": 1.5980420867673516e-05,
+      "loss": 1.0459,
       "step": 4805
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 15.635001182556152,
-      "learning_rate": 1.0367846045905583e-05,
-      "loss": 1.4803,
+      "epoch": 0.6,
+      "grad_norm": 28.971946716308594,
+      "learning_rate": 1.5979584152616826e-05,
+      "loss": 2.5234,
       "step": 4806
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 12.778668403625488,
-      "learning_rate": 1.0365841435301194e-05,
-      "loss": 1.837,
+      "epoch": 0.6,
+      "grad_norm": 16.047687530517578,
+      "learning_rate": 1.597874743756014e-05,
+      "loss": 1.0227,
       "step": 4807
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 19.424400329589844,
-      "learning_rate": 1.0363836824696803e-05,
-      "loss": 1.6641,
+      "epoch": 0.6,
+      "grad_norm": 13.568920135498047,
+      "learning_rate": 1.5977910722503453e-05,
+      "loss": 2.0609,
       "step": 4808
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 9.217851638793945,
-      "learning_rate": 1.0361832214092413e-05,
-      "loss": 0.925,
+      "epoch": 0.6,
+      "grad_norm": 11.14881420135498,
+      "learning_rate": 1.5977074007446763e-05,
+      "loss": 1.15,
       "step": 4809
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 26.674348831176758,
-      "learning_rate": 1.0359827603488025e-05,
-      "loss": 1.5977,
+      "epoch": 0.6,
+      "grad_norm": 13.92988109588623,
+      "learning_rate": 1.5976237292390077e-05,
+      "loss": 3.2552,
       "step": 4810
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 13.42351245880127,
-      "learning_rate": 1.0357822992883633e-05,
-      "loss": 1.3927,
+      "epoch": 0.6,
+      "grad_norm": 10.029170036315918,
+      "learning_rate": 1.597540057733339e-05,
+      "loss": 1.4787,
       "step": 4811
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 36.85655975341797,
-      "learning_rate": 1.0355818382279243e-05,
-      "loss": 1.8373,
+      "epoch": 0.6,
+      "grad_norm": 16.906906127929688,
+      "learning_rate": 1.5974563862276704e-05,
+      "loss": 1.2747,
       "step": 4812
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 49.87083435058594,
-      "learning_rate": 1.0353813771674851e-05,
-      "loss": 1.6954,
+      "epoch": 0.6,
+      "grad_norm": 28.421478271484375,
+      "learning_rate": 1.5973727147220015e-05,
+      "loss": 2.1904,
       "step": 4813
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 30.802719116210938,
-      "learning_rate": 1.0351809161070463e-05,
-      "loss": 2.0961,
+      "epoch": 0.6,
+      "grad_norm": 7.58961296081543,
+      "learning_rate": 1.5972890432163328e-05,
+      "loss": 1.2808,
       "step": 4814
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 22.26266860961914,
-      "learning_rate": 1.0349804550466073e-05,
-      "loss": 1.3555,
+      "epoch": 0.6,
+      "grad_norm": 19.453166961669922,
+      "learning_rate": 1.5972053717106642e-05,
+      "loss": 2.7334,
       "step": 4815
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 18.78736114501953,
-      "learning_rate": 1.0347799939861682e-05,
-      "loss": 1.4008,
+      "epoch": 0.6,
+      "grad_norm": 12.650189399719238,
+      "learning_rate": 1.5971217002049952e-05,
+      "loss": 2.8295,
       "step": 4816
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 10.189375877380371,
-      "learning_rate": 1.0345795329257293e-05,
-      "loss": 0.6236,
+      "epoch": 0.6,
+      "grad_norm": 14.469470977783203,
+      "learning_rate": 1.5970380286993266e-05,
+      "loss": 1.7788,
       "step": 4817
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 14.57485580444336,
-      "learning_rate": 1.0343790718652902e-05,
-      "loss": 1.3009,
+      "epoch": 0.6,
+      "grad_norm": 10.258919715881348,
+      "learning_rate": 1.596954357193658e-05,
+      "loss": 0.8568,
       "step": 4818
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 16.516605377197266,
-      "learning_rate": 1.0341786108048512e-05,
-      "loss": 1.3581,
+      "epoch": 0.6,
+      "grad_norm": 11.894331932067871,
+      "learning_rate": 1.596870685687989e-05,
+      "loss": 1.9034,
       "step": 4819
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 12.826788902282715,
-      "learning_rate": 1.0339781497444122e-05,
-      "loss": 1.9941,
+      "epoch": 0.6,
+      "grad_norm": 8.998530387878418,
+      "learning_rate": 1.5967870141823203e-05,
+      "loss": 1.0493,
       "step": 4820
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 43.45224380493164,
-      "learning_rate": 1.0337776886839732e-05,
-      "loss": 1.8419,
+      "epoch": 0.61,
+      "grad_norm": 14.476089477539062,
+      "learning_rate": 1.5967033426766517e-05,
+      "loss": 1.9384,
       "step": 4821
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 33.118247985839844,
-      "learning_rate": 1.0335772276235342e-05,
-      "loss": 1.7156,
+      "epoch": 0.61,
+      "grad_norm": 23.7738094329834,
+      "learning_rate": 1.596619671170983e-05,
+      "loss": 0.8177,
       "step": 4822
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 29.060731887817383,
-      "learning_rate": 1.0333767665630952e-05,
-      "loss": 1.5526,
+      "epoch": 0.61,
+      "grad_norm": 5.879161357879639,
+      "learning_rate": 1.596535999665314e-05,
+      "loss": 1.3281,
       "step": 4823
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 30.88375473022461,
-      "learning_rate": 1.0331763055026562e-05,
-      "loss": 1.9801,
+      "epoch": 0.61,
+      "grad_norm": 6.781850814819336,
+      "learning_rate": 1.5964523281596455e-05,
+      "loss": 2.4006,
       "step": 4824
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 26.168588638305664,
-      "learning_rate": 1.032975844442217e-05,
-      "loss": 1.63,
+      "epoch": 0.61,
+      "grad_norm": 13.156648635864258,
+      "learning_rate": 1.5963686566539768e-05,
+      "loss": 1.1909,
       "step": 4825
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 20.537208557128906,
-      "learning_rate": 1.0327753833817783e-05,
-      "loss": 1.582,
+      "epoch": 0.61,
+      "grad_norm": 51.71833801269531,
+      "learning_rate": 1.596284985148308e-05,
+      "loss": 2.3185,
       "step": 4826
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 15.51192855834961,
-      "learning_rate": 1.0325749223213393e-05,
-      "loss": 2.2813,
+      "epoch": 0.61,
+      "grad_norm": 15.952625274658203,
+      "learning_rate": 1.5962013136426392e-05,
+      "loss": 2.0639,
       "step": 4827
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 18.50126838684082,
-      "learning_rate": 1.0323744612609001e-05,
-      "loss": 1.5379,
+      "epoch": 0.61,
+      "grad_norm": 13.1610107421875,
+      "learning_rate": 1.5961176421369702e-05,
+      "loss": 1.4908,
       "step": 4828
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 19.359989166259766,
-      "learning_rate": 1.0321740002004613e-05,
-      "loss": 1.4196,
+      "epoch": 0.61,
+      "grad_norm": 7.100407600402832,
+      "learning_rate": 1.5960339706313016e-05,
+      "loss": 0.9432,
       "step": 4829
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 71.47821807861328,
-      "learning_rate": 1.0319735391400221e-05,
-      "loss": 2.627,
+      "epoch": 0.61,
+      "grad_norm": 11.82643985748291,
+      "learning_rate": 1.595950299125633e-05,
+      "loss": 1.1546,
       "step": 4830
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 8.43783950805664,
-      "learning_rate": 1.0317730780795831e-05,
-      "loss": 1.1434,
+      "epoch": 0.61,
+      "grad_norm": 38.60078811645508,
+      "learning_rate": 1.595866627619964e-05,
+      "loss": 2.5733,
       "step": 4831
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 49.501930236816406,
-      "learning_rate": 1.031572617019144e-05,
-      "loss": 2.0673,
+      "epoch": 0.61,
+      "grad_norm": 16.149267196655273,
+      "learning_rate": 1.5957829561142954e-05,
+      "loss": 2.1974,
       "step": 4832
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 11.1619234085083,
-      "learning_rate": 1.0313721559587051e-05,
-      "loss": 1.8033,
+      "epoch": 0.61,
+      "grad_norm": 15.542160987854004,
+      "learning_rate": 1.5956992846086264e-05,
+      "loss": 2.3165,
       "step": 4833
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 7.6200408935546875,
-      "learning_rate": 1.0311716948982662e-05,
-      "loss": 0.8928,
+      "epoch": 0.61,
+      "grad_norm": 11.998748779296875,
+      "learning_rate": 1.5956156131029578e-05,
+      "loss": 2.2605,
       "step": 4834
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 19.349905014038086,
-      "learning_rate": 1.030971233837827e-05,
-      "loss": 1.6486,
+      "epoch": 0.61,
+      "grad_norm": 9.115984916687012,
+      "learning_rate": 1.595531941597289e-05,
+      "loss": 1.1619,
       "step": 4835
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 8.616739273071289,
-      "learning_rate": 1.0307707727773882e-05,
-      "loss": 1.2562,
+      "epoch": 0.61,
+      "grad_norm": 33.47035217285156,
+      "learning_rate": 1.5954482700916205e-05,
+      "loss": 2.7648,
       "step": 4836
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 30.071773529052734,
-      "learning_rate": 1.030570311716949e-05,
-      "loss": 1.4047,
+      "epoch": 0.61,
+      "grad_norm": 8.653242111206055,
+      "learning_rate": 1.5953645985859515e-05,
+      "loss": 3.0944,
       "step": 4837
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 12.02312183380127,
-      "learning_rate": 1.03036985065651e-05,
-      "loss": 1.5015,
+      "epoch": 0.61,
+      "grad_norm": 6.583014965057373,
+      "learning_rate": 1.595280927080283e-05,
+      "loss": 2.3757,
       "step": 4838
     },
     {
-      "epoch": 1.45,
-      "grad_norm": 13.285801887512207,
-      "learning_rate": 1.0301693895960712e-05,
-      "loss": 1.4643,
+      "epoch": 0.61,
+      "grad_norm": 45.90184020996094,
+      "learning_rate": 1.5951972555746142e-05,
+      "loss": 2.0398,
       "step": 4839
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 20.811077117919922,
-      "learning_rate": 1.029968928535632e-05,
-      "loss": 1.8669,
+      "epoch": 0.61,
+      "grad_norm": 7.804754734039307,
+      "learning_rate": 1.5951135840689453e-05,
+      "loss": 2.4884,
       "step": 4840
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 21.803550720214844,
-      "learning_rate": 1.029768467475193e-05,
-      "loss": 2.0773,
+      "epoch": 0.61,
+      "grad_norm": 4.901381492614746,
+      "learning_rate": 1.5950299125632766e-05,
+      "loss": 0.7774,
       "step": 4841
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 35.751792907714844,
-      "learning_rate": 1.0295680064147539e-05,
-      "loss": 2.3893,
+      "epoch": 0.61,
+      "grad_norm": 20.78880500793457,
+      "learning_rate": 1.594946241057608e-05,
+      "loss": 1.8401,
       "step": 4842
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 14.943572044372559,
-      "learning_rate": 1.029367545354315e-05,
-      "loss": 1.9117,
+      "epoch": 0.61,
+      "grad_norm": 12.154851913452148,
+      "learning_rate": 1.5948625695519394e-05,
+      "loss": 2.0679,
       "step": 4843
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 19.128864288330078,
-      "learning_rate": 1.0291670842938759e-05,
-      "loss": 1.9138,
+      "epoch": 0.61,
+      "grad_norm": 16.952661514282227,
+      "learning_rate": 1.5947788980462704e-05,
+      "loss": 1.7988,
       "step": 4844
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 17.63079833984375,
-      "learning_rate": 1.028966623233437e-05,
-      "loss": 1.2719,
+      "epoch": 0.61,
+      "grad_norm": 6.197364330291748,
+      "learning_rate": 1.5946952265406018e-05,
+      "loss": 1.1393,
       "step": 4845
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 40.6061897277832,
-      "learning_rate": 1.028766162172998e-05,
-      "loss": 1.8636,
+      "epoch": 0.61,
+      "grad_norm": 61.77532196044922,
+      "learning_rate": 1.594611555034933e-05,
+      "loss": 3.0282,
       "step": 4846
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 11.293282508850098,
-      "learning_rate": 1.028565701112559e-05,
-      "loss": 1.1617,
+      "epoch": 0.61,
+      "grad_norm": 38.84717559814453,
+      "learning_rate": 1.594527883529264e-05,
+      "loss": 0.968,
       "step": 4847
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 7.563941955566406,
-      "learning_rate": 1.0283652400521201e-05,
-      "loss": 1.266,
+      "epoch": 0.61,
+      "grad_norm": 10.309555053710938,
+      "learning_rate": 1.5944442120235955e-05,
+      "loss": 2.0344,
       "step": 4848
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 37.318321228027344,
-      "learning_rate": 1.028164778991681e-05,
-      "loss": 4.1312,
+      "epoch": 0.61,
+      "grad_norm": 13.139561653137207,
+      "learning_rate": 1.594360540517927e-05,
+      "loss": 3.2425,
       "step": 4849
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 15.308874130249023,
-      "learning_rate": 1.027964317931242e-05,
-      "loss": 1.563,
+      "epoch": 0.61,
+      "grad_norm": 16.187753677368164,
+      "learning_rate": 1.5942768690122582e-05,
+      "loss": 2.3351,
       "step": 4850
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 12.4258451461792,
-      "learning_rate": 1.0277638568708028e-05,
-      "loss": 1.7674,
+      "epoch": 0.61,
+      "grad_norm": 9.564990997314453,
+      "learning_rate": 1.5941931975065893e-05,
+      "loss": 1.5682,
       "step": 4851
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 12.700484275817871,
-      "learning_rate": 1.027563395810364e-05,
-      "loss": 1.2176,
+      "epoch": 0.61,
+      "grad_norm": 19.982685089111328,
+      "learning_rate": 1.5941095260009206e-05,
+      "loss": 1.553,
       "step": 4852
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 26.99094581604004,
-      "learning_rate": 1.027362934749925e-05,
-      "loss": 1.5912,
+      "epoch": 0.61,
+      "grad_norm": 25.18014907836914,
+      "learning_rate": 1.5940258544952517e-05,
+      "loss": 3.4746,
       "step": 4853
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 52.29587936401367,
-      "learning_rate": 1.0271624736894858e-05,
-      "loss": 2.6716,
+      "epoch": 0.61,
+      "grad_norm": 20.708629608154297,
+      "learning_rate": 1.593942182989583e-05,
+      "loss": 1.016,
       "step": 4854
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 28.816272735595703,
-      "learning_rate": 1.026962012629047e-05,
-      "loss": 2.7605,
+      "epoch": 0.61,
+      "grad_norm": 11.345169067382812,
+      "learning_rate": 1.5938585114839144e-05,
+      "loss": 1.6501,
       "step": 4855
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 28.791126251220703,
-      "learning_rate": 1.0267615515686078e-05,
-      "loss": 1.8248,
+      "epoch": 0.61,
+      "grad_norm": 2.4988794326782227,
+      "learning_rate": 1.5937748399782454e-05,
+      "loss": 0.2278,
       "step": 4856
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 17.190807342529297,
-      "learning_rate": 1.0265610905081688e-05,
-      "loss": 1.6728,
+      "epoch": 0.61,
+      "grad_norm": 11.859201431274414,
+      "learning_rate": 1.5936911684725768e-05,
+      "loss": 1.9389,
       "step": 4857
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 17.33432960510254,
-      "learning_rate": 1.02636062944773e-05,
-      "loss": 1.427,
+      "epoch": 0.61,
+      "grad_norm": 6.510630130767822,
+      "learning_rate": 1.593607496966908e-05,
+      "loss": 1.506,
       "step": 4858
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 18.81448745727539,
-      "learning_rate": 1.0261601683872909e-05,
-      "loss": 1.4079,
+      "epoch": 0.61,
+      "grad_norm": 14.640543937683105,
+      "learning_rate": 1.593523825461239e-05,
+      "loss": 0.8049,
       "step": 4859
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 26.18571662902832,
-      "learning_rate": 1.0259597073268519e-05,
-      "loss": 1.3066,
+      "epoch": 0.61,
+      "grad_norm": 24.115324020385742,
+      "learning_rate": 1.5934401539555705e-05,
+      "loss": 2.1795,
       "step": 4860
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 22.120250701904297,
-      "learning_rate": 1.0257592462664127e-05,
-      "loss": 1.8271,
+      "epoch": 0.61,
+      "grad_norm": 23.66679573059082,
+      "learning_rate": 1.5933564824499016e-05,
+      "loss": 1.728,
       "step": 4861
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 19.113744735717773,
-      "learning_rate": 1.0255587852059739e-05,
-      "loss": 0.9674,
+      "epoch": 0.61,
+      "grad_norm": 14.935659408569336,
+      "learning_rate": 1.593272810944233e-05,
+      "loss": 1.4521,
       "step": 4862
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 15.18375301361084,
-      "learning_rate": 1.0253583241455347e-05,
-      "loss": 1.3229,
+      "epoch": 0.61,
+      "grad_norm": 16.17740821838379,
+      "learning_rate": 1.5931891394385643e-05,
+      "loss": 1.7848,
       "step": 4863
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 14.799633979797363,
-      "learning_rate": 1.0251578630850957e-05,
-      "loss": 1.1113,
+      "epoch": 0.61,
+      "grad_norm": 17.31148910522461,
+      "learning_rate": 1.5931054679328957e-05,
+      "loss": 0.8621,
       "step": 4864
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 13.80958366394043,
-      "learning_rate": 1.0249574020246569e-05,
-      "loss": 1.6729,
+      "epoch": 0.61,
+      "grad_norm": 13.355034828186035,
+      "learning_rate": 1.5930217964272267e-05,
+      "loss": 2.0393,
       "step": 4865
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 38.37821578979492,
-      "learning_rate": 1.0247569409642177e-05,
-      "loss": 1.5533,
+      "epoch": 0.61,
+      "grad_norm": 8.642890930175781,
+      "learning_rate": 1.592938124921558e-05,
+      "loss": 3.2727,
       "step": 4866
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 28.763893127441406,
-      "learning_rate": 1.0245564799037788e-05,
-      "loss": 1.699,
+      "epoch": 0.61,
+      "grad_norm": 26.613035202026367,
+      "learning_rate": 1.5928544534158894e-05,
+      "loss": 1.9293,
       "step": 4867
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 20.876445770263672,
-      "learning_rate": 1.0243560188433398e-05,
-      "loss": 1.3333,
+      "epoch": 0.61,
+      "grad_norm": 38.60970687866211,
+      "learning_rate": 1.5927707819102204e-05,
+      "loss": 3.1967,
       "step": 4868
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 52.730857849121094,
-      "learning_rate": 1.0241555577829008e-05,
-      "loss": 1.4987,
+      "epoch": 0.61,
+      "grad_norm": 14.778573989868164,
+      "learning_rate": 1.5926871104045518e-05,
+      "loss": 0.7463,
       "step": 4869
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 13.893447875976562,
-      "learning_rate": 1.0239550967224618e-05,
-      "loss": 1.6037,
+      "epoch": 0.61,
+      "grad_norm": 22.723390579223633,
+      "learning_rate": 1.592603438898883e-05,
+      "loss": 2.1331,
       "step": 4870
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 28.038808822631836,
-      "learning_rate": 1.0237546356620228e-05,
-      "loss": 3.0782,
+      "epoch": 0.61,
+      "grad_norm": 14.271210670471191,
+      "learning_rate": 1.5925197673932145e-05,
+      "loss": 2.4401,
       "step": 4871
     },
     {
-      "epoch": 1.46,
-      "grad_norm": 26.272857666015625,
-      "learning_rate": 1.0235541746015838e-05,
-      "loss": 1.7294,
+      "epoch": 0.61,
+      "grad_norm": 13.761286735534668,
+      "learning_rate": 1.5924360958875456e-05,
+      "loss": 1.13,
       "step": 4872
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 87.54938507080078,
-      "learning_rate": 1.0233537135411446e-05,
-      "loss": 2.8123,
+      "epoch": 0.61,
+      "grad_norm": 14.678537368774414,
+      "learning_rate": 1.592352424381877e-05,
+      "loss": 1.6376,
       "step": 4873
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 25.847455978393555,
-      "learning_rate": 1.0231532524807058e-05,
-      "loss": 1.3259,
+      "epoch": 0.61,
+      "grad_norm": 20.866357803344727,
+      "learning_rate": 1.5922687528762083e-05,
+      "loss": 1.9563,
       "step": 4874
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 10.328527450561523,
-      "learning_rate": 1.0229527914202666e-05,
-      "loss": 0.9254,
+      "epoch": 0.61,
+      "grad_norm": 8.459383964538574,
+      "learning_rate": 1.5921850813705393e-05,
+      "loss": 0.4451,
       "step": 4875
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 19.659297943115234,
-      "learning_rate": 1.0227523303598277e-05,
-      "loss": 1.1272,
+      "epoch": 0.61,
+      "grad_norm": 11.555710792541504,
+      "learning_rate": 1.5921014098648707e-05,
+      "loss": 2.7204,
       "step": 4876
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 25.877544403076172,
-      "learning_rate": 1.0225518692993888e-05,
-      "loss": 1.8762,
+      "epoch": 0.61,
+      "grad_norm": 5.4839253425598145,
+      "learning_rate": 1.592017738359202e-05,
+      "loss": 0.143,
       "step": 4877
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 22.409603118896484,
-      "learning_rate": 1.0223514082389497e-05,
-      "loss": 2.1701,
+      "epoch": 0.61,
+      "grad_norm": 15.43350887298584,
+      "learning_rate": 1.5919340668535334e-05,
+      "loss": 3.0729,
       "step": 4878
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 48.738155364990234,
-      "learning_rate": 1.0221509471785107e-05,
-      "loss": 2.1954,
+      "epoch": 0.61,
+      "grad_norm": 8.133819580078125,
+      "learning_rate": 1.5918503953478644e-05,
+      "loss": 1.1949,
       "step": 4879
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 13.183887481689453,
-      "learning_rate": 1.0219504861180715e-05,
-      "loss": 1.5571,
+      "epoch": 0.61,
+      "grad_norm": 17.95276641845703,
+      "learning_rate": 1.5917667238421958e-05,
+      "loss": 3.1655,
       "step": 4880
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 18.5463809967041,
-      "learning_rate": 1.0217500250576327e-05,
-      "loss": 1.5617,
+      "epoch": 0.61,
+      "grad_norm": 15.375577926635742,
+      "learning_rate": 1.5916830523365268e-05,
+      "loss": 2.6146,
       "step": 4881
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 22.341135025024414,
-      "learning_rate": 1.0215495639971937e-05,
-      "loss": 1.0747,
+      "epoch": 0.61,
+      "grad_norm": 24.520431518554688,
+      "learning_rate": 1.5915993808308582e-05,
+      "loss": 2.0791,
       "step": 4882
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 10.601929664611816,
-      "learning_rate": 1.0213491029367545e-05,
-      "loss": 1.539,
+      "epoch": 0.61,
+      "grad_norm": 8.88187313079834,
+      "learning_rate": 1.5915157093251896e-05,
+      "loss": 1.4884,
       "step": 4883
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 12.355496406555176,
-      "learning_rate": 1.0211486418763157e-05,
-      "loss": 1.3754,
+      "epoch": 0.61,
+      "grad_norm": 10.687667846679688,
+      "learning_rate": 1.5914320378195206e-05,
+      "loss": 0.5388,
       "step": 4884
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 11.556707382202148,
-      "learning_rate": 1.0209481808158766e-05,
-      "loss": 1.5293,
+      "epoch": 0.61,
+      "grad_norm": 10.427061080932617,
+      "learning_rate": 1.591348366313852e-05,
+      "loss": 1.0136,
       "step": 4885
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 19.798812866210938,
-      "learning_rate": 1.0207477197554376e-05,
-      "loss": 0.9688,
+      "epoch": 0.61,
+      "grad_norm": 11.740730285644531,
+      "learning_rate": 1.591264694808183e-05,
+      "loss": 1.8335,
       "step": 4886
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 13.712343215942383,
-      "learning_rate": 1.0205472586949984e-05,
-      "loss": 1.5041,
+      "epoch": 0.61,
+      "grad_norm": 13.261872291564941,
+      "learning_rate": 1.5911810233025143e-05,
+      "loss": 2.3499,
       "step": 4887
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 10.897645950317383,
-      "learning_rate": 1.0203467976345596e-05,
-      "loss": 1.0905,
+      "epoch": 0.61,
+      "grad_norm": 28.052698135375977,
+      "learning_rate": 1.5910973517968457e-05,
+      "loss": 2.8952,
       "step": 4888
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 28.161396026611328,
-      "learning_rate": 1.0201463365741206e-05,
-      "loss": 1.7225,
+      "epoch": 0.61,
+      "grad_norm": 7.432820796966553,
+      "learning_rate": 1.5910136802911767e-05,
+      "loss": 1.5737,
       "step": 4889
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 34.16568374633789,
-      "learning_rate": 1.0199458755136814e-05,
-      "loss": 2.0862,
+      "epoch": 0.61,
+      "grad_norm": 6.619985103607178,
+      "learning_rate": 1.590930008785508e-05,
+      "loss": 0.5595,
       "step": 4890
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 19.398710250854492,
-      "learning_rate": 1.0197454144532426e-05,
-      "loss": 1.5471,
+      "epoch": 0.61,
+      "grad_norm": 11.153505325317383,
+      "learning_rate": 1.5908463372798395e-05,
+      "loss": 2.432,
       "step": 4891
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 24.547252655029297,
-      "learning_rate": 1.0195449533928035e-05,
-      "loss": 3.0927,
+      "epoch": 0.61,
+      "grad_norm": 11.692172050476074,
+      "learning_rate": 1.5907626657741708e-05,
+      "loss": 0.3043,
       "step": 4892
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 11.176749229431152,
-      "learning_rate": 1.0193444923323645e-05,
-      "loss": 1.1639,
+      "epoch": 0.61,
+      "grad_norm": 8.810521125793457,
+      "learning_rate": 1.590678994268502e-05,
+      "loss": 1.3017,
       "step": 4893
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 50.76742172241211,
-      "learning_rate": 1.0191440312719255e-05,
-      "loss": 2.6678,
+      "epoch": 0.61,
+      "grad_norm": 12.059951782226562,
+      "learning_rate": 1.5905953227628332e-05,
+      "loss": 1.153,
       "step": 4894
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 9.156387329101562,
-      "learning_rate": 1.0189435702114865e-05,
-      "loss": 0.9565,
+      "epoch": 0.61,
+      "grad_norm": 9.708504676818848,
+      "learning_rate": 1.5905116512571646e-05,
+      "loss": 1.6417,
       "step": 4895
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 23.076967239379883,
-      "learning_rate": 1.0187431091510475e-05,
-      "loss": 1.6497,
+      "epoch": 0.61,
+      "grad_norm": 7.3110833168029785,
+      "learning_rate": 1.5904279797514956e-05,
+      "loss": 0.9502,
       "step": 4896
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 16.60446548461914,
-      "learning_rate": 1.0185426480906085e-05,
-      "loss": 1.5919,
+      "epoch": 0.61,
+      "grad_norm": 4.893747329711914,
+      "learning_rate": 1.590344308245827e-05,
+      "loss": 0.491,
       "step": 4897
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 12.027603149414062,
-      "learning_rate": 1.0183421870301695e-05,
-      "loss": 2.2724,
+      "epoch": 0.61,
+      "grad_norm": 9.931058883666992,
+      "learning_rate": 1.5902606367401583e-05,
+      "loss": 1.3111,
       "step": 4898
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 18.373804092407227,
-      "learning_rate": 1.0181417259697303e-05,
-      "loss": 0.8864,
+      "epoch": 0.61,
+      "grad_norm": 10.12717056274414,
+      "learning_rate": 1.5901769652344897e-05,
+      "loss": 1.4998,
       "step": 4899
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 33.86744689941406,
-      "learning_rate": 1.0179412649092915e-05,
-      "loss": 2.4096,
+      "epoch": 0.61,
+      "grad_norm": 9.866714477539062,
+      "learning_rate": 1.5900932937288207e-05,
+      "loss": 2.3386,
       "step": 4900
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 19.00617218017578,
-      "learning_rate": 1.0177408038488525e-05,
-      "loss": 1.4766,
+      "epoch": 0.62,
+      "grad_norm": 5.476425647735596,
+      "learning_rate": 1.590009622223152e-05,
+      "loss": 1.6813,
       "step": 4901
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 10.875265121459961,
-      "learning_rate": 1.0175403427884134e-05,
-      "loss": 1.6248,
+      "epoch": 0.62,
+      "grad_norm": 9.888223648071289,
+      "learning_rate": 1.5899259507174835e-05,
+      "loss": 2.1962,
       "step": 4902
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 18.423274993896484,
-      "learning_rate": 1.0173398817279745e-05,
-      "loss": 2.1238,
+      "epoch": 0.62,
+      "grad_norm": 17.479103088378906,
+      "learning_rate": 1.5898422792118145e-05,
+      "loss": 2.7827,
       "step": 4903
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 14.276087760925293,
-      "learning_rate": 1.0171394206675354e-05,
-      "loss": 1.471,
+      "epoch": 0.62,
+      "grad_norm": 23.746028900146484,
+      "learning_rate": 1.589758607706146e-05,
+      "loss": 2.0941,
       "step": 4904
     },
     {
-      "epoch": 1.47,
-      "grad_norm": 8.41922664642334,
-      "learning_rate": 1.0169389596070964e-05,
-      "loss": 0.8771,
+      "epoch": 0.62,
+      "grad_norm": 8.534255981445312,
+      "learning_rate": 1.5896749362004772e-05,
+      "loss": 1.9096,
       "step": 4905
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 17.14394760131836,
-      "learning_rate": 1.0167384985466572e-05,
-      "loss": 1.0387,
+      "epoch": 0.62,
+      "grad_norm": 8.96469783782959,
+      "learning_rate": 1.5895912646948082e-05,
+      "loss": 1.7232,
       "step": 4906
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 20.45330047607422,
-      "learning_rate": 1.0165380374862184e-05,
-      "loss": 1.9674,
+      "epoch": 0.62,
+      "grad_norm": 17.049060821533203,
+      "learning_rate": 1.5895075931891396e-05,
+      "loss": 1.6946,
       "step": 4907
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 6.636265754699707,
-      "learning_rate": 1.0163375764257794e-05,
-      "loss": 0.3816,
+      "epoch": 0.62,
+      "grad_norm": 38.09623718261719,
+      "learning_rate": 1.589423921683471e-05,
+      "loss": 4.0612,
       "step": 4908
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 15.284008979797363,
-      "learning_rate": 1.0161371153653403e-05,
-      "loss": 1.6652,
+      "epoch": 0.62,
+      "grad_norm": 12.76835823059082,
+      "learning_rate": 1.589340250177802e-05,
+      "loss": 3.0702,
       "step": 4909
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 16.42159080505371,
-      "learning_rate": 1.0159366543049014e-05,
-      "loss": 2.3129,
+      "epoch": 0.62,
+      "grad_norm": 79.78744506835938,
+      "learning_rate": 1.5892565786721334e-05,
+      "loss": 2.443,
       "step": 4910
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 19.10989761352539,
-      "learning_rate": 1.0157361932444623e-05,
-      "loss": 1.8581,
+      "epoch": 0.62,
+      "grad_norm": 102.937744140625,
+      "learning_rate": 1.5891729071664647e-05,
+      "loss": 1.7817,
       "step": 4911
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 30.188232421875,
-      "learning_rate": 1.0155357321840233e-05,
-      "loss": 1.5959,
+      "epoch": 0.62,
+      "grad_norm": 12.582338333129883,
+      "learning_rate": 1.5890892356607957e-05,
+      "loss": 1.7731,
       "step": 4912
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 12.307696342468262,
-      "learning_rate": 1.0153352711235845e-05,
-      "loss": 1.6547,
+      "epoch": 0.62,
+      "grad_norm": 14.361742973327637,
+      "learning_rate": 1.589005564155127e-05,
+      "loss": 1.8241,
       "step": 4913
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 12.962136268615723,
-      "learning_rate": 1.0151348100631453e-05,
-      "loss": 1.8137,
+      "epoch": 0.62,
+      "grad_norm": 13.8981294631958,
+      "learning_rate": 1.588921892649458e-05,
+      "loss": 0.2828,
       "step": 4914
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 10.3368558883667,
-      "learning_rate": 1.0149343490027063e-05,
-      "loss": 0.9236,
+      "epoch": 0.62,
+      "grad_norm": 73.4322280883789,
+      "learning_rate": 1.5888382211437895e-05,
+      "loss": 3.2836,
       "step": 4915
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 20.91614532470703,
-      "learning_rate": 1.0147338879422673e-05,
-      "loss": 1.7155,
+      "epoch": 0.62,
+      "grad_norm": 18.202783584594727,
+      "learning_rate": 1.588754549638121e-05,
+      "loss": 1.3835,
       "step": 4916
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 38.6691780090332,
-      "learning_rate": 1.0145334268818283e-05,
-      "loss": 2.1508,
+      "epoch": 0.62,
+      "grad_norm": 15.362526893615723,
+      "learning_rate": 1.588670878132452e-05,
+      "loss": 2.5685,
       "step": 4917
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 9.685796737670898,
-      "learning_rate": 1.0143329658213892e-05,
-      "loss": 1.7707,
+      "epoch": 0.62,
+      "grad_norm": 24.820188522338867,
+      "learning_rate": 1.5885872066267833e-05,
+      "loss": 2.3165,
       "step": 4918
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 14.471162796020508,
-      "learning_rate": 1.0141325047609503e-05,
-      "loss": 1.5946,
+      "epoch": 0.62,
+      "grad_norm": 110.93339538574219,
+      "learning_rate": 1.5885035351211146e-05,
+      "loss": 3.0482,
       "step": 4919
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 18.22927474975586,
-      "learning_rate": 1.0139320437005114e-05,
-      "loss": 1.2375,
-      "step": 4920
-    },
-    {
-      "epoch": 1.48,
-      "eval_loss": 0.2065747231245041,
-      "eval_runtime": 43.2855,
-      "eval_samples_per_second": 34.169,
-      "eval_steps_per_second": 34.169,
+      "epoch": 0.62,
+      "grad_norm": 9.03571891784668,
+      "learning_rate": 1.588419863615446e-05,
+      "loss": 1.4118,
       "step": 4920
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 51.773658752441406,
-      "learning_rate": 1.0137315826400722e-05,
-      "loss": 0.986,
+      "epoch": 0.62,
+      "grad_norm": 49.11772155761719,
+      "learning_rate": 1.588336192109777e-05,
+      "loss": 1.68,
       "step": 4921
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 105.13095092773438,
-      "learning_rate": 1.0135311215796334e-05,
-      "loss": 2.8983,
+      "epoch": 0.62,
+      "grad_norm": 9.715534210205078,
+      "learning_rate": 1.5882525206041084e-05,
+      "loss": 0.9393,
       "step": 4922
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 14.783707618713379,
-      "learning_rate": 1.0133306605191942e-05,
-      "loss": 1.9219,
+      "epoch": 0.62,
+      "grad_norm": 11.484861373901367,
+      "learning_rate": 1.5881688490984397e-05,
+      "loss": 2.2139,
       "step": 4923
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 8.431492805480957,
-      "learning_rate": 1.0131301994587552e-05,
-      "loss": 0.9575,
+      "epoch": 0.62,
+      "grad_norm": 6.814791202545166,
+      "learning_rate": 1.5880851775927708e-05,
+      "loss": 1.5921,
       "step": 4924
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 14.085335731506348,
-      "learning_rate": 1.0129297383983164e-05,
-      "loss": 1.2251,
+      "epoch": 0.62,
+      "grad_norm": 21.735166549682617,
+      "learning_rate": 1.588001506087102e-05,
+      "loss": 1.5475,
       "step": 4925
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 17.102575302124023,
-      "learning_rate": 1.0127292773378772e-05,
-      "loss": 1.4003,
+      "epoch": 0.62,
+      "grad_norm": 21.243627548217773,
+      "learning_rate": 1.5879178345814335e-05,
+      "loss": 2.0279,
       "step": 4926
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 13.873455047607422,
-      "learning_rate": 1.0125288162774382e-05,
-      "loss": 1.3603,
+      "epoch": 0.62,
+      "grad_norm": 11.919015884399414,
+      "learning_rate": 1.587834163075765e-05,
+      "loss": 1.765,
       "step": 4927
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 8.495047569274902,
-      "learning_rate": 1.012328355216999e-05,
-      "loss": 1.8705,
+      "epoch": 0.62,
+      "grad_norm": 33.100257873535156,
+      "learning_rate": 1.587750491570096e-05,
+      "loss": 1.8073,
       "step": 4928
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 12.480233192443848,
-      "learning_rate": 1.0121278941565603e-05,
-      "loss": 1.6846,
+      "epoch": 0.62,
+      "grad_norm": 11.556069374084473,
+      "learning_rate": 1.5876668200644273e-05,
+      "loss": 1.0835,
       "step": 4929
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 9.334129333496094,
-      "learning_rate": 1.0119274330961211e-05,
-      "loss": 1.924,
+      "epoch": 0.62,
+      "grad_norm": 26.59724998474121,
+      "learning_rate": 1.5875831485587586e-05,
+      "loss": 4.0394,
       "step": 4930
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 12.200242042541504,
-      "learning_rate": 1.0117269720356821e-05,
-      "loss": 1.4727,
+      "epoch": 0.62,
+      "grad_norm": 23.04380989074707,
+      "learning_rate": 1.5874994770530896e-05,
+      "loss": 2.3475,
       "step": 4931
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 10.241484642028809,
-      "learning_rate": 1.0115265109752433e-05,
-      "loss": 1.2527,
+      "epoch": 0.62,
+      "grad_norm": 14.799032211303711,
+      "learning_rate": 1.587415805547421e-05,
+      "loss": 1.123,
       "step": 4932
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 18.393444061279297,
-      "learning_rate": 1.0113260499148041e-05,
-      "loss": 1.9137,
+      "epoch": 0.62,
+      "grad_norm": 3.6003031730651855,
+      "learning_rate": 1.5873321340417524e-05,
+      "loss": 0.2666,
       "step": 4933
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 30.996809005737305,
-      "learning_rate": 1.0111255888543651e-05,
-      "loss": 1.9626,
+      "epoch": 0.62,
+      "grad_norm": 18.916271209716797,
+      "learning_rate": 1.5872484625360834e-05,
+      "loss": 3.6269,
       "step": 4934
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 22.243946075439453,
-      "learning_rate": 1.010925127793926e-05,
-      "loss": 1.4565,
+      "epoch": 0.62,
+      "grad_norm": 10.74314022064209,
+      "learning_rate": 1.5871647910304148e-05,
+      "loss": 1.4515,
       "step": 4935
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 10.407564163208008,
-      "learning_rate": 1.0107246667334871e-05,
-      "loss": 1.0015,
+      "epoch": 0.62,
+      "grad_norm": 24.510046005249023,
+      "learning_rate": 1.587081119524746e-05,
+      "loss": 3.0281,
       "step": 4936
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 17.139793395996094,
-      "learning_rate": 1.010524205673048e-05,
-      "loss": 1.3928,
+      "epoch": 0.62,
+      "grad_norm": 7.19443416595459,
+      "learning_rate": 1.586997448019077e-05,
+      "loss": 0.3961,
       "step": 4937
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 34.3438606262207,
-      "learning_rate": 1.010323744612609e-05,
-      "loss": 1.5863,
+      "epoch": 0.62,
+      "grad_norm": 8.771178245544434,
+      "learning_rate": 1.5869137765134085e-05,
+      "loss": 2.8149,
       "step": 4938
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 14.947382926940918,
-      "learning_rate": 1.0101232835521702e-05,
-      "loss": 1.3544,
+      "epoch": 0.62,
+      "grad_norm": 7.240764141082764,
+      "learning_rate": 1.5868301050077395e-05,
+      "loss": 1.5047,
       "step": 4939
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 11.249651908874512,
-      "learning_rate": 1.009922822491731e-05,
-      "loss": 1.3051,
+      "epoch": 0.62,
+      "grad_norm": 43.71602249145508,
+      "learning_rate": 1.586746433502071e-05,
+      "loss": 2.9675,
       "step": 4940
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 16.109973907470703,
-      "learning_rate": 1.009722361431292e-05,
-      "loss": 2.0421,
+      "epoch": 0.62,
+      "grad_norm": 31.66825294494629,
+      "learning_rate": 1.5866627619964023e-05,
+      "loss": 4.119,
       "step": 4941
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 14.892985343933105,
-      "learning_rate": 1.009521900370853e-05,
-      "loss": 1.1327,
+      "epoch": 0.62,
+      "grad_norm": 10.924080848693848,
+      "learning_rate": 1.5865790904907333e-05,
+      "loss": 1.1588,
       "step": 4942
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 27.952659606933594,
-      "learning_rate": 1.009321439310414e-05,
-      "loss": 2.3613,
+      "epoch": 0.62,
+      "grad_norm": 7.768029689788818,
+      "learning_rate": 1.5864954189850647e-05,
+      "loss": 1.9436,
       "step": 4943
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 27.172393798828125,
-      "learning_rate": 1.009120978249975e-05,
-      "loss": 2.1473,
+      "epoch": 0.62,
+      "grad_norm": 18.65514373779297,
+      "learning_rate": 1.586411747479396e-05,
+      "loss": 3.5025,
       "step": 4944
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 9.170857429504395,
-      "learning_rate": 1.008920517189536e-05,
-      "loss": 1.2239,
+      "epoch": 0.62,
+      "grad_norm": 17.466522216796875,
+      "learning_rate": 1.586328075973727e-05,
+      "loss": 2.2984,
       "step": 4945
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 11.424710273742676,
-      "learning_rate": 1.008720056129097e-05,
-      "loss": 0.8955,
+      "epoch": 0.62,
+      "grad_norm": 5.136081695556641,
+      "learning_rate": 1.5862444044680584e-05,
+      "loss": 0.314,
       "step": 4946
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 13.28514289855957,
-      "learning_rate": 1.0085195950686579e-05,
-      "loss": 1.6111,
+      "epoch": 0.62,
+      "grad_norm": 9.889122009277344,
+      "learning_rate": 1.5861607329623898e-05,
+      "loss": 1.0442,
       "step": 4947
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 32.51649475097656,
-      "learning_rate": 1.008319134008219e-05,
-      "loss": 1.9871,
+      "epoch": 0.62,
+      "grad_norm": 30.280052185058594,
+      "learning_rate": 1.586077061456721e-05,
+      "loss": 1.293,
       "step": 4948
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 15.511556625366211,
-      "learning_rate": 1.00811867294778e-05,
-      "loss": 1.37,
+      "epoch": 0.62,
+      "grad_norm": 25.252389907836914,
+      "learning_rate": 1.5859933899510522e-05,
+      "loss": 3.7356,
       "step": 4949
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 16.7358341217041,
-      "learning_rate": 1.007918211887341e-05,
-      "loss": 1.6897,
+      "epoch": 0.62,
+      "grad_norm": 20.295284271240234,
+      "learning_rate": 1.5859097184453835e-05,
+      "loss": 1.943,
       "step": 4950
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 13.070718765258789,
-      "learning_rate": 1.0077177508269021e-05,
-      "loss": 0.666,
+      "epoch": 0.62,
+      "grad_norm": 12.399702072143555,
+      "learning_rate": 1.585826046939715e-05,
+      "loss": 1.38,
       "step": 4951
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 22.743227005004883,
-      "learning_rate": 1.007517289766463e-05,
-      "loss": 1.8756,
+      "epoch": 0.62,
+      "grad_norm": 16.72775650024414,
+      "learning_rate": 1.585742375434046e-05,
+      "loss": 1.3178,
       "step": 4952
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 20.257579803466797,
-      "learning_rate": 1.007316828706024e-05,
-      "loss": 1.9218,
+      "epoch": 0.62,
+      "grad_norm": 10.180647850036621,
+      "learning_rate": 1.5856587039283773e-05,
+      "loss": 0.878,
       "step": 4953
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 32.02715301513672,
-      "learning_rate": 1.0071163676455848e-05,
-      "loss": 1.2685,
+      "epoch": 0.62,
+      "grad_norm": 30.216543197631836,
+      "learning_rate": 1.5855750324227087e-05,
+      "loss": 2.0368,
       "step": 4954
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 15.203670501708984,
-      "learning_rate": 1.006915906585146e-05,
-      "loss": 1.5685,
+      "epoch": 0.62,
+      "grad_norm": 17.116146087646484,
+      "learning_rate": 1.58549136091704e-05,
+      "loss": 1.4108,
       "step": 4955
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 7.677938461303711,
-      "learning_rate": 1.006715445524707e-05,
-      "loss": 1.0445,
+      "epoch": 0.62,
+      "grad_norm": 6.7107768058776855,
+      "learning_rate": 1.585407689411371e-05,
+      "loss": 1.7707,
       "step": 4956
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 24.447038650512695,
-      "learning_rate": 1.0065149844642678e-05,
-      "loss": 1.7323,
+      "epoch": 0.62,
+      "grad_norm": 10.454689025878906,
+      "learning_rate": 1.5853240179057024e-05,
+      "loss": 1.6508,
       "step": 4957
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 12.981391906738281,
-      "learning_rate": 1.006314523403829e-05,
-      "loss": 1.9296,
+      "epoch": 0.62,
+      "grad_norm": 9.240266799926758,
+      "learning_rate": 1.5852403464000338e-05,
+      "loss": 1.6171,
       "step": 4958
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 38.580570220947266,
-      "learning_rate": 1.0061140623433898e-05,
-      "loss": 1.909,
+      "epoch": 0.62,
+      "grad_norm": 34.222286224365234,
+      "learning_rate": 1.5851566748943648e-05,
+      "loss": 1.233,
       "step": 4959
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 11.870725631713867,
-      "learning_rate": 1.0059136012829508e-05,
-      "loss": 1.3958,
+      "epoch": 0.62,
+      "grad_norm": 13.324649810791016,
+      "learning_rate": 1.5850730033886962e-05,
+      "loss": 1.1105,
       "step": 4960
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 7.389049530029297,
-      "learning_rate": 1.0057131402225117e-05,
-      "loss": 0.8741,
+      "epoch": 0.62,
+      "grad_norm": 15.83174991607666,
+      "learning_rate": 1.5849893318830275e-05,
+      "loss": 1.5953,
       "step": 4961
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 12.032258033752441,
-      "learning_rate": 1.0055126791620729e-05,
-      "loss": 1.4206,
+      "epoch": 0.62,
+      "grad_norm": 16.952129364013672,
+      "learning_rate": 1.5849056603773586e-05,
+      "loss": 2.568,
       "step": 4962
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 22.42498207092285,
-      "learning_rate": 1.0053122181016339e-05,
-      "loss": 2.5246,
+      "epoch": 0.62,
+      "grad_norm": 6.830140590667725,
+      "learning_rate": 1.58482198887169e-05,
+      "loss": 1.5096,
       "step": 4963
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 12.639444351196289,
-      "learning_rate": 1.0051117570411949e-05,
-      "loss": 1.8031,
+      "epoch": 0.62,
+      "grad_norm": 9.433883666992188,
+      "learning_rate": 1.584738317366021e-05,
+      "loss": 2.3211,
       "step": 4964
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 33.977386474609375,
-      "learning_rate": 1.0049112959807559e-05,
-      "loss": 1.595,
+      "epoch": 0.62,
+      "grad_norm": 33.27182388305664,
+      "learning_rate": 1.5846546458603523e-05,
+      "loss": 4.0639,
       "step": 4965
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 13.614667892456055,
-      "learning_rate": 1.0047108349203167e-05,
-      "loss": 0.975,
+      "epoch": 0.62,
+      "grad_norm": 15.093009948730469,
+      "learning_rate": 1.5845709743546837e-05,
+      "loss": 1.7505,
       "step": 4966
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 18.450183868408203,
-      "learning_rate": 1.0045103738598779e-05,
-      "loss": 2.5177,
+      "epoch": 0.62,
+      "grad_norm": 5.886056900024414,
+      "learning_rate": 1.5844873028490147e-05,
+      "loss": 0.6074,
       "step": 4967
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 15.293846130371094,
-      "learning_rate": 1.0043099127994389e-05,
-      "loss": 1.472,
+      "epoch": 0.62,
+      "grad_norm": 22.552860260009766,
+      "learning_rate": 1.584403631343346e-05,
+      "loss": 1.9228,
       "step": 4968
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 16.424936294555664,
-      "learning_rate": 1.0041094517389997e-05,
-      "loss": 1.477,
+      "epoch": 0.62,
+      "grad_norm": 13.729313850402832,
+      "learning_rate": 1.5843199598376774e-05,
+      "loss": 2.583,
       "step": 4969
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 13.09062385559082,
-      "learning_rate": 1.003908990678561e-05,
-      "loss": 1.414,
+      "epoch": 0.62,
+      "grad_norm": 8.290722846984863,
+      "learning_rate": 1.5842362883320085e-05,
+      "loss": 0.8759,
       "step": 4970
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 21.331588745117188,
-      "learning_rate": 1.0037085296181218e-05,
-      "loss": 2.6522,
+      "epoch": 0.62,
+      "grad_norm": 10.95855712890625,
+      "learning_rate": 1.58415261682634e-05,
+      "loss": 1.096,
       "step": 4971
     },
     {
-      "epoch": 1.49,
-      "grad_norm": 44.23732376098633,
-      "learning_rate": 1.0035080685576828e-05,
-      "loss": 1.6369,
+      "epoch": 0.62,
+      "grad_norm": 17.684823989868164,
+      "learning_rate": 1.5840689453206712e-05,
+      "loss": 2.0879,
       "step": 4972
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 13.58181095123291,
-      "learning_rate": 1.0033076074972436e-05,
-      "loss": 1.6649,
+      "epoch": 0.62,
+      "grad_norm": 7.833134174346924,
+      "learning_rate": 1.5839852738150022e-05,
+      "loss": 1.6442,
       "step": 4973
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 25.254730224609375,
-      "learning_rate": 1.0031071464368048e-05,
-      "loss": 2.0687,
+      "epoch": 0.62,
+      "grad_norm": 22.88947296142578,
+      "learning_rate": 1.5839016023093336e-05,
+      "loss": 2.8722,
       "step": 4974
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 10.236112594604492,
-      "learning_rate": 1.0029066853763658e-05,
-      "loss": 1.3942,
+      "epoch": 0.62,
+      "grad_norm": 14.389993667602539,
+      "learning_rate": 1.583817930803665e-05,
+      "loss": 1.0751,
       "step": 4975
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 13.936944961547852,
-      "learning_rate": 1.0027062243159266e-05,
-      "loss": 2.3225,
+      "epoch": 0.62,
+      "grad_norm": 6.895519733428955,
+      "learning_rate": 1.5837342592979963e-05,
+      "loss": 0.6522,
       "step": 4976
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 11.376901626586914,
-      "learning_rate": 1.0025057632554878e-05,
-      "loss": 1.3294,
+      "epoch": 0.62,
+      "grad_norm": 9.349720001220703,
+      "learning_rate": 1.5836505877923273e-05,
+      "loss": 0.8121,
       "step": 4977
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 15.195333480834961,
-      "learning_rate": 1.0023053021950487e-05,
-      "loss": 2.3572,
+      "epoch": 0.62,
+      "grad_norm": 16.215843200683594,
+      "learning_rate": 1.5835669162866587e-05,
+      "loss": 1.9194,
       "step": 4978
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 35.94542694091797,
-      "learning_rate": 1.0021048411346097e-05,
-      "loss": 3.3368,
+      "epoch": 0.62,
+      "grad_norm": 10.783334732055664,
+      "learning_rate": 1.58348324478099e-05,
+      "loss": 0.9586,
       "step": 4979
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 21.131452560424805,
-      "learning_rate": 1.0019043800741705e-05,
-      "loss": 1.618,
+      "epoch": 0.62,
+      "grad_norm": 18.04204750061035,
+      "learning_rate": 1.583399573275321e-05,
+      "loss": 1.4249,
       "step": 4980
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 22.635162353515625,
-      "learning_rate": 1.0017039190137317e-05,
-      "loss": 2.0817,
+      "epoch": 0.63,
+      "grad_norm": 16.274057388305664,
+      "learning_rate": 1.5833159017696525e-05,
+      "loss": 0.9971,
       "step": 4981
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 22.40682601928711,
-      "learning_rate": 1.0015034579532927e-05,
-      "loss": 1.2649,
+      "epoch": 0.63,
+      "grad_norm": 12.840714454650879,
+      "learning_rate": 1.583232230263984e-05,
+      "loss": 1.802,
       "step": 4982
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 7.878605365753174,
-      "learning_rate": 1.0013029968928535e-05,
-      "loss": 1.3573,
+      "epoch": 0.63,
+      "grad_norm": 9.13722038269043,
+      "learning_rate": 1.5831485587583152e-05,
+      "loss": 1.0917,
       "step": 4983
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 19.825239181518555,
-      "learning_rate": 1.0011025358324147e-05,
-      "loss": 1.9824,
+      "epoch": 0.63,
+      "grad_norm": 19.6453857421875,
+      "learning_rate": 1.5830648872526462e-05,
+      "loss": 3.3396,
       "step": 4984
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 14.39383602142334,
-      "learning_rate": 1.0009020747719755e-05,
-      "loss": 1.8926,
+      "epoch": 0.63,
+      "grad_norm": 13.792831420898438,
+      "learning_rate": 1.5829812157469776e-05,
+      "loss": 1.3786,
       "step": 4985
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 21.175561904907227,
-      "learning_rate": 1.0007016137115366e-05,
-      "loss": 1.114,
+      "epoch": 0.63,
+      "grad_norm": 9.151402473449707,
+      "learning_rate": 1.582897544241309e-05,
+      "loss": 0.5139,
       "step": 4986
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 17.892271041870117,
-      "learning_rate": 1.0005011526510977e-05,
-      "loss": 1.2682,
+      "epoch": 0.63,
+      "grad_norm": 13.77298355102539,
+      "learning_rate": 1.58281387273564e-05,
+      "loss": 2.9223,
       "step": 4987
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 14.41032600402832,
-      "learning_rate": 1.0003006915906586e-05,
-      "loss": 1.3076,
+      "epoch": 0.63,
+      "grad_norm": 13.152591705322266,
+      "learning_rate": 1.5827302012299713e-05,
+      "loss": 1.7311,
       "step": 4988
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 10.802579879760742,
-      "learning_rate": 1.0001002305302196e-05,
-      "loss": 1.1787,
+      "epoch": 0.63,
+      "grad_norm": 13.780109405517578,
+      "learning_rate": 1.5826465297243027e-05,
+      "loss": 0.9852,
       "step": 4989
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 16.22612762451172,
-      "learning_rate": 9.998997694697806e-06,
-      "loss": 1.8036,
+      "epoch": 0.63,
+      "grad_norm": 18.66058921813965,
+      "learning_rate": 1.5825628582186337e-05,
+      "loss": 3.3515,
       "step": 4990
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 9.207633018493652,
-      "learning_rate": 9.996993084093416e-06,
-      "loss": 1.2887,
+      "epoch": 0.63,
+      "grad_norm": 16.108327865600586,
+      "learning_rate": 1.582479186712965e-05,
+      "loss": 0.7064,
       "step": 4991
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 109.87674713134766,
-      "learning_rate": 9.994988473489026e-06,
-      "loss": 2.7811,
+      "epoch": 0.63,
+      "grad_norm": 33.89423751831055,
+      "learning_rate": 1.582395515207296e-05,
+      "loss": 2.1219,
       "step": 4992
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 10.483904838562012,
-      "learning_rate": 9.992983862884636e-06,
-      "loss": 2.3548,
+      "epoch": 0.63,
+      "grad_norm": 20.117361068725586,
+      "learning_rate": 1.5823118437016275e-05,
+      "loss": 2.2392,
       "step": 4993
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 8.791914939880371,
-      "learning_rate": 9.990979252280245e-06,
-      "loss": 1.0692,
+      "epoch": 0.63,
+      "grad_norm": 2.8925938606262207,
+      "learning_rate": 1.582228172195959e-05,
+      "loss": 0.053,
       "step": 4994
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 12.697139739990234,
-      "learning_rate": 9.988974641675855e-06,
-      "loss": 1.2917,
+      "epoch": 0.63,
+      "grad_norm": 11.56701946258545,
+      "learning_rate": 1.58214450069029e-05,
+      "loss": 1.919,
       "step": 4995
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 15.190261840820312,
-      "learning_rate": 9.986970031071466e-06,
-      "loss": 1.7266,
+      "epoch": 0.63,
+      "grad_norm": 12.960175514221191,
+      "learning_rate": 1.5820608291846212e-05,
+      "loss": 1.4824,
       "step": 4996
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 13.911056518554688,
-      "learning_rate": 9.984965420467075e-06,
-      "loss": 1.2805,
+      "epoch": 0.63,
+      "grad_norm": 28.917999267578125,
+      "learning_rate": 1.5819771576789526e-05,
+      "loss": 2.58,
       "step": 4997
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 10.715677261352539,
-      "learning_rate": 9.982960809862685e-06,
-      "loss": 2.3035,
+      "epoch": 0.63,
+      "grad_norm": 15.101088523864746,
+      "learning_rate": 1.5818934861732836e-05,
+      "loss": 1.1439,
       "step": 4998
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 23.649269104003906,
-      "learning_rate": 9.980956199258295e-06,
-      "loss": 1.6413,
+      "epoch": 0.63,
+      "grad_norm": 17.131656646728516,
+      "learning_rate": 1.581809814667615e-05,
+      "loss": 3.1353,
       "step": 4999
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 9.821858406066895,
-      "learning_rate": 9.978951588653905e-06,
-      "loss": 1.1268,
+      "epoch": 0.63,
+      "grad_norm": 27.308841705322266,
+      "learning_rate": 1.5817261431619464e-05,
+      "loss": 1.8908,
       "step": 5000
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 11.555136680603027,
-      "learning_rate": 9.976946978049515e-06,
-      "loss": 1.2356,
+      "epoch": 0.63,
+      "grad_norm": 10.716662406921387,
+      "learning_rate": 1.5816424716562774e-05,
+      "loss": 1.1753,
       "step": 5001
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 15.840431213378906,
-      "learning_rate": 9.974942367445125e-06,
-      "loss": 1.7111,
+      "epoch": 0.63,
+      "grad_norm": 17.947704315185547,
+      "learning_rate": 1.5815588001506088e-05,
+      "loss": 1.5808,
       "step": 5002
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 13.110334396362305,
-      "learning_rate": 9.972937756840735e-06,
-      "loss": 1.3668,
+      "epoch": 0.63,
+      "grad_norm": 9.132237434387207,
+      "learning_rate": 1.58147512864494e-05,
+      "loss": 0.5576,
       "step": 5003
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 15.277511596679688,
-      "learning_rate": 9.970933146236345e-06,
-      "loss": 1.524,
+      "epoch": 0.63,
+      "grad_norm": 6.144209861755371,
+      "learning_rate": 1.5813914571392715e-05,
+      "loss": 0.64,
       "step": 5004
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 22.332592010498047,
-      "learning_rate": 9.968928535631954e-06,
-      "loss": 1.8393,
+      "epoch": 0.63,
+      "grad_norm": 32.854312896728516,
+      "learning_rate": 1.5813077856336025e-05,
+      "loss": 1.5851,
       "step": 5005
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 10.41042709350586,
-      "learning_rate": 9.966923925027564e-06,
-      "loss": 0.9121,
+      "epoch": 0.63,
+      "grad_norm": 19.065019607543945,
+      "learning_rate": 1.581224114127934e-05,
+      "loss": 1.7912,
       "step": 5006
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 17.307802200317383,
-      "learning_rate": 9.964919314423174e-06,
-      "loss": 1.2376,
+      "epoch": 0.63,
+      "grad_norm": 10.907378196716309,
+      "learning_rate": 1.5811404426222652e-05,
+      "loss": 2.303,
       "step": 5007
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 28.895647048950195,
-      "learning_rate": 9.962914703818784e-06,
-      "loss": 2.9159,
+      "epoch": 0.63,
+      "grad_norm": 88.3462142944336,
+      "learning_rate": 1.5810567711165963e-05,
+      "loss": 3.0653,
       "step": 5008
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 54.32232666015625,
-      "learning_rate": 9.960910093214394e-06,
-      "loss": 2.1614,
+      "epoch": 0.63,
+      "grad_norm": 14.676846504211426,
+      "learning_rate": 1.5809730996109276e-05,
+      "loss": 2.3928,
       "step": 5009
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 13.556760787963867,
-      "learning_rate": 9.958905482610004e-06,
-      "loss": 0.914,
+      "epoch": 0.63,
+      "grad_norm": 15.923056602478027,
+      "learning_rate": 1.580889428105259e-05,
+      "loss": 0.9603,
       "step": 5010
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 16.66999053955078,
-      "learning_rate": 9.956900872005614e-06,
-      "loss": 1.303,
+      "epoch": 0.63,
+      "grad_norm": 10.484317779541016,
+      "learning_rate": 1.5808057565995904e-05,
+      "loss": 3.553,
       "step": 5011
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 23.729116439819336,
-      "learning_rate": 9.954896261401223e-06,
-      "loss": 1.1319,
+      "epoch": 0.63,
+      "grad_norm": 8.377695083618164,
+      "learning_rate": 1.5807220850939214e-05,
+      "loss": 2.7543,
       "step": 5012
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 14.170336723327637,
-      "learning_rate": 9.952891650796833e-06,
-      "loss": 1.2095,
+      "epoch": 0.63,
+      "grad_norm": 12.684469223022461,
+      "learning_rate": 1.5806384135882528e-05,
+      "loss": 1.6304,
       "step": 5013
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 8.873191833496094,
-      "learning_rate": 9.950887040192443e-06,
-      "loss": 1.0473,
+      "epoch": 0.63,
+      "grad_norm": 33.287166595458984,
+      "learning_rate": 1.580554742082584e-05,
+      "loss": 2.6802,
       "step": 5014
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 19.656044006347656,
-      "learning_rate": 9.948882429588053e-06,
-      "loss": 1.9538,
+      "epoch": 0.63,
+      "grad_norm": 38.08937454223633,
+      "learning_rate": 1.580471070576915e-05,
+      "loss": 0.912,
       "step": 5015
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 7.630486011505127,
-      "learning_rate": 9.946877818983663e-06,
-      "loss": 1.2515,
+      "epoch": 0.63,
+      "grad_norm": 13.854416847229004,
+      "learning_rate": 1.5803873990712465e-05,
+      "loss": 1.3063,
       "step": 5016
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 15.626066207885742,
-      "learning_rate": 9.944873208379273e-06,
-      "loss": 1.7024,
+      "epoch": 0.63,
+      "grad_norm": 11.320171356201172,
+      "learning_rate": 1.5803037275655775e-05,
+      "loss": 1.699,
       "step": 5017
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 11.820615768432617,
-      "learning_rate": 9.942868597774883e-06,
-      "loss": 1.5467,
+      "epoch": 0.63,
+      "grad_norm": 6.064815044403076,
+      "learning_rate": 1.580220056059909e-05,
+      "loss": 0.5363,
       "step": 5018
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 26.351261138916016,
-      "learning_rate": 9.940863987170493e-06,
-      "loss": 1.0137,
+      "epoch": 0.63,
+      "grad_norm": 33.121761322021484,
+      "learning_rate": 1.5801363845542403e-05,
+      "loss": 1.9679,
       "step": 5019
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 15.984896659851074,
-      "learning_rate": 9.938859376566102e-06,
-      "loss": 1.2989,
+      "epoch": 0.63,
+      "grad_norm": 11.829465866088867,
+      "learning_rate": 1.5800527130485713e-05,
+      "loss": 0.7114,
       "step": 5020
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 23.20572853088379,
-      "learning_rate": 9.936854765961713e-06,
-      "loss": 1.8128,
+      "epoch": 0.63,
+      "grad_norm": 26.17995262145996,
+      "learning_rate": 1.5799690415429027e-05,
+      "loss": 2.2246,
       "step": 5021
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 12.120548248291016,
-      "learning_rate": 9.934850155357323e-06,
-      "loss": 1.295,
+      "epoch": 0.63,
+      "grad_norm": 13.5386323928833,
+      "learning_rate": 1.579885370037234e-05,
+      "loss": 2.1091,
       "step": 5022
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 14.345947265625,
-      "learning_rate": 9.932845544752932e-06,
-      "loss": 1.0609,
+      "epoch": 0.63,
+      "grad_norm": 23.967710494995117,
+      "learning_rate": 1.579801698531565e-05,
+      "loss": 1.9515,
       "step": 5023
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 10.384604454040527,
-      "learning_rate": 9.930840934148542e-06,
-      "loss": 0.8389,
+      "epoch": 0.63,
+      "grad_norm": 19.2731876373291,
+      "learning_rate": 1.5797180270258964e-05,
+      "loss": 2.5988,
       "step": 5024
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 50.97624588012695,
-      "learning_rate": 9.928836323544152e-06,
-      "loss": 2.2094,
+      "epoch": 0.63,
+      "grad_norm": 9.11064338684082,
+      "learning_rate": 1.5796343555202278e-05,
+      "loss": 0.5122,
       "step": 5025
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 15.930320739746094,
-      "learning_rate": 9.926831712939762e-06,
-      "loss": 1.0719,
+      "epoch": 0.63,
+      "grad_norm": 11.065043449401855,
+      "learning_rate": 1.5795506840145588e-05,
+      "loss": 0.9982,
       "step": 5026
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 15.929656982421875,
-      "learning_rate": 9.924827102335372e-06,
-      "loss": 1.4516,
+      "epoch": 0.63,
+      "grad_norm": 17.819351196289062,
+      "learning_rate": 1.5794670125088902e-05,
+      "loss": 1.8422,
       "step": 5027
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 20.207595825195312,
-      "learning_rate": 9.922822491730982e-06,
-      "loss": 1.122,
+      "epoch": 0.63,
+      "grad_norm": 22.37653350830078,
+      "learning_rate": 1.5793833410032215e-05,
+      "loss": 1.7192,
       "step": 5028
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 10.85169506072998,
-      "learning_rate": 9.920817881126592e-06,
-      "loss": 1.3796,
+      "epoch": 0.63,
+      "grad_norm": 12.90748405456543,
+      "learning_rate": 1.5792996694975526e-05,
+      "loss": 1.1331,
       "step": 5029
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 40.899234771728516,
-      "learning_rate": 9.918813270522202e-06,
-      "loss": 2.79,
+      "epoch": 0.63,
+      "grad_norm": 20.128108978271484,
+      "learning_rate": 1.579215997991884e-05,
+      "loss": 1.3026,
       "step": 5030
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 77.47624206542969,
-      "learning_rate": 9.91680865991781e-06,
-      "loss": 2.6326,
+      "epoch": 0.63,
+      "grad_norm": 6.115049839019775,
+      "learning_rate": 1.5791323264862153e-05,
+      "loss": 2.1955,
       "step": 5031
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 11.343374252319336,
-      "learning_rate": 9.914804049313421e-06,
-      "loss": 1.2866,
+      "epoch": 0.63,
+      "grad_norm": 16.82707977294922,
+      "learning_rate": 1.5790486549805467e-05,
+      "loss": 3.8006,
       "step": 5032
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 35.656246185302734,
-      "learning_rate": 9.912799438709033e-06,
-      "loss": 2.6495,
+      "epoch": 0.63,
+      "grad_norm": 9.201262474060059,
+      "learning_rate": 1.5789649834748777e-05,
+      "loss": 1.1929,
       "step": 5033
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 17.106098175048828,
-      "learning_rate": 9.910794828104641e-06,
-      "loss": 1.4173,
+      "epoch": 0.63,
+      "grad_norm": 24.340532302856445,
+      "learning_rate": 1.578881311969209e-05,
+      "loss": 2.5925,
       "step": 5034
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 25.03550148010254,
-      "learning_rate": 9.908790217500251e-06,
-      "loss": 1.5695,
+      "epoch": 0.63,
+      "grad_norm": 21.014192581176758,
+      "learning_rate": 1.5787976404635404e-05,
+      "loss": 2.3361,
       "step": 5035
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 14.62669563293457,
-      "learning_rate": 9.906785606895861e-06,
-      "loss": 1.6762,
+      "epoch": 0.63,
+      "grad_norm": 11.501837730407715,
+      "learning_rate": 1.5787139689578714e-05,
+      "loss": 1.1706,
       "step": 5036
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 98.68997192382812,
-      "learning_rate": 9.904780996291471e-06,
-      "loss": 3.2678,
+      "epoch": 0.63,
+      "grad_norm": 15.21597957611084,
+      "learning_rate": 1.5786302974522028e-05,
+      "loss": 2.0659,
       "step": 5037
     },
     {
-      "epoch": 1.51,
-      "grad_norm": 10.68436050415039,
-      "learning_rate": 9.902776385687081e-06,
-      "loss": 1.6819,
+      "epoch": 0.63,
+      "grad_norm": 5.9182047843933105,
+      "learning_rate": 1.578546625946534e-05,
+      "loss": 0.4091,
       "step": 5038
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 17.10354232788086,
-      "learning_rate": 9.900771775082692e-06,
-      "loss": 1.5759,
+      "epoch": 0.63,
+      "grad_norm": 14.510945320129395,
+      "learning_rate": 1.5784629544408655e-05,
+      "loss": 2.0028,
       "step": 5039
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 24.183143615722656,
-      "learning_rate": 9.898767164478302e-06,
-      "loss": 1.51,
-      "step": 5040
-    },
-    {
-      "epoch": 1.52,
-      "eval_loss": 0.1983981430530548,
-      "eval_runtime": 43.7634,
-      "eval_samples_per_second": 33.795,
-      "eval_steps_per_second": 33.795,
+      "epoch": 0.63,
+      "grad_norm": 8.41043758392334,
+      "learning_rate": 1.5783792829351966e-05,
+      "loss": 0.6456,
       "step": 5040
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 11.218318939208984,
-      "learning_rate": 9.896762553873912e-06,
-      "loss": 2.6052,
+      "epoch": 0.63,
+      "grad_norm": 14.400217056274414,
+      "learning_rate": 1.578295611429528e-05,
+      "loss": 2.1407,
       "step": 5041
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 28.35023307800293,
-      "learning_rate": 9.89475794326952e-06,
-      "loss": 1.5494,
+      "epoch": 0.63,
+      "grad_norm": 9.780742645263672,
+      "learning_rate": 1.5782119399238593e-05,
+      "loss": 1.5189,
       "step": 5042
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 39.44877624511719,
-      "learning_rate": 9.89275333266513e-06,
-      "loss": 2.2204,
+      "epoch": 0.63,
+      "grad_norm": 4.016148567199707,
+      "learning_rate": 1.5781282684181903e-05,
+      "loss": 0.2529,
       "step": 5043
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 16.99179458618164,
-      "learning_rate": 9.89074872206074e-06,
-      "loss": 1.9028,
+      "epoch": 0.63,
+      "grad_norm": 15.589259147644043,
+      "learning_rate": 1.5780445969125217e-05,
+      "loss": 2.1329,
       "step": 5044
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 22.564708709716797,
-      "learning_rate": 9.88874411145635e-06,
-      "loss": 2.0602,
+      "epoch": 0.63,
+      "grad_norm": 12.96193790435791,
+      "learning_rate": 1.5779609254068527e-05,
+      "loss": 2.2834,
       "step": 5045
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 10.778221130371094,
-      "learning_rate": 9.88673950085196e-06,
-      "loss": 1.071,
+      "epoch": 0.63,
+      "grad_norm": 30.995908737182617,
+      "learning_rate": 1.577877253901184e-05,
+      "loss": 1.8793,
       "step": 5046
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 25.14896583557129,
-      "learning_rate": 9.88473489024757e-06,
-      "loss": 1.4653,
+      "epoch": 0.63,
+      "grad_norm": 15.587225914001465,
+      "learning_rate": 1.5777935823955154e-05,
+      "loss": 2.889,
       "step": 5047
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 24.445661544799805,
-      "learning_rate": 9.88273027964318e-06,
-      "loss": 1.2668,
+      "epoch": 0.63,
+      "grad_norm": 10.563833236694336,
+      "learning_rate": 1.5777099108898465e-05,
+      "loss": 2.2132,
       "step": 5048
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 19.382781982421875,
-      "learning_rate": 9.88072566903879e-06,
-      "loss": 1.6534,
+      "epoch": 0.63,
+      "grad_norm": 9.568826675415039,
+      "learning_rate": 1.5776262393841778e-05,
+      "loss": 1.0254,
       "step": 5049
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 48.97084426879883,
-      "learning_rate": 9.878721058434399e-06,
-      "loss": 1.6285,
+      "epoch": 0.63,
+      "grad_norm": 15.084373474121094,
+      "learning_rate": 1.577542567878509e-05,
+      "loss": 1.1353,
       "step": 5050
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 23.401174545288086,
-      "learning_rate": 9.87671644783001e-06,
-      "loss": 2.24,
+      "epoch": 0.63,
+      "grad_norm": 12.499186515808105,
+      "learning_rate": 1.5774588963728402e-05,
+      "loss": 1.9012,
       "step": 5051
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 7.404333591461182,
-      "learning_rate": 9.874711837225621e-06,
-      "loss": 0.8004,
+      "epoch": 0.63,
+      "grad_norm": 12.14394474029541,
+      "learning_rate": 1.5773752248671716e-05,
+      "loss": 2.1066,
       "step": 5052
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 14.808720588684082,
-      "learning_rate": 9.87270722662123e-06,
-      "loss": 1.193,
+      "epoch": 0.63,
+      "grad_norm": 7.123768329620361,
+      "learning_rate": 1.577291553361503e-05,
+      "loss": 1.5745,
       "step": 5053
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 11.823633193969727,
-      "learning_rate": 9.87070261601684e-06,
-      "loss": 1.1512,
+      "epoch": 0.63,
+      "grad_norm": 11.73341178894043,
+      "learning_rate": 1.577207881855834e-05,
+      "loss": 1.5408,
       "step": 5054
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 20.415996551513672,
-      "learning_rate": 9.86869800541245e-06,
-      "loss": 1.7406,
+      "epoch": 0.63,
+      "grad_norm": 15.225319862365723,
+      "learning_rate": 1.5771242103501653e-05,
+      "loss": 0.7718,
       "step": 5055
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 584.0230712890625,
-      "learning_rate": 9.86669339480806e-06,
-      "loss": 1.3769,
+      "epoch": 0.63,
+      "grad_norm": 16.591278076171875,
+      "learning_rate": 1.5770405388444967e-05,
+      "loss": 2.4582,
       "step": 5056
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 14.067730903625488,
-      "learning_rate": 9.864688784203668e-06,
-      "loss": 1.1319,
+      "epoch": 0.63,
+      "grad_norm": 34.30024719238281,
+      "learning_rate": 1.5769568673388277e-05,
+      "loss": 3.0588,
       "step": 5057
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 27.009614944458008,
-      "learning_rate": 9.86268417359928e-06,
-      "loss": 1.6301,
+      "epoch": 0.63,
+      "grad_norm": 14.780279159545898,
+      "learning_rate": 1.576873195833159e-05,
+      "loss": 1.0337,
       "step": 5058
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 12.77624225616455,
-      "learning_rate": 9.86067956299489e-06,
-      "loss": 1.4832,
+      "epoch": 0.63,
+      "grad_norm": 16.732141494750977,
+      "learning_rate": 1.5767895243274905e-05,
+      "loss": 2.9683,
       "step": 5059
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 14.48194408416748,
-      "learning_rate": 9.858674952390498e-06,
-      "loss": 1.3063,
+      "epoch": 0.64,
+      "grad_norm": 18.878597259521484,
+      "learning_rate": 1.5767058528218218e-05,
+      "loss": 3.4954,
       "step": 5060
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 23.552724838256836,
-      "learning_rate": 9.856670341786108e-06,
-      "loss": 1.726,
+      "epoch": 0.64,
+      "grad_norm": 22.96034812927246,
+      "learning_rate": 1.576622181316153e-05,
+      "loss": 1.2982,
       "step": 5061
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 35.44766616821289,
-      "learning_rate": 9.854665731181718e-06,
-      "loss": 1.7173,
+      "epoch": 0.64,
+      "grad_norm": 19.98801040649414,
+      "learning_rate": 1.5765385098104842e-05,
+      "loss": 2.2953,
       "step": 5062
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 17.609161376953125,
-      "learning_rate": 9.852661120577328e-06,
-      "loss": 1.2193,
+      "epoch": 0.64,
+      "grad_norm": 25.850393295288086,
+      "learning_rate": 1.5764548383048156e-05,
+      "loss": 3.2962,
       "step": 5063
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 13.460494041442871,
-      "learning_rate": 9.850656509972939e-06,
-      "loss": 1.2271,
+      "epoch": 0.64,
+      "grad_norm": 15.799878120422363,
+      "learning_rate": 1.5763711667991466e-05,
+      "loss": 1.8083,
       "step": 5064
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 19.18136215209961,
-      "learning_rate": 9.848651899368549e-06,
-      "loss": 1.0003,
+      "epoch": 0.64,
+      "grad_norm": 9.549363136291504,
+      "learning_rate": 1.576287495293478e-05,
+      "loss": 1.6983,
       "step": 5065
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 33.11497497558594,
-      "learning_rate": 9.846647288764159e-06,
-      "loss": 1.9836,
+      "epoch": 0.64,
+      "grad_norm": 16.10846519470215,
+      "learning_rate": 1.5762038237878093e-05,
+      "loss": 0.9541,
       "step": 5066
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 17.282258987426758,
-      "learning_rate": 9.844642678159769e-06,
-      "loss": 1.7075,
+      "epoch": 0.64,
+      "grad_norm": 13.372329711914062,
+      "learning_rate": 1.5761201522821407e-05,
+      "loss": 0.8804,
       "step": 5067
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 12.306963920593262,
-      "learning_rate": 9.842638067555377e-06,
-      "loss": 0.9739,
+      "epoch": 0.64,
+      "grad_norm": 16.088727951049805,
+      "learning_rate": 1.5760364807764717e-05,
+      "loss": 1.3293,
       "step": 5068
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 18.901031494140625,
-      "learning_rate": 9.840633456950987e-06,
-      "loss": 1.4667,
+      "epoch": 0.64,
+      "grad_norm": 42.135169982910156,
+      "learning_rate": 1.575952809270803e-05,
+      "loss": 2.4092,
       "step": 5069
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 22.379085540771484,
-      "learning_rate": 9.838628846346599e-06,
-      "loss": 1.1675,
+      "epoch": 0.64,
+      "grad_norm": 19.614219665527344,
+      "learning_rate": 1.575869137765134e-05,
+      "loss": 3.0553,
       "step": 5070
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 14.647932052612305,
-      "learning_rate": 9.836624235742207e-06,
-      "loss": 1.809,
+      "epoch": 0.64,
+      "grad_norm": 10.226930618286133,
+      "learning_rate": 1.5757854662594655e-05,
+      "loss": 1.9401,
       "step": 5071
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 8.400578498840332,
-      "learning_rate": 9.834619625137818e-06,
-      "loss": 1.8459,
+      "epoch": 0.64,
+      "grad_norm": 16.097923278808594,
+      "learning_rate": 1.575701794753797e-05,
+      "loss": 1.559,
       "step": 5072
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 14.352058410644531,
-      "learning_rate": 9.832615014533428e-06,
-      "loss": 1.6325,
+      "epoch": 0.64,
+      "grad_norm": 22.118406295776367,
+      "learning_rate": 1.575618123248128e-05,
+      "loss": 1.5106,
       "step": 5073
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 38.508724212646484,
-      "learning_rate": 9.830610403929038e-06,
-      "loss": 1.4109,
+      "epoch": 0.64,
+      "grad_norm": 11.076290130615234,
+      "learning_rate": 1.5755344517424592e-05,
+      "loss": 1.292,
       "step": 5074
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 24.62671661376953,
-      "learning_rate": 9.828605793324648e-06,
-      "loss": 2.3143,
+      "epoch": 0.64,
+      "grad_norm": 7.768183708190918,
+      "learning_rate": 1.5754507802367903e-05,
+      "loss": 2.8704,
       "step": 5075
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 32.014671325683594,
-      "learning_rate": 9.826601182720258e-06,
-      "loss": 2.2224,
+      "epoch": 0.64,
+      "grad_norm": 15.45603084564209,
+      "learning_rate": 1.5753671087311216e-05,
+      "loss": 2.2323,
       "step": 5076
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 16.45624351501465,
-      "learning_rate": 9.824596572115868e-06,
-      "loss": 1.3855,
+      "epoch": 0.64,
+      "grad_norm": 16.193885803222656,
+      "learning_rate": 1.575283437225453e-05,
+      "loss": 2.3428,
       "step": 5077
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 24.989152908325195,
-      "learning_rate": 9.822591961511478e-06,
-      "loss": 1.8901,
+      "epoch": 0.64,
+      "grad_norm": 15.888275146484375,
+      "learning_rate": 1.575199765719784e-05,
+      "loss": 1.1974,
       "step": 5078
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 16.87116241455078,
-      "learning_rate": 9.820587350907086e-06,
-      "loss": 2.1003,
+      "epoch": 0.64,
+      "grad_norm": 17.932859420776367,
+      "learning_rate": 1.5751160942141154e-05,
+      "loss": 1.7937,
       "step": 5079
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 19.88626480102539,
-      "learning_rate": 9.818582740302697e-06,
-      "loss": 0.9223,
+      "epoch": 0.64,
+      "grad_norm": 32.422584533691406,
+      "learning_rate": 1.5750324227084468e-05,
+      "loss": 2.737,
       "step": 5080
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 29.52408218383789,
-      "learning_rate": 9.816578129698307e-06,
-      "loss": 2.1227,
+      "epoch": 0.64,
+      "grad_norm": 5.988121509552002,
+      "learning_rate": 1.574948751202778e-05,
+      "loss": 0.3623,
       "step": 5081
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 6.786428451538086,
-      "learning_rate": 9.814573519093917e-06,
-      "loss": 1.1044,
+      "epoch": 0.64,
+      "grad_norm": 5.981936931610107,
+      "learning_rate": 1.574865079697109e-05,
+      "loss": 0.7407,
       "step": 5082
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 11.187318801879883,
-      "learning_rate": 9.812568908489527e-06,
-      "loss": 1.0142,
+      "epoch": 0.64,
+      "grad_norm": 26.017316818237305,
+      "learning_rate": 1.5747814081914405e-05,
+      "loss": 2.8178,
       "step": 5083
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 9.082967758178711,
-      "learning_rate": 9.810564297885137e-06,
-      "loss": 0.8634,
+      "epoch": 0.64,
+      "grad_norm": 10.025772094726562,
+      "learning_rate": 1.574697736685772e-05,
+      "loss": 3.3694,
       "step": 5084
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 38.15378189086914,
-      "learning_rate": 9.808559687280747e-06,
-      "loss": 1.4613,
+      "epoch": 0.64,
+      "grad_norm": 17.240604400634766,
+      "learning_rate": 1.574614065180103e-05,
+      "loss": 1.836,
       "step": 5085
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 11.708313941955566,
-      "learning_rate": 9.806555076676357e-06,
-      "loss": 2.1042,
+      "epoch": 0.64,
+      "grad_norm": 10.621012687683105,
+      "learning_rate": 1.5745303936744343e-05,
+      "loss": 1.0921,
       "step": 5086
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 15.158965110778809,
-      "learning_rate": 9.804550466071965e-06,
-      "loss": 1.2576,
+      "epoch": 0.64,
+      "grad_norm": 4.24019718170166,
+      "learning_rate": 1.5744467221687656e-05,
+      "loss": 0.2206,
       "step": 5087
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 15.897801399230957,
-      "learning_rate": 9.802545855467577e-06,
-      "loss": 1.5028,
+      "epoch": 0.64,
+      "grad_norm": 25.50225067138672,
+      "learning_rate": 1.574363050663097e-05,
+      "loss": 1.9584,
       "step": 5088
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 58.225669860839844,
-      "learning_rate": 9.800541244863187e-06,
-      "loss": 1.0376,
+      "epoch": 0.64,
+      "grad_norm": 8.65950870513916,
+      "learning_rate": 1.574279379157428e-05,
+      "loss": 0.7873,
       "step": 5089
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 26.382015228271484,
-      "learning_rate": 9.798536634258796e-06,
-      "loss": 1.2807,
+      "epoch": 0.64,
+      "grad_norm": 16.24616813659668,
+      "learning_rate": 1.5741957076517594e-05,
+      "loss": 1.4247,
       "step": 5090
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 11.506767272949219,
-      "learning_rate": 9.796532023654406e-06,
-      "loss": 1.4462,
+      "epoch": 0.64,
+      "grad_norm": 15.66484260559082,
+      "learning_rate": 1.5741120361460907e-05,
+      "loss": 2.0374,
       "step": 5091
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 36.81036376953125,
-      "learning_rate": 9.794527413050016e-06,
-      "loss": 2.7945,
+      "epoch": 0.64,
+      "grad_norm": 14.813385963439941,
+      "learning_rate": 1.5740283646404218e-05,
+      "loss": 2.5328,
       "step": 5092
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 13.376030921936035,
-      "learning_rate": 9.792522802445626e-06,
-      "loss": 1.562,
+      "epoch": 0.64,
+      "grad_norm": 10.920860290527344,
+      "learning_rate": 1.573944693134753e-05,
+      "loss": 0.7747,
       "step": 5093
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 17.786754608154297,
-      "learning_rate": 9.790518191841234e-06,
-      "loss": 2.5748,
+      "epoch": 0.64,
+      "grad_norm": 25.48909568786621,
+      "learning_rate": 1.5738610216290845e-05,
+      "loss": 1.3409,
       "step": 5094
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 22.214862823486328,
-      "learning_rate": 9.788513581236846e-06,
-      "loss": 1.8133,
+      "epoch": 0.64,
+      "grad_norm": 82.24311828613281,
+      "learning_rate": 1.573777350123416e-05,
+      "loss": 2.8238,
       "step": 5095
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 7.965408802032471,
-      "learning_rate": 9.786508970632456e-06,
-      "loss": 1.2572,
+      "epoch": 0.64,
+      "grad_norm": 14.695939064025879,
+      "learning_rate": 1.573693678617747e-05,
+      "loss": 1.3314,
       "step": 5096
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 16.189109802246094,
-      "learning_rate": 9.784504360028065e-06,
-      "loss": 1.1491,
+      "epoch": 0.64,
+      "grad_norm": 14.964838027954102,
+      "learning_rate": 1.5736100071120783e-05,
+      "loss": 1.2088,
       "step": 5097
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 21.910438537597656,
-      "learning_rate": 9.782499749423675e-06,
-      "loss": 1.8276,
+      "epoch": 0.64,
+      "grad_norm": 11.288729667663574,
+      "learning_rate": 1.5735263356064093e-05,
+      "loss": 1.1307,
       "step": 5098
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 29.36257553100586,
-      "learning_rate": 9.780495138819285e-06,
-      "loss": 1.69,
+      "epoch": 0.64,
+      "grad_norm": 20.675094604492188,
+      "learning_rate": 1.5734426641007407e-05,
+      "loss": 3.8026,
       "step": 5099
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 15.409547805786133,
-      "learning_rate": 9.778490528214895e-06,
-      "loss": 1.83,
+      "epoch": 0.64,
+      "grad_norm": 7.097488880157471,
+      "learning_rate": 1.573358992595072e-05,
+      "loss": 0.9689,
       "step": 5100
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 12.414823532104492,
-      "learning_rate": 9.776485917610505e-06,
-      "loss": 1.3998,
+      "epoch": 0.64,
+      "grad_norm": 5.893520832061768,
+      "learning_rate": 1.573275321089403e-05,
+      "loss": 0.6925,
       "step": 5101
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 25.707353591918945,
-      "learning_rate": 9.774481307006115e-06,
-      "loss": 2.12,
+      "epoch": 0.64,
+      "grad_norm": 19.848941802978516,
+      "learning_rate": 1.5731916495837344e-05,
+      "loss": 3.517,
       "step": 5102
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 43.882991790771484,
-      "learning_rate": 9.772476696401725e-06,
-      "loss": 1.7967,
+      "epoch": 0.64,
+      "grad_norm": 13.46828556060791,
+      "learning_rate": 1.5731079780780654e-05,
+      "loss": 0.8219,
       "step": 5103
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 12.999314308166504,
-      "learning_rate": 9.770472085797335e-06,
-      "loss": 1.1704,
+      "epoch": 0.64,
+      "grad_norm": 41.50696563720703,
+      "learning_rate": 1.5730243065723968e-05,
+      "loss": 1.7747,
       "step": 5104
     },
     {
-      "epoch": 1.53,
-      "grad_norm": 30.975318908691406,
-      "learning_rate": 9.768467475192944e-06,
-      "loss": 2.4516,
+      "epoch": 0.64,
+      "grad_norm": 9.662480354309082,
+      "learning_rate": 1.572940635066728e-05,
+      "loss": 1.5009,
       "step": 5105
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 10.727395057678223,
-      "learning_rate": 9.766462864588554e-06,
-      "loss": 0.5141,
+      "epoch": 0.64,
+      "grad_norm": 24.629592895507812,
+      "learning_rate": 1.5728569635610592e-05,
+      "loss": 1.3831,
       "step": 5106
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 16.37921142578125,
-      "learning_rate": 9.764458253984165e-06,
-      "loss": 1.9043,
+      "epoch": 0.64,
+      "grad_norm": 34.82011413574219,
+      "learning_rate": 1.5727732920553906e-05,
+      "loss": 1.8308,
       "step": 5107
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 11.777350425720215,
-      "learning_rate": 9.762453643379774e-06,
-      "loss": 1.3563,
+      "epoch": 0.64,
+      "grad_norm": 8.111581802368164,
+      "learning_rate": 1.572689620549722e-05,
+      "loss": 0.6872,
       "step": 5108
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 10.583935737609863,
-      "learning_rate": 9.760449032775384e-06,
-      "loss": 1.3274,
+      "epoch": 0.64,
+      "grad_norm": 13.667208671569824,
+      "learning_rate": 1.5726059490440533e-05,
+      "loss": 2.047,
       "step": 5109
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 24.881576538085938,
-      "learning_rate": 9.758444422170994e-06,
-      "loss": 1.0478,
+      "epoch": 0.64,
+      "grad_norm": 8.9616060256958,
+      "learning_rate": 1.5725222775383843e-05,
+      "loss": 0.8753,
       "step": 5110
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 24.082473754882812,
-      "learning_rate": 9.756439811566604e-06,
-      "loss": 0.8794,
+      "epoch": 0.64,
+      "grad_norm": 43.2788200378418,
+      "learning_rate": 1.5724386060327157e-05,
+      "loss": 1.0503,
       "step": 5111
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 15.496257781982422,
-      "learning_rate": 9.754435200962214e-06,
-      "loss": 0.8897,
+      "epoch": 0.64,
+      "grad_norm": 15.115678787231445,
+      "learning_rate": 1.572354934527047e-05,
+      "loss": 2.3047,
       "step": 5112
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 27.82330894470215,
-      "learning_rate": 9.752430590357824e-06,
-      "loss": 2.5503,
+      "epoch": 0.64,
+      "grad_norm": 11.667755126953125,
+      "learning_rate": 1.572271263021378e-05,
+      "loss": 0.7404,
       "step": 5113
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 14.794507026672363,
-      "learning_rate": 9.750425979753434e-06,
-      "loss": 2.2607,
+      "epoch": 0.64,
+      "grad_norm": 16.43598175048828,
+      "learning_rate": 1.5721875915157094e-05,
+      "loss": 2.3748,
       "step": 5114
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 25.15207862854004,
-      "learning_rate": 9.748421369149044e-06,
-      "loss": 1.2937,
+      "epoch": 0.64,
+      "grad_norm": 6.232161045074463,
+      "learning_rate": 1.5721039200100408e-05,
+      "loss": 0.5916,
       "step": 5115
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 28.609935760498047,
-      "learning_rate": 9.746416758544653e-06,
-      "loss": 1.8648,
+      "epoch": 0.64,
+      "grad_norm": 20.68305015563965,
+      "learning_rate": 1.572020248504372e-05,
+      "loss": 3.2559,
       "step": 5116
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 19.665058135986328,
-      "learning_rate": 9.744412147940263e-06,
-      "loss": 1.3485,
+      "epoch": 0.64,
+      "grad_norm": 32.31875991821289,
+      "learning_rate": 1.5719365769987032e-05,
+      "loss": 1.1103,
       "step": 5117
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 16.362789154052734,
-      "learning_rate": 9.742407537335873e-06,
-      "loss": 1.6568,
+      "epoch": 0.64,
+      "grad_norm": 7.220429420471191,
+      "learning_rate": 1.5718529054930346e-05,
+      "loss": 0.9726,
       "step": 5118
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 43.66654586791992,
-      "learning_rate": 9.740402926731483e-06,
-      "loss": 1.5276,
+      "epoch": 0.64,
+      "grad_norm": 15.526352882385254,
+      "learning_rate": 1.571769233987366e-05,
+      "loss": 1.332,
       "step": 5119
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 11.123966217041016,
-      "learning_rate": 9.738398316127093e-06,
-      "loss": 1.4542,
+      "epoch": 0.64,
+      "grad_norm": 12.444489479064941,
+      "learning_rate": 1.571685562481697e-05,
+      "loss": 2.893,
       "step": 5120
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 10.916337966918945,
-      "learning_rate": 9.736393705522703e-06,
-      "loss": 1.7105,
+      "epoch": 0.64,
+      "grad_norm": 16.54176902770996,
+      "learning_rate": 1.5716018909760283e-05,
+      "loss": 2.7116,
       "step": 5121
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 38.372467041015625,
-      "learning_rate": 9.734389094918313e-06,
-      "loss": 2.5102,
+      "epoch": 0.64,
+      "grad_norm": 8.2006254196167,
+      "learning_rate": 1.5715182194703597e-05,
+      "loss": 0.7196,
       "step": 5122
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 12.8078031539917,
-      "learning_rate": 9.732384484313923e-06,
-      "loss": 1.573,
+      "epoch": 0.64,
+      "grad_norm": 9.685171127319336,
+      "learning_rate": 1.5714345479646907e-05,
+      "loss": 1.8595,
       "step": 5123
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 22.485212326049805,
-      "learning_rate": 9.730379873709532e-06,
-      "loss": 1.986,
+      "epoch": 0.64,
+      "grad_norm": 19.2276554107666,
+      "learning_rate": 1.571350876459022e-05,
+      "loss": 2.0104,
       "step": 5124
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 16.489240646362305,
-      "learning_rate": 9.728375263105144e-06,
-      "loss": 1.5365,
+      "epoch": 0.64,
+      "grad_norm": 12.441829681396484,
+      "learning_rate": 1.5712672049533534e-05,
+      "loss": 1.353,
       "step": 5125
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 48.157371520996094,
-      "learning_rate": 9.726370652500754e-06,
-      "loss": 2.6245,
+      "epoch": 0.64,
+      "grad_norm": 27.83458137512207,
+      "learning_rate": 1.5711835334476845e-05,
+      "loss": 1.2713,
       "step": 5126
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 19.086997985839844,
-      "learning_rate": 9.724366041896362e-06,
-      "loss": 2.8709,
+      "epoch": 0.64,
+      "grad_norm": 23.49789047241211,
+      "learning_rate": 1.5710998619420158e-05,
+      "loss": 1.3408,
       "step": 5127
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 16.59282112121582,
-      "learning_rate": 9.722361431291972e-06,
-      "loss": 1.6568,
+      "epoch": 0.64,
+      "grad_norm": 11.275956153869629,
+      "learning_rate": 1.571016190436347e-05,
+      "loss": 1.6182,
       "step": 5128
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 77.64139556884766,
-      "learning_rate": 9.720356820687582e-06,
-      "loss": 2.4053,
+      "epoch": 0.64,
+      "grad_norm": 8.767870903015137,
+      "learning_rate": 1.5709325189306782e-05,
+      "loss": 1.2063,
       "step": 5129
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 32.31578063964844,
-      "learning_rate": 9.718352210083192e-06,
-      "loss": 2.4112,
+      "epoch": 0.64,
+      "grad_norm": 6.207084655761719,
+      "learning_rate": 1.5708488474250096e-05,
+      "loss": 0.3131,
       "step": 5130
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 16.772424697875977,
-      "learning_rate": 9.716347599478802e-06,
-      "loss": 1.8791,
+      "epoch": 0.64,
+      "grad_norm": 24.9721622467041,
+      "learning_rate": 1.5707651759193406e-05,
+      "loss": 1.3211,
       "step": 5131
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 18.102991104125977,
-      "learning_rate": 9.714342988874412e-06,
-      "loss": 1.3216,
+      "epoch": 0.64,
+      "grad_norm": 23.742414474487305,
+      "learning_rate": 1.570681504413672e-05,
+      "loss": 1.5197,
       "step": 5132
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 11.130638122558594,
-      "learning_rate": 9.712338378270023e-06,
-      "loss": 1.4781,
+      "epoch": 0.64,
+      "grad_norm": 15.931519508361816,
+      "learning_rate": 1.5705978329080033e-05,
+      "loss": 2.3792,
       "step": 5133
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 23.45066261291504,
-      "learning_rate": 9.710333767665631e-06,
-      "loss": 2.065,
+      "epoch": 0.64,
+      "grad_norm": 8.295441627502441,
+      "learning_rate": 1.5705141614023344e-05,
+      "loss": 0.5454,
       "step": 5134
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 24.560604095458984,
-      "learning_rate": 9.708329157061241e-06,
-      "loss": 1.383,
+      "epoch": 0.64,
+      "grad_norm": 63.2232551574707,
+      "learning_rate": 1.5704304898966657e-05,
+      "loss": 2.3334,
       "step": 5135
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 12.12047290802002,
-      "learning_rate": 9.706324546456851e-06,
-      "loss": 1.5988,
+      "epoch": 0.64,
+      "grad_norm": 14.293220520019531,
+      "learning_rate": 1.570346818390997e-05,
+      "loss": 1.2896,
       "step": 5136
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 10.139289855957031,
-      "learning_rate": 9.704319935852461e-06,
-      "loss": 0.8568,
+      "epoch": 0.64,
+      "grad_norm": 12.947236061096191,
+      "learning_rate": 1.5702631468853285e-05,
+      "loss": 1.4245,
       "step": 5137
     },
     {
-      "epoch": 1.54,
-      "grad_norm": 10.667194366455078,
-      "learning_rate": 9.702315325248071e-06,
-      "loss": 1.2382,
+      "epoch": 0.64,
+      "grad_norm": 6.499480724334717,
+      "learning_rate": 1.5701794753796595e-05,
+      "loss": 0.6759,
       "step": 5138
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 14.842784881591797,
-      "learning_rate": 9.700310714643681e-06,
-      "loss": 1.474,
+      "epoch": 0.64,
+      "grad_norm": 22.969093322753906,
+      "learning_rate": 1.570095803873991e-05,
+      "loss": 1.924,
       "step": 5139
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 28.56570816040039,
-      "learning_rate": 9.698306104039291e-06,
-      "loss": 1.9052,
+      "epoch": 0.65,
+      "grad_norm": 14.942154884338379,
+      "learning_rate": 1.5700121323683222e-05,
+      "loss": 1.3929,
       "step": 5140
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 14.900588035583496,
-      "learning_rate": 9.696301493434901e-06,
-      "loss": 1.2868,
+      "epoch": 0.65,
+      "grad_norm": 8.734114646911621,
+      "learning_rate": 1.5699284608626532e-05,
+      "loss": 2.4714,
       "step": 5141
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 11.330522537231445,
-      "learning_rate": 9.69429688283051e-06,
-      "loss": 1.1221,
+      "epoch": 0.65,
+      "grad_norm": 16.237415313720703,
+      "learning_rate": 1.5698447893569846e-05,
+      "loss": 1.1233,
       "step": 5142
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 10.978798866271973,
-      "learning_rate": 9.69229227222612e-06,
-      "loss": 1.2924,
+      "epoch": 0.65,
+      "grad_norm": 8.782207489013672,
+      "learning_rate": 1.569761117851316e-05,
+      "loss": 0.5252,
       "step": 5143
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 10.9464111328125,
-      "learning_rate": 9.690287661621732e-06,
-      "loss": 0.9553,
+      "epoch": 0.65,
+      "grad_norm": 31.56356430053711,
+      "learning_rate": 1.5696774463456473e-05,
+      "loss": 3.5507,
       "step": 5144
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 19.434829711914062,
-      "learning_rate": 9.68828305101734e-06,
-      "loss": 1.6625,
+      "epoch": 0.65,
+      "grad_norm": 33.98700714111328,
+      "learning_rate": 1.5695937748399784e-05,
+      "loss": 2.9867,
       "step": 5145
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 13.196100234985352,
-      "learning_rate": 9.68627844041295e-06,
-      "loss": 1.5143,
+      "epoch": 0.65,
+      "grad_norm": 19.101417541503906,
+      "learning_rate": 1.5695101033343097e-05,
+      "loss": 1.3417,
       "step": 5146
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 13.529244422912598,
-      "learning_rate": 9.68427382980856e-06,
-      "loss": 1.6467,
+      "epoch": 0.65,
+      "grad_norm": 7.179376125335693,
+      "learning_rate": 1.569426431828641e-05,
+      "loss": 1.3175,
       "step": 5147
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 37.84720993041992,
-      "learning_rate": 9.68226921920417e-06,
-      "loss": 3.0862,
+      "epoch": 0.65,
+      "grad_norm": 12.053905487060547,
+      "learning_rate": 1.569342760322972e-05,
+      "loss": 1.1468,
       "step": 5148
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 11.348771095275879,
-      "learning_rate": 9.68026460859978e-06,
-      "loss": 1.3557,
+      "epoch": 0.65,
+      "grad_norm": 31.480859756469727,
+      "learning_rate": 1.5692590888173035e-05,
+      "loss": 2.4152,
       "step": 5149
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 11.032979965209961,
-      "learning_rate": 9.67825999799539e-06,
-      "loss": 0.9113,
+      "epoch": 0.65,
+      "grad_norm": 22.707080841064453,
+      "learning_rate": 1.569175417311635e-05,
+      "loss": 1.7221,
       "step": 5150
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 15.761959075927734,
-      "learning_rate": 9.676255387391e-06,
-      "loss": 1.7351,
+      "epoch": 0.65,
+      "grad_norm": 22.629104614257812,
+      "learning_rate": 1.569091745805966e-05,
+      "loss": 2.7781,
       "step": 5151
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 53.275428771972656,
-      "learning_rate": 9.67425077678661e-06,
-      "loss": 2.9133,
+      "epoch": 0.65,
+      "grad_norm": 15.60866641998291,
+      "learning_rate": 1.5690080743002972e-05,
+      "loss": 1.1647,
       "step": 5152
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 18.882171630859375,
-      "learning_rate": 9.672246166182219e-06,
-      "loss": 2.1493,
+      "epoch": 0.65,
+      "grad_norm": 10.045681953430176,
+      "learning_rate": 1.5689244027946286e-05,
+      "loss": 0.8843,
       "step": 5153
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 13.768794059753418,
-      "learning_rate": 9.67024155557783e-06,
-      "loss": 1.1085,
+      "epoch": 0.65,
+      "grad_norm": 9.390864372253418,
+      "learning_rate": 1.5688407312889596e-05,
+      "loss": 1.3782,
       "step": 5154
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 15.78554916381836,
-      "learning_rate": 9.66823694497344e-06,
-      "loss": 1.7214,
+      "epoch": 0.65,
+      "grad_norm": 15.955741882324219,
+      "learning_rate": 1.568757059783291e-05,
+      "loss": 1.3725,
       "step": 5155
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 18.663970947265625,
-      "learning_rate": 9.66623233436905e-06,
-      "loss": 1.5341,
+      "epoch": 0.65,
+      "grad_norm": 4.5042724609375,
+      "learning_rate": 1.568673388277622e-05,
+      "loss": 1.396,
       "step": 5156
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 91.2535629272461,
-      "learning_rate": 9.66422772376466e-06,
-      "loss": 2.4091,
+      "epoch": 0.65,
+      "grad_norm": 7.471918106079102,
+      "learning_rate": 1.5685897167719534e-05,
+      "loss": 1.0309,
       "step": 5157
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 30.0625057220459,
-      "learning_rate": 9.66222311316027e-06,
-      "loss": 1.2433,
+      "epoch": 0.65,
+      "grad_norm": 5.654822826385498,
+      "learning_rate": 1.5685060452662847e-05,
+      "loss": 0.134,
       "step": 5158
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 10.793164253234863,
-      "learning_rate": 9.66021850255588e-06,
-      "loss": 1.3247,
+      "epoch": 0.65,
+      "grad_norm": 26.877819061279297,
+      "learning_rate": 1.5684223737606158e-05,
+      "loss": 2.7742,
       "step": 5159
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 14.865962028503418,
-      "learning_rate": 9.65821389195149e-06,
-      "loss": 0.7533,
-      "step": 5160
-    },
-    {
-      "epoch": 1.55,
-      "eval_loss": 0.20735611021518707,
-      "eval_runtime": 43.7679,
-      "eval_samples_per_second": 33.792,
-      "eval_steps_per_second": 33.792,
+      "epoch": 0.65,
+      "grad_norm": 7.091485500335693,
+      "learning_rate": 1.568338702254947e-05,
+      "loss": 0.4301,
       "step": 5160
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 18.602514266967773,
-      "learning_rate": 9.656209281347098e-06,
-      "loss": 1.5636,
+      "epoch": 0.65,
+      "grad_norm": 68.3599624633789,
+      "learning_rate": 1.5682550307492785e-05,
+      "loss": 0.8968,
       "step": 5161
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 14.882292747497559,
-      "learning_rate": 9.65420467074271e-06,
-      "loss": 1.1038,
+      "epoch": 0.65,
+      "grad_norm": 7.607362747192383,
+      "learning_rate": 1.5681713592436095e-05,
+      "loss": 1.3871,
       "step": 5162
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 56.73546600341797,
-      "learning_rate": 9.65220006013832e-06,
-      "loss": 1.9049,
+      "epoch": 0.65,
+      "grad_norm": 16.349225997924805,
+      "learning_rate": 1.568087687737941e-05,
+      "loss": 1.6503,
       "step": 5163
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 13.342903137207031,
-      "learning_rate": 9.650195449533928e-06,
-      "loss": 1.6425,
+      "epoch": 0.65,
+      "grad_norm": 13.676576614379883,
+      "learning_rate": 1.5680040162322723e-05,
+      "loss": 1.7194,
       "step": 5164
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 13.89185905456543,
-      "learning_rate": 9.648190838929538e-06,
-      "loss": 1.0887,
+      "epoch": 0.65,
+      "grad_norm": 18.912063598632812,
+      "learning_rate": 1.5679203447266036e-05,
+      "loss": 2.1749,
       "step": 5165
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 72.23283386230469,
-      "learning_rate": 9.646186228325149e-06,
-      "loss": 2.1655,
+      "epoch": 0.65,
+      "grad_norm": 19.127605438232422,
+      "learning_rate": 1.5678366732209346e-05,
+      "loss": 1.3759,
       "step": 5166
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 17.974084854125977,
-      "learning_rate": 9.644181617720759e-06,
-      "loss": 2.2678,
+      "epoch": 0.65,
+      "grad_norm": 12.958504676818848,
+      "learning_rate": 1.567753001715266e-05,
+      "loss": 0.8994,
       "step": 5167
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 69.75627136230469,
-      "learning_rate": 9.642177007116369e-06,
-      "loss": 1.7682,
+      "epoch": 0.65,
+      "grad_norm": 16.317556381225586,
+      "learning_rate": 1.5676693302095974e-05,
+      "loss": 1.1109,
       "step": 5168
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 11.242698669433594,
-      "learning_rate": 9.640172396511979e-06,
-      "loss": 1.3168,
+      "epoch": 0.65,
+      "grad_norm": 11.528014183044434,
+      "learning_rate": 1.5675856587039284e-05,
+      "loss": 1.4751,
       "step": 5169
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 66.39893341064453,
-      "learning_rate": 9.638167785907589e-06,
-      "loss": 2.6081,
+      "epoch": 0.65,
+      "grad_norm": 3.7973978519439697,
+      "learning_rate": 1.5675019871982598e-05,
+      "loss": 0.4032,
       "step": 5170
     },
     {
-      "epoch": 1.55,
-      "grad_norm": 9.396846771240234,
-      "learning_rate": 9.636163175303199e-06,
-      "loss": 0.8918,
+      "epoch": 0.65,
+      "grad_norm": 17.930723190307617,
+      "learning_rate": 1.567418315692591e-05,
+      "loss": 1.8452,
       "step": 5171
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 30.329811096191406,
-      "learning_rate": 9.634158564698807e-06,
-      "loss": 2.5263,
+      "epoch": 0.65,
+      "grad_norm": 21.808406829833984,
+      "learning_rate": 1.5673346441869225e-05,
+      "loss": 3.4642,
       "step": 5172
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 8.505033493041992,
-      "learning_rate": 9.632153954094417e-06,
-      "loss": 0.9187,
+      "epoch": 0.65,
+      "grad_norm": 12.586520195007324,
+      "learning_rate": 1.5672509726812535e-05,
+      "loss": 0.6569,
       "step": 5173
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 15.874688148498535,
-      "learning_rate": 9.63014934349003e-06,
-      "loss": 1.2034,
+      "epoch": 0.65,
+      "grad_norm": 12.065384864807129,
+      "learning_rate": 1.567167301175585e-05,
+      "loss": 1.2497,
       "step": 5174
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 12.436075210571289,
-      "learning_rate": 9.628144732885638e-06,
-      "loss": 1.2702,
+      "epoch": 0.65,
+      "grad_norm": 26.136062622070312,
+      "learning_rate": 1.5670836296699162e-05,
+      "loss": 0.9877,
       "step": 5175
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 16.405839920043945,
-      "learning_rate": 9.626140122281248e-06,
-      "loss": 1.2595,
+      "epoch": 0.65,
+      "grad_norm": 29.510662078857422,
+      "learning_rate": 1.5669999581642473e-05,
+      "loss": 2.4064,
       "step": 5176
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 6.489515781402588,
-      "learning_rate": 9.624135511676858e-06,
-      "loss": 0.699,
+      "epoch": 0.65,
+      "grad_norm": 7.213874340057373,
+      "learning_rate": 1.5669162866585786e-05,
+      "loss": 0.9525,
       "step": 5177
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 15.752748489379883,
-      "learning_rate": 9.622130901072468e-06,
-      "loss": 1.8403,
+      "epoch": 0.65,
+      "grad_norm": 24.652938842773438,
+      "learning_rate": 1.56683261515291e-05,
+      "loss": 1.4499,
       "step": 5178
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 21.755474090576172,
-      "learning_rate": 9.620126290468076e-06,
-      "loss": 1.9319,
+      "epoch": 0.65,
+      "grad_norm": 9.178464889526367,
+      "learning_rate": 1.566748943647241e-05,
+      "loss": 0.7875,
       "step": 5179
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 14.760236740112305,
-      "learning_rate": 9.618121679863686e-06,
-      "loss": 1.3553,
+      "epoch": 0.65,
+      "grad_norm": 14.615882873535156,
+      "learning_rate": 1.5666652721415724e-05,
+      "loss": 0.9102,
       "step": 5180
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 20.644493103027344,
-      "learning_rate": 9.616117069259298e-06,
-      "loss": 1.2051,
+      "epoch": 0.65,
+      "grad_norm": 16.932817459106445,
+      "learning_rate": 1.5665816006359034e-05,
+      "loss": 1.8659,
       "step": 5181
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 18.111047744750977,
-      "learning_rate": 9.614112458654906e-06,
-      "loss": 1.1888,
+      "epoch": 0.65,
+      "grad_norm": 13.214982032775879,
+      "learning_rate": 1.5664979291302348e-05,
+      "loss": 1.4819,
       "step": 5182
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 9.958390235900879,
-      "learning_rate": 9.612107848050517e-06,
-      "loss": 1.2644,
+      "epoch": 0.65,
+      "grad_norm": 14.489808082580566,
+      "learning_rate": 1.566414257624566e-05,
+      "loss": 2.7115,
       "step": 5183
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 16.58241081237793,
-      "learning_rate": 9.610103237446127e-06,
-      "loss": 0.8966,
+      "epoch": 0.65,
+      "grad_norm": 32.23086166381836,
+      "learning_rate": 1.5663305861188972e-05,
+      "loss": 2.2479,
       "step": 5184
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 27.23797607421875,
-      "learning_rate": 9.608098626841737e-06,
-      "loss": 2.5713,
+      "epoch": 0.65,
+      "grad_norm": 13.670648574829102,
+      "learning_rate": 1.5662469146132285e-05,
+      "loss": 1.5881,
       "step": 5185
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 28.486438751220703,
-      "learning_rate": 9.606094016237347e-06,
-      "loss": 2.6203,
+      "epoch": 0.65,
+      "grad_norm": 23.302671432495117,
+      "learning_rate": 1.5661632431075596e-05,
+      "loss": 2.2353,
       "step": 5186
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 15.266054153442383,
-      "learning_rate": 9.604089405632957e-06,
-      "loss": 1.8284,
+      "epoch": 0.65,
+      "grad_norm": 22.597736358642578,
+      "learning_rate": 1.566079571601891e-05,
+      "loss": 2.3094,
       "step": 5187
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 27.08075714111328,
-      "learning_rate": 9.602084795028567e-06,
-      "loss": 1.749,
+      "epoch": 0.65,
+      "grad_norm": 12.448516845703125,
+      "learning_rate": 1.5659959000962223e-05,
+      "loss": 2.5301,
       "step": 5188
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 15.239933013916016,
-      "learning_rate": 9.600080184424177e-06,
-      "loss": 1.4726,
+      "epoch": 0.65,
+      "grad_norm": 6.709007740020752,
+      "learning_rate": 1.5659122285905537e-05,
+      "loss": 0.5511,
       "step": 5189
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 29.341596603393555,
-      "learning_rate": 9.598075573819785e-06,
-      "loss": 1.697,
+      "epoch": 0.65,
+      "grad_norm": 10.571820259094238,
+      "learning_rate": 1.5658285570848847e-05,
+      "loss": 1.4207,
       "step": 5190
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 17.9268856048584,
-      "learning_rate": 9.596070963215396e-06,
-      "loss": 1.7262,
+      "epoch": 0.65,
+      "grad_norm": 14.686131477355957,
+      "learning_rate": 1.565744885579216e-05,
+      "loss": 0.6408,
       "step": 5191
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 18.277809143066406,
-      "learning_rate": 9.594066352611006e-06,
-      "loss": 2.4779,
+      "epoch": 0.65,
+      "grad_norm": 8.774033546447754,
+      "learning_rate": 1.5656612140735474e-05,
+      "loss": 1.8296,
       "step": 5192
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 24.198942184448242,
-      "learning_rate": 9.592061742006616e-06,
-      "loss": 1.2228,
+      "epoch": 0.65,
+      "grad_norm": 13.244007110595703,
+      "learning_rate": 1.5655775425678784e-05,
+      "loss": 2.2862,
       "step": 5193
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 10.138649940490723,
-      "learning_rate": 9.590057131402226e-06,
-      "loss": 2.3378,
+      "epoch": 0.65,
+      "grad_norm": 4.894765853881836,
+      "learning_rate": 1.5654938710622098e-05,
+      "loss": 0.178,
       "step": 5194
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 15.209259986877441,
-      "learning_rate": 9.588052520797836e-06,
-      "loss": 1.8108,
+      "epoch": 0.65,
+      "grad_norm": 4.754327774047852,
+      "learning_rate": 1.5654101995565412e-05,
+      "loss": 0.6165,
       "step": 5195
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 22.023107528686523,
-      "learning_rate": 9.586047910193446e-06,
-      "loss": 1.5428,
+      "epoch": 0.65,
+      "grad_norm": 1.4888056516647339,
+      "learning_rate": 1.5653265280508725e-05,
+      "loss": 0.0392,
       "step": 5196
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 18.7374210357666,
-      "learning_rate": 9.584043299589056e-06,
-      "loss": 1.6044,
+      "epoch": 0.65,
+      "grad_norm": 15.606212615966797,
+      "learning_rate": 1.5652428565452036e-05,
+      "loss": 1.7273,
       "step": 5197
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 21.584501266479492,
-      "learning_rate": 9.582038688984664e-06,
-      "loss": 2.0156,
+      "epoch": 0.65,
+      "grad_norm": 10.794861793518066,
+      "learning_rate": 1.565159185039535e-05,
+      "loss": 1.3038,
       "step": 5198
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 17.4039363861084,
-      "learning_rate": 9.580034078380276e-06,
-      "loss": 1.3307,
+      "epoch": 0.65,
+      "grad_norm": 15.449944496154785,
+      "learning_rate": 1.5650755135338663e-05,
+      "loss": 2.1364,
       "step": 5199
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 10.785429000854492,
-      "learning_rate": 9.578029467775886e-06,
-      "loss": 1.2009,
+      "epoch": 0.65,
+      "grad_norm": 17.74704933166504,
+      "learning_rate": 1.5649918420281973e-05,
+      "loss": 2.6271,
       "step": 5200
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 13.932561874389648,
-      "learning_rate": 9.576024857171495e-06,
-      "loss": 1.6818,
+      "epoch": 0.65,
+      "eval_loss": 0.11776099354028702,
+      "eval_runtime": 95.0543,
+      "eval_samples_per_second": 37.263,
+      "eval_steps_per_second": 37.263,
+      "step": 5200
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.526359558105469,
+      "learning_rate": 1.5649081705225287e-05,
+      "loss": 1.1783,
       "step": 5201
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 13.384397506713867,
-      "learning_rate": 9.574020246567105e-06,
-      "loss": 1.7232,
+      "epoch": 0.65,
+      "grad_norm": 8.998695373535156,
+      "learning_rate": 1.56482449901686e-05,
+      "loss": 1.1888,
       "step": 5202
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 32.59089279174805,
-      "learning_rate": 9.572015635962715e-06,
-      "loss": 2.1975,
+      "epoch": 0.65,
+      "grad_norm": 34.16201400756836,
+      "learning_rate": 1.5647408275111914e-05,
+      "loss": 1.4792,
       "step": 5203
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 19.552274703979492,
-      "learning_rate": 9.570011025358325e-06,
-      "loss": 2.2128,
+      "epoch": 0.65,
+      "grad_norm": 22.636789321899414,
+      "learning_rate": 1.5646571560055224e-05,
+      "loss": 1.4457,
       "step": 5204
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 15.523249626159668,
-      "learning_rate": 9.568006414753935e-06,
-      "loss": 1.1143,
+      "epoch": 0.65,
+      "grad_norm": 32.693355560302734,
+      "learning_rate": 1.5645734844998538e-05,
+      "loss": 2.9091,
       "step": 5205
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 35.84797286987305,
-      "learning_rate": 9.566001804149545e-06,
-      "loss": 2.5374,
+      "epoch": 0.65,
+      "grad_norm": 16.178911209106445,
+      "learning_rate": 1.5644898129941852e-05,
+      "loss": 1.4922,
       "step": 5206
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 14.734626770019531,
-      "learning_rate": 9.563997193545155e-06,
-      "loss": 1.2287,
+      "epoch": 0.65,
+      "grad_norm": 15.090814590454102,
+      "learning_rate": 1.5644061414885162e-05,
+      "loss": 1.7392,
       "step": 5207
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 13.902031898498535,
-      "learning_rate": 9.561992582940765e-06,
-      "loss": 1.4145,
+      "epoch": 0.65,
+      "grad_norm": 10.362354278564453,
+      "learning_rate": 1.5643224699828476e-05,
+      "loss": 0.7582,
       "step": 5208
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 9.233687400817871,
-      "learning_rate": 9.559987972336374e-06,
-      "loss": 1.2157,
+      "epoch": 0.65,
+      "grad_norm": 12.946538925170898,
+      "learning_rate": 1.5642387984771786e-05,
+      "loss": 0.9911,
       "step": 5209
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 47.99122619628906,
-      "learning_rate": 9.557983361731984e-06,
-      "loss": 1.6352,
+      "epoch": 0.65,
+      "grad_norm": 19.269906997680664,
+      "learning_rate": 1.56415512697151e-05,
+      "loss": 3.0072,
       "step": 5210
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 18.855663299560547,
-      "learning_rate": 9.555978751127596e-06,
-      "loss": 2.1963,
+      "epoch": 0.65,
+      "grad_norm": 13.436691284179688,
+      "learning_rate": 1.5640714554658413e-05,
+      "loss": 1.9378,
       "step": 5211
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 25.103059768676758,
-      "learning_rate": 9.553974140523204e-06,
-      "loss": 2.7753,
+      "epoch": 0.65,
+      "grad_norm": 11.440930366516113,
+      "learning_rate": 1.5639877839601723e-05,
+      "loss": 0.5874,
       "step": 5212
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 19.241830825805664,
-      "learning_rate": 9.551969529918814e-06,
-      "loss": 1.5551,
+      "epoch": 0.65,
+      "grad_norm": 16.88869857788086,
+      "learning_rate": 1.5639041124545037e-05,
+      "loss": 1.3259,
       "step": 5213
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 13.835005760192871,
-      "learning_rate": 9.549964919314424e-06,
-      "loss": 1.4738,
+      "epoch": 0.65,
+      "grad_norm": 26.100208282470703,
+      "learning_rate": 1.5638204409488347e-05,
+      "loss": 2.6398,
       "step": 5214
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 8.090837478637695,
-      "learning_rate": 9.547960308710034e-06,
-      "loss": 1.7791,
+      "epoch": 0.65,
+      "grad_norm": 7.826216697692871,
+      "learning_rate": 1.563736769443166e-05,
+      "loss": 1.6174,
       "step": 5215
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 22.887216567993164,
-      "learning_rate": 9.545955698105643e-06,
-      "loss": 1.4865,
+      "epoch": 0.65,
+      "grad_norm": 15.352578163146973,
+      "learning_rate": 1.5636530979374975e-05,
+      "loss": 1.6649,
       "step": 5216
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 18.71986961364746,
-      "learning_rate": 9.543951087501254e-06,
-      "loss": 1.7043,
+      "epoch": 0.65,
+      "grad_norm": 7.8712077140808105,
+      "learning_rate": 1.563569426431829e-05,
+      "loss": 1.7075,
       "step": 5217
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 9.452502250671387,
-      "learning_rate": 9.541946476896864e-06,
-      "loss": 1.9759,
+      "epoch": 0.65,
+      "grad_norm": 19.553529739379883,
+      "learning_rate": 1.56348575492616e-05,
+      "loss": 1.2477,
       "step": 5218
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 18.228004455566406,
-      "learning_rate": 9.539941866292473e-06,
-      "loss": 1.4368,
+      "epoch": 0.65,
+      "grad_norm": 30.20794105529785,
+      "learning_rate": 1.5634020834204912e-05,
+      "loss": 2.1334,
       "step": 5219
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 27.472970962524414,
-      "learning_rate": 9.537937255688083e-06,
-      "loss": 1.5878,
+      "epoch": 0.66,
+      "grad_norm": 17.59926414489746,
+      "learning_rate": 1.5633184119148226e-05,
+      "loss": 2.1509,
       "step": 5220
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 34.94017028808594,
-      "learning_rate": 9.535932645083693e-06,
-      "loss": 1.7284,
+      "epoch": 0.66,
+      "grad_norm": 14.72649097442627,
+      "learning_rate": 1.5632347404091536e-05,
+      "loss": 1.6159,
       "step": 5221
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 19.611160278320312,
-      "learning_rate": 9.533928034479303e-06,
-      "loss": 1.9294,
+      "epoch": 0.66,
+      "grad_norm": 7.848047733306885,
+      "learning_rate": 1.563151068903485e-05,
+      "loss": 1.0181,
       "step": 5222
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 181.42205810546875,
-      "learning_rate": 9.531923423874913e-06,
-      "loss": 1.5287,
+      "epoch": 0.66,
+      "grad_norm": 10.326218605041504,
+      "learning_rate": 1.5630673973978163e-05,
+      "loss": 1.5618,
       "step": 5223
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 27.391780853271484,
-      "learning_rate": 9.529918813270523e-06,
-      "loss": 1.0979,
+      "epoch": 0.66,
+      "grad_norm": 5.257458686828613,
+      "learning_rate": 1.5629837258921477e-05,
+      "loss": 1.1577,
       "step": 5224
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 19.48676872253418,
-      "learning_rate": 9.527914202666133e-06,
-      "loss": 1.9428,
+      "epoch": 0.66,
+      "grad_norm": 14.704678535461426,
+      "learning_rate": 1.5629000543864787e-05,
+      "loss": 1.8026,
       "step": 5225
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 60.94241714477539,
-      "learning_rate": 9.525909592061743e-06,
-      "loss": 2.5457,
+      "epoch": 0.66,
+      "grad_norm": 24.467130661010742,
+      "learning_rate": 1.56281638288081e-05,
+      "loss": 2.5593,
       "step": 5226
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 39.271629333496094,
-      "learning_rate": 9.523904981457352e-06,
-      "loss": 1.834,
+      "epoch": 0.66,
+      "grad_norm": 21.703311920166016,
+      "learning_rate": 1.5627327113751415e-05,
+      "loss": 3.3068,
       "step": 5227
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 7.752985000610352,
-      "learning_rate": 9.521900370852962e-06,
-      "loss": 1.3035,
+      "epoch": 0.66,
+      "grad_norm": 19.474443435668945,
+      "learning_rate": 1.5626490398694725e-05,
+      "loss": 1.4674,
       "step": 5228
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 11.235089302062988,
-      "learning_rate": 9.519895760248572e-06,
-      "loss": 0.659,
+      "epoch": 0.66,
+      "grad_norm": 14.166336059570312,
+      "learning_rate": 1.562565368363804e-05,
+      "loss": 2.139,
       "step": 5229
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 22.830162048339844,
-      "learning_rate": 9.517891149644182e-06,
-      "loss": 2.1836,
+      "epoch": 0.66,
+      "grad_norm": 8.847391128540039,
+      "learning_rate": 1.5624816968581352e-05,
+      "loss": 1.7414,
       "step": 5230
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 40.63211441040039,
-      "learning_rate": 9.515886539039792e-06,
-      "loss": 2.0237,
+      "epoch": 0.66,
+      "grad_norm": 26.29215431213379,
+      "learning_rate": 1.5623980253524666e-05,
+      "loss": 2.1403,
       "step": 5231
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 15.20790958404541,
-      "learning_rate": 9.513881928435402e-06,
-      "loss": 0.9827,
+      "epoch": 0.66,
+      "grad_norm": 35.24619674682617,
+      "learning_rate": 1.5623143538467976e-05,
+      "loss": 2.2782,
       "step": 5232
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 24.22981071472168,
-      "learning_rate": 9.511877317831012e-06,
-      "loss": 1.6325,
+      "epoch": 0.66,
+      "grad_norm": 9.792732238769531,
+      "learning_rate": 1.562230682341129e-05,
+      "loss": 3.4117,
       "step": 5233
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 69.4743423461914,
-      "learning_rate": 9.509872707226622e-06,
-      "loss": 2.8366,
+      "epoch": 0.66,
+      "grad_norm": 8.300969123840332,
+      "learning_rate": 1.56214701083546e-05,
+      "loss": 0.9535,
       "step": 5234
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 16.03273582458496,
-      "learning_rate": 9.50786809662223e-06,
-      "loss": 1.8027,
+      "epoch": 0.66,
+      "grad_norm": 22.07040023803711,
+      "learning_rate": 1.5620633393297914e-05,
+      "loss": 2.7798,
       "step": 5235
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 27.365671157836914,
-      "learning_rate": 9.505863486017843e-06,
-      "loss": 1.7576,
+      "epoch": 0.66,
+      "grad_norm": 12.015377044677734,
+      "learning_rate": 1.5619796678241227e-05,
+      "loss": 2.0949,
       "step": 5236
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 24.969091415405273,
-      "learning_rate": 9.503858875413453e-06,
-      "loss": 2.0772,
+      "epoch": 0.66,
+      "grad_norm": 38.503719329833984,
+      "learning_rate": 1.5618959963184538e-05,
+      "loss": 2.7961,
       "step": 5237
     },
     {
-      "epoch": 1.57,
-      "grad_norm": 17.248416900634766,
-      "learning_rate": 9.501854264809061e-06,
-      "loss": 1.3726,
+      "epoch": 0.66,
+      "grad_norm": 8.067269325256348,
+      "learning_rate": 1.561812324812785e-05,
+      "loss": 1.1209,
       "step": 5238
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 15.464237213134766,
-      "learning_rate": 9.499849654204671e-06,
-      "loss": 1.4722,
+      "epoch": 0.66,
+      "grad_norm": 36.50723648071289,
+      "learning_rate": 1.561728653307116e-05,
+      "loss": 3.2759,
       "step": 5239
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 24.053791046142578,
-      "learning_rate": 9.497845043600281e-06,
-      "loss": 1.4818,
+      "epoch": 0.66,
+      "grad_norm": 8.340876579284668,
+      "learning_rate": 1.5616449818014475e-05,
+      "loss": 0.8212,
       "step": 5240
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 21.699386596679688,
-      "learning_rate": 9.495840432995891e-06,
-      "loss": 2.1324,
+      "epoch": 0.66,
+      "grad_norm": 18.946107864379883,
+      "learning_rate": 1.561561310295779e-05,
+      "loss": 1.8706,
       "step": 5241
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 32.680335998535156,
-      "learning_rate": 9.493835822391501e-06,
-      "loss": 2.5121,
+      "epoch": 0.66,
+      "grad_norm": 13.62232780456543,
+      "learning_rate": 1.56147763879011e-05,
+      "loss": 2.333,
       "step": 5242
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 9.323308944702148,
-      "learning_rate": 9.491831211787111e-06,
-      "loss": 1.7795,
+      "epoch": 0.66,
+      "grad_norm": 5.632523536682129,
+      "learning_rate": 1.5613939672844413e-05,
+      "loss": 1.7993,
       "step": 5243
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 9.845460891723633,
-      "learning_rate": 9.489826601182722e-06,
-      "loss": 1.5512,
+      "epoch": 0.66,
+      "grad_norm": 49.13154602050781,
+      "learning_rate": 1.5613102957787726e-05,
+      "loss": 2.3888,
       "step": 5244
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 12.808528900146484,
-      "learning_rate": 9.487821990578332e-06,
-      "loss": 1.6268,
+      "epoch": 0.66,
+      "grad_norm": 40.47331619262695,
+      "learning_rate": 1.561226624273104e-05,
+      "loss": 1.5035,
       "step": 5245
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 30.653648376464844,
-      "learning_rate": 9.48581737997394e-06,
-      "loss": 2.2517,
+      "epoch": 0.66,
+      "grad_norm": 33.18256378173828,
+      "learning_rate": 1.561142952767435e-05,
+      "loss": 1.3711,
       "step": 5246
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 37.43178939819336,
-      "learning_rate": 9.48381276936955e-06,
-      "loss": 1.6323,
+      "epoch": 0.66,
+      "grad_norm": 13.688959121704102,
+      "learning_rate": 1.5610592812617664e-05,
+      "loss": 1.282,
       "step": 5247
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 23.361366271972656,
-      "learning_rate": 9.481808158765162e-06,
-      "loss": 1.6171,
+      "epoch": 0.66,
+      "grad_norm": 15.67768383026123,
+      "learning_rate": 1.5609756097560978e-05,
+      "loss": 2.3848,
       "step": 5248
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 43.43690872192383,
-      "learning_rate": 9.47980354816077e-06,
-      "loss": 1.8716,
+      "epoch": 0.66,
+      "grad_norm": 8.087139129638672,
+      "learning_rate": 1.5608919382504288e-05,
+      "loss": 1.9862,
       "step": 5249
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 16.062522888183594,
-      "learning_rate": 9.47779893755638e-06,
-      "loss": 2.9676,
+      "epoch": 0.66,
+      "grad_norm": 19.833606719970703,
+      "learning_rate": 1.56080826674476e-05,
+      "loss": 2.536,
       "step": 5250
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 34.8533935546875,
-      "learning_rate": 9.47579432695199e-06,
-      "loss": 1.6051,
+      "epoch": 0.66,
+      "grad_norm": 11.725749969482422,
+      "learning_rate": 1.5607245952390915e-05,
+      "loss": 2.0792,
       "step": 5251
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 20.79990577697754,
-      "learning_rate": 9.4737897163476e-06,
-      "loss": 2.5814,
+      "epoch": 0.66,
+      "grad_norm": 23.482179641723633,
+      "learning_rate": 1.560640923733423e-05,
+      "loss": 2.3725,
       "step": 5252
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 94.41461181640625,
-      "learning_rate": 9.47178510574321e-06,
-      "loss": 3.8194,
+      "epoch": 0.66,
+      "grad_norm": 13.740874290466309,
+      "learning_rate": 1.560557252227754e-05,
+      "loss": 2.5001,
       "step": 5253
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 9.540099143981934,
-      "learning_rate": 9.46978049513882e-06,
-      "loss": 2.1172,
+      "epoch": 0.66,
+      "grad_norm": 19.1165714263916,
+      "learning_rate": 1.5604735807220853e-05,
+      "loss": 2.192,
       "step": 5254
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 18.652847290039062,
-      "learning_rate": 9.46777588453443e-06,
-      "loss": 1.3757,
+      "epoch": 0.66,
+      "grad_norm": 13.846579551696777,
+      "learning_rate": 1.5603899092164166e-05,
+      "loss": 2.5799,
       "step": 5255
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 6.206528663635254,
-      "learning_rate": 9.465771273930041e-06,
-      "loss": 0.92,
+      "epoch": 0.66,
+      "grad_norm": 12.332139015197754,
+      "learning_rate": 1.5603062377107477e-05,
+      "loss": 0.978,
       "step": 5256
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 18.400524139404297,
-      "learning_rate": 9.46376666332565e-06,
-      "loss": 1.8978,
+      "epoch": 0.66,
+      "grad_norm": 20.706724166870117,
+      "learning_rate": 1.560222566205079e-05,
+      "loss": 3.2421,
       "step": 5257
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 22.74957275390625,
-      "learning_rate": 9.46176205272126e-06,
-      "loss": 1.8973,
+      "epoch": 0.66,
+      "grad_norm": 15.696081161499023,
+      "learning_rate": 1.5601388946994104e-05,
+      "loss": 1.1646,
       "step": 5258
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 15.965211868286133,
-      "learning_rate": 9.45975744211687e-06,
-      "loss": 1.8043,
+      "epoch": 0.66,
+      "grad_norm": 17.897335052490234,
+      "learning_rate": 1.5600552231937418e-05,
+      "loss": 2.4383,
       "step": 5259
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 19.99260139465332,
-      "learning_rate": 9.45775283151248e-06,
-      "loss": 1.6596,
+      "epoch": 0.66,
+      "grad_norm": 10.454360961914062,
+      "learning_rate": 1.5599715516880728e-05,
+      "loss": 1.5693,
       "step": 5260
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 9.259965896606445,
-      "learning_rate": 9.45574822090809e-06,
-      "loss": 1.4668,
+      "epoch": 0.66,
+      "grad_norm": 13.172468185424805,
+      "learning_rate": 1.559887880182404e-05,
+      "loss": 1.4832,
       "step": 5261
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 49.869110107421875,
-      "learning_rate": 9.4537436103037e-06,
-      "loss": 3.4404,
+      "epoch": 0.66,
+      "grad_norm": 19.44193458557129,
+      "learning_rate": 1.559804208676735e-05,
+      "loss": 3.0477,
       "step": 5262
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 26.978092193603516,
-      "learning_rate": 9.45173899969931e-06,
-      "loss": 2.3785,
+      "epoch": 0.66,
+      "grad_norm": 52.53721618652344,
+      "learning_rate": 1.5597205371710665e-05,
+      "loss": 1.1252,
       "step": 5263
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 37.760353088378906,
-      "learning_rate": 9.449734389094918e-06,
-      "loss": 1.1802,
+      "epoch": 0.66,
+      "grad_norm": 10.68852710723877,
+      "learning_rate": 1.559636865665398e-05,
+      "loss": 1.5092,
       "step": 5264
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 24.364702224731445,
-      "learning_rate": 9.447729778490528e-06,
-      "loss": 2.288,
+      "epoch": 0.66,
+      "grad_norm": 1.36933171749115,
+      "learning_rate": 1.559553194159729e-05,
+      "loss": 0.0572,
       "step": 5265
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 17.613224029541016,
-      "learning_rate": 9.445725167886138e-06,
-      "loss": 1.8934,
+      "epoch": 0.66,
+      "grad_norm": 14.17722225189209,
+      "learning_rate": 1.5594695226540603e-05,
+      "loss": 0.3962,
       "step": 5266
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 25.8133544921875,
-      "learning_rate": 9.443720557281748e-06,
-      "loss": 1.6238,
+      "epoch": 0.66,
+      "grad_norm": 10.241806030273438,
+      "learning_rate": 1.5593858511483913e-05,
+      "loss": 2.3737,
       "step": 5267
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 32.17334747314453,
-      "learning_rate": 9.441715946677358e-06,
-      "loss": 1.8157,
+      "epoch": 0.66,
+      "grad_norm": 28.391183853149414,
+      "learning_rate": 1.5593021796427227e-05,
+      "loss": 1.0734,
       "step": 5268
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 16.442485809326172,
-      "learning_rate": 9.439711336072969e-06,
-      "loss": 1.1689,
+      "epoch": 0.66,
+      "grad_norm": 14.059039115905762,
+      "learning_rate": 1.559218508137054e-05,
+      "loss": 1.8246,
       "step": 5269
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 11.316102027893066,
-      "learning_rate": 9.437706725468579e-06,
-      "loss": 2.063,
+      "epoch": 0.66,
+      "grad_norm": 6.69702672958374,
+      "learning_rate": 1.559134836631385e-05,
+      "loss": 0.1278,
       "step": 5270
     },
     {
-      "epoch": 1.58,
-      "grad_norm": 11.197502136230469,
-      "learning_rate": 9.435702114864189e-06,
-      "loss": 1.4149,
+      "epoch": 0.66,
+      "grad_norm": 8.767797470092773,
+      "learning_rate": 1.5590511651257164e-05,
+      "loss": 0.8445,
       "step": 5271
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 15.829837799072266,
-      "learning_rate": 9.433697504259797e-06,
-      "loss": 1.6842,
+      "epoch": 0.66,
+      "grad_norm": 6.164175987243652,
+      "learning_rate": 1.5589674936200478e-05,
+      "loss": 0.355,
       "step": 5272
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 44.153804779052734,
-      "learning_rate": 9.431692893655409e-06,
-      "loss": 1.8206,
+      "epoch": 0.66,
+      "grad_norm": 28.865331649780273,
+      "learning_rate": 1.558883822114379e-05,
+      "loss": 3.0153,
       "step": 5273
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 8.531728744506836,
-      "learning_rate": 9.429688283051019e-06,
-      "loss": 0.9972,
+      "epoch": 0.66,
+      "grad_norm": 12.738309860229492,
+      "learning_rate": 1.5588001506087102e-05,
+      "loss": 1.9587,
       "step": 5274
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 16.26192283630371,
-      "learning_rate": 9.427683672446627e-06,
-      "loss": 1.8538,
+      "epoch": 0.66,
+      "grad_norm": 30.209096908569336,
+      "learning_rate": 1.5587164791030416e-05,
+      "loss": 1.3462,
       "step": 5275
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 10.455742835998535,
-      "learning_rate": 9.425679061842237e-06,
-      "loss": 1.0675,
+      "epoch": 0.66,
+      "grad_norm": 26.268959045410156,
+      "learning_rate": 1.558632807597373e-05,
+      "loss": 2.0748,
       "step": 5276
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 19.000938415527344,
-      "learning_rate": 9.423674451237848e-06,
-      "loss": 1.9213,
+      "epoch": 0.66,
+      "grad_norm": 7.454601287841797,
+      "learning_rate": 1.558549136091704e-05,
+      "loss": 0.8775,
       "step": 5277
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 17.04540252685547,
-      "learning_rate": 9.421669840633458e-06,
-      "loss": 1.4436,
+      "epoch": 0.66,
+      "grad_norm": 6.933131217956543,
+      "learning_rate": 1.5584654645860353e-05,
+      "loss": 0.4548,
       "step": 5278
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 15.419578552246094,
-      "learning_rate": 9.419665230029068e-06,
-      "loss": 1.9087,
+      "epoch": 0.66,
+      "grad_norm": 6.293020725250244,
+      "learning_rate": 1.5583817930803667e-05,
+      "loss": 2.0141,
       "step": 5279
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 18.85577392578125,
-      "learning_rate": 9.417660619424678e-06,
-      "loss": 2.2146,
-      "step": 5280
-    },
-    {
-      "epoch": 1.59,
-      "eval_loss": 0.1917702704668045,
-      "eval_runtime": 43.7028,
-      "eval_samples_per_second": 33.842,
-      "eval_steps_per_second": 33.842,
+      "epoch": 0.66,
+      "grad_norm": 45.84587478637695,
+      "learning_rate": 1.558298121574698e-05,
+      "loss": 1.7475,
       "step": 5280
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 24.73578453063965,
-      "learning_rate": 9.415656008820288e-06,
-      "loss": 2.0709,
+      "epoch": 0.66,
+      "grad_norm": 12.852069854736328,
+      "learning_rate": 1.558214450069029e-05,
+      "loss": 1.1722,
       "step": 5281
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 10.795376777648926,
-      "learning_rate": 9.413651398215898e-06,
-      "loss": 0.9204,
+      "epoch": 0.66,
+      "grad_norm": 9.643006324768066,
+      "learning_rate": 1.5581307785633604e-05,
+      "loss": 2.4848,
       "step": 5282
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 14.635207176208496,
-      "learning_rate": 9.411646787611506e-06,
-      "loss": 1.7222,
+      "epoch": 0.66,
+      "grad_norm": 11.011727333068848,
+      "learning_rate": 1.5580471070576918e-05,
+      "loss": 2.8387,
       "step": 5283
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 12.989263534545898,
-      "learning_rate": 9.409642177007116e-06,
-      "loss": 1.3394,
+      "epoch": 0.66,
+      "grad_norm": 13.584528923034668,
+      "learning_rate": 1.5579634355520228e-05,
+      "loss": 1.7407,
       "step": 5284
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 12.602009773254395,
-      "learning_rate": 9.407637566402728e-06,
-      "loss": 1.7059,
+      "epoch": 0.66,
+      "grad_norm": 12.662556648254395,
+      "learning_rate": 1.5578797640463542e-05,
+      "loss": 2.0174,
       "step": 5285
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 20.175142288208008,
-      "learning_rate": 9.405632955798337e-06,
-      "loss": 1.7654,
+      "epoch": 0.66,
+      "grad_norm": 16.633058547973633,
+      "learning_rate": 1.5577960925406856e-05,
+      "loss": 2.0047,
       "step": 5286
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 18.197296142578125,
-      "learning_rate": 9.403628345193947e-06,
-      "loss": 1.7068,
+      "epoch": 0.66,
+      "grad_norm": 12.066131591796875,
+      "learning_rate": 1.5577124210350166e-05,
+      "loss": 1.0126,
       "step": 5287
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 34.1553955078125,
-      "learning_rate": 9.401623734589557e-06,
-      "loss": 1.3868,
+      "epoch": 0.66,
+      "grad_norm": 35.76958084106445,
+      "learning_rate": 1.557628749529348e-05,
+      "loss": 2.378,
       "step": 5288
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 27.550334930419922,
-      "learning_rate": 9.399619123985167e-06,
-      "loss": 1.2817,
+      "epoch": 0.66,
+      "grad_norm": 8.880400657653809,
+      "learning_rate": 1.5575450780236793e-05,
+      "loss": 1.9348,
       "step": 5289
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 19.8583984375,
-      "learning_rate": 9.397614513380777e-06,
-      "loss": 1.9351,
+      "epoch": 0.66,
+      "grad_norm": 3.5895142555236816,
+      "learning_rate": 1.5574614065180103e-05,
+      "loss": 0.1277,
       "step": 5290
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 6.953277587890625,
-      "learning_rate": 9.395609902776387e-06,
-      "loss": 1.1993,
+      "epoch": 0.66,
+      "grad_norm": 4.755842208862305,
+      "learning_rate": 1.5573777350123417e-05,
+      "loss": 0.3882,
       "step": 5291
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 15.303847312927246,
-      "learning_rate": 9.393605292171997e-06,
-      "loss": 1.6499,
+      "epoch": 0.66,
+      "grad_norm": 12.279382705688477,
+      "learning_rate": 1.5572940635066727e-05,
+      "loss": 2.5931,
       "step": 5292
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 28.220230102539062,
-      "learning_rate": 9.391600681567607e-06,
-      "loss": 1.7453,
+      "epoch": 0.66,
+      "grad_norm": 4.911112308502197,
+      "learning_rate": 1.557210392001004e-05,
+      "loss": 0.6729,
       "step": 5293
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 40.280029296875,
-      "learning_rate": 9.389596070963216e-06,
-      "loss": 1.9758,
+      "epoch": 0.66,
+      "grad_norm": 11.579627990722656,
+      "learning_rate": 1.5571267204953355e-05,
+      "loss": 3.0826,
       "step": 5294
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 15.754435539245605,
-      "learning_rate": 9.387591460358826e-06,
-      "loss": 1.1859,
+      "epoch": 0.66,
+      "grad_norm": 8.23723030090332,
+      "learning_rate": 1.5570430489896665e-05,
+      "loss": 1.0335,
       "step": 5295
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 64.05276489257812,
-      "learning_rate": 9.385586849754436e-06,
-      "loss": 2.8985,
+      "epoch": 0.66,
+      "grad_norm": 23.425081253051758,
+      "learning_rate": 1.556959377483998e-05,
+      "loss": 1.8962,
       "step": 5296
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 12.505768775939941,
-      "learning_rate": 9.383582239150046e-06,
-      "loss": 1.7193,
+      "epoch": 0.66,
+      "grad_norm": 22.175535202026367,
+      "learning_rate": 1.5568757059783292e-05,
+      "loss": 2.0819,
       "step": 5297
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 22.20297622680664,
-      "learning_rate": 9.381577628545656e-06,
-      "loss": 1.8241,
+      "epoch": 0.66,
+      "grad_norm": 13.097886085510254,
+      "learning_rate": 1.5567920344726602e-05,
+      "loss": 1.8502,
       "step": 5298
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 17.095487594604492,
-      "learning_rate": 9.379573017941266e-06,
-      "loss": 1.7652,
+      "epoch": 0.67,
+      "grad_norm": 10.502854347229004,
+      "learning_rate": 1.5567083629669916e-05,
+      "loss": 1.0218,
       "step": 5299
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 30.220666885375977,
-      "learning_rate": 9.377568407336876e-06,
-      "loss": 2.0137,
+      "epoch": 0.67,
+      "grad_norm": 9.6867036819458,
+      "learning_rate": 1.556624691461323e-05,
+      "loss": 0.9032,
       "step": 5300
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 16.760820388793945,
-      "learning_rate": 9.375563796732484e-06,
-      "loss": 1.4244,
+      "epoch": 0.67,
+      "grad_norm": 21.5103702545166,
+      "learning_rate": 1.5565410199556543e-05,
+      "loss": 2.3803,
       "step": 5301
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 74.30367279052734,
-      "learning_rate": 9.373559186128095e-06,
-      "loss": 1.818,
+      "epoch": 0.67,
+      "grad_norm": 13.037554740905762,
+      "learning_rate": 1.5564573484499854e-05,
+      "loss": 1.0351,
       "step": 5302
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 8.43071460723877,
-      "learning_rate": 9.371554575523705e-06,
-      "loss": 0.9715,
+      "epoch": 0.67,
+      "grad_norm": 19.55864143371582,
+      "learning_rate": 1.5563736769443167e-05,
+      "loss": 0.9527,
       "step": 5303
     },
     {
-      "epoch": 1.59,
-      "grad_norm": 19.98003387451172,
-      "learning_rate": 9.369549964919315e-06,
-      "loss": 1.5193,
+      "epoch": 0.67,
+      "grad_norm": 4.488838195800781,
+      "learning_rate": 1.556290005438648e-05,
+      "loss": 0.074,
       "step": 5304
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 15.74586009979248,
-      "learning_rate": 9.367545354314925e-06,
-      "loss": 1.7498,
+      "epoch": 0.67,
+      "grad_norm": 7.514483451843262,
+      "learning_rate": 1.556206333932979e-05,
+      "loss": 0.5455,
       "step": 5305
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 16.798616409301758,
-      "learning_rate": 9.365540743710535e-06,
-      "loss": 0.9751,
+      "epoch": 0.67,
+      "grad_norm": 16.274606704711914,
+      "learning_rate": 1.5561226624273105e-05,
+      "loss": 1.7984,
       "step": 5306
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 15.125486373901367,
-      "learning_rate": 9.363536133106145e-06,
-      "loss": 2.1381,
+      "epoch": 0.67,
+      "grad_norm": 9.492721557617188,
+      "learning_rate": 1.556038990921642e-05,
+      "loss": 1.2578,
       "step": 5307
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 26.87381362915039,
-      "learning_rate": 9.361531522501755e-06,
-      "loss": 1.6783,
+      "epoch": 0.67,
+      "grad_norm": 11.834115982055664,
+      "learning_rate": 1.5559553194159732e-05,
+      "loss": 1.3758,
       "step": 5308
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 23.354263305664062,
-      "learning_rate": 9.359526911897363e-06,
-      "loss": 2.2363,
+      "epoch": 0.67,
+      "grad_norm": 23.507463455200195,
+      "learning_rate": 1.5558716479103042e-05,
+      "loss": 2.7738,
       "step": 5309
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 11.7379789352417,
-      "learning_rate": 9.357522301292975e-06,
-      "loss": 1.6394,
+      "epoch": 0.67,
+      "grad_norm": 18.14545440673828,
+      "learning_rate": 1.5557879764046356e-05,
+      "loss": 1.151,
       "step": 5310
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 16.918331146240234,
-      "learning_rate": 9.355517690688585e-06,
-      "loss": 2.0194,
+      "epoch": 0.67,
+      "grad_norm": 9.639022827148438,
+      "learning_rate": 1.555704304898967e-05,
+      "loss": 2.6962,
       "step": 5311
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 10.654472351074219,
-      "learning_rate": 9.353513080084194e-06,
-      "loss": 2.1258,
+      "epoch": 0.67,
+      "grad_norm": 37.715087890625,
+      "learning_rate": 1.555620633393298e-05,
+      "loss": 1.3331,
       "step": 5312
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 23.184656143188477,
-      "learning_rate": 9.351508469479804e-06,
-      "loss": 2.9148,
+      "epoch": 0.67,
+      "grad_norm": 16.375370025634766,
+      "learning_rate": 1.5555369618876294e-05,
+      "loss": 3.3425,
       "step": 5313
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 10.799365997314453,
-      "learning_rate": 9.349503858875414e-06,
-      "loss": 0.941,
+      "epoch": 0.67,
+      "grad_norm": 16.068984985351562,
+      "learning_rate": 1.5554532903819607e-05,
+      "loss": 2.0363,
       "step": 5314
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 12.17577838897705,
-      "learning_rate": 9.347499248271024e-06,
-      "loss": 0.8023,
+      "epoch": 0.67,
+      "grad_norm": 5.682156085968018,
+      "learning_rate": 1.5553696188762917e-05,
+      "loss": 2.2618,
       "step": 5315
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 25.213882446289062,
-      "learning_rate": 9.345494637666634e-06,
-      "loss": 1.6508,
+      "epoch": 0.67,
+      "grad_norm": 15.031339645385742,
+      "learning_rate": 1.555285947370623e-05,
+      "loss": 2.9795,
       "step": 5316
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 11.12057113647461,
-      "learning_rate": 9.343490027062244e-06,
-      "loss": 1.589,
+      "epoch": 0.67,
+      "grad_norm": 13.337397575378418,
+      "learning_rate": 1.5552022758649545e-05,
+      "loss": 0.792,
       "step": 5317
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 27.033601760864258,
-      "learning_rate": 9.341485416457854e-06,
-      "loss": 1.8537,
+      "epoch": 0.67,
+      "grad_norm": 27.936046600341797,
+      "learning_rate": 1.5551186043592855e-05,
+      "loss": 2.1488,
       "step": 5318
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 16.07554054260254,
-      "learning_rate": 9.339480805853464e-06,
-      "loss": 1.4078,
+      "epoch": 0.67,
+      "grad_norm": 19.554880142211914,
+      "learning_rate": 1.555034932853617e-05,
+      "loss": 3.7691,
       "step": 5319
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 16.774127960205078,
-      "learning_rate": 9.337476195249073e-06,
-      "loss": 1.2248,
+      "epoch": 0.67,
+      "grad_norm": 7.329078197479248,
+      "learning_rate": 1.554951261347948e-05,
+      "loss": 2.1147,
       "step": 5320
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 12.099883079528809,
-      "learning_rate": 9.335471584644683e-06,
-      "loss": 0.5273,
+      "epoch": 0.67,
+      "grad_norm": 24.679201126098633,
+      "learning_rate": 1.5548675898422793e-05,
+      "loss": 3.1108,
       "step": 5321
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 14.563944816589355,
-      "learning_rate": 9.333466974040295e-06,
-      "loss": 1.2972,
+      "epoch": 0.67,
+      "grad_norm": 15.159878730773926,
+      "learning_rate": 1.5547839183366106e-05,
+      "loss": 1.4298,
       "step": 5322
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 6.988048553466797,
-      "learning_rate": 9.331462363435903e-06,
-      "loss": 1.2299,
+      "epoch": 0.67,
+      "grad_norm": 85.50730895996094,
+      "learning_rate": 1.5547002468309417e-05,
+      "loss": 2.3678,
       "step": 5323
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 15.345603942871094,
-      "learning_rate": 9.329457752831513e-06,
-      "loss": 1.8004,
+      "epoch": 0.67,
+      "grad_norm": 9.848220825195312,
+      "learning_rate": 1.554616575325273e-05,
+      "loss": 0.7269,
       "step": 5324
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 8.920988082885742,
-      "learning_rate": 9.327453142227123e-06,
-      "loss": 1.799,
+      "epoch": 0.67,
+      "grad_norm": 49.288997650146484,
+      "learning_rate": 1.5545329038196044e-05,
+      "loss": 2.2035,
       "step": 5325
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 13.415757179260254,
-      "learning_rate": 9.325448531622733e-06,
-      "loss": 1.3893,
+      "epoch": 0.67,
+      "grad_norm": 14.30613899230957,
+      "learning_rate": 1.5544492323139354e-05,
+      "loss": 1.5891,
       "step": 5326
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 14.817873001098633,
-      "learning_rate": 9.323443921018343e-06,
-      "loss": 2.0269,
+      "epoch": 0.67,
+      "grad_norm": 17.46340560913086,
+      "learning_rate": 1.5543655608082668e-05,
+      "loss": 1.3967,
       "step": 5327
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 17.05929183959961,
-      "learning_rate": 9.321439310413953e-06,
-      "loss": 1.9584,
+      "epoch": 0.67,
+      "grad_norm": 12.865313529968262,
+      "learning_rate": 1.554281889302598e-05,
+      "loss": 0.9915,
       "step": 5328
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 10.261931419372559,
-      "learning_rate": 9.319434699809563e-06,
-      "loss": 1.5321,
+      "epoch": 0.67,
+      "grad_norm": 10.885964393615723,
+      "learning_rate": 1.5541982177969295e-05,
+      "loss": 1.491,
       "step": 5329
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 18.25226402282715,
-      "learning_rate": 9.317430089205174e-06,
-      "loss": 1.5914,
+      "epoch": 0.67,
+      "grad_norm": 52.34687423706055,
+      "learning_rate": 1.5541145462912605e-05,
+      "loss": 2.7669,
       "step": 5330
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 13.496363639831543,
-      "learning_rate": 9.315425478600782e-06,
-      "loss": 1.6435,
+      "epoch": 0.67,
+      "grad_norm": 17.977941513061523,
+      "learning_rate": 1.554030874785592e-05,
+      "loss": 1.8955,
       "step": 5331
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 26.062244415283203,
-      "learning_rate": 9.313420867996392e-06,
-      "loss": 1.892,
+      "epoch": 0.67,
+      "grad_norm": 22.670303344726562,
+      "learning_rate": 1.5539472032799233e-05,
+      "loss": 2.4956,
       "step": 5332
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 28.14862632751465,
-      "learning_rate": 9.311416257392002e-06,
-      "loss": 0.9718,
+      "epoch": 0.67,
+      "grad_norm": 10.357001304626465,
+      "learning_rate": 1.5538635317742543e-05,
+      "loss": 0.5737,
       "step": 5333
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 9.369063377380371,
-      "learning_rate": 9.309411646787612e-06,
-      "loss": 1.6377,
+      "epoch": 0.67,
+      "grad_norm": 9.255528450012207,
+      "learning_rate": 1.5537798602685856e-05,
+      "loss": 1.5152,
       "step": 5334
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 15.546137809753418,
-      "learning_rate": 9.307407036183222e-06,
-      "loss": 1.9211,
+      "epoch": 0.67,
+      "grad_norm": 5.881645679473877,
+      "learning_rate": 1.553696188762917e-05,
+      "loss": 1.94,
       "step": 5335
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 32.82224655151367,
-      "learning_rate": 9.305402425578832e-06,
-      "loss": 1.2737,
+      "epoch": 0.67,
+      "grad_norm": 7.286981105804443,
+      "learning_rate": 1.5536125172572484e-05,
+      "loss": 0.8161,
       "step": 5336
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 58.94491958618164,
-      "learning_rate": 9.303397814974442e-06,
-      "loss": 2.3802,
+      "epoch": 0.67,
+      "grad_norm": 22.43034553527832,
+      "learning_rate": 1.5535288457515794e-05,
+      "loss": 2.1878,
       "step": 5337
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 17.832128524780273,
-      "learning_rate": 9.30139320437005e-06,
-      "loss": 2.0701,
+      "epoch": 0.67,
+      "grad_norm": 11.97895622253418,
+      "learning_rate": 1.5534451742459108e-05,
+      "loss": 0.579,
       "step": 5338
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 25.756505966186523,
-      "learning_rate": 9.299388593765661e-06,
-      "loss": 1.6935,
+      "epoch": 0.67,
+      "grad_norm": 9.473199844360352,
+      "learning_rate": 1.553361502740242e-05,
+      "loss": 1.7777,
       "step": 5339
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 12.220959663391113,
-      "learning_rate": 9.297383983161273e-06,
-      "loss": 2.3572,
+      "epoch": 0.67,
+      "grad_norm": 15.786806106567383,
+      "learning_rate": 1.553277831234573e-05,
+      "loss": 1.3267,
       "step": 5340
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 28.716758728027344,
-      "learning_rate": 9.295379372556883e-06,
-      "loss": 1.8092,
+      "epoch": 0.67,
+      "grad_norm": 15.628312110900879,
+      "learning_rate": 1.5531941597289045e-05,
+      "loss": 1.7803,
       "step": 5341
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 17.98668670654297,
-      "learning_rate": 9.293374761952491e-06,
-      "loss": 1.61,
+      "epoch": 0.67,
+      "grad_norm": 7.371047019958496,
+      "learning_rate": 1.553110488223236e-05,
+      "loss": 0.9534,
       "step": 5342
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 15.28249454498291,
-      "learning_rate": 9.291370151348101e-06,
-      "loss": 2.0471,
+      "epoch": 0.67,
+      "grad_norm": 10.50656795501709,
+      "learning_rate": 1.553026816717567e-05,
+      "loss": 1.1232,
       "step": 5343
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 10.611080169677734,
-      "learning_rate": 9.289365540743711e-06,
-      "loss": 0.9532,
+      "epoch": 0.67,
+      "grad_norm": 14.262672424316406,
+      "learning_rate": 1.5529431452118983e-05,
+      "loss": 0.9357,
       "step": 5344
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 18.82200813293457,
-      "learning_rate": 9.287360930139321e-06,
-      "loss": 1.315,
+      "epoch": 0.67,
+      "grad_norm": 16.33648681640625,
+      "learning_rate": 1.5528594737062293e-05,
+      "loss": 2.7881,
       "step": 5345
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 19.420570373535156,
-      "learning_rate": 9.28535631953493e-06,
-      "loss": 1.0041,
+      "epoch": 0.67,
+      "grad_norm": 18.935623168945312,
+      "learning_rate": 1.5527758022005607e-05,
+      "loss": 0.9888,
       "step": 5346
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 16.25090980529785,
-      "learning_rate": 9.283351708930542e-06,
-      "loss": 2.2406,
+      "epoch": 0.67,
+      "grad_norm": 10.97852611541748,
+      "learning_rate": 1.552692130694892e-05,
+      "loss": 1.9534,
       "step": 5347
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 36.150394439697266,
-      "learning_rate": 9.281347098326152e-06,
-      "loss": 2.0748,
+      "epoch": 0.67,
+      "grad_norm": 6.660601615905762,
+      "learning_rate": 1.552608459189223e-05,
+      "loss": 2.1308,
       "step": 5348
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 69.37711334228516,
-      "learning_rate": 9.27934248772176e-06,
-      "loss": 2.1891,
+      "epoch": 0.67,
+      "grad_norm": 13.707566261291504,
+      "learning_rate": 1.5525247876835544e-05,
+      "loss": 0.9985,
       "step": 5349
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 9.345280647277832,
-      "learning_rate": 9.27733787711737e-06,
-      "loss": 1.2469,
+      "epoch": 0.67,
+      "grad_norm": 8.461606979370117,
+      "learning_rate": 1.5524411161778858e-05,
+      "loss": 1.68,
       "step": 5350
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 51.229942321777344,
-      "learning_rate": 9.27533326651298e-06,
-      "loss": 2.5097,
+      "epoch": 0.67,
+      "grad_norm": 8.554165840148926,
+      "learning_rate": 1.5523574446722168e-05,
+      "loss": 1.0881,
       "step": 5351
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 33.678749084472656,
-      "learning_rate": 9.27332865590859e-06,
-      "loss": 2.5323,
+      "epoch": 0.67,
+      "grad_norm": 25.228971481323242,
+      "learning_rate": 1.5522737731665482e-05,
+      "loss": 2.156,
       "step": 5352
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 20.67754364013672,
-      "learning_rate": 9.2713240453042e-06,
-      "loss": 3.1563,
+      "epoch": 0.67,
+      "grad_norm": 5.438741207122803,
+      "learning_rate": 1.5521901016608795e-05,
+      "loss": 2.7531,
       "step": 5353
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 19.53748321533203,
-      "learning_rate": 9.26931943469981e-06,
-      "loss": 1.1978,
+      "epoch": 0.67,
+      "grad_norm": 20.67116928100586,
+      "learning_rate": 1.5521064301552106e-05,
+      "loss": 1.821,
       "step": 5354
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 16.690845489501953,
-      "learning_rate": 9.26731482409542e-06,
-      "loss": 1.1068,
+      "epoch": 0.67,
+      "grad_norm": 8.502362251281738,
+      "learning_rate": 1.552022758649542e-05,
+      "loss": 0.7838,
       "step": 5355
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 19.17826271057129,
-      "learning_rate": 9.26531021349103e-06,
-      "loss": 1.8296,
+      "epoch": 0.67,
+      "grad_norm": 24.416032791137695,
+      "learning_rate": 1.5519390871438733e-05,
+      "loss": 3.4488,
       "step": 5356
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 15.507776260375977,
-      "learning_rate": 9.263305602886639e-06,
-      "loss": 1.6507,
+      "epoch": 0.67,
+      "grad_norm": 17.0931396484375,
+      "learning_rate": 1.5518554156382047e-05,
+      "loss": 1.8967,
       "step": 5357
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 129.35165405273438,
-      "learning_rate": 9.261300992282249e-06,
-      "loss": 1.632,
+      "epoch": 0.67,
+      "grad_norm": 20.132095336914062,
+      "learning_rate": 1.5517717441325357e-05,
+      "loss": 0.9988,
       "step": 5358
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 9.657642364501953,
-      "learning_rate": 9.259296381677861e-06,
-      "loss": 1.315,
+      "epoch": 0.67,
+      "grad_norm": 19.004472732543945,
+      "learning_rate": 1.551688072626867e-05,
+      "loss": 1.2706,
       "step": 5359
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 18.087684631347656,
-      "learning_rate": 9.25729177107347e-06,
-      "loss": 1.2448,
+      "epoch": 0.67,
+      "grad_norm": 14.730697631835938,
+      "learning_rate": 1.5516044011211984e-05,
+      "loss": 2.3345,
       "step": 5360
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 16.07610511779785,
-      "learning_rate": 9.25528716046908e-06,
-      "loss": 1.2188,
+      "epoch": 0.67,
+      "grad_norm": 14.693830490112305,
+      "learning_rate": 1.5515207296155295e-05,
+      "loss": 1.0428,
       "step": 5361
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 14.000444412231445,
-      "learning_rate": 9.25328254986469e-06,
-      "loss": 1.9099,
+      "epoch": 0.67,
+      "grad_norm": 20.921735763549805,
+      "learning_rate": 1.5514370581098608e-05,
+      "loss": 3.0207,
       "step": 5362
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 22.69326400756836,
-      "learning_rate": 9.2512779392603e-06,
-      "loss": 2.2418,
+      "epoch": 0.67,
+      "grad_norm": 10.459497451782227,
+      "learning_rate": 1.5513533866041922e-05,
+      "loss": 1.7045,
       "step": 5363
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 41.782066345214844,
-      "learning_rate": 9.24927332865591e-06,
-      "loss": 1.8902,
+      "epoch": 0.67,
+      "grad_norm": 26.556926727294922,
+      "learning_rate": 1.5512697150985235e-05,
+      "loss": 0.7821,
       "step": 5364
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 9.715392112731934,
-      "learning_rate": 9.24726871805152e-06,
-      "loss": 1.1494,
+      "epoch": 0.67,
+      "grad_norm": 54.429847717285156,
+      "learning_rate": 1.5511860435928546e-05,
+      "loss": 2.2101,
       "step": 5365
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 13.382871627807617,
-      "learning_rate": 9.24526410744713e-06,
-      "loss": 1.6138,
+      "epoch": 0.67,
+      "grad_norm": 7.08843994140625,
+      "learning_rate": 1.551102372087186e-05,
+      "loss": 1.3268,
       "step": 5366
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 33.94945526123047,
-      "learning_rate": 9.24325949684274e-06,
-      "loss": 1.562,
+      "epoch": 0.67,
+      "grad_norm": 11.765007972717285,
+      "learning_rate": 1.5510187005815173e-05,
+      "loss": 1.4397,
       "step": 5367
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 24.688888549804688,
-      "learning_rate": 9.241254886238348e-06,
-      "loss": 1.3487,
+      "epoch": 0.67,
+      "grad_norm": 15.495498657226562,
+      "learning_rate": 1.5509350290758483e-05,
+      "loss": 1.4547,
       "step": 5368
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 19.392419815063477,
-      "learning_rate": 9.239250275633958e-06,
-      "loss": 1.7019,
+      "epoch": 0.67,
+      "grad_norm": 6.075366497039795,
+      "learning_rate": 1.5508513575701797e-05,
+      "loss": 2.4579,
       "step": 5369
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 9.527002334594727,
-      "learning_rate": 9.237245665029568e-06,
-      "loss": 1.6555,
+      "epoch": 0.67,
+      "grad_norm": 10.94253921508789,
+      "learning_rate": 1.550767686064511e-05,
+      "loss": 0.4965,
       "step": 5370
     },
     {
-      "epoch": 1.61,
-      "grad_norm": 11.718170166015625,
-      "learning_rate": 9.235241054425179e-06,
-      "loss": 1.143,
+      "epoch": 0.67,
+      "grad_norm": 19.25712013244629,
+      "learning_rate": 1.550684014558842e-05,
+      "loss": 1.6914,
       "step": 5371
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 14.71898078918457,
-      "learning_rate": 9.233236443820789e-06,
-      "loss": 1.3483,
+      "epoch": 0.67,
+      "grad_norm": 26.579479217529297,
+      "learning_rate": 1.5506003430531734e-05,
+      "loss": 2.3986,
       "step": 5372
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 12.641324996948242,
-      "learning_rate": 9.231231833216399e-06,
-      "loss": 1.3701,
+      "epoch": 0.67,
+      "grad_norm": 16.759775161743164,
+      "learning_rate": 1.5505166715475045e-05,
+      "loss": 0.5726,
       "step": 5373
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 26.643348693847656,
-      "learning_rate": 9.229227222612009e-06,
-      "loss": 1.8719,
+      "epoch": 0.67,
+      "grad_norm": 8.281192779541016,
+      "learning_rate": 1.550433000041836e-05,
+      "loss": 0.3718,
       "step": 5374
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 11.661134719848633,
-      "learning_rate": 9.227222612007619e-06,
-      "loss": 1.5889,
+      "epoch": 0.67,
+      "grad_norm": 9.185592651367188,
+      "learning_rate": 1.5503493285361672e-05,
+      "loss": 0.5656,
       "step": 5375
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 14.931244850158691,
-      "learning_rate": 9.225218001403227e-06,
-      "loss": 1.7465,
+      "epoch": 0.67,
+      "grad_norm": 7.616875648498535,
+      "learning_rate": 1.5502656570304982e-05,
+      "loss": 2.2267,
       "step": 5376
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 11.898720741271973,
-      "learning_rate": 9.223213390798839e-06,
-      "loss": 1.6241,
+      "epoch": 0.67,
+      "grad_norm": 8.69092082977295,
+      "learning_rate": 1.5501819855248296e-05,
+      "loss": 1.4501,
       "step": 5377
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 16.589155197143555,
-      "learning_rate": 9.221208780194449e-06,
-      "loss": 1.4509,
+      "epoch": 0.67,
+      "grad_norm": 42.72675704956055,
+      "learning_rate": 1.550098314019161e-05,
+      "loss": 4.0754,
       "step": 5378
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 23.514259338378906,
-      "learning_rate": 9.219204169590058e-06,
-      "loss": 2.5493,
+      "epoch": 0.68,
+      "grad_norm": 21.543594360351562,
+      "learning_rate": 1.550014642513492e-05,
+      "loss": 2.0531,
       "step": 5379
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 17.52058982849121,
-      "learning_rate": 9.217199558985668e-06,
-      "loss": 1.2928,
+      "epoch": 0.68,
+      "grad_norm": 10.249034881591797,
+      "learning_rate": 1.5499309710078234e-05,
+      "loss": 2.3686,
       "step": 5380
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 28.47098159790039,
-      "learning_rate": 9.215194948381278e-06,
-      "loss": 1.2744,
+      "epoch": 0.68,
+      "grad_norm": 5.479506969451904,
+      "learning_rate": 1.5498472995021547e-05,
+      "loss": 0.1317,
       "step": 5381
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 15.93428897857666,
-      "learning_rate": 9.213190337776888e-06,
-      "loss": 1.5216,
+      "epoch": 0.68,
+      "grad_norm": 5.757992267608643,
+      "learning_rate": 1.5497636279964857e-05,
+      "loss": 0.3811,
       "step": 5382
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 73.04627227783203,
-      "learning_rate": 9.211185727172498e-06,
-      "loss": 0.9938,
+      "epoch": 0.68,
+      "grad_norm": 13.579237937927246,
+      "learning_rate": 1.549679956490817e-05,
+      "loss": 1.3426,
       "step": 5383
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 26.94357681274414,
-      "learning_rate": 9.209181116568108e-06,
-      "loss": 1.0955,
+      "epoch": 0.68,
+      "grad_norm": 10.298381805419922,
+      "learning_rate": 1.5495962849851485e-05,
+      "loss": 0.8953,
       "step": 5384
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 15.998095512390137,
-      "learning_rate": 9.207176505963718e-06,
-      "loss": 1.2675,
+      "epoch": 0.68,
+      "grad_norm": 8.469018936157227,
+      "learning_rate": 1.54951261347948e-05,
+      "loss": 1.9812,
       "step": 5385
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 6.676560401916504,
-      "learning_rate": 9.205171895359326e-06,
-      "loss": 0.8448,
+      "epoch": 0.68,
+      "grad_norm": 8.103954315185547,
+      "learning_rate": 1.549428941973811e-05,
+      "loss": 1.631,
       "step": 5386
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 17.63335609436035,
-      "learning_rate": 9.203167284754936e-06,
-      "loss": 1.771,
+      "epoch": 0.68,
+      "grad_norm": 17.763341903686523,
+      "learning_rate": 1.5493452704681422e-05,
+      "loss": 3.0545,
       "step": 5387
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 15.837037086486816,
-      "learning_rate": 9.201162674150547e-06,
-      "loss": 1.1493,
+      "epoch": 0.68,
+      "grad_norm": 9.169461250305176,
+      "learning_rate": 1.5492615989624736e-05,
+      "loss": 0.6142,
       "step": 5388
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 15.462651252746582,
-      "learning_rate": 9.199158063546157e-06,
-      "loss": 1.4158,
+      "epoch": 0.68,
+      "grad_norm": 12.485591888427734,
+      "learning_rate": 1.5491779274568046e-05,
+      "loss": 3.1047,
       "step": 5389
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 14.927474975585938,
-      "learning_rate": 9.197153452941767e-06,
-      "loss": 2.7956,
+      "epoch": 0.68,
+      "grad_norm": 49.102455139160156,
+      "learning_rate": 1.549094255951136e-05,
+      "loss": 1.9242,
       "step": 5390
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 25.812740325927734,
-      "learning_rate": 9.195148842337377e-06,
-      "loss": 2.2199,
+      "epoch": 0.68,
+      "grad_norm": 9.454066276550293,
+      "learning_rate": 1.5490105844454673e-05,
+      "loss": 1.6279,
       "step": 5391
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 17.734699249267578,
-      "learning_rate": 9.193144231732987e-06,
-      "loss": 2.1131,
+      "epoch": 0.68,
+      "grad_norm": 13.472455024719238,
+      "learning_rate": 1.5489269129397987e-05,
+      "loss": 1.9243,
       "step": 5392
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 15.593416213989258,
-      "learning_rate": 9.191139621128597e-06,
-      "loss": 1.3466,
+      "epoch": 0.68,
+      "grad_norm": 20.5775203704834,
+      "learning_rate": 1.5488432414341297e-05,
+      "loss": 1.5563,
       "step": 5393
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 14.880626678466797,
-      "learning_rate": 9.189135010524205e-06,
-      "loss": 1.4476,
+      "epoch": 0.68,
+      "grad_norm": 12.55504035949707,
+      "learning_rate": 1.548759569928461e-05,
+      "loss": 0.8326,
       "step": 5394
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 17.891508102416992,
-      "learning_rate": 9.187130399919815e-06,
-      "loss": 1.5905,
+      "epoch": 0.68,
+      "grad_norm": 31.129262924194336,
+      "learning_rate": 1.5486758984227925e-05,
+      "loss": 1.5193,
       "step": 5395
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 20.343963623046875,
-      "learning_rate": 9.185125789315427e-06,
-      "loss": 2.0189,
+      "epoch": 0.68,
+      "grad_norm": 14.942036628723145,
+      "learning_rate": 1.5485922269171235e-05,
+      "loss": 2.8091,
       "step": 5396
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 19.60910415649414,
-      "learning_rate": 9.183121178711036e-06,
-      "loss": 2.392,
+      "epoch": 0.68,
+      "grad_norm": 7.471428394317627,
+      "learning_rate": 1.548508555411455e-05,
+      "loss": 0.8795,
       "step": 5397
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 19.910167694091797,
-      "learning_rate": 9.181116568106646e-06,
-      "loss": 1.713,
+      "epoch": 0.68,
+      "grad_norm": 22.331401824951172,
+      "learning_rate": 1.548424883905786e-05,
+      "loss": 1.7552,
       "step": 5398
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 13.35189151763916,
-      "learning_rate": 9.179111957502256e-06,
-      "loss": 1.2514,
+      "epoch": 0.68,
+      "grad_norm": 4.722949028015137,
+      "learning_rate": 1.5483412124001173e-05,
+      "loss": 0.1993,
       "step": 5399
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 27.091657638549805,
-      "learning_rate": 9.177107346897866e-06,
-      "loss": 1.4003,
-      "step": 5400
-    },
-    {
-      "epoch": 1.62,
-      "eval_loss": 0.18608930706977844,
-      "eval_runtime": 43.4962,
-      "eval_samples_per_second": 34.003,
-      "eval_steps_per_second": 34.003,
+      "epoch": 0.68,
+      "grad_norm": 14.18830680847168,
+      "learning_rate": 1.5482575408944486e-05,
+      "loss": 0.393,
       "step": 5400
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 38.99711608886719,
-      "learning_rate": 9.175102736293476e-06,
-      "loss": 2.452,
+      "epoch": 0.68,
+      "grad_norm": 70.21976470947266,
+      "learning_rate": 1.5481738693887796e-05,
+      "loss": 3.0059,
       "step": 5401
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 17.14046859741211,
-      "learning_rate": 9.173098125689086e-06,
-      "loss": 1.0333,
+      "epoch": 0.68,
+      "grad_norm": 46.213504791259766,
+      "learning_rate": 1.548090197883111e-05,
+      "loss": 1.9276,
       "step": 5402
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 14.545239448547363,
-      "learning_rate": 9.171093515084696e-06,
-      "loss": 1.1729,
+      "epoch": 0.68,
+      "grad_norm": 12.647377014160156,
+      "learning_rate": 1.548006526377442e-05,
+      "loss": 2.4336,
       "step": 5403
     },
     {
-      "epoch": 1.62,
-      "grad_norm": 21.479055404663086,
-      "learning_rate": 9.169088904480306e-06,
-      "loss": 2.0552,
+      "epoch": 0.68,
+      "grad_norm": 12.181181907653809,
+      "learning_rate": 1.5479228548717734e-05,
+      "loss": 0.9303,
       "step": 5404
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 52.615394592285156,
-      "learning_rate": 9.167084293875915e-06,
-      "loss": 1.6672,
+      "epoch": 0.68,
+      "grad_norm": 19.40496063232422,
+      "learning_rate": 1.5478391833661048e-05,
+      "loss": 1.0818,
       "step": 5405
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 12.193754196166992,
-      "learning_rate": 9.165079683271525e-06,
-      "loss": 1.3845,
+      "epoch": 0.68,
+      "grad_norm": 15.307136535644531,
+      "learning_rate": 1.547755511860436e-05,
+      "loss": 1.7573,
       "step": 5406
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 10.990750312805176,
-      "learning_rate": 9.163075072667135e-06,
-      "loss": 1.3259,
+      "epoch": 0.68,
+      "grad_norm": 17.155710220336914,
+      "learning_rate": 1.547671840354767e-05,
+      "loss": 2.1602,
       "step": 5407
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 14.663102149963379,
-      "learning_rate": 9.161070462062745e-06,
-      "loss": 1.432,
+      "epoch": 0.68,
+      "grad_norm": 25.457571029663086,
+      "learning_rate": 1.5475881688490985e-05,
+      "loss": 1.6301,
       "step": 5408
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 8.971065521240234,
-      "learning_rate": 9.159065851458355e-06,
-      "loss": 0.9139,
+      "epoch": 0.68,
+      "grad_norm": 24.962495803833008,
+      "learning_rate": 1.54750449734343e-05,
+      "loss": 2.5019,
       "step": 5409
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 12.251411437988281,
-      "learning_rate": 9.157061240853965e-06,
-      "loss": 1.2786,
+      "epoch": 0.68,
+      "grad_norm": 11.651535034179688,
+      "learning_rate": 1.547420825837761e-05,
+      "loss": 0.8768,
       "step": 5410
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 10.386260032653809,
-      "learning_rate": 9.155056630249575e-06,
-      "loss": 1.8492,
+      "epoch": 0.68,
+      "grad_norm": 17.485122680664062,
+      "learning_rate": 1.5473371543320923e-05,
+      "loss": 3.7041,
       "step": 5411
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 25.967205047607422,
-      "learning_rate": 9.153052019645185e-06,
-      "loss": 1.8648,
+      "epoch": 0.68,
+      "grad_norm": 14.24410343170166,
+      "learning_rate": 1.5472534828264236e-05,
+      "loss": 0.8224,
       "step": 5412
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 15.101587295532227,
-      "learning_rate": 9.151047409040794e-06,
-      "loss": 1.3361,
+      "epoch": 0.68,
+      "grad_norm": 21.170881271362305,
+      "learning_rate": 1.547169811320755e-05,
+      "loss": 3.0605,
       "step": 5413
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 21.25626564025879,
-      "learning_rate": 9.149042798436405e-06,
-      "loss": 1.624,
+      "epoch": 0.68,
+      "grad_norm": 128.70204162597656,
+      "learning_rate": 1.547086139815086e-05,
+      "loss": 2.0236,
       "step": 5414
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 12.912954330444336,
-      "learning_rate": 9.147038187832015e-06,
-      "loss": 1.2722,
+      "epoch": 0.68,
+      "grad_norm": 6.5688323974609375,
+      "learning_rate": 1.5470024683094174e-05,
+      "loss": 1.5321,
       "step": 5415
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 19.924049377441406,
-      "learning_rate": 9.145033577227624e-06,
-      "loss": 1.8115,
+      "epoch": 0.68,
+      "grad_norm": 27.743898391723633,
+      "learning_rate": 1.5469187968037488e-05,
+      "loss": 3.4685,
       "step": 5416
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 10.341268539428711,
-      "learning_rate": 9.143028966623234e-06,
-      "loss": 1.1745,
+      "epoch": 0.68,
+      "grad_norm": 11.474124908447266,
+      "learning_rate": 1.5468351252980798e-05,
+      "loss": 1.1201,
       "step": 5417
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 19.03986167907715,
-      "learning_rate": 9.141024356018844e-06,
-      "loss": 1.8332,
+      "epoch": 0.68,
+      "grad_norm": 10.405303001403809,
+      "learning_rate": 1.546751453792411e-05,
+      "loss": 1.8472,
       "step": 5418
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 10.237374305725098,
-      "learning_rate": 9.139019745414454e-06,
-      "loss": 1.1372,
+      "epoch": 0.68,
+      "grad_norm": 5.412685394287109,
+      "learning_rate": 1.5466677822867425e-05,
+      "loss": 0.5036,
       "step": 5419
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 12.311965942382812,
-      "learning_rate": 9.137015134810064e-06,
-      "loss": 1.3524,
+      "epoch": 0.68,
+      "grad_norm": 8.794434547424316,
+      "learning_rate": 1.546584110781074e-05,
+      "loss": 0.9806,
       "step": 5420
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 17.118541717529297,
-      "learning_rate": 9.135010524205674e-06,
-      "loss": 1.4631,
+      "epoch": 0.68,
+      "grad_norm": 8.287175178527832,
+      "learning_rate": 1.546500439275405e-05,
+      "loss": 0.843,
       "step": 5421
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 13.466850280761719,
-      "learning_rate": 9.133005913601284e-06,
-      "loss": 2.0476,
+      "epoch": 0.68,
+      "grad_norm": 17.346330642700195,
+      "learning_rate": 1.5464167677697363e-05,
+      "loss": 1.9287,
       "step": 5422
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 22.777727127075195,
-      "learning_rate": 9.131001302996893e-06,
-      "loss": 1.6859,
+      "epoch": 0.68,
+      "grad_norm": 14.11860466003418,
+      "learning_rate": 1.5463330962640676e-05,
+      "loss": 2.1149,
       "step": 5423
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 24.398303985595703,
-      "learning_rate": 9.128996692392503e-06,
-      "loss": 2.0839,
+      "epoch": 0.68,
+      "grad_norm": 11.32366943359375,
+      "learning_rate": 1.5462494247583987e-05,
+      "loss": 2.592,
       "step": 5424
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 22.156652450561523,
-      "learning_rate": 9.126992081788113e-06,
-      "loss": 1.3133,
+      "epoch": 0.68,
+      "grad_norm": 19.463321685791016,
+      "learning_rate": 1.54616575325273e-05,
+      "loss": 2.1927,
       "step": 5425
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 10.810490608215332,
-      "learning_rate": 9.124987471183725e-06,
-      "loss": 0.9388,
+      "epoch": 0.68,
+      "grad_norm": 11.99179458618164,
+      "learning_rate": 1.546082081747061e-05,
+      "loss": 1.5235,
       "step": 5426
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 17.22280502319336,
-      "learning_rate": 9.122982860579333e-06,
-      "loss": 2.081,
+      "epoch": 0.68,
+      "grad_norm": 10.898519515991211,
+      "learning_rate": 1.5459984102413924e-05,
+      "loss": 1.3556,
       "step": 5427
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 11.593786239624023,
-      "learning_rate": 9.120978249974943e-06,
-      "loss": 1.6318,
+      "epoch": 0.68,
+      "grad_norm": 52.73088455200195,
+      "learning_rate": 1.5459147387357238e-05,
+      "loss": 2.1083,
       "step": 5428
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 111.95779418945312,
-      "learning_rate": 9.118973639370553e-06,
-      "loss": 2.2157,
+      "epoch": 0.68,
+      "grad_norm": 36.21796798706055,
+      "learning_rate": 1.5458310672300548e-05,
+      "loss": 2.343,
       "step": 5429
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 10.570384979248047,
-      "learning_rate": 9.116969028766163e-06,
-      "loss": 1.8122,
+      "epoch": 0.68,
+      "grad_norm": 22.104034423828125,
+      "learning_rate": 1.5457473957243862e-05,
+      "loss": 1.1986,
       "step": 5430
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 47.375701904296875,
-      "learning_rate": 9.114964418161772e-06,
-      "loss": 2.1275,
+      "epoch": 0.68,
+      "grad_norm": 14.936227798461914,
+      "learning_rate": 1.5456637242187172e-05,
+      "loss": 2.8078,
       "step": 5431
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 15.159667015075684,
-      "learning_rate": 9.112959807557382e-06,
-      "loss": 2.2646,
+      "epoch": 0.68,
+      "grad_norm": 21.90635108947754,
+      "learning_rate": 1.5455800527130486e-05,
+      "loss": 2.5358,
       "step": 5432
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 10.478726387023926,
-      "learning_rate": 9.110955196952994e-06,
-      "loss": 1.4377,
+      "epoch": 0.68,
+      "grad_norm": 25.158336639404297,
+      "learning_rate": 1.54549638120738e-05,
+      "loss": 1.835,
       "step": 5433
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 27.897777557373047,
-      "learning_rate": 9.108950586348602e-06,
-      "loss": 2.1688,
+      "epoch": 0.68,
+      "grad_norm": 9.340642929077148,
+      "learning_rate": 1.5454127097017113e-05,
+      "loss": 1.394,
       "step": 5434
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 26.705507278442383,
-      "learning_rate": 9.106945975744212e-06,
-      "loss": 1.7822,
+      "epoch": 0.68,
+      "grad_norm": 13.902207374572754,
+      "learning_rate": 1.5453290381960423e-05,
+      "loss": 0.6865,
       "step": 5435
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 23.994443893432617,
-      "learning_rate": 9.104941365139822e-06,
-      "loss": 2.0833,
+      "epoch": 0.68,
+      "grad_norm": 10.27506160736084,
+      "learning_rate": 1.5452453666903737e-05,
+      "loss": 0.6639,
       "step": 5436
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 66.67716217041016,
-      "learning_rate": 9.102936754535432e-06,
-      "loss": 1.3532,
+      "epoch": 0.68,
+      "grad_norm": 13.183588027954102,
+      "learning_rate": 1.545161695184705e-05,
+      "loss": 1.5091,
       "step": 5437
     },
     {
-      "epoch": 1.63,
-      "grad_norm": 30.51035499572754,
-      "learning_rate": 9.100932143931042e-06,
-      "loss": 2.0045,
+      "epoch": 0.68,
+      "grad_norm": 31.602766036987305,
+      "learning_rate": 1.545078023679036e-05,
+      "loss": 2.0934,
       "step": 5438
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 10.716803550720215,
-      "learning_rate": 9.098927533326652e-06,
-      "loss": 1.6747,
+      "epoch": 0.68,
+      "grad_norm": 3.6912848949432373,
+      "learning_rate": 1.5449943521733674e-05,
+      "loss": 0.0864,
       "step": 5439
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 13.023651123046875,
-      "learning_rate": 9.096922922722262e-06,
-      "loss": 0.7083,
+      "epoch": 0.68,
+      "grad_norm": 22.57712173461914,
+      "learning_rate": 1.5449106806676988e-05,
+      "loss": 1.5905,
       "step": 5440
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 25.49052619934082,
-      "learning_rate": 9.094918312117873e-06,
-      "loss": 1.0234,
+      "epoch": 0.68,
+      "grad_norm": 21.62580108642578,
+      "learning_rate": 1.5448270091620302e-05,
+      "loss": 1.5551,
       "step": 5441
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 33.57606887817383,
-      "learning_rate": 9.092913701513481e-06,
-      "loss": 2.2643,
+      "epoch": 0.68,
+      "grad_norm": 18.682220458984375,
+      "learning_rate": 1.5447433376563612e-05,
+      "loss": 2.4257,
       "step": 5442
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 11.915518760681152,
-      "learning_rate": 9.090909090909091e-06,
-      "loss": 1.0967,
+      "epoch": 0.68,
+      "grad_norm": 15.763568878173828,
+      "learning_rate": 1.5446596661506926e-05,
+      "loss": 0.9516,
       "step": 5443
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 15.158368110656738,
-      "learning_rate": 9.088904480304701e-06,
-      "loss": 1.534,
+      "epoch": 0.68,
+      "grad_norm": 18.485803604125977,
+      "learning_rate": 1.544575994645024e-05,
+      "loss": 3.442,
       "step": 5444
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 11.216906547546387,
-      "learning_rate": 9.086899869700311e-06,
-      "loss": 1.4261,
+      "epoch": 0.68,
+      "grad_norm": 11.463976860046387,
+      "learning_rate": 1.544492323139355e-05,
+      "loss": 1.0116,
       "step": 5445
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 26.098888397216797,
-      "learning_rate": 9.084895259095921e-06,
-      "loss": 2.0426,
+      "epoch": 0.68,
+      "grad_norm": 12.108723640441895,
+      "learning_rate": 1.5444086516336863e-05,
+      "loss": 1.7329,
       "step": 5446
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 13.751211166381836,
-      "learning_rate": 9.082890648491531e-06,
-      "loss": 1.5604,
+      "epoch": 0.68,
+      "grad_norm": 8.408315658569336,
+      "learning_rate": 1.5443249801280177e-05,
+      "loss": 1.3185,
       "step": 5447
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 12.62765884399414,
-      "learning_rate": 9.080886037887141e-06,
-      "loss": 1.7308,
+      "epoch": 0.68,
+      "grad_norm": 8.830940246582031,
+      "learning_rate": 1.544241308622349e-05,
+      "loss": 1.1114,
       "step": 5448
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 20.71200942993164,
-      "learning_rate": 9.078881427282752e-06,
-      "loss": 0.9927,
+      "epoch": 0.68,
+      "grad_norm": 9.236875534057617,
+      "learning_rate": 1.54415763711668e-05,
+      "loss": 0.9462,
       "step": 5449
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 23.00042152404785,
-      "learning_rate": 9.07687681667836e-06,
-      "loss": 2.4304,
+      "epoch": 0.68,
+      "grad_norm": 19.101408004760742,
+      "learning_rate": 1.5440739656110114e-05,
+      "loss": 2.1817,
       "step": 5450
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 28.70808982849121,
-      "learning_rate": 9.074872206073972e-06,
-      "loss": 1.5417,
+      "epoch": 0.68,
+      "grad_norm": 9.254010200500488,
+      "learning_rate": 1.5439902941053425e-05,
+      "loss": 0.8988,
       "step": 5451
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 9.605061531066895,
-      "learning_rate": 9.072867595469582e-06,
-      "loss": 1.3111,
+      "epoch": 0.68,
+      "grad_norm": 7.230218887329102,
+      "learning_rate": 1.5439066225996738e-05,
+      "loss": 0.9233,
       "step": 5452
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 7.0347490310668945,
-      "learning_rate": 9.07086298486519e-06,
-      "loss": 0.7902,
+      "epoch": 0.68,
+      "grad_norm": 15.310093879699707,
+      "learning_rate": 1.5438229510940052e-05,
+      "loss": 1.1652,
       "step": 5453
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 7.641110420227051,
-      "learning_rate": 9.0688583742608e-06,
-      "loss": 0.5109,
+      "epoch": 0.68,
+      "grad_norm": 7.4407782554626465,
+      "learning_rate": 1.5437392795883362e-05,
+      "loss": 0.5707,
       "step": 5454
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 39.103248596191406,
-      "learning_rate": 9.06685376365641e-06,
-      "loss": 1.863,
+      "epoch": 0.68,
+      "grad_norm": 16.615798950195312,
+      "learning_rate": 1.5436556080826676e-05,
+      "loss": 0.9137,
       "step": 5455
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 10.603463172912598,
-      "learning_rate": 9.06484915305202e-06,
-      "loss": 1.5076,
+      "epoch": 0.68,
+      "grad_norm": 45.70905303955078,
+      "learning_rate": 1.5435719365769986e-05,
+      "loss": 1.9834,
       "step": 5456
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 53.90522766113281,
-      "learning_rate": 9.06284454244763e-06,
-      "loss": 3.2397,
+      "epoch": 0.68,
+      "grad_norm": 11.287595748901367,
+      "learning_rate": 1.54348826507133e-05,
+      "loss": 1.7566,
       "step": 5457
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 55.33717727661133,
-      "learning_rate": 9.06083993184324e-06,
-      "loss": 2.0653,
+      "epoch": 0.68,
+      "grad_norm": 29.066984176635742,
+      "learning_rate": 1.5434045935656613e-05,
+      "loss": 3.2238,
       "step": 5458
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 18.05723762512207,
-      "learning_rate": 9.05883532123885e-06,
-      "loss": 1.8223,
+      "epoch": 0.69,
+      "grad_norm": 3.7322309017181396,
+      "learning_rate": 1.5433209220599924e-05,
+      "loss": 0.3103,
       "step": 5459
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 45.47984313964844,
-      "learning_rate": 9.05683071063446e-06,
-      "loss": 1.3857,
+      "epoch": 0.69,
+      "grad_norm": 22.915382385253906,
+      "learning_rate": 1.5432372505543237e-05,
+      "loss": 1.743,
       "step": 5460
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 64.19039916992188,
-      "learning_rate": 9.05482610003007e-06,
-      "loss": 1.9055,
+      "epoch": 0.69,
+      "grad_norm": 26.823711395263672,
+      "learning_rate": 1.543153579048655e-05,
+      "loss": 3.0431,
       "step": 5461
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 16.403478622436523,
-      "learning_rate": 9.05282148942568e-06,
-      "loss": 1.2061,
+      "epoch": 0.69,
+      "grad_norm": 12.175504684448242,
+      "learning_rate": 1.5430699075429865e-05,
+      "loss": 0.8336,
       "step": 5462
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 15.951025009155273,
-      "learning_rate": 9.050816878821291e-06,
-      "loss": 2.1111,
+      "epoch": 0.69,
+      "grad_norm": 11.215303421020508,
+      "learning_rate": 1.5429862360373175e-05,
+      "loss": 1.9734,
       "step": 5463
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 32.44859313964844,
-      "learning_rate": 9.0488122682169e-06,
-      "loss": 1.3111,
+      "epoch": 0.69,
+      "grad_norm": 26.189634323120117,
+      "learning_rate": 1.542902564531649e-05,
+      "loss": 3.3735,
       "step": 5464
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 22.92570686340332,
-      "learning_rate": 9.04680765761251e-06,
-      "loss": 1.4222,
+      "epoch": 0.69,
+      "grad_norm": 38.86238098144531,
+      "learning_rate": 1.5428188930259802e-05,
+      "loss": 1.7573,
       "step": 5465
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 11.310336112976074,
-      "learning_rate": 9.04480304700812e-06,
-      "loss": 1.0347,
+      "epoch": 0.69,
+      "grad_norm": 8.166007041931152,
+      "learning_rate": 1.5427352215203112e-05,
+      "loss": 1.7582,
       "step": 5466
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 48.00858688354492,
-      "learning_rate": 9.04279843640373e-06,
-      "loss": 1.5265,
+      "epoch": 0.69,
+      "grad_norm": 3.437197685241699,
+      "learning_rate": 1.5426515500146426e-05,
+      "loss": 0.0896,
       "step": 5467
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 22.504119873046875,
-      "learning_rate": 9.040793825799338e-06,
-      "loss": 1.2273,
+      "epoch": 0.69,
+      "grad_norm": 6.311569690704346,
+      "learning_rate": 1.542567878508974e-05,
+      "loss": 1.5704,
       "step": 5468
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 23.860111236572266,
-      "learning_rate": 9.03878921519495e-06,
-      "loss": 1.4192,
+      "epoch": 0.69,
+      "grad_norm": 9.943201065063477,
+      "learning_rate": 1.5424842070033053e-05,
+      "loss": 1.1738,
       "step": 5469
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 11.303675651550293,
-      "learning_rate": 9.03678460459056e-06,
-      "loss": 1.0257,
+      "epoch": 0.69,
+      "grad_norm": 5.164342880249023,
+      "learning_rate": 1.5424005354976364e-05,
+      "loss": 0.8256,
       "step": 5470
     },
     {
-      "epoch": 1.64,
-      "grad_norm": 12.571292877197266,
-      "learning_rate": 9.034779993986168e-06,
-      "loss": 1.4753,
+      "epoch": 0.69,
+      "grad_norm": 26.206350326538086,
+      "learning_rate": 1.5423168639919677e-05,
+      "loss": 2.0694,
       "step": 5471
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 45.509620666503906,
-      "learning_rate": 9.032775383381778e-06,
-      "loss": 2.5814,
+      "epoch": 0.69,
+      "grad_norm": 29.577566146850586,
+      "learning_rate": 1.542233192486299e-05,
+      "loss": 1.9165,
       "step": 5472
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 11.296418190002441,
-      "learning_rate": 9.030770772777388e-06,
-      "loss": 0.8345,
+      "epoch": 0.69,
+      "grad_norm": 10.16978931427002,
+      "learning_rate": 1.54214952098063e-05,
+      "loss": 1.3248,
       "step": 5473
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 19.737348556518555,
-      "learning_rate": 9.028766162172999e-06,
-      "loss": 2.0533,
+      "epoch": 0.69,
+      "grad_norm": 16.154563903808594,
+      "learning_rate": 1.5420658494749615e-05,
+      "loss": 0.6503,
       "step": 5474
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 32.28429412841797,
-      "learning_rate": 9.026761551568609e-06,
-      "loss": 1.0527,
+      "epoch": 0.69,
+      "grad_norm": 18.64064598083496,
+      "learning_rate": 1.541982177969293e-05,
+      "loss": 2.6751,
       "step": 5475
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 5.6402106285095215,
-      "learning_rate": 9.024756940964219e-06,
-      "loss": 0.7558,
+      "epoch": 0.69,
+      "grad_norm": 17.2528133392334,
+      "learning_rate": 1.541898506463624e-05,
+      "loss": 1.8936,
       "step": 5476
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 11.547632217407227,
-      "learning_rate": 9.022752330359829e-06,
-      "loss": 1.089,
+      "epoch": 0.69,
+      "grad_norm": 17.381118774414062,
+      "learning_rate": 1.5418148349579552e-05,
+      "loss": 1.9031,
       "step": 5477
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 7.933753967285156,
-      "learning_rate": 9.020747719755439e-06,
-      "loss": 1.1976,
+      "epoch": 0.69,
+      "grad_norm": 31.099803924560547,
+      "learning_rate": 1.5417311634522866e-05,
+      "loss": 2.3239,
       "step": 5478
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 25.738367080688477,
-      "learning_rate": 9.018743109151047e-06,
-      "loss": 2.3856,
+      "epoch": 0.69,
+      "grad_norm": 11.734057426452637,
+      "learning_rate": 1.5416474919466176e-05,
+      "loss": 0.5819,
       "step": 5479
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 14.972805976867676,
-      "learning_rate": 9.016738498546657e-06,
-      "loss": 2.7914,
+      "epoch": 0.69,
+      "grad_norm": 16.285499572753906,
+      "learning_rate": 1.541563820440949e-05,
+      "loss": 2.2578,
       "step": 5480
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 23.992773056030273,
-      "learning_rate": 9.014733887942267e-06,
-      "loss": 2.568,
+      "epoch": 0.69,
+      "grad_norm": 6.641711711883545,
+      "learning_rate": 1.5414801489352804e-05,
+      "loss": 1.3813,
       "step": 5481
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 12.318902015686035,
-      "learning_rate": 9.012729277337878e-06,
-      "loss": 1.2272,
+      "epoch": 0.69,
+      "grad_norm": 4.232131481170654,
+      "learning_rate": 1.5413964774296114e-05,
+      "loss": 0.6922,
       "step": 5482
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 13.769071578979492,
-      "learning_rate": 9.010724666733488e-06,
-      "loss": 1.4336,
+      "epoch": 0.69,
+      "grad_norm": 14.072823524475098,
+      "learning_rate": 1.5413128059239428e-05,
+      "loss": 1.7136,
       "step": 5483
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 29.345212936401367,
-      "learning_rate": 9.008720056129098e-06,
-      "loss": 1.8268,
+      "epoch": 0.69,
+      "grad_norm": 18.54960823059082,
+      "learning_rate": 1.5412291344182738e-05,
+      "loss": 2.7273,
       "step": 5484
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 10.752758026123047,
-      "learning_rate": 9.006715445524708e-06,
-      "loss": 0.75,
+      "epoch": 0.69,
+      "grad_norm": 12.27863597869873,
+      "learning_rate": 1.541145462912605e-05,
+      "loss": 1.326,
       "step": 5485
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 11.773067474365234,
-      "learning_rate": 9.004710834920318e-06,
-      "loss": 1.5268,
+      "epoch": 0.69,
+      "grad_norm": 19.024850845336914,
+      "learning_rate": 1.5410617914069365e-05,
+      "loss": 1.7985,
       "step": 5486
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 20.230098724365234,
-      "learning_rate": 9.002706224315926e-06,
-      "loss": 1.2181,
+      "epoch": 0.69,
+      "grad_norm": 8.920555114746094,
+      "learning_rate": 1.5409781199012675e-05,
+      "loss": 0.5712,
       "step": 5487
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 19.085262298583984,
-      "learning_rate": 9.000701613711538e-06,
-      "loss": 2.0583,
+      "epoch": 0.69,
+      "grad_norm": 12.335217475891113,
+      "learning_rate": 1.540894448395599e-05,
+      "loss": 1.1266,
       "step": 5488
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 15.87958812713623,
-      "learning_rate": 8.998697003107148e-06,
-      "loss": 1.2279,
+      "epoch": 0.69,
+      "grad_norm": 6.6275739669799805,
+      "learning_rate": 1.5408107768899303e-05,
+      "loss": 0.9102,
       "step": 5489
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 10.969903945922852,
-      "learning_rate": 8.996692392502757e-06,
-      "loss": 1.4018,
+      "epoch": 0.69,
+      "grad_norm": 9.099052429199219,
+      "learning_rate": 1.5407271053842616e-05,
+      "loss": 1.3513,
       "step": 5490
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 16.4309024810791,
-      "learning_rate": 8.994687781898367e-06,
-      "loss": 1.7592,
+      "epoch": 0.69,
+      "grad_norm": 11.68352222442627,
+      "learning_rate": 1.5406434338785927e-05,
+      "loss": 2.1072,
       "step": 5491
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 17.328758239746094,
-      "learning_rate": 8.992683171293977e-06,
-      "loss": 1.3829,
+      "epoch": 0.69,
+      "grad_norm": 6.421557903289795,
+      "learning_rate": 1.540559762372924e-05,
+      "loss": 0.6181,
       "step": 5492
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 23.75958824157715,
-      "learning_rate": 8.990678560689587e-06,
-      "loss": 1.7726,
+      "epoch": 0.69,
+      "grad_norm": 13.317495346069336,
+      "learning_rate": 1.5404760908672554e-05,
+      "loss": 1.9661,
       "step": 5493
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 7.305182933807373,
-      "learning_rate": 8.988673950085197e-06,
-      "loss": 0.9196,
+      "epoch": 0.69,
+      "grad_norm": 5.760676383972168,
+      "learning_rate": 1.5403924193615864e-05,
+      "loss": 0.1942,
       "step": 5494
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 23.182376861572266,
-      "learning_rate": 8.986669339480807e-06,
-      "loss": 2.1978,
+      "epoch": 0.69,
+      "grad_norm": 21.07034683227539,
+      "learning_rate": 1.5403087478559178e-05,
+      "loss": 2.1126,
       "step": 5495
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 16.009490966796875,
-      "learning_rate": 8.984664728876417e-06,
-      "loss": 1.3805,
+      "epoch": 0.69,
+      "grad_norm": 15.3811674118042,
+      "learning_rate": 1.540225076350249e-05,
+      "loss": 1.292,
       "step": 5496
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 25.598642349243164,
-      "learning_rate": 8.982660118272027e-06,
-      "loss": 1.7893,
+      "epoch": 0.69,
+      "grad_norm": 13.275308609008789,
+      "learning_rate": 1.5401414048445805e-05,
+      "loss": 1.9513,
       "step": 5497
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 14.513004302978516,
-      "learning_rate": 8.980655507667636e-06,
-      "loss": 1.0957,
+      "epoch": 0.69,
+      "grad_norm": 14.056856155395508,
+      "learning_rate": 1.5400577333389115e-05,
+      "loss": 1.4046,
       "step": 5498
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 33.75980758666992,
-      "learning_rate": 8.978650897063246e-06,
-      "loss": 1.9849,
+      "epoch": 0.69,
+      "grad_norm": 4.988726615905762,
+      "learning_rate": 1.539974061833243e-05,
+      "loss": 0.4338,
       "step": 5499
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 17.729082107543945,
-      "learning_rate": 8.976646286458857e-06,
-      "loss": 1.2108,
+      "epoch": 0.69,
+      "grad_norm": 36.79734420776367,
+      "learning_rate": 1.5398903903275743e-05,
+      "loss": 2.014,
       "step": 5500
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 15.00300121307373,
-      "learning_rate": 8.974641675854466e-06,
-      "loss": 1.9234,
+      "epoch": 0.69,
+      "grad_norm": 6.316336631774902,
+      "learning_rate": 1.5398067188219053e-05,
+      "loss": 0.7961,
       "step": 5501
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 37.787193298339844,
-      "learning_rate": 8.972637065250076e-06,
-      "loss": 1.5762,
+      "epoch": 0.69,
+      "grad_norm": 11.95064926147461,
+      "learning_rate": 1.5397230473162367e-05,
+      "loss": 2.1392,
       "step": 5502
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 18.156917572021484,
-      "learning_rate": 8.970632454645686e-06,
-      "loss": 1.2841,
+      "epoch": 0.69,
+      "grad_norm": 8.447251319885254,
+      "learning_rate": 1.539639375810568e-05,
+      "loss": 1.3479,
       "step": 5503
     },
     {
-      "epoch": 1.65,
-      "grad_norm": 17.250186920166016,
-      "learning_rate": 8.968627844041296e-06,
-      "loss": 1.4742,
+      "epoch": 0.69,
+      "grad_norm": 16.203733444213867,
+      "learning_rate": 1.539555704304899e-05,
+      "loss": 2.0525,
       "step": 5504
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 27.10199737548828,
-      "learning_rate": 8.966623233436904e-06,
-      "loss": 1.9449,
+      "epoch": 0.69,
+      "grad_norm": 28.032142639160156,
+      "learning_rate": 1.5394720327992304e-05,
+      "loss": 3.0286,
       "step": 5505
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 17.242645263671875,
-      "learning_rate": 8.964618622832516e-06,
-      "loss": 1.3392,
+      "epoch": 0.69,
+      "grad_norm": 10.23719310760498,
+      "learning_rate": 1.5393883612935618e-05,
+      "loss": 0.5594,
       "step": 5506
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 12.527267456054688,
-      "learning_rate": 8.962614012228126e-06,
-      "loss": 1.559,
+      "epoch": 0.69,
+      "grad_norm": 20.76209831237793,
+      "learning_rate": 1.5393046897878928e-05,
+      "loss": 2.1147,
       "step": 5507
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 12.160088539123535,
-      "learning_rate": 8.960609401623735e-06,
-      "loss": 1.2504,
+      "epoch": 0.69,
+      "grad_norm": 11.435663223266602,
+      "learning_rate": 1.539221018282224e-05,
+      "loss": 1.9703,
       "step": 5508
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 9.42102336883545,
-      "learning_rate": 8.958604791019345e-06,
-      "loss": 0.9488,
+      "epoch": 0.69,
+      "grad_norm": 16.902219772338867,
+      "learning_rate": 1.5391373467765552e-05,
+      "loss": 1.7572,
       "step": 5509
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 16.88702964782715,
-      "learning_rate": 8.956600180414955e-06,
-      "loss": 1.5164,
+      "epoch": 0.69,
+      "grad_norm": 36.9378662109375,
+      "learning_rate": 1.5390536752708866e-05,
+      "loss": 2.7558,
       "step": 5510
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 8.981550216674805,
-      "learning_rate": 8.954595569810565e-06,
-      "loss": 1.2683,
+      "epoch": 0.69,
+      "grad_norm": 34.30980682373047,
+      "learning_rate": 1.538970003765218e-05,
+      "loss": 1.6774,
       "step": 5511
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 14.901409149169922,
-      "learning_rate": 8.952590959206175e-06,
-      "loss": 1.3492,
+      "epoch": 0.69,
+      "grad_norm": 11.313263893127441,
+      "learning_rate": 1.538886332259549e-05,
+      "loss": 1.1251,
       "step": 5512
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 38.2581787109375,
-      "learning_rate": 8.950586348601785e-06,
-      "loss": 1.4274,
+      "epoch": 0.69,
+      "grad_norm": 80.6723861694336,
+      "learning_rate": 1.5388026607538803e-05,
+      "loss": 2.4211,
       "step": 5513
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 11.266697883605957,
-      "learning_rate": 8.948581737997395e-06,
-      "loss": 1.2673,
+      "epoch": 0.69,
+      "grad_norm": 20.437837600708008,
+      "learning_rate": 1.5387189892482117e-05,
+      "loss": 2.8406,
       "step": 5514
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 12.022287368774414,
-      "learning_rate": 8.946577127393005e-06,
-      "loss": 1.7178,
+      "epoch": 0.69,
+      "grad_norm": 7.128717422485352,
+      "learning_rate": 1.5386353177425427e-05,
+      "loss": 1.1533,
       "step": 5515
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 18.94673728942871,
-      "learning_rate": 8.944572516788614e-06,
-      "loss": 1.3636,
+      "epoch": 0.69,
+      "grad_norm": 12.50041675567627,
+      "learning_rate": 1.538551646236874e-05,
+      "loss": 2.2954,
       "step": 5516
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 30.44729232788086,
-      "learning_rate": 8.942567906184224e-06,
-      "loss": 1.623,
+      "epoch": 0.69,
+      "grad_norm": 23.601917266845703,
+      "learning_rate": 1.5384679747312054e-05,
+      "loss": 3.0042,
       "step": 5517
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 13.971253395080566,
-      "learning_rate": 8.940563295579834e-06,
-      "loss": 1.3185,
+      "epoch": 0.69,
+      "grad_norm": 15.071398735046387,
+      "learning_rate": 1.5383843032255368e-05,
+      "loss": 0.8468,
       "step": 5518
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 10.269094467163086,
-      "learning_rate": 8.938558684975444e-06,
-      "loss": 1.2367,
+      "epoch": 0.69,
+      "grad_norm": 15.801284790039062,
+      "learning_rate": 1.5383006317198678e-05,
+      "loss": 2.6363,
       "step": 5519
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 19.084022521972656,
-      "learning_rate": 8.936554074371054e-06,
-      "loss": 0.9847,
-      "step": 5520
-    },
-    {
-      "epoch": 1.66,
-      "eval_loss": 0.1914350539445877,
-      "eval_runtime": 43.4865,
-      "eval_samples_per_second": 34.011,
-      "eval_steps_per_second": 34.011,
+      "epoch": 0.69,
+      "grad_norm": 9.802026748657227,
+      "learning_rate": 1.5382169602141992e-05,
+      "loss": 0.9632,
       "step": 5520
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 11.31253719329834,
-      "learning_rate": 8.934549463766664e-06,
-      "loss": 1.363,
+      "epoch": 0.69,
+      "grad_norm": 7.131788730621338,
+      "learning_rate": 1.5381332887085306e-05,
+      "loss": 0.6508,
       "step": 5521
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 22.407804489135742,
-      "learning_rate": 8.932544853162274e-06,
-      "loss": 1.953,
+      "epoch": 0.69,
+      "grad_norm": 15.925463676452637,
+      "learning_rate": 1.5380496172028616e-05,
+      "loss": 1.7814,
       "step": 5522
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 36.859561920166016,
-      "learning_rate": 8.930540242557884e-06,
-      "loss": 2.4004,
+      "epoch": 0.69,
+      "grad_norm": 5.8094000816345215,
+      "learning_rate": 1.537965945697193e-05,
+      "loss": 0.3943,
       "step": 5523
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 28.108509063720703,
-      "learning_rate": 8.928535631953493e-06,
-      "loss": 1.4061,
+      "epoch": 0.69,
+      "grad_norm": 20.999128341674805,
+      "learning_rate": 1.5378822741915243e-05,
+      "loss": 1.2808,
       "step": 5524
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 20.431781768798828,
-      "learning_rate": 8.926531021349104e-06,
-      "loss": 1.4469,
+      "epoch": 0.69,
+      "grad_norm": 10.400137901306152,
+      "learning_rate": 1.5377986026858557e-05,
+      "loss": 1.3163,
       "step": 5525
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 20.850435256958008,
-      "learning_rate": 8.924526410744714e-06,
-      "loss": 1.789,
+      "epoch": 0.69,
+      "grad_norm": 30.212005615234375,
+      "learning_rate": 1.5377149311801867e-05,
+      "loss": 1.9231,
       "step": 5526
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 11.911688804626465,
-      "learning_rate": 8.922521800140323e-06,
-      "loss": 2.0367,
+      "epoch": 0.69,
+      "grad_norm": 9.523595809936523,
+      "learning_rate": 1.537631259674518e-05,
+      "loss": 2.1191,
       "step": 5527
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 10.329282760620117,
-      "learning_rate": 8.920517189535933e-06,
-      "loss": 1.4678,
+      "epoch": 0.69,
+      "grad_norm": 18.021493911743164,
+      "learning_rate": 1.5375475881688494e-05,
+      "loss": 3.7398,
       "step": 5528
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 12.712068557739258,
-      "learning_rate": 8.918512578931543e-06,
-      "loss": 1.6763,
+      "epoch": 0.69,
+      "grad_norm": 21.648252487182617,
+      "learning_rate": 1.5374639166631805e-05,
+      "loss": 1.3649,
       "step": 5529
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 58.54856872558594,
-      "learning_rate": 8.916507968327153e-06,
-      "loss": 2.0221,
+      "epoch": 0.69,
+      "grad_norm": 13.179224014282227,
+      "learning_rate": 1.5373802451575118e-05,
+      "loss": 0.5187,
       "step": 5530
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 13.266801834106445,
-      "learning_rate": 8.914503357722763e-06,
-      "loss": 1.7743,
+      "epoch": 0.69,
+      "grad_norm": 5.153297424316406,
+      "learning_rate": 1.5372965736518432e-05,
+      "loss": 1.0926,
       "step": 5531
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 27.546993255615234,
-      "learning_rate": 8.912498747118373e-06,
-      "loss": 2.6575,
+      "epoch": 0.69,
+      "grad_norm": 15.431589126586914,
+      "learning_rate": 1.5372129021461742e-05,
+      "loss": 3.4108,
       "step": 5532
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 11.523126602172852,
-      "learning_rate": 8.910494136513983e-06,
-      "loss": 1.1144,
+      "epoch": 0.69,
+      "grad_norm": 25.87183952331543,
+      "learning_rate": 1.5371292306405056e-05,
+      "loss": 1.4335,
       "step": 5533
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 11.785201072692871,
-      "learning_rate": 8.908489525909593e-06,
-      "loss": 1.1003,
+      "epoch": 0.69,
+      "grad_norm": 6.297723770141602,
+      "learning_rate": 1.537045559134837e-05,
+      "loss": 0.313,
       "step": 5534
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 14.82313346862793,
-      "learning_rate": 8.906484915305202e-06,
-      "loss": 1.8222,
+      "epoch": 0.69,
+      "grad_norm": 24.791057586669922,
+      "learning_rate": 1.536961887629168e-05,
+      "loss": 2.463,
       "step": 5535
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 14.647685050964355,
-      "learning_rate": 8.904480304700812e-06,
-      "loss": 0.8845,
+      "epoch": 0.69,
+      "grad_norm": 13.483685493469238,
+      "learning_rate": 1.5368782161234993e-05,
+      "loss": 1.0357,
       "step": 5536
     },
     {
-      "epoch": 1.66,
-      "grad_norm": 8.299193382263184,
-      "learning_rate": 8.902475694096424e-06,
-      "loss": 1.3426,
+      "epoch": 0.69,
+      "grad_norm": 15.028130531311035,
+      "learning_rate": 1.5367945446178304e-05,
+      "loss": 1.5467,
       "step": 5537
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 31.66522979736328,
-      "learning_rate": 8.900471083492032e-06,
-      "loss": 1.5596,
+      "epoch": 0.7,
+      "grad_norm": 7.930243492126465,
+      "learning_rate": 1.5367108731121617e-05,
+      "loss": 0.9488,
       "step": 5538
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 10.633186340332031,
-      "learning_rate": 8.898466472887642e-06,
-      "loss": 0.8703,
+      "epoch": 0.7,
+      "grad_norm": 29.752031326293945,
+      "learning_rate": 1.536627201606493e-05,
+      "loss": 1.6596,
       "step": 5539
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 12.473995208740234,
-      "learning_rate": 8.896461862283252e-06,
-      "loss": 1.1009,
+      "epoch": 0.7,
+      "grad_norm": 14.493793487548828,
+      "learning_rate": 1.536543530100824e-05,
+      "loss": 2.9665,
       "step": 5540
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 16.58210563659668,
-      "learning_rate": 8.894457251678862e-06,
-      "loss": 0.7829,
+      "epoch": 0.7,
+      "grad_norm": 12.4276762008667,
+      "learning_rate": 1.5364598585951555e-05,
+      "loss": 0.8027,
       "step": 5541
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 9.069561004638672,
-      "learning_rate": 8.89245264107447e-06,
-      "loss": 0.5217,
+      "epoch": 0.7,
+      "grad_norm": 39.77607345581055,
+      "learning_rate": 1.536376187089487e-05,
+      "loss": 2.4362,
       "step": 5542
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 14.045775413513184,
-      "learning_rate": 8.890448030470083e-06,
-      "loss": 1.2927,
+      "epoch": 0.7,
+      "grad_norm": 13.369621276855469,
+      "learning_rate": 1.536292515583818e-05,
+      "loss": 2.7699,
       "step": 5543
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 10.430898666381836,
-      "learning_rate": 8.888443419865693e-06,
-      "loss": 1.0182,
+      "epoch": 0.7,
+      "grad_norm": 7.263457298278809,
+      "learning_rate": 1.5362088440781492e-05,
+      "loss": 0.5554,
       "step": 5544
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 28.892423629760742,
-      "learning_rate": 8.886438809261303e-06,
-      "loss": 2.1135,
+      "epoch": 0.7,
+      "grad_norm": 6.9838337898254395,
+      "learning_rate": 1.5361251725724806e-05,
+      "loss": 0.9054,
       "step": 5545
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 33.67451095581055,
-      "learning_rate": 8.884434198656911e-06,
-      "loss": 2.2505,
+      "epoch": 0.7,
+      "grad_norm": 20.616817474365234,
+      "learning_rate": 1.5360415010668116e-05,
+      "loss": 2.8528,
       "step": 5546
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 22.55853271484375,
-      "learning_rate": 8.882429588052521e-06,
-      "loss": 1.275,
+      "epoch": 0.7,
+      "grad_norm": 4.747254371643066,
+      "learning_rate": 1.535957829561143e-05,
+      "loss": 1.3521,
       "step": 5547
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 11.746838569641113,
-      "learning_rate": 8.880424977448131e-06,
-      "loss": 1.0598,
+      "epoch": 0.7,
+      "grad_norm": 8.651074409484863,
+      "learning_rate": 1.5358741580554744e-05,
+      "loss": 1.9477,
       "step": 5548
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 18.632312774658203,
-      "learning_rate": 8.878420366843741e-06,
-      "loss": 2.3075,
+      "epoch": 0.7,
+      "grad_norm": 10.826473236083984,
+      "learning_rate": 1.5357904865498057e-05,
+      "loss": 0.6565,
       "step": 5549
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 23.912031173706055,
-      "learning_rate": 8.876415756239351e-06,
-      "loss": 1.5192,
+      "epoch": 0.7,
+      "grad_norm": 16.596824645996094,
+      "learning_rate": 1.5357068150441367e-05,
+      "loss": 1.9037,
       "step": 5550
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 19.950305938720703,
-      "learning_rate": 8.874411145634962e-06,
-      "loss": 1.5378,
+      "epoch": 0.7,
+      "grad_norm": 20.1361141204834,
+      "learning_rate": 1.535623143538468e-05,
+      "loss": 1.6692,
       "step": 5551
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 15.393779754638672,
-      "learning_rate": 8.872406535030572e-06,
-      "loss": 1.4361,
+      "epoch": 0.7,
+      "grad_norm": 9.58398151397705,
+      "learning_rate": 1.5355394720327995e-05,
+      "loss": 1.0216,
       "step": 5552
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 11.668610572814941,
-      "learning_rate": 8.87040192442618e-06,
-      "loss": 1.1552,
+      "epoch": 0.7,
+      "grad_norm": 21.33419418334961,
+      "learning_rate": 1.5354558005271305e-05,
+      "loss": 1.6206,
       "step": 5553
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 11.19984245300293,
-      "learning_rate": 8.86839731382179e-06,
-      "loss": 1.3071,
+      "epoch": 0.7,
+      "grad_norm": 15.278912544250488,
+      "learning_rate": 1.535372129021462e-05,
+      "loss": 2.2082,
       "step": 5554
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 12.745835304260254,
-      "learning_rate": 8.8663927032174e-06,
-      "loss": 0.7712,
+      "epoch": 0.7,
+      "grad_norm": 8.61217212677002,
+      "learning_rate": 1.5352884575157932e-05,
+      "loss": 1.9911,
       "step": 5555
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 32.49069595336914,
-      "learning_rate": 8.86438809261301e-06,
-      "loss": 1.9268,
+      "epoch": 0.7,
+      "grad_norm": 17.128591537475586,
+      "learning_rate": 1.5352047860101246e-05,
+      "loss": 1.4462,
       "step": 5556
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 8.485462188720703,
-      "learning_rate": 8.86238348200862e-06,
-      "loss": 1.0959,
+      "epoch": 0.7,
+      "grad_norm": 6.570550441741943,
+      "learning_rate": 1.5351211145044556e-05,
+      "loss": 0.5636,
       "step": 5557
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 14.367705345153809,
-      "learning_rate": 8.86037887140423e-06,
-      "loss": 1.0967,
+      "epoch": 0.7,
+      "grad_norm": 14.50590991973877,
+      "learning_rate": 1.535037442998787e-05,
+      "loss": 3.6256,
       "step": 5558
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 22.95987892150879,
-      "learning_rate": 8.85837426079984e-06,
-      "loss": 1.6232,
+      "epoch": 0.7,
+      "grad_norm": 13.506613731384277,
+      "learning_rate": 1.5349537714931184e-05,
+      "loss": 1.0145,
       "step": 5559
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 15.968416213989258,
-      "learning_rate": 8.85636965019545e-06,
-      "loss": 1.2239,
+      "epoch": 0.7,
+      "grad_norm": 7.986975193023682,
+      "learning_rate": 1.5348700999874494e-05,
+      "loss": 1.4675,
       "step": 5560
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 10.333566665649414,
-      "learning_rate": 8.854365039591059e-06,
-      "loss": 1.248,
+      "epoch": 0.7,
+      "grad_norm": 10.020309448242188,
+      "learning_rate": 1.5347864284817807e-05,
+      "loss": 1.5019,
       "step": 5561
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 45.59640884399414,
-      "learning_rate": 8.85236042898667e-06,
-      "loss": 1.6009,
+      "epoch": 0.7,
+      "grad_norm": 21.0300350189209,
+      "learning_rate": 1.5347027569761118e-05,
+      "loss": 0.9188,
       "step": 5562
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 28.932910919189453,
-      "learning_rate": 8.85035581838228e-06,
-      "loss": 1.88,
+      "epoch": 0.7,
+      "grad_norm": 8.697385787963867,
+      "learning_rate": 1.534619085470443e-05,
+      "loss": 0.604,
       "step": 5563
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 12.517439842224121,
-      "learning_rate": 8.84835120777789e-06,
-      "loss": 1.8985,
+      "epoch": 0.7,
+      "grad_norm": 9.769986152648926,
+      "learning_rate": 1.5345354139647745e-05,
+      "loss": 1.6472,
       "step": 5564
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 17.193233489990234,
-      "learning_rate": 8.8463465971735e-06,
-      "loss": 1.1834,
+      "epoch": 0.7,
+      "grad_norm": 13.042610168457031,
+      "learning_rate": 1.5344517424591055e-05,
+      "loss": 1.7165,
       "step": 5565
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 26.31612205505371,
-      "learning_rate": 8.84434198656911e-06,
-      "loss": 2.4826,
+      "epoch": 0.7,
+      "grad_norm": 38.07878112792969,
+      "learning_rate": 1.534368070953437e-05,
+      "loss": 3.2106,
       "step": 5566
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 16.692577362060547,
-      "learning_rate": 8.84233737596472e-06,
-      "loss": 1.3551,
+      "epoch": 0.7,
+      "grad_norm": 45.39943313598633,
+      "learning_rate": 1.534284399447768e-05,
+      "loss": 4.0076,
       "step": 5567
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 19.141315460205078,
-      "learning_rate": 8.84033276536033e-06,
-      "loss": 1.5564,
+      "epoch": 0.7,
+      "grad_norm": 6.458475112915039,
+      "learning_rate": 1.5342007279420993e-05,
+      "loss": 1.7158,
       "step": 5568
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 12.95341682434082,
-      "learning_rate": 8.83832815475594e-06,
-      "loss": 1.3491,
+      "epoch": 0.7,
+      "grad_norm": 19.190696716308594,
+      "learning_rate": 1.5341170564364306e-05,
+      "loss": 0.9015,
       "step": 5569
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 11.91671085357666,
-      "learning_rate": 8.83632354415155e-06,
-      "loss": 1.4844,
+      "epoch": 0.7,
+      "grad_norm": 101.54644012451172,
+      "learning_rate": 1.534033384930762e-05,
+      "loss": 1.944,
       "step": 5570
     },
     {
-      "epoch": 1.67,
-      "grad_norm": 17.263015747070312,
-      "learning_rate": 8.83431893354716e-06,
-      "loss": 1.5285,
+      "epoch": 0.7,
+      "grad_norm": 23.400814056396484,
+      "learning_rate": 1.533949713425093e-05,
+      "loss": 1.376,
       "step": 5571
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 10.586352348327637,
-      "learning_rate": 8.832314322942768e-06,
-      "loss": 1.4477,
+      "epoch": 0.7,
+      "grad_norm": 16.705427169799805,
+      "learning_rate": 1.5338660419194244e-05,
+      "loss": 2.3008,
       "step": 5572
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 17.991832733154297,
-      "learning_rate": 8.830309712338378e-06,
-      "loss": 1.3879,
+      "epoch": 0.7,
+      "grad_norm": 16.300222396850586,
+      "learning_rate": 1.5337823704137558e-05,
+      "loss": 2.3399,
       "step": 5573
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 19.859966278076172,
-      "learning_rate": 8.82830510173399e-06,
-      "loss": 1.8104,
+      "epoch": 0.7,
+      "grad_norm": 38.15534973144531,
+      "learning_rate": 1.5336986989080868e-05,
+      "loss": 1.3535,
       "step": 5574
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 436.9852294921875,
-      "learning_rate": 8.826300491129598e-06,
-      "loss": 2.6175,
+      "epoch": 0.7,
+      "grad_norm": 26.14240264892578,
+      "learning_rate": 1.533615027402418e-05,
+      "loss": 2.1262,
       "step": 5575
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 12.303122520446777,
-      "learning_rate": 8.824295880525209e-06,
-      "loss": 0.9817,
+      "epoch": 0.7,
+      "grad_norm": 10.343050956726074,
+      "learning_rate": 1.5335313558967495e-05,
+      "loss": 0.8729,
       "step": 5576
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 13.081160545349121,
-      "learning_rate": 8.822291269920819e-06,
-      "loss": 2.0876,
+      "epoch": 0.7,
+      "grad_norm": 13.817024230957031,
+      "learning_rate": 1.533447684391081e-05,
+      "loss": 1.0332,
       "step": 5577
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 9.616783142089844,
-      "learning_rate": 8.820286659316429e-06,
-      "loss": 1.7439,
+      "epoch": 0.7,
+      "grad_norm": 22.90940284729004,
+      "learning_rate": 1.533364012885412e-05,
+      "loss": 0.6995,
       "step": 5578
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 15.427207946777344,
-      "learning_rate": 8.818282048712039e-06,
-      "loss": 1.205,
+      "epoch": 0.7,
+      "grad_norm": 9.916383743286133,
+      "learning_rate": 1.5332803413797433e-05,
+      "loss": 1.0828,
       "step": 5579
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 17.407861709594727,
-      "learning_rate": 8.816277438107649e-06,
-      "loss": 1.9737,
+      "epoch": 0.7,
+      "grad_norm": 16.110549926757812,
+      "learning_rate": 1.5331966698740746e-05,
+      "loss": 1.7157,
       "step": 5580
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 30.608501434326172,
-      "learning_rate": 8.814272827503259e-06,
-      "loss": 1.4772,
+      "epoch": 0.7,
+      "grad_norm": 10.246378898620605,
+      "learning_rate": 1.5331129983684057e-05,
+      "loss": 2.2646,
       "step": 5581
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 17.050554275512695,
-      "learning_rate": 8.812268216898869e-06,
-      "loss": 1.2003,
+      "epoch": 0.7,
+      "grad_norm": 88.1216049194336,
+      "learning_rate": 1.533029326862737e-05,
+      "loss": 1.5729,
       "step": 5582
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 21.977340698242188,
-      "learning_rate": 8.810263606294477e-06,
-      "loss": 2.4612,
+      "epoch": 0.7,
+      "grad_norm": 9.350086212158203,
+      "learning_rate": 1.5329456553570684e-05,
+      "loss": 1.5943,
       "step": 5583
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 48.15430450439453,
-      "learning_rate": 8.808258995690088e-06,
-      "loss": 1.7486,
+      "epoch": 0.7,
+      "grad_norm": 8.791582107543945,
+      "learning_rate": 1.5328619838513998e-05,
+      "loss": 1.1793,
       "step": 5584
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 15.752915382385254,
-      "learning_rate": 8.806254385085698e-06,
-      "loss": 1.6484,
+      "epoch": 0.7,
+      "grad_norm": 20.162771224975586,
+      "learning_rate": 1.5327783123457308e-05,
+      "loss": 1.5224,
       "step": 5585
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 37.53594207763672,
-      "learning_rate": 8.804249774481308e-06,
-      "loss": 2.1715,
+      "epoch": 0.7,
+      "grad_norm": 25.328039169311523,
+      "learning_rate": 1.532694640840062e-05,
+      "loss": 2.0069,
       "step": 5586
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 10.964682579040527,
-      "learning_rate": 8.802245163876918e-06,
-      "loss": 1.4082,
+      "epoch": 0.7,
+      "grad_norm": 128.4274444580078,
+      "learning_rate": 1.5326109693343935e-05,
+      "loss": 1.6119,
       "step": 5587
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 23.61287498474121,
-      "learning_rate": 8.800240553272528e-06,
-      "loss": 0.7415,
+      "epoch": 0.7,
+      "grad_norm": 19.298948287963867,
+      "learning_rate": 1.5325272978287245e-05,
+      "loss": 0.7291,
       "step": 5588
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 15.340435981750488,
-      "learning_rate": 8.798235942668138e-06,
-      "loss": 1.5422,
+      "epoch": 0.7,
+      "grad_norm": 8.975295066833496,
+      "learning_rate": 1.532443626323056e-05,
+      "loss": 0.8676,
       "step": 5589
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 12.990081787109375,
-      "learning_rate": 8.796231332063746e-06,
-      "loss": 1.4462,
+      "epoch": 0.7,
+      "grad_norm": 42.80747985839844,
+      "learning_rate": 1.532359954817387e-05,
+      "loss": 1.6301,
       "step": 5590
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 10.152392387390137,
-      "learning_rate": 8.794226721459356e-06,
-      "loss": 0.9929,
+      "epoch": 0.7,
+      "grad_norm": 11.307854652404785,
+      "learning_rate": 1.5322762833117183e-05,
+      "loss": 0.9856,
       "step": 5591
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 13.210673332214355,
-      "learning_rate": 8.792222110854968e-06,
-      "loss": 1.7244,
+      "epoch": 0.7,
+      "grad_norm": 13.756830215454102,
+      "learning_rate": 1.5321926118060497e-05,
+      "loss": 1.5287,
       "step": 5592
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 13.106926918029785,
-      "learning_rate": 8.790217500250577e-06,
-      "loss": 1.1276,
+      "epoch": 0.7,
+      "grad_norm": 7.556723117828369,
+      "learning_rate": 1.5321089403003807e-05,
+      "loss": 0.9806,
       "step": 5593
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 11.7565336227417,
-      "learning_rate": 8.788212889646187e-06,
-      "loss": 1.1467,
+      "epoch": 0.7,
+      "grad_norm": 11.78661060333252,
+      "learning_rate": 1.532025268794712e-05,
+      "loss": 1.3586,
       "step": 5594
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 32.24980163574219,
-      "learning_rate": 8.786208279041797e-06,
-      "loss": 1.8956,
+      "epoch": 0.7,
+      "grad_norm": 29.15213966369629,
+      "learning_rate": 1.531941597289043e-05,
+      "loss": 2.4593,
       "step": 5595
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 15.186212539672852,
-      "learning_rate": 8.784203668437407e-06,
-      "loss": 1.8275,
+      "epoch": 0.7,
+      "grad_norm": 13.034687995910645,
+      "learning_rate": 1.5318579257833744e-05,
+      "loss": 1.1098,
       "step": 5596
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 26.510732650756836,
-      "learning_rate": 8.782199057833017e-06,
-      "loss": 1.2493,
+      "epoch": 0.7,
+      "grad_norm": 21.663742065429688,
+      "learning_rate": 1.5317742542777058e-05,
+      "loss": 1.2553,
       "step": 5597
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 9.903314590454102,
-      "learning_rate": 8.780194447228625e-06,
-      "loss": 0.8917,
+      "epoch": 0.7,
+      "grad_norm": 19.978364944458008,
+      "learning_rate": 1.5316905827720372e-05,
+      "loss": 1.5179,
       "step": 5598
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 9.444249153137207,
-      "learning_rate": 8.778189836624237e-06,
-      "loss": 1.073,
+      "epoch": 0.7,
+      "grad_norm": 15.038065910339355,
+      "learning_rate": 1.5316069112663682e-05,
+      "loss": 1.4449,
       "step": 5599
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 17.22249412536621,
-      "learning_rate": 8.776185226019847e-06,
-      "loss": 1.2148,
+      "epoch": 0.7,
+      "grad_norm": 14.578551292419434,
+      "learning_rate": 1.5315232397606996e-05,
+      "loss": 1.1441,
       "step": 5600
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 9.89676284790039,
-      "learning_rate": 8.774180615415456e-06,
-      "loss": 1.2275,
+      "epoch": 0.7,
+      "eval_loss": 0.11903589963912964,
+      "eval_runtime": 94.7477,
+      "eval_samples_per_second": 37.383,
+      "eval_steps_per_second": 37.383,
+      "step": 5600
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 23.72501564025879,
+      "learning_rate": 1.531439568255031e-05,
+      "loss": 3.6484,
       "step": 5601
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 12.96465015411377,
-      "learning_rate": 8.772176004811066e-06,
-      "loss": 1.0976,
+      "epoch": 0.7,
+      "grad_norm": 15.129220008850098,
+      "learning_rate": 1.531355896749362e-05,
+      "loss": 1.1805,
       "step": 5602
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 15.916191101074219,
-      "learning_rate": 8.770171394206676e-06,
-      "loss": 1.7115,
+      "epoch": 0.7,
+      "grad_norm": 10.363043785095215,
+      "learning_rate": 1.5312722252436933e-05,
+      "loss": 1.2322,
       "step": 5603
     },
     {
-      "epoch": 1.68,
-      "grad_norm": 18.037708282470703,
-      "learning_rate": 8.768166783602286e-06,
-      "loss": 1.7032,
+      "epoch": 0.7,
+      "grad_norm": 11.52414321899414,
+      "learning_rate": 1.5311885537380247e-05,
+      "loss": 1.0481,
       "step": 5604
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 15.334985733032227,
-      "learning_rate": 8.766162172997896e-06,
-      "loss": 1.3752,
+      "epoch": 0.7,
+      "grad_norm": 8.691742897033691,
+      "learning_rate": 1.531104882232356e-05,
+      "loss": 1.5519,
       "step": 5605
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 9.522956848144531,
-      "learning_rate": 8.764157562393506e-06,
-      "loss": 0.623,
+      "epoch": 0.7,
+      "grad_norm": 15.305723190307617,
+      "learning_rate": 1.531021210726687e-05,
+      "loss": 1.3309,
       "step": 5606
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 21.579017639160156,
-      "learning_rate": 8.762152951789116e-06,
-      "loss": 1.5285,
+      "epoch": 0.7,
+      "grad_norm": 22.525123596191406,
+      "learning_rate": 1.5309375392210184e-05,
+      "loss": 1.3004,
       "step": 5607
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 10.721216201782227,
-      "learning_rate": 8.760148341184726e-06,
-      "loss": 0.893,
+      "epoch": 0.7,
+      "grad_norm": 18.87114906311035,
+      "learning_rate": 1.5308538677153498e-05,
+      "loss": 1.4569,
       "step": 5608
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 29.23563575744629,
-      "learning_rate": 8.758143730580335e-06,
-      "loss": 2.0828,
+      "epoch": 0.7,
+      "grad_norm": 17.632736206054688,
+      "learning_rate": 1.530770196209681e-05,
+      "loss": 1.9694,
       "step": 5609
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 5.085595607757568,
-      "learning_rate": 8.756139119975945e-06,
-      "loss": 0.6403,
+      "epoch": 0.7,
+      "grad_norm": 16.05190086364746,
+      "learning_rate": 1.5306865247040122e-05,
+      "loss": 0.5519,
       "step": 5610
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 16.123058319091797,
-      "learning_rate": 8.754134509371556e-06,
-      "loss": 1.7075,
+      "epoch": 0.7,
+      "grad_norm": 15.205567359924316,
+      "learning_rate": 1.5306028531983436e-05,
+      "loss": 0.8385,
       "step": 5611
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 54.834938049316406,
-      "learning_rate": 8.752129898767165e-06,
-      "loss": 1.6164,
+      "epoch": 0.7,
+      "grad_norm": 10.732412338256836,
+      "learning_rate": 1.530519181692675e-05,
+      "loss": 1.9857,
       "step": 5612
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 15.420845031738281,
-      "learning_rate": 8.750125288162775e-06,
-      "loss": 1.3926,
+      "epoch": 0.7,
+      "grad_norm": 3.120633840560913,
+      "learning_rate": 1.530435510187006e-05,
+      "loss": 0.0964,
       "step": 5613
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 14.784860610961914,
-      "learning_rate": 8.748120677558385e-06,
-      "loss": 1.175,
+      "epoch": 0.7,
+      "grad_norm": 17.112985610961914,
+      "learning_rate": 1.5303518386813373e-05,
+      "loss": 2.0094,
       "step": 5614
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 32.64680862426758,
-      "learning_rate": 8.746116066953995e-06,
-      "loss": 1.5805,
+      "epoch": 0.7,
+      "grad_norm": 7.553732395172119,
+      "learning_rate": 1.5302681671756683e-05,
+      "loss": 0.5445,
       "step": 5615
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 28.060972213745117,
-      "learning_rate": 8.744111456349605e-06,
-      "loss": 1.803,
+      "epoch": 0.7,
+      "grad_norm": 7.010446071624756,
+      "learning_rate": 1.5301844956699997e-05,
+      "loss": 1.3869,
       "step": 5616
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 12.223137855529785,
-      "learning_rate": 8.742106845745215e-06,
-      "loss": 1.149,
+      "epoch": 0.7,
+      "grad_norm": 16.693117141723633,
+      "learning_rate": 1.530100824164331e-05,
+      "loss": 1.4811,
       "step": 5617
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 14.413814544677734,
-      "learning_rate": 8.740102235140825e-06,
-      "loss": 1.5615,
+      "epoch": 0.71,
+      "grad_norm": 11.959110260009766,
+      "learning_rate": 1.530017152658662e-05,
+      "loss": 1.2947,
       "step": 5618
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 24.743038177490234,
-      "learning_rate": 8.738097624536435e-06,
-      "loss": 1.5725,
+      "epoch": 0.71,
+      "grad_norm": 9.126626968383789,
+      "learning_rate": 1.5299334811529935e-05,
+      "loss": 1.7426,
       "step": 5619
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 37.43949508666992,
-      "learning_rate": 8.736093013932044e-06,
-      "loss": 2.383,
+      "epoch": 0.71,
+      "grad_norm": 134.4025115966797,
+      "learning_rate": 1.5298498096473245e-05,
+      "loss": 1.8938,
       "step": 5620
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 19.173534393310547,
-      "learning_rate": 8.734088403327654e-06,
-      "loss": 2.1207,
+      "epoch": 0.71,
+      "grad_norm": 15.066232681274414,
+      "learning_rate": 1.529766138141656e-05,
+      "loss": 2.3855,
       "step": 5621
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 17.49724006652832,
-      "learning_rate": 8.732083792723264e-06,
-      "loss": 1.7264,
+      "epoch": 0.71,
+      "grad_norm": 24.542264938354492,
+      "learning_rate": 1.5296824666359872e-05,
+      "loss": 0.9356,
       "step": 5622
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 9.76134967803955,
-      "learning_rate": 8.730079182118874e-06,
-      "loss": 1.5944,
+      "epoch": 0.71,
+      "grad_norm": 7.537159442901611,
+      "learning_rate": 1.5295987951303183e-05,
+      "loss": 1.6194,
       "step": 5623
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 11.717310905456543,
-      "learning_rate": 8.728074571514484e-06,
-      "loss": 1.756,
+      "epoch": 0.71,
+      "grad_norm": 8.826299667358398,
+      "learning_rate": 1.5295151236246496e-05,
+      "loss": 2.2357,
       "step": 5624
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 21.17192840576172,
-      "learning_rate": 8.726069960910094e-06,
-      "loss": 2.2556,
+      "epoch": 0.71,
+      "grad_norm": 16.212858200073242,
+      "learning_rate": 1.529431452118981e-05,
+      "loss": 1.3535,
       "step": 5625
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 16.35768699645996,
-      "learning_rate": 8.724065350305704e-06,
-      "loss": 1.2782,
+      "epoch": 0.71,
+      "grad_norm": 23.37677574157715,
+      "learning_rate": 1.5293477806133123e-05,
+      "loss": 1.8374,
       "step": 5626
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 12.22645378112793,
-      "learning_rate": 8.722060739701313e-06,
-      "loss": 1.3418,
+      "epoch": 0.71,
+      "grad_norm": 7.744858264923096,
+      "learning_rate": 1.5292641091076434e-05,
+      "loss": 2.3618,
       "step": 5627
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 19.546850204467773,
-      "learning_rate": 8.720056129096923e-06,
-      "loss": 1.2375,
+      "epoch": 0.71,
+      "grad_norm": 55.11528778076172,
+      "learning_rate": 1.5291804376019747e-05,
+      "loss": 2.4706,
       "step": 5628
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 14.505806922912598,
-      "learning_rate": 8.718051518492535e-06,
-      "loss": 1.5247,
+      "epoch": 0.71,
+      "grad_norm": 13.760666847229004,
+      "learning_rate": 1.529096766096306e-05,
+      "loss": 0.7654,
       "step": 5629
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 11.707478523254395,
-      "learning_rate": 8.716046907888145e-06,
-      "loss": 1.2686,
+      "epoch": 0.71,
+      "grad_norm": 13.986364364624023,
+      "learning_rate": 1.529013094590637e-05,
+      "loss": 1.6457,
       "step": 5630
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 12.525168418884277,
-      "learning_rate": 8.714042297283753e-06,
-      "loss": 2.1298,
+      "epoch": 0.71,
+      "grad_norm": 5.984301567077637,
+      "learning_rate": 1.5289294230849685e-05,
+      "loss": 0.6163,
       "step": 5631
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 13.70942497253418,
-      "learning_rate": 8.712037686679363e-06,
-      "loss": 0.8167,
+      "epoch": 0.71,
+      "grad_norm": 11.783964157104492,
+      "learning_rate": 1.5288457515793e-05,
+      "loss": 0.8994,
       "step": 5632
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 18.232576370239258,
-      "learning_rate": 8.710033076074973e-06,
-      "loss": 1.0261,
+      "epoch": 0.71,
+      "grad_norm": 73.12659454345703,
+      "learning_rate": 1.5287620800736312e-05,
+      "loss": 2.9345,
       "step": 5633
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 19.63440704345703,
-      "learning_rate": 8.708028465470583e-06,
-      "loss": 2.1018,
+      "epoch": 0.71,
+      "grad_norm": 10.865703582763672,
+      "learning_rate": 1.5286784085679622e-05,
+      "loss": 2.3683,
       "step": 5634
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 17.513425827026367,
-      "learning_rate": 8.706023854866193e-06,
-      "loss": 1.2959,
+      "epoch": 0.71,
+      "grad_norm": 12.446228981018066,
+      "learning_rate": 1.5285947370622936e-05,
+      "loss": 2.5152,
       "step": 5635
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 15.64637565612793,
-      "learning_rate": 8.704019244261803e-06,
-      "loss": 1.8533,
+      "epoch": 0.71,
+      "grad_norm": 13.962808609008789,
+      "learning_rate": 1.528511065556625e-05,
+      "loss": 1.0881,
       "step": 5636
     },
     {
-      "epoch": 1.69,
-      "grad_norm": 7.151846885681152,
-      "learning_rate": 8.702014633657414e-06,
-      "loss": 0.8085,
+      "epoch": 0.71,
+      "grad_norm": 15.042387962341309,
+      "learning_rate": 1.528427394050956e-05,
+      "loss": 2.8747,
       "step": 5637
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 9.188348770141602,
-      "learning_rate": 8.700010023053022e-06,
-      "loss": 1.1322,
+      "epoch": 0.71,
+      "grad_norm": 104.28298950195312,
+      "learning_rate": 1.5283437225452874e-05,
+      "loss": 3.2383,
       "step": 5638
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 9.800728797912598,
-      "learning_rate": 8.698005412448632e-06,
-      "loss": 0.9624,
+      "epoch": 0.71,
+      "grad_norm": 37.05125045776367,
+      "learning_rate": 1.5282600510396187e-05,
+      "loss": 1.7988,
       "step": 5639
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 14.96066951751709,
-      "learning_rate": 8.696000801844242e-06,
-      "loss": 1.3465,
-      "step": 5640
-    },
-    {
-      "epoch": 1.7,
-      "eval_loss": 0.1950412392616272,
-      "eval_runtime": 43.6602,
-      "eval_samples_per_second": 33.875,
-      "eval_steps_per_second": 33.875,
+      "epoch": 0.71,
+      "grad_norm": 22.914669036865234,
+      "learning_rate": 1.5281763795339498e-05,
+      "loss": 1.1929,
       "step": 5640
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 21.22214698791504,
-      "learning_rate": 8.693996191239852e-06,
-      "loss": 1.9302,
+      "epoch": 0.71,
+      "grad_norm": 12.188714981079102,
+      "learning_rate": 1.528092708028281e-05,
+      "loss": 2.225,
       "step": 5641
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 11.030116081237793,
-      "learning_rate": 8.691991580635462e-06,
-      "loss": 0.9403,
+      "epoch": 0.71,
+      "grad_norm": 12.646594047546387,
+      "learning_rate": 1.5280090365226125e-05,
+      "loss": 1.5027,
       "step": 5642
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 10.79010009765625,
-      "learning_rate": 8.689986970031072e-06,
-      "loss": 1.2434,
+      "epoch": 0.71,
+      "grad_norm": 6.229799270629883,
+      "learning_rate": 1.5279253650169435e-05,
+      "loss": 0.515,
       "step": 5643
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 17.979541778564453,
-      "learning_rate": 8.687982359426682e-06,
-      "loss": 1.4883,
+      "epoch": 0.71,
+      "grad_norm": 12.691096305847168,
+      "learning_rate": 1.527841693511275e-05,
+      "loss": 1.5411,
       "step": 5644
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 11.965887069702148,
-      "learning_rate": 8.685977748822292e-06,
-      "loss": 1.381,
+      "epoch": 0.71,
+      "grad_norm": 15.420011520385742,
+      "learning_rate": 1.5277580220056062e-05,
+      "loss": 2.482,
       "step": 5645
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 16.491119384765625,
-      "learning_rate": 8.683973138217901e-06,
-      "loss": 1.1538,
+      "epoch": 0.71,
+      "grad_norm": 13.545212745666504,
+      "learning_rate": 1.5276743504999373e-05,
+      "loss": 1.7458,
       "step": 5646
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 12.865142822265625,
-      "learning_rate": 8.681968527613511e-06,
-      "loss": 1.2,
+      "epoch": 0.71,
+      "grad_norm": 1.1677732467651367,
+      "learning_rate": 1.5275906789942686e-05,
+      "loss": 0.0206,
       "step": 5647
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 22.270700454711914,
-      "learning_rate": 8.679963917009123e-06,
-      "loss": 1.3727,
+      "epoch": 0.71,
+      "grad_norm": 5.918126583099365,
+      "learning_rate": 1.5275070074885997e-05,
+      "loss": 2.0381,
       "step": 5648
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 31.27298927307129,
-      "learning_rate": 8.677959306404731e-06,
-      "loss": 2.6597,
+      "epoch": 0.71,
+      "grad_norm": 42.72283172607422,
+      "learning_rate": 1.527423335982931e-05,
+      "loss": 2.3005,
       "step": 5649
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 10.323209762573242,
-      "learning_rate": 8.675954695800341e-06,
-      "loss": 0.9057,
+      "epoch": 0.71,
+      "grad_norm": 22.049144744873047,
+      "learning_rate": 1.5273396644772624e-05,
+      "loss": 1.9597,
       "step": 5650
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 14.9830961227417,
-      "learning_rate": 8.673950085195951e-06,
-      "loss": 2.7499,
+      "epoch": 0.71,
+      "grad_norm": 6.75792121887207,
+      "learning_rate": 1.5272559929715934e-05,
+      "loss": 1.9141,
       "step": 5651
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 17.780193328857422,
-      "learning_rate": 8.671945474591561e-06,
-      "loss": 1.4829,
+      "epoch": 0.71,
+      "grad_norm": 8.55809211730957,
+      "learning_rate": 1.5271723214659248e-05,
+      "loss": 2.7011,
       "step": 5652
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 14.081412315368652,
-      "learning_rate": 8.669940863987171e-06,
-      "loss": 1.159,
+      "epoch": 0.71,
+      "grad_norm": 13.690023422241211,
+      "learning_rate": 1.527088649960256e-05,
+      "loss": 0.4655,
       "step": 5653
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 16.137582778930664,
-      "learning_rate": 8.667936253382782e-06,
-      "loss": 1.0004,
+      "epoch": 0.71,
+      "grad_norm": 17.889263153076172,
+      "learning_rate": 1.5270049784545875e-05,
+      "loss": 2.3118,
       "step": 5654
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 21.85881805419922,
-      "learning_rate": 8.665931642778392e-06,
-      "loss": 1.4793,
+      "epoch": 0.71,
+      "grad_norm": 18.358047485351562,
+      "learning_rate": 1.5269213069489185e-05,
+      "loss": 2.4082,
       "step": 5655
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 21.70241355895996,
-      "learning_rate": 8.663927032174002e-06,
-      "loss": 1.6126,
+      "epoch": 0.71,
+      "grad_norm": 8.954587936401367,
+      "learning_rate": 1.52683763544325e-05,
+      "loss": 1.8922,
       "step": 5656
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 16.00739288330078,
-      "learning_rate": 8.66192242156961e-06,
-      "loss": 1.7453,
+      "epoch": 0.71,
+      "grad_norm": 27.990930557250977,
+      "learning_rate": 1.5267539639375813e-05,
+      "loss": 2.135,
       "step": 5657
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 41.264095306396484,
-      "learning_rate": 8.65991781096522e-06,
-      "loss": 1.8027,
+      "epoch": 0.71,
+      "grad_norm": 30.18874168395996,
+      "learning_rate": 1.5266702924319123e-05,
+      "loss": 1.9133,
       "step": 5658
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 59.908199310302734,
-      "learning_rate": 8.65791320036083e-06,
-      "loss": 1.4352,
+      "epoch": 0.71,
+      "grad_norm": 34.31163024902344,
+      "learning_rate": 1.5265866209262437e-05,
+      "loss": 2.7876,
       "step": 5659
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 26.252933502197266,
-      "learning_rate": 8.65590858975644e-06,
-      "loss": 1.1851,
+      "epoch": 0.71,
+      "grad_norm": 6.2728447914123535,
+      "learning_rate": 1.526502949420575e-05,
+      "loss": 0.7026,
       "step": 5660
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 14.983129501342773,
-      "learning_rate": 8.65390397915205e-06,
-      "loss": 1.1969,
+      "epoch": 0.71,
+      "grad_norm": 14.44756031036377,
+      "learning_rate": 1.5264192779149064e-05,
+      "loss": 1.9564,
       "step": 5661
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 7.037642478942871,
-      "learning_rate": 8.65189936854766e-06,
-      "loss": 0.6684,
+      "epoch": 0.71,
+      "grad_norm": 10.850642204284668,
+      "learning_rate": 1.5263356064092374e-05,
+      "loss": 0.6897,
       "step": 5662
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 12.585198402404785,
-      "learning_rate": 8.64989475794327e-06,
-      "loss": 1.4334,
+      "epoch": 0.71,
+      "grad_norm": 8.339897155761719,
+      "learning_rate": 1.5262519349035688e-05,
+      "loss": 0.6974,
       "step": 5663
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 21.677915573120117,
-      "learning_rate": 8.64789014733888e-06,
-      "loss": 2.5165,
+      "epoch": 0.71,
+      "grad_norm": 53.20794677734375,
+      "learning_rate": 1.5261682633979e-05,
+      "loss": 4.7937,
       "step": 5664
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 23.204702377319336,
-      "learning_rate": 8.645885536734489e-06,
-      "loss": 1.5472,
+      "epoch": 0.71,
+      "grad_norm": 20.38677978515625,
+      "learning_rate": 1.5260845918922312e-05,
+      "loss": 1.6178,
       "step": 5665
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 10.986197471618652,
-      "learning_rate": 8.643880926130101e-06,
-      "loss": 1.6953,
+      "epoch": 0.71,
+      "grad_norm": 10.413582801818848,
+      "learning_rate": 1.5260009203865625e-05,
+      "loss": 0.7548,
       "step": 5666
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 11.328323364257812,
-      "learning_rate": 8.641876315525711e-06,
-      "loss": 0.5602,
+      "epoch": 0.71,
+      "grad_norm": 21.674497604370117,
+      "learning_rate": 1.525917248880894e-05,
+      "loss": 2.4653,
       "step": 5667
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 9.922806739807129,
-      "learning_rate": 8.63987170492132e-06,
-      "loss": 1.3713,
+      "epoch": 0.71,
+      "grad_norm": 45.409671783447266,
+      "learning_rate": 1.5258335773752251e-05,
+      "loss": 1.736,
       "step": 5668
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 8.560861587524414,
-      "learning_rate": 8.63786709431693e-06,
-      "loss": 1.0511,
+      "epoch": 0.71,
+      "grad_norm": 13.167337417602539,
+      "learning_rate": 1.5257499058695561e-05,
+      "loss": 1.1981,
       "step": 5669
     },
     {
-      "epoch": 1.7,
-      "grad_norm": 38.06193923950195,
-      "learning_rate": 8.63586248371254e-06,
-      "loss": 1.5991,
+      "epoch": 0.71,
+      "grad_norm": 7.016934394836426,
+      "learning_rate": 1.5256662343638875e-05,
+      "loss": 0.4368,
       "step": 5670
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 17.680063247680664,
-      "learning_rate": 8.63385787310815e-06,
-      "loss": 2.7477,
+      "epoch": 0.71,
+      "grad_norm": 22.98343849182129,
+      "learning_rate": 1.5255825628582189e-05,
+      "loss": 1.8847,
       "step": 5671
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 29.49445915222168,
-      "learning_rate": 8.63185326250376e-06,
-      "loss": 1.7613,
+      "epoch": 0.71,
+      "grad_norm": 8.879514694213867,
+      "learning_rate": 1.5254988913525499e-05,
+      "loss": 1.9898,
       "step": 5672
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 32.89970397949219,
-      "learning_rate": 8.62984865189937e-06,
-      "loss": 1.6423,
+      "epoch": 0.71,
+      "grad_norm": 6.920591354370117,
+      "learning_rate": 1.5254152198468812e-05,
+      "loss": 0.8359,
       "step": 5673
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 40.36264419555664,
-      "learning_rate": 8.62784404129498e-06,
-      "loss": 1.2661,
+      "epoch": 0.71,
+      "grad_norm": 14.733323097229004,
+      "learning_rate": 1.5253315483412126e-05,
+      "loss": 2.9059,
       "step": 5674
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 15.551639556884766,
-      "learning_rate": 8.625839430690588e-06,
-      "loss": 1.6711,
+      "epoch": 0.71,
+      "grad_norm": 9.401959419250488,
+      "learning_rate": 1.5252478768355438e-05,
+      "loss": 1.1228,
       "step": 5675
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 30.854095458984375,
-      "learning_rate": 8.623834820086198e-06,
-      "loss": 1.0708,
+      "epoch": 0.71,
+      "grad_norm": 14.805521011352539,
+      "learning_rate": 1.525164205329875e-05,
+      "loss": 0.7959,
       "step": 5676
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 30.149341583251953,
-      "learning_rate": 8.621830209481808e-06,
-      "loss": 1.7262,
+      "epoch": 0.71,
+      "grad_norm": 17.671751022338867,
+      "learning_rate": 1.5250805338242064e-05,
+      "loss": 1.9533,
       "step": 5677
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 17.061527252197266,
-      "learning_rate": 8.619825598877418e-06,
-      "loss": 1.4248,
+      "epoch": 0.71,
+      "grad_norm": 9.388689994812012,
+      "learning_rate": 1.5249968623185376e-05,
+      "loss": 2.1568,
       "step": 5678
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 78.73602294921875,
-      "learning_rate": 8.617820988273029e-06,
-      "loss": 2.1561,
+      "epoch": 0.71,
+      "grad_norm": 22.36805534362793,
+      "learning_rate": 1.5249131908128688e-05,
+      "loss": 2.0653,
       "step": 5679
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 51.467491149902344,
-      "learning_rate": 8.615816377668639e-06,
-      "loss": 1.7134,
+      "epoch": 0.71,
+      "grad_norm": 6.249520778656006,
+      "learning_rate": 1.5248295193072e-05,
+      "loss": 2.0328,
       "step": 5680
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 17.935325622558594,
-      "learning_rate": 8.613811767064249e-06,
-      "loss": 1.576,
+      "epoch": 0.71,
+      "grad_norm": 12.663107872009277,
+      "learning_rate": 1.5247458478015313e-05,
+      "loss": 1.3492,
       "step": 5681
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 14.966944694519043,
-      "learning_rate": 8.611807156459859e-06,
-      "loss": 2.0311,
+      "epoch": 0.71,
+      "grad_norm": 7.885650157928467,
+      "learning_rate": 1.5246621762958627e-05,
+      "loss": 0.5711,
       "step": 5682
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 14.564214706420898,
-      "learning_rate": 8.609802545855467e-06,
-      "loss": 1.7562,
+      "epoch": 0.71,
+      "grad_norm": 7.987274646759033,
+      "learning_rate": 1.5245785047901937e-05,
+      "loss": 1.8003,
       "step": 5683
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 11.899775505065918,
-      "learning_rate": 8.607797935251077e-06,
-      "loss": 1.2177,
+      "epoch": 0.71,
+      "grad_norm": 12.844233512878418,
+      "learning_rate": 1.524494833284525e-05,
+      "loss": 1.0966,
       "step": 5684
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 14.998151779174805,
-      "learning_rate": 8.605793324646689e-06,
-      "loss": 1.6503,
+      "epoch": 0.71,
+      "grad_norm": 20.380704879760742,
+      "learning_rate": 1.5244111617788564e-05,
+      "loss": 1.5229,
       "step": 5685
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 18.94357681274414,
-      "learning_rate": 8.603788714042297e-06,
-      "loss": 1.411,
+      "epoch": 0.71,
+      "grad_norm": 5.888026714324951,
+      "learning_rate": 1.5243274902731875e-05,
+      "loss": 0.4774,
       "step": 5686
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 14.1841459274292,
-      "learning_rate": 8.601784103437908e-06,
-      "loss": 1.5919,
+      "epoch": 0.71,
+      "grad_norm": 5.787362098693848,
+      "learning_rate": 1.5242438187675188e-05,
+      "loss": 2.1025,
       "step": 5687
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 32.609100341796875,
-      "learning_rate": 8.599779492833518e-06,
-      "loss": 1.7305,
+      "epoch": 0.71,
+      "grad_norm": 41.53786087036133,
+      "learning_rate": 1.5241601472618502e-05,
+      "loss": 3.4078,
       "step": 5688
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 35.52297592163086,
-      "learning_rate": 8.597774882229128e-06,
-      "loss": 2.0784,
+      "epoch": 0.71,
+      "grad_norm": 29.634103775024414,
+      "learning_rate": 1.5240764757561814e-05,
+      "loss": 2.0694,
       "step": 5689
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 13.570123672485352,
-      "learning_rate": 8.595770271624738e-06,
-      "loss": 1.0387,
+      "epoch": 0.71,
+      "grad_norm": 7.678778171539307,
+      "learning_rate": 1.5239928042505126e-05,
+      "loss": 1.2133,
       "step": 5690
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 12.182998657226562,
-      "learning_rate": 8.593765661020348e-06,
-      "loss": 1.042,
+      "epoch": 0.71,
+      "grad_norm": 7.2586798667907715,
+      "learning_rate": 1.523909132744844e-05,
+      "loss": 0.7017,
       "step": 5691
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 56.89901351928711,
-      "learning_rate": 8.591761050415958e-06,
-      "loss": 2.2758,
+      "epoch": 0.71,
+      "grad_norm": 10.308073997497559,
+      "learning_rate": 1.5238254612391751e-05,
+      "loss": 1.672,
       "step": 5692
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 16.903789520263672,
-      "learning_rate": 8.589756439811568e-06,
-      "loss": 1.0774,
+      "epoch": 0.71,
+      "grad_norm": 8.929986000061035,
+      "learning_rate": 1.5237417897335063e-05,
+      "loss": 0.3281,
       "step": 5693
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 36.77326583862305,
-      "learning_rate": 8.587751829207176e-06,
-      "loss": 2.6291,
+      "epoch": 0.71,
+      "grad_norm": 18.599157333374023,
+      "learning_rate": 1.5236581182278375e-05,
+      "loss": 1.7576,
       "step": 5694
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 24.694618225097656,
-      "learning_rate": 8.585747218602787e-06,
-      "loss": 1.4135,
+      "epoch": 0.71,
+      "grad_norm": 46.4240608215332,
+      "learning_rate": 1.5235744467221689e-05,
+      "loss": 1.8176,
       "step": 5695
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 44.71474075317383,
-      "learning_rate": 8.583742607998397e-06,
-      "loss": 1.5254,
+      "epoch": 0.71,
+      "grad_norm": 12.64277458190918,
+      "learning_rate": 1.5234907752165003e-05,
+      "loss": 1.5387,
       "step": 5696
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 27.899707794189453,
-      "learning_rate": 8.581737997394007e-06,
-      "loss": 1.8512,
+      "epoch": 0.71,
+      "grad_norm": 9.800168991088867,
+      "learning_rate": 1.5234071037108313e-05,
+      "loss": 0.7847,
       "step": 5697
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 8.52084732055664,
-      "learning_rate": 8.579733386789617e-06,
-      "loss": 1.3119,
+      "epoch": 0.72,
+      "grad_norm": 9.509275436401367,
+      "learning_rate": 1.5233234322051627e-05,
+      "loss": 0.6097,
       "step": 5698
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 12.418741226196289,
-      "learning_rate": 8.577728776185227e-06,
-      "loss": 1.7178,
+      "epoch": 0.72,
+      "grad_norm": 16.417707443237305,
+      "learning_rate": 1.523239760699494e-05,
+      "loss": 1.5009,
       "step": 5699
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 11.793305397033691,
-      "learning_rate": 8.575724165580837e-06,
-      "loss": 1.0568,
+      "epoch": 0.72,
+      "grad_norm": 10.363656997680664,
+      "learning_rate": 1.523156089193825e-05,
+      "loss": 0.9842,
       "step": 5700
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 9.70165729522705,
-      "learning_rate": 8.573719554976447e-06,
-      "loss": 1.6653,
+      "epoch": 0.72,
+      "grad_norm": 6.770571231842041,
+      "learning_rate": 1.5230724176881564e-05,
+      "loss": 0.5578,
       "step": 5701
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 15.392131805419922,
-      "learning_rate": 8.571714944372055e-06,
-      "loss": 2.2157,
+      "epoch": 0.72,
+      "grad_norm": 39.45873260498047,
+      "learning_rate": 1.5229887461824878e-05,
+      "loss": 1.904,
       "step": 5702
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 35.93635177612305,
-      "learning_rate": 8.569710333767667e-06,
-      "loss": 1.2638,
+      "epoch": 0.72,
+      "grad_norm": 9.785941123962402,
+      "learning_rate": 1.522905074676819e-05,
+      "loss": 0.9517,
       "step": 5703
     },
     {
-      "epoch": 1.71,
-      "grad_norm": 12.030016899108887,
-      "learning_rate": 8.567705723163277e-06,
-      "loss": 1.3019,
+      "epoch": 0.72,
+      "grad_norm": 8.684469223022461,
+      "learning_rate": 1.5228214031711502e-05,
+      "loss": 0.7899,
       "step": 5704
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 39.288814544677734,
-      "learning_rate": 8.565701112558886e-06,
-      "loss": 2.3544,
+      "epoch": 0.72,
+      "grad_norm": 14.175126075744629,
+      "learning_rate": 1.5227377316654814e-05,
+      "loss": 1.3251,
       "step": 5705
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 23.307493209838867,
-      "learning_rate": 8.563696501954496e-06,
-      "loss": 1.5503,
+      "epoch": 0.72,
+      "grad_norm": 11.509045600891113,
+      "learning_rate": 1.5226540601598127e-05,
+      "loss": 1.8926,
       "step": 5706
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 26.393465042114258,
-      "learning_rate": 8.561691891350106e-06,
-      "loss": 2.5251,
+      "epoch": 0.72,
+      "grad_norm": 44.07933807373047,
+      "learning_rate": 1.522570388654144e-05,
+      "loss": 1.7037,
       "step": 5707
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 10.045159339904785,
-      "learning_rate": 8.559687280745716e-06,
-      "loss": 1.1675,
+      "epoch": 0.72,
+      "grad_norm": 27.919408798217773,
+      "learning_rate": 1.5224867171484751e-05,
+      "loss": 1.7059,
       "step": 5708
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 35.52949142456055,
-      "learning_rate": 8.557682670141326e-06,
-      "loss": 2.1613,
+      "epoch": 0.72,
+      "grad_norm": 16.3609619140625,
+      "learning_rate": 1.5224030456428065e-05,
+      "loss": 3.2524,
       "step": 5709
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 19.822158813476562,
-      "learning_rate": 8.555678059536936e-06,
-      "loss": 1.5757,
+      "epoch": 0.72,
+      "grad_norm": 7.351317405700684,
+      "learning_rate": 1.5223193741371378e-05,
+      "loss": 1.9986,
       "step": 5710
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 41.30690383911133,
-      "learning_rate": 8.553673448932546e-06,
-      "loss": 2.7388,
+      "epoch": 0.72,
+      "grad_norm": 8.646843910217285,
+      "learning_rate": 1.5222357026314689e-05,
+      "loss": 1.075,
       "step": 5711
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 20.42682456970215,
-      "learning_rate": 8.551668838328155e-06,
-      "loss": 2.3481,
+      "epoch": 0.72,
+      "grad_norm": 47.738895416259766,
+      "learning_rate": 1.5221520311258002e-05,
+      "loss": 2.951,
       "step": 5712
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 9.312350273132324,
-      "learning_rate": 8.549664227723765e-06,
-      "loss": 0.9268,
+      "epoch": 0.72,
+      "grad_norm": 28.56525230407715,
+      "learning_rate": 1.5220683596201316e-05,
+      "loss": 1.5992,
       "step": 5713
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 11.393967628479004,
-      "learning_rate": 8.547659617119375e-06,
-      "loss": 1.6853,
+      "epoch": 0.72,
+      "grad_norm": 11.733442306518555,
+      "learning_rate": 1.5219846881144626e-05,
+      "loss": 1.3493,
       "step": 5714
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 22.255510330200195,
-      "learning_rate": 8.545655006514987e-06,
-      "loss": 1.0106,
+      "epoch": 0.72,
+      "grad_norm": 9.42813491821289,
+      "learning_rate": 1.521901016608794e-05,
+      "loss": 0.8787,
       "step": 5715
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 12.787086486816406,
-      "learning_rate": 8.543650395910595e-06,
-      "loss": 1.6355,
+      "epoch": 0.72,
+      "grad_norm": 14.754581451416016,
+      "learning_rate": 1.5218173451031254e-05,
+      "loss": 0.965,
       "step": 5716
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 39.999839782714844,
-      "learning_rate": 8.541645785306205e-06,
-      "loss": 1.8123,
+      "epoch": 0.72,
+      "grad_norm": 21.997575759887695,
+      "learning_rate": 1.5217336735974566e-05,
+      "loss": 1.7967,
       "step": 5717
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 7.7507853507995605,
-      "learning_rate": 8.539641174701815e-06,
-      "loss": 1.4238,
+      "epoch": 0.72,
+      "grad_norm": 9.579668045043945,
+      "learning_rate": 1.5216500020917878e-05,
+      "loss": 2.0504,
       "step": 5718
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 15.439056396484375,
-      "learning_rate": 8.537636564097425e-06,
-      "loss": 1.7531,
+      "epoch": 0.72,
+      "grad_norm": 15.060192108154297,
+      "learning_rate": 1.521566330586119e-05,
+      "loss": 1.9743,
       "step": 5719
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 17.616294860839844,
-      "learning_rate": 8.535631953493034e-06,
-      "loss": 1.8694,
+      "epoch": 0.72,
+      "grad_norm": 18.11211395263672,
+      "learning_rate": 1.5214826590804503e-05,
+      "loss": 2.1337,
       "step": 5720
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 10.567305564880371,
-      "learning_rate": 8.533627342888644e-06,
-      "loss": 1.0895,
+      "epoch": 0.72,
+      "grad_norm": 10.534466743469238,
+      "learning_rate": 1.5213989875747815e-05,
+      "loss": 1.7234,
       "step": 5721
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 14.490511894226074,
-      "learning_rate": 8.531622732284255e-06,
-      "loss": 1.4669,
+      "epoch": 0.72,
+      "grad_norm": 15.524293899536133,
+      "learning_rate": 1.5213153160691127e-05,
+      "loss": 0.9391,
       "step": 5722
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 9.218794822692871,
-      "learning_rate": 8.529618121679864e-06,
-      "loss": 1.0658,
+      "epoch": 0.72,
+      "grad_norm": 27.32461929321289,
+      "learning_rate": 1.521231644563444e-05,
+      "loss": 3.1592,
       "step": 5723
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 10.850655555725098,
-      "learning_rate": 8.527613511075474e-06,
-      "loss": 1.0692,
+      "epoch": 0.72,
+      "grad_norm": 16.06114387512207,
+      "learning_rate": 1.5211479730577754e-05,
+      "loss": 2.1683,
       "step": 5724
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 29.35923194885254,
-      "learning_rate": 8.525608900471084e-06,
-      "loss": 1.4617,
+      "epoch": 0.72,
+      "grad_norm": 5.4787211418151855,
+      "learning_rate": 1.5210643015521065e-05,
+      "loss": 0.2532,
       "step": 5725
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 19.219905853271484,
-      "learning_rate": 8.523604289866694e-06,
-      "loss": 2.7337,
+      "epoch": 0.72,
+      "grad_norm": 14.662065505981445,
+      "learning_rate": 1.5209806300464378e-05,
+      "loss": 2.8309,
       "step": 5726
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 17.148277282714844,
-      "learning_rate": 8.521599679262304e-06,
-      "loss": 1.4809,
+      "epoch": 0.72,
+      "grad_norm": 13.122373580932617,
+      "learning_rate": 1.5208969585407692e-05,
+      "loss": 1.8849,
       "step": 5727
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 55.859092712402344,
-      "learning_rate": 8.519595068657914e-06,
-      "loss": 1.3051,
+      "epoch": 0.72,
+      "grad_norm": 33.124778747558594,
+      "learning_rate": 1.5208132870351002e-05,
+      "loss": 2.8645,
       "step": 5728
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 12.13316822052002,
-      "learning_rate": 8.517590458053524e-06,
-      "loss": 1.3944,
+      "epoch": 0.72,
+      "grad_norm": 13.683943748474121,
+      "learning_rate": 1.5207296155294316e-05,
+      "loss": 1.1949,
       "step": 5729
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 26.604719161987305,
-      "learning_rate": 8.515585847449134e-06,
-      "loss": 1.6258,
+      "epoch": 0.72,
+      "grad_norm": 9.916946411132812,
+      "learning_rate": 1.520645944023763e-05,
+      "loss": 2.8687,
       "step": 5730
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 13.429365158081055,
-      "learning_rate": 8.513581236844743e-06,
-      "loss": 1.0267,
+      "epoch": 0.72,
+      "grad_norm": 15.309993743896484,
+      "learning_rate": 1.5205622725180941e-05,
+      "loss": 2.193,
       "step": 5731
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 13.195096969604492,
-      "learning_rate": 8.511576626240353e-06,
-      "loss": 1.1112,
+      "epoch": 0.72,
+      "grad_norm": 11.65296745300293,
+      "learning_rate": 1.5204786010124253e-05,
+      "loss": 1.0531,
       "step": 5732
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 19.855567932128906,
-      "learning_rate": 8.509572015635963e-06,
-      "loss": 1.1505,
+      "epoch": 0.72,
+      "grad_norm": 27.579500198364258,
+      "learning_rate": 1.5203949295067565e-05,
+      "loss": 1.1293,
       "step": 5733
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 13.332557678222656,
-      "learning_rate": 8.507567405031573e-06,
-      "loss": 1.5326,
+      "epoch": 0.72,
+      "grad_norm": 6.373600006103516,
+      "learning_rate": 1.5203112580010879e-05,
+      "loss": 1.607,
       "step": 5734
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 11.699455261230469,
-      "learning_rate": 8.505562794427183e-06,
-      "loss": 1.3027,
+      "epoch": 0.72,
+      "grad_norm": 12.031510353088379,
+      "learning_rate": 1.5202275864954191e-05,
+      "loss": 1.845,
       "step": 5735
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 55.0339469909668,
-      "learning_rate": 8.503558183822793e-06,
-      "loss": 1.8905,
+      "epoch": 0.72,
+      "grad_norm": 6.446319103240967,
+      "learning_rate": 1.5201439149897503e-05,
+      "loss": 1.699,
       "step": 5736
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 14.223106384277344,
-      "learning_rate": 8.501553573218403e-06,
-      "loss": 1.1121,
+      "epoch": 0.72,
+      "grad_norm": 18.43304443359375,
+      "learning_rate": 1.5200602434840817e-05,
+      "loss": 1.6499,
       "step": 5737
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 19.943954467773438,
-      "learning_rate": 8.499548962614013e-06,
-      "loss": 1.9677,
+      "epoch": 0.72,
+      "grad_norm": 21.51383399963379,
+      "learning_rate": 1.519976571978413e-05,
+      "loss": 1.3497,
       "step": 5738
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 30.25005340576172,
-      "learning_rate": 8.497544352009622e-06,
-      "loss": 1.7023,
+      "epoch": 0.72,
+      "grad_norm": 11.108257293701172,
+      "learning_rate": 1.519892900472744e-05,
+      "loss": 0.6843,
       "step": 5739
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 19.0448055267334,
-      "learning_rate": 8.495539741405234e-06,
-      "loss": 1.3845,
+      "epoch": 0.72,
+      "grad_norm": 13.190654754638672,
+      "learning_rate": 1.5198092289670754e-05,
+      "loss": 2.054,
       "step": 5740
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 16.908838272094727,
-      "learning_rate": 8.493535130800844e-06,
-      "loss": 1.7335,
+      "epoch": 0.72,
+      "grad_norm": 12.344770431518555,
+      "learning_rate": 1.5197255574614068e-05,
+      "loss": 1.5029,
       "step": 5741
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 24.786378860473633,
-      "learning_rate": 8.491530520196452e-06,
-      "loss": 1.7752,
+      "epoch": 0.72,
+      "grad_norm": 10.223804473876953,
+      "learning_rate": 1.5196418859557378e-05,
+      "loss": 1.5523,
       "step": 5742
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 14.871647834777832,
-      "learning_rate": 8.489525909592062e-06,
-      "loss": 1.7707,
+      "epoch": 0.72,
+      "grad_norm": 16.279478073120117,
+      "learning_rate": 1.5195582144500692e-05,
+      "loss": 2.2957,
       "step": 5743
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 9.237364768981934,
-      "learning_rate": 8.487521298987672e-06,
-      "loss": 1.2148,
+      "epoch": 0.72,
+      "grad_norm": 11.446799278259277,
+      "learning_rate": 1.5194745429444005e-05,
+      "loss": 1.5505,
       "step": 5744
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 20.291860580444336,
-      "learning_rate": 8.485516688383282e-06,
-      "loss": 1.3439,
+      "epoch": 0.72,
+      "grad_norm": 29.410280227661133,
+      "learning_rate": 1.5193908714387317e-05,
+      "loss": 2.2996,
       "step": 5745
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 23.027454376220703,
-      "learning_rate": 8.483512077778892e-06,
-      "loss": 1.6966,
+      "epoch": 0.72,
+      "grad_norm": 34.11577606201172,
+      "learning_rate": 1.519307199933063e-05,
+      "loss": 1.5296,
       "step": 5746
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 14.931220054626465,
-      "learning_rate": 8.481507467174502e-06,
-      "loss": 1.8485,
+      "epoch": 0.72,
+      "grad_norm": 22.040115356445312,
+      "learning_rate": 1.5192235284273941e-05,
+      "loss": 1.6252,
       "step": 5747
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 8.827281951904297,
-      "learning_rate": 8.479502856570113e-06,
-      "loss": 0.8949,
+      "epoch": 0.72,
+      "grad_norm": 37.460208892822266,
+      "learning_rate": 1.5191398569217255e-05,
+      "loss": 2.2588,
       "step": 5748
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 13.313647270202637,
-      "learning_rate": 8.477498245965723e-06,
-      "loss": 1.3746,
+      "epoch": 0.72,
+      "grad_norm": 11.320125579833984,
+      "learning_rate": 1.5190561854160567e-05,
+      "loss": 2.044,
       "step": 5749
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 40.448280334472656,
-      "learning_rate": 8.475493635361331e-06,
-      "loss": 2.6908,
+      "epoch": 0.72,
+      "grad_norm": 15.098015785217285,
+      "learning_rate": 1.5189725139103879e-05,
+      "loss": 2.4267,
       "step": 5750
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 15.871477127075195,
-      "learning_rate": 8.473489024756941e-06,
-      "loss": 1.6084,
+      "epoch": 0.72,
+      "grad_norm": 12.515155792236328,
+      "learning_rate": 1.5188888424047192e-05,
+      "loss": 1.7228,
       "step": 5751
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 38.03894805908203,
-      "learning_rate": 8.471484414152553e-06,
-      "loss": 1.693,
+      "epoch": 0.72,
+      "grad_norm": 44.248958587646484,
+      "learning_rate": 1.5188051708990506e-05,
+      "loss": 2.0615,
       "step": 5752
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 53.909751892089844,
-      "learning_rate": 8.469479803548161e-06,
-      "loss": 1.3998,
+      "epoch": 0.72,
+      "grad_norm": 17.942691802978516,
+      "learning_rate": 1.5187214993933816e-05,
+      "loss": 2.337,
       "step": 5753
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 23.2731876373291,
-      "learning_rate": 8.467475192943771e-06,
-      "loss": 1.1304,
+      "epoch": 0.72,
+      "grad_norm": 9.37904167175293,
+      "learning_rate": 1.518637827887713e-05,
+      "loss": 1.5438,
       "step": 5754
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 16.913698196411133,
-      "learning_rate": 8.465470582339381e-06,
-      "loss": 1.2357,
+      "epoch": 0.72,
+      "grad_norm": 21.882314682006836,
+      "learning_rate": 1.5185541563820444e-05,
+      "loss": 2.6402,
       "step": 5755
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 25.4580135345459,
-      "learning_rate": 8.463465971734992e-06,
-      "loss": 1.669,
+      "epoch": 0.72,
+      "grad_norm": 19.666954040527344,
+      "learning_rate": 1.5184704848763754e-05,
+      "loss": 1.7297,
       "step": 5756
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 17.684297561645508,
-      "learning_rate": 8.4614613611306e-06,
-      "loss": 1.8284,
+      "epoch": 0.72,
+      "grad_norm": 14.530044555664062,
+      "learning_rate": 1.5183868133707067e-05,
+      "loss": 1.3628,
       "step": 5757
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 11.82136058807373,
-      "learning_rate": 8.459456750526212e-06,
-      "loss": 1.5129,
+      "epoch": 0.72,
+      "grad_norm": 34.499019622802734,
+      "learning_rate": 1.518303141865038e-05,
+      "loss": 2.3917,
       "step": 5758
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 19.27686882019043,
-      "learning_rate": 8.457452139921822e-06,
-      "loss": 2.0577,
+      "epoch": 0.72,
+      "grad_norm": 10.190333366394043,
+      "learning_rate": 1.5182194703593693e-05,
+      "loss": 3.1606,
       "step": 5759
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 19.104829788208008,
-      "learning_rate": 8.45544752931743e-06,
-      "loss": 1.8845,
-      "step": 5760
-    },
-    {
-      "epoch": 1.73,
-      "eval_loss": 0.19915775954723358,
-      "eval_runtime": 43.571,
-      "eval_samples_per_second": 33.945,
-      "eval_steps_per_second": 33.945,
+      "epoch": 0.72,
+      "grad_norm": 20.318904876708984,
+      "learning_rate": 1.5181357988537005e-05,
+      "loss": 2.2891,
       "step": 5760
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 30.6223201751709,
-      "learning_rate": 8.45344291871304e-06,
-      "loss": 1.4588,
+      "epoch": 0.72,
+      "grad_norm": 40.29243469238281,
+      "learning_rate": 1.5180521273480317e-05,
+      "loss": 0.8837,
       "step": 5761
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 20.711313247680664,
-      "learning_rate": 8.45143830810865e-06,
-      "loss": 1.9552,
+      "epoch": 0.72,
+      "grad_norm": 9.452630996704102,
+      "learning_rate": 1.517968455842363e-05,
+      "loss": 1.8831,
       "step": 5762
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 9.937006950378418,
-      "learning_rate": 8.44943369750426e-06,
-      "loss": 1.4769,
+      "epoch": 0.72,
+      "grad_norm": 5.90389347076416,
+      "learning_rate": 1.5178847843366943e-05,
+      "loss": 0.5783,
       "step": 5763
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 11.456831932067871,
-      "learning_rate": 8.44742908689987e-06,
-      "loss": 1.486,
+      "epoch": 0.72,
+      "grad_norm": 26.681222915649414,
+      "learning_rate": 1.5178011128310255e-05,
+      "loss": 1.5196,
       "step": 5764
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 11.295050621032715,
-      "learning_rate": 8.44542447629548e-06,
-      "loss": 1.1309,
+      "epoch": 0.72,
+      "grad_norm": 14.583979606628418,
+      "learning_rate": 1.5177174413253568e-05,
+      "loss": 1.7832,
       "step": 5765
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 10.123924255371094,
-      "learning_rate": 8.44341986569109e-06,
-      "loss": 0.8497,
+      "epoch": 0.72,
+      "grad_norm": 6.687341690063477,
+      "learning_rate": 1.5176337698196882e-05,
+      "loss": 0.9506,
       "step": 5766
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 23.44426727294922,
-      "learning_rate": 8.4414152550867e-06,
-      "loss": 1.9237,
+      "epoch": 0.72,
+      "grad_norm": 7.795732498168945,
+      "learning_rate": 1.5175500983140192e-05,
+      "loss": 0.8163,
       "step": 5767
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 23.85992431640625,
-      "learning_rate": 8.439410644482309e-06,
-      "loss": 2.8773,
+      "epoch": 0.72,
+      "grad_norm": 62.16205978393555,
+      "learning_rate": 1.5174664268083506e-05,
+      "loss": 1.8991,
       "step": 5768
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 12.136680603027344,
-      "learning_rate": 8.43740603387792e-06,
-      "loss": 1.9736,
+      "epoch": 0.72,
+      "grad_norm": 43.09776306152344,
+      "learning_rate": 1.517382755302682e-05,
+      "loss": 1.6252,
       "step": 5769
     },
     {
-      "epoch": 1.73,
-      "grad_norm": 13.099090576171875,
-      "learning_rate": 8.43540142327353e-06,
-      "loss": 1.9405,
+      "epoch": 0.72,
+      "grad_norm": 7.008608818054199,
+      "learning_rate": 1.517299083797013e-05,
+      "loss": 0.7518,
       "step": 5770
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 18.678722381591797,
-      "learning_rate": 8.43339681266914e-06,
-      "loss": 2.2382,
+      "epoch": 0.72,
+      "grad_norm": 4.863678932189941,
+      "learning_rate": 1.5172154122913443e-05,
+      "loss": 0.9405,
       "step": 5771
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 14.240285873413086,
-      "learning_rate": 8.43139220206475e-06,
-      "loss": 1.3103,
+      "epoch": 0.72,
+      "grad_norm": 11.31428337097168,
+      "learning_rate": 1.5171317407856755e-05,
+      "loss": 1.3956,
       "step": 5772
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 21.930389404296875,
-      "learning_rate": 8.42938759146036e-06,
-      "loss": 1.4554,
+      "epoch": 0.72,
+      "grad_norm": 18.717117309570312,
+      "learning_rate": 1.5170480692800069e-05,
+      "loss": 1.6175,
       "step": 5773
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 20.750886917114258,
-      "learning_rate": 8.42738298085597e-06,
-      "loss": 1.1572,
+      "epoch": 0.72,
+      "grad_norm": 23.560714721679688,
+      "learning_rate": 1.5169643977743381e-05,
+      "loss": 2.3305,
       "step": 5774
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 13.55035400390625,
-      "learning_rate": 8.42537837025158e-06,
-      "loss": 1.0804,
+      "epoch": 0.72,
+      "grad_norm": 17.379194259643555,
+      "learning_rate": 1.5168807262686693e-05,
+      "loss": 1.2868,
       "step": 5775
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 12.210824966430664,
-      "learning_rate": 8.423373759647188e-06,
-      "loss": 1.526,
+      "epoch": 0.72,
+      "grad_norm": 13.342092514038086,
+      "learning_rate": 1.5167970547630006e-05,
+      "loss": 2.5117,
       "step": 5776
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 9.945907592773438,
-      "learning_rate": 8.4213691490428e-06,
-      "loss": 1.1015,
+      "epoch": 0.73,
+      "grad_norm": 10.251347541809082,
+      "learning_rate": 1.5167133832573317e-05,
+      "loss": 1.5929,
       "step": 5777
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 16.482894897460938,
-      "learning_rate": 8.41936453843841e-06,
-      "loss": 1.9265,
+      "epoch": 0.73,
+      "grad_norm": 19.485706329345703,
+      "learning_rate": 1.516629711751663e-05,
+      "loss": 1.989,
       "step": 5778
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 16.412376403808594,
-      "learning_rate": 8.417359927834018e-06,
-      "loss": 1.9719,
+      "epoch": 0.73,
+      "grad_norm": 17.89291763305664,
+      "learning_rate": 1.5165460402459944e-05,
+      "loss": 1.2587,
       "step": 5779
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 49.32164764404297,
-      "learning_rate": 8.415355317229628e-06,
-      "loss": 2.7723,
+      "epoch": 0.73,
+      "grad_norm": 20.66066551208496,
+      "learning_rate": 1.5164623687403258e-05,
+      "loss": 1.8046,
       "step": 5780
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 27.84273910522461,
-      "learning_rate": 8.413350706625239e-06,
-      "loss": 1.1557,
+      "epoch": 0.73,
+      "grad_norm": 25.348182678222656,
+      "learning_rate": 1.5163786972346568e-05,
+      "loss": 1.994,
       "step": 5781
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 25.465152740478516,
-      "learning_rate": 8.411346096020849e-06,
-      "loss": 1.241,
+      "epoch": 0.73,
+      "grad_norm": 18.08916664123535,
+      "learning_rate": 1.5162950257289882e-05,
+      "loss": 3.1051,
       "step": 5782
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 21.347864151000977,
-      "learning_rate": 8.409341485416459e-06,
-      "loss": 2.8817,
+      "epoch": 0.73,
+      "grad_norm": 26.334613800048828,
+      "learning_rate": 1.5162113542233195e-05,
+      "loss": 0.6722,
       "step": 5783
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 17.83620834350586,
-      "learning_rate": 8.407336874812069e-06,
-      "loss": 1.4303,
+      "epoch": 0.73,
+      "grad_norm": 3.4977641105651855,
+      "learning_rate": 1.5161276827176505e-05,
+      "loss": 0.4504,
       "step": 5784
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 24.558000564575195,
-      "learning_rate": 8.405332264207679e-06,
-      "loss": 0.9856,
+      "epoch": 0.73,
+      "grad_norm": 24.30824089050293,
+      "learning_rate": 1.5160440112119819e-05,
+      "loss": 1.6743,
       "step": 5785
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 16.54327392578125,
-      "learning_rate": 8.403327653603289e-06,
-      "loss": 0.702,
+      "epoch": 0.73,
+      "grad_norm": 8.112894058227539,
+      "learning_rate": 1.5159603397063131e-05,
+      "loss": 1.1564,
       "step": 5786
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 21.651714324951172,
-      "learning_rate": 8.401323042998897e-06,
-      "loss": 1.6273,
+      "epoch": 0.73,
+      "grad_norm": 4.117024898529053,
+      "learning_rate": 1.5158766682006445e-05,
+      "loss": 0.3427,
       "step": 5787
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 13.498676300048828,
-      "learning_rate": 8.399318432394507e-06,
-      "loss": 0.949,
+      "epoch": 0.73,
+      "grad_norm": 16.71449851989746,
+      "learning_rate": 1.5157929966949757e-05,
+      "loss": 2.5273,
       "step": 5788
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 10.88339900970459,
-      "learning_rate": 8.39731382179012e-06,
-      "loss": 1.7288,
+      "epoch": 0.73,
+      "grad_norm": 12.165425300598145,
+      "learning_rate": 1.5157093251893069e-05,
+      "loss": 0.888,
       "step": 5789
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 10.3096284866333,
-      "learning_rate": 8.395309211185728e-06,
-      "loss": 1.1514,
+      "epoch": 0.73,
+      "grad_norm": 18.057558059692383,
+      "learning_rate": 1.5156256536836382e-05,
+      "loss": 1.0489,
       "step": 5790
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 17.805374145507812,
-      "learning_rate": 8.393304600581338e-06,
-      "loss": 1.5093,
+      "epoch": 0.73,
+      "grad_norm": 39.79581069946289,
+      "learning_rate": 1.5155419821779693e-05,
+      "loss": 2.0024,
       "step": 5791
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 16.813570022583008,
-      "learning_rate": 8.391299989976948e-06,
-      "loss": 0.8186,
+      "epoch": 0.73,
+      "grad_norm": 17.75973129272461,
+      "learning_rate": 1.5154583106723006e-05,
+      "loss": 2.2979,
       "step": 5792
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 20.913787841796875,
-      "learning_rate": 8.389295379372558e-06,
-      "loss": 1.7035,
+      "epoch": 0.73,
+      "grad_norm": 10.372922897338867,
+      "learning_rate": 1.515374639166632e-05,
+      "loss": 0.681,
       "step": 5793
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 52.88319778442383,
-      "learning_rate": 8.387290768768166e-06,
-      "loss": 1.7603,
+      "epoch": 0.73,
+      "grad_norm": 15.179542541503906,
+      "learning_rate": 1.5152909676609633e-05,
+      "loss": 1.9187,
       "step": 5794
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 15.752192497253418,
-      "learning_rate": 8.385286158163778e-06,
-      "loss": 1.2862,
+      "epoch": 0.73,
+      "grad_norm": 12.520872116088867,
+      "learning_rate": 1.5152072961552944e-05,
+      "loss": 1.2136,
       "step": 5795
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 8.408364295959473,
-      "learning_rate": 8.383281547559388e-06,
-      "loss": 0.9196,
+      "epoch": 0.73,
+      "grad_norm": 33.91791915893555,
+      "learning_rate": 1.5151236246496257e-05,
+      "loss": 1.4088,
       "step": 5796
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 18.868669509887695,
-      "learning_rate": 8.381276936954997e-06,
-      "loss": 1.5121,
+      "epoch": 0.73,
+      "grad_norm": 23.570266723632812,
+      "learning_rate": 1.515039953143957e-05,
+      "loss": 1.559,
       "step": 5797
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 20.32513999938965,
-      "learning_rate": 8.379272326350607e-06,
-      "loss": 2.3994,
+      "epoch": 0.73,
+      "grad_norm": 18.323856353759766,
+      "learning_rate": 1.5149562816382881e-05,
+      "loss": 1.7075,
       "step": 5798
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 19.855003356933594,
-      "learning_rate": 8.377267715746217e-06,
-      "loss": 1.8385,
+      "epoch": 0.73,
+      "grad_norm": 8.110604286193848,
+      "learning_rate": 1.5148726101326195e-05,
+      "loss": 0.1192,
       "step": 5799
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 15.905160903930664,
-      "learning_rate": 8.375263105141827e-06,
-      "loss": 1.1867,
+      "epoch": 0.73,
+      "grad_norm": 17.01755142211914,
+      "learning_rate": 1.5147889386269507e-05,
+      "loss": 2.3614,
       "step": 5800
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 8.156031608581543,
-      "learning_rate": 8.373258494537437e-06,
-      "loss": 0.9248,
+      "epoch": 0.73,
+      "grad_norm": 12.760758399963379,
+      "learning_rate": 1.514705267121282e-05,
+      "loss": 2.2628,
       "step": 5801
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 20.510618209838867,
-      "learning_rate": 8.371253883933047e-06,
-      "loss": 1.6939,
+      "epoch": 0.73,
+      "grad_norm": 27.24703598022461,
+      "learning_rate": 1.5146215956156133e-05,
+      "loss": 1.8465,
       "step": 5802
     },
     {
-      "epoch": 1.74,
-      "grad_norm": 39.78818130493164,
-      "learning_rate": 8.369249273328657e-06,
-      "loss": 1.3656,
+      "epoch": 0.73,
+      "grad_norm": 11.42320728302002,
+      "learning_rate": 1.5145379241099444e-05,
+      "loss": 1.4728,
       "step": 5803
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 33.16683578491211,
-      "learning_rate": 8.367244662724267e-06,
-      "loss": 1.5406,
+      "epoch": 0.73,
+      "grad_norm": 15.267542839050293,
+      "learning_rate": 1.5144542526042758e-05,
+      "loss": 1.6918,
       "step": 5804
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 21.53314971923828,
-      "learning_rate": 8.365240052119875e-06,
-      "loss": 1.0033,
+      "epoch": 0.73,
+      "grad_norm": 11.409517288208008,
+      "learning_rate": 1.5143705810986068e-05,
+      "loss": 1.9686,
       "step": 5805
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 24.56745719909668,
-      "learning_rate": 8.363235441515486e-06,
-      "loss": 1.7757,
+      "epoch": 0.73,
+      "grad_norm": 23.77625274658203,
+      "learning_rate": 1.5142869095929382e-05,
+      "loss": 1.0783,
       "step": 5806
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 15.24342155456543,
-      "learning_rate": 8.361230830911096e-06,
-      "loss": 1.292,
+      "epoch": 0.73,
+      "grad_norm": 20.124948501586914,
+      "learning_rate": 1.5142032380872696e-05,
+      "loss": 0.981,
       "step": 5807
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 27.435335159301758,
-      "learning_rate": 8.359226220306706e-06,
-      "loss": 1.9828,
+      "epoch": 0.73,
+      "grad_norm": 12.48090934753418,
+      "learning_rate": 1.514119566581601e-05,
+      "loss": 1.0114,
       "step": 5808
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 20.47377586364746,
-      "learning_rate": 8.357221609702316e-06,
-      "loss": 1.9325,
+      "epoch": 0.73,
+      "grad_norm": 11.107354164123535,
+      "learning_rate": 1.514035895075932e-05,
+      "loss": 0.9579,
       "step": 5809
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 6.70631742477417,
-      "learning_rate": 8.355216999097926e-06,
-      "loss": 1.1601,
+      "epoch": 0.73,
+      "grad_norm": 10.767329216003418,
+      "learning_rate": 1.5139522235702633e-05,
+      "loss": 2.4955,
       "step": 5810
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 13.876242637634277,
-      "learning_rate": 8.353212388493536e-06,
-      "loss": 1.3469,
+      "epoch": 0.73,
+      "grad_norm": 8.370363235473633,
+      "learning_rate": 1.5138685520645945e-05,
+      "loss": 2.1176,
       "step": 5811
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 22.673198699951172,
-      "learning_rate": 8.351207777889146e-06,
-      "loss": 2.9192,
+      "epoch": 0.73,
+      "grad_norm": 17.333980560302734,
+      "learning_rate": 1.5137848805589257e-05,
+      "loss": 1.7624,
       "step": 5812
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 9.959436416625977,
-      "learning_rate": 8.349203167284754e-06,
-      "loss": 1.4124,
+      "epoch": 0.73,
+      "grad_norm": 13.138062477111816,
+      "learning_rate": 1.513701209053257e-05,
+      "loss": 1.5703,
       "step": 5813
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 30.46356773376465,
-      "learning_rate": 8.347198556680366e-06,
-      "loss": 2.1221,
+      "epoch": 0.73,
+      "grad_norm": 9.520326614379883,
+      "learning_rate": 1.5136175375475883e-05,
+      "loss": 0.9554,
       "step": 5814
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 12.394341468811035,
-      "learning_rate": 8.345193946075976e-06,
-      "loss": 1.2795,
+      "epoch": 0.73,
+      "grad_norm": 21.114534378051758,
+      "learning_rate": 1.5135338660419196e-05,
+      "loss": 2.0713,
       "step": 5815
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 71.86808013916016,
-      "learning_rate": 8.343189335471585e-06,
-      "loss": 3.5364,
+      "epoch": 0.73,
+      "grad_norm": 10.720865249633789,
+      "learning_rate": 1.5134501945362507e-05,
+      "loss": 1.742,
       "step": 5816
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 13.528796195983887,
-      "learning_rate": 8.341184724867195e-06,
-      "loss": 1.1612,
+      "epoch": 0.73,
+      "grad_norm": 27.921756744384766,
+      "learning_rate": 1.513366523030582e-05,
+      "loss": 1.9351,
       "step": 5817
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 38.51426696777344,
-      "learning_rate": 8.339180114262805e-06,
-      "loss": 1.6762,
+      "epoch": 0.73,
+      "grad_norm": 27.85314178466797,
+      "learning_rate": 1.5132828515249134e-05,
+      "loss": 1.4395,
       "step": 5818
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 15.68415355682373,
-      "learning_rate": 8.337175503658415e-06,
-      "loss": 2.1607,
+      "epoch": 0.73,
+      "grad_norm": 10.19604778289795,
+      "learning_rate": 1.5131991800192444e-05,
+      "loss": 0.805,
       "step": 5819
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 11.136553764343262,
-      "learning_rate": 8.335170893054025e-06,
-      "loss": 0.7781,
+      "epoch": 0.73,
+      "grad_norm": 71.00228118896484,
+      "learning_rate": 1.5131155085135758e-05,
+      "loss": 3.1386,
       "step": 5820
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 15.836930274963379,
-      "learning_rate": 8.333166282449635e-06,
-      "loss": 1.9072,
+      "epoch": 0.73,
+      "grad_norm": 19.628244400024414,
+      "learning_rate": 1.5130318370079072e-05,
+      "loss": 1.9822,
       "step": 5821
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 27.627634048461914,
-      "learning_rate": 8.331161671845245e-06,
-      "loss": 1.4755,
+      "epoch": 0.73,
+      "grad_norm": 15.459436416625977,
+      "learning_rate": 1.5129481655022385e-05,
+      "loss": 1.6848,
       "step": 5822
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 11.358156204223633,
-      "learning_rate": 8.329157061240855e-06,
-      "loss": 1.3935,
+      "epoch": 0.73,
+      "grad_norm": 7.346459865570068,
+      "learning_rate": 1.5128644939965695e-05,
+      "loss": 2.9543,
       "step": 5823
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 11.681863784790039,
-      "learning_rate": 8.327152450636464e-06,
-      "loss": 1.2521,
+      "epoch": 0.73,
+      "grad_norm": 35.106449127197266,
+      "learning_rate": 1.5127808224909009e-05,
+      "loss": 2.701,
       "step": 5824
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 37.17042541503906,
-      "learning_rate": 8.325147840032074e-06,
-      "loss": 1.7879,
+      "epoch": 0.73,
+      "grad_norm": 67.77783966064453,
+      "learning_rate": 1.5126971509852321e-05,
+      "loss": 2.7705,
       "step": 5825
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 17.618371963500977,
-      "learning_rate": 8.323143229427686e-06,
-      "loss": 1.283,
+      "epoch": 0.73,
+      "grad_norm": 9.533352851867676,
+      "learning_rate": 1.5126134794795633e-05,
+      "loss": 2.0982,
       "step": 5826
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 13.247238159179688,
-      "learning_rate": 8.321138618823294e-06,
-      "loss": 0.9504,
+      "epoch": 0.73,
+      "grad_norm": 12.30368423461914,
+      "learning_rate": 1.5125298079738947e-05,
+      "loss": 1.8838,
       "step": 5827
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 22.750028610229492,
-      "learning_rate": 8.319134008218904e-06,
-      "loss": 1.7133,
+      "epoch": 0.73,
+      "grad_norm": 7.546562194824219,
+      "learning_rate": 1.5124461364682259e-05,
+      "loss": 1.8823,
       "step": 5828
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 23.75035285949707,
-      "learning_rate": 8.317129397614514e-06,
-      "loss": 1.4273,
+      "epoch": 0.73,
+      "grad_norm": 18.230085372924805,
+      "learning_rate": 1.5123624649625572e-05,
+      "loss": 1.4497,
       "step": 5829
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 17.772930145263672,
-      "learning_rate": 8.315124787010124e-06,
-      "loss": 1.4838,
+      "epoch": 0.73,
+      "grad_norm": 11.704885482788086,
+      "learning_rate": 1.5122787934568883e-05,
+      "loss": 0.8476,
       "step": 5830
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 14.174525260925293,
-      "learning_rate": 8.313120176405733e-06,
-      "loss": 1.5128,
+      "epoch": 0.73,
+      "grad_norm": 13.442259788513184,
+      "learning_rate": 1.5121951219512196e-05,
+      "loss": 2.872,
       "step": 5831
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 59.5333366394043,
-      "learning_rate": 8.311115565801344e-06,
-      "loss": 1.0576,
+      "epoch": 0.73,
+      "grad_norm": 14.087244033813477,
+      "learning_rate": 1.512111450445551e-05,
+      "loss": 2.4998,
       "step": 5832
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 20.53312873840332,
-      "learning_rate": 8.309110955196954e-06,
-      "loss": 1.0399,
+      "epoch": 0.73,
+      "grad_norm": 2.390308141708374,
+      "learning_rate": 1.512027778939882e-05,
+      "loss": 0.0709,
       "step": 5833
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 31.05029296875,
-      "learning_rate": 8.307106344592565e-06,
-      "loss": 1.5115,
+      "epoch": 0.73,
+      "grad_norm": 41.68463897705078,
+      "learning_rate": 1.5119441074342134e-05,
+      "loss": 3.2485,
       "step": 5834
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 12.972631454467773,
-      "learning_rate": 8.305101733988173e-06,
-      "loss": 1.2375,
+      "epoch": 0.73,
+      "grad_norm": 16.793846130371094,
+      "learning_rate": 1.5118604359285447e-05,
+      "loss": 0.8453,
       "step": 5835
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 13.963759422302246,
-      "learning_rate": 8.303097123383783e-06,
-      "loss": 1.5626,
+      "epoch": 0.73,
+      "grad_norm": 5.7606611251831055,
+      "learning_rate": 1.5117767644228761e-05,
+      "loss": 1.9174,
       "step": 5836
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 11.903511047363281,
-      "learning_rate": 8.301092512779393e-06,
-      "loss": 1.4446,
+      "epoch": 0.73,
+      "grad_norm": 10.618186950683594,
+      "learning_rate": 1.5116930929172071e-05,
+      "loss": 1.0161,
       "step": 5837
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 18.711076736450195,
-      "learning_rate": 8.299087902175003e-06,
-      "loss": 1.3406,
+      "epoch": 0.73,
+      "grad_norm": 21.706491470336914,
+      "learning_rate": 1.5116094214115385e-05,
+      "loss": 2.4378,
       "step": 5838
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 84.2789535522461,
-      "learning_rate": 8.297083291570613e-06,
-      "loss": 2.1487,
+      "epoch": 0.73,
+      "grad_norm": 11.850515365600586,
+      "learning_rate": 1.5115257499058697e-05,
+      "loss": 1.4123,
       "step": 5839
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 22.546630859375,
-      "learning_rate": 8.295078680966223e-06,
-      "loss": 1.1753,
+      "epoch": 0.73,
+      "grad_norm": 19.344482421875,
+      "learning_rate": 1.5114420784002009e-05,
+      "loss": 2.6149,
       "step": 5840
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 12.994787216186523,
-      "learning_rate": 8.293074070361833e-06,
-      "loss": 1.6728,
+      "epoch": 0.73,
+      "grad_norm": 35.159759521484375,
+      "learning_rate": 1.5113584068945322e-05,
+      "loss": 1.4108,
       "step": 5841
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 8.531490325927734,
-      "learning_rate": 8.291069459757442e-06,
-      "loss": 0.8257,
+      "epoch": 0.73,
+      "grad_norm": 14.665122985839844,
+      "learning_rate": 1.5112747353888634e-05,
+      "loss": 2.0348,
       "step": 5842
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 21.086061477661133,
-      "learning_rate": 8.289064849153052e-06,
-      "loss": 1.1679,
+      "epoch": 0.73,
+      "grad_norm": 12.41344165802002,
+      "learning_rate": 1.5111910638831948e-05,
+      "loss": 2.1254,
       "step": 5843
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 23.762374877929688,
-      "learning_rate": 8.287060238548664e-06,
-      "loss": 3.049,
+      "epoch": 0.73,
+      "grad_norm": 42.95769119262695,
+      "learning_rate": 1.5111073923775258e-05,
+      "loss": 3.2697,
       "step": 5844
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 18.857986450195312,
-      "learning_rate": 8.285055627944272e-06,
-      "loss": 1.5555,
+      "epoch": 0.73,
+      "grad_norm": 25.262025833129883,
+      "learning_rate": 1.5110237208718572e-05,
+      "loss": 1.1157,
       "step": 5845
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 11.890861511230469,
-      "learning_rate": 8.283051017339882e-06,
-      "loss": 1.1098,
+      "epoch": 0.73,
+      "grad_norm": 19.907249450683594,
+      "learning_rate": 1.5109400493661886e-05,
+      "loss": 2.3685,
       "step": 5846
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 8.958333969116211,
-      "learning_rate": 8.281046406735492e-06,
-      "loss": 0.8418,
+      "epoch": 0.73,
+      "grad_norm": 16.224332809448242,
+      "learning_rate": 1.5108563778605196e-05,
+      "loss": 2.0695,
       "step": 5847
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 15.934253692626953,
-      "learning_rate": 8.279041796131102e-06,
-      "loss": 1.3008,
+      "epoch": 0.73,
+      "grad_norm": 9.057453155517578,
+      "learning_rate": 1.510772706354851e-05,
+      "loss": 2.1087,
       "step": 5848
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 21.336219787597656,
-      "learning_rate": 8.277037185526712e-06,
-      "loss": 1.1645,
+      "epoch": 0.73,
+      "grad_norm": 33.40324401855469,
+      "learning_rate": 1.5106890348491823e-05,
+      "loss": 0.9799,
       "step": 5849
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 12.621298789978027,
-      "learning_rate": 8.27503257492232e-06,
-      "loss": 2.0196,
+      "epoch": 0.73,
+      "grad_norm": 8.36249828338623,
+      "learning_rate": 1.5106053633435135e-05,
+      "loss": 0.6396,
       "step": 5850
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 13.761375427246094,
-      "learning_rate": 8.273027964317933e-06,
-      "loss": 2.1065,
+      "epoch": 0.73,
+      "grad_norm": 13.222352027893066,
+      "learning_rate": 1.5105216918378447e-05,
+      "loss": 2.3163,
       "step": 5851
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 35.57234191894531,
-      "learning_rate": 8.271023353713543e-06,
-      "loss": 1.1288,
+      "epoch": 0.73,
+      "grad_norm": 11.83804988861084,
+      "learning_rate": 1.510438020332176e-05,
+      "loss": 1.0425,
       "step": 5852
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 10.00034236907959,
-      "learning_rate": 8.269018743109151e-06,
-      "loss": 1.3954,
+      "epoch": 0.73,
+      "grad_norm": 8.717399597167969,
+      "learning_rate": 1.5103543488265073e-05,
+      "loss": 0.9513,
       "step": 5853
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 37.05139923095703,
-      "learning_rate": 8.267014132504761e-06,
-      "loss": 1.4784,
+      "epoch": 0.73,
+      "grad_norm": 16.511960983276367,
+      "learning_rate": 1.5102706773208385e-05,
+      "loss": 1.6451,
       "step": 5854
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 22.268341064453125,
-      "learning_rate": 8.265009521900371e-06,
-      "loss": 1.3877,
+      "epoch": 0.73,
+      "grad_norm": 20.263565063476562,
+      "learning_rate": 1.5101870058151698e-05,
+      "loss": 0.8851,
       "step": 5855
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 13.753389358520508,
-      "learning_rate": 8.263004911295981e-06,
-      "loss": 1.5012,
+      "epoch": 0.73,
+      "grad_norm": 11.619148254394531,
+      "learning_rate": 1.510103334309501e-05,
+      "loss": 3.0982,
       "step": 5856
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 18.250154495239258,
-      "learning_rate": 8.261000300691591e-06,
-      "loss": 1.7476,
+      "epoch": 0.74,
+      "grad_norm": 17.539045333862305,
+      "learning_rate": 1.5100196628038324e-05,
+      "loss": 1.0885,
       "step": 5857
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 14.758729934692383,
-      "learning_rate": 8.258995690087201e-06,
-      "loss": 1.5117,
+      "epoch": 0.74,
+      "grad_norm": 21.50510597229004,
+      "learning_rate": 1.5099359912981634e-05,
+      "loss": 1.0739,
       "step": 5858
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 20.81943130493164,
-      "learning_rate": 8.256991079482812e-06,
-      "loss": 1.363,
+      "epoch": 0.74,
+      "grad_norm": 33.4230842590332,
+      "learning_rate": 1.5098523197924948e-05,
+      "loss": 2.9656,
       "step": 5859
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 26.64165496826172,
-      "learning_rate": 8.254986468878422e-06,
-      "loss": 1.7073,
+      "epoch": 0.74,
+      "grad_norm": 15.180930137634277,
+      "learning_rate": 1.5097686482868261e-05,
+      "loss": 1.5781,
       "step": 5860
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 14.987895965576172,
-      "learning_rate": 8.25298185827403e-06,
-      "loss": 0.7878,
+      "epoch": 0.74,
+      "grad_norm": 13.385412216186523,
+      "learning_rate": 1.5096849767811572e-05,
+      "loss": 2.7543,
       "step": 5861
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 17.91753387451172,
-      "learning_rate": 8.25097724766964e-06,
-      "loss": 1.4102,
+      "epoch": 0.74,
+      "grad_norm": 18.107589721679688,
+      "learning_rate": 1.5096013052754885e-05,
+      "loss": 2.256,
       "step": 5862
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 6.570464134216309,
-      "learning_rate": 8.248972637065252e-06,
-      "loss": 0.7017,
+      "epoch": 0.74,
+      "grad_norm": 25.71607780456543,
+      "learning_rate": 1.5095176337698199e-05,
+      "loss": 2.8298,
       "step": 5863
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 21.44278335571289,
-      "learning_rate": 8.24696802646086e-06,
-      "loss": 1.4487,
+      "epoch": 0.74,
+      "grad_norm": 92.62855529785156,
+      "learning_rate": 1.5094339622641511e-05,
+      "loss": 0.4783,
       "step": 5864
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 15.875109672546387,
-      "learning_rate": 8.24496341585647e-06,
-      "loss": 1.178,
+      "epoch": 0.74,
+      "grad_norm": 34.5122184753418,
+      "learning_rate": 1.5093502907584823e-05,
+      "loss": 1.4871,
       "step": 5865
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 36.79011154174805,
-      "learning_rate": 8.24295880525208e-06,
-      "loss": 1.9915,
+      "epoch": 0.74,
+      "grad_norm": 9.95934772491455,
+      "learning_rate": 1.5092666192528137e-05,
+      "loss": 1.7922,
       "step": 5866
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 21.93402862548828,
-      "learning_rate": 8.24095419464769e-06,
-      "loss": 1.4122,
+      "epoch": 0.74,
+      "grad_norm": 25.549142837524414,
+      "learning_rate": 1.5091829477471449e-05,
+      "loss": 1.5132,
       "step": 5867
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 23.80208969116211,
-      "learning_rate": 8.2389495840433e-06,
-      "loss": 2.1488,
+      "epoch": 0.74,
+      "grad_norm": 5.921211242675781,
+      "learning_rate": 1.509099276241476e-05,
+      "loss": 0.9918,
       "step": 5868
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 10.25802993774414,
-      "learning_rate": 8.23694497343891e-06,
-      "loss": 1.7158,
+      "epoch": 0.74,
+      "grad_norm": 47.82905197143555,
+      "learning_rate": 1.5090156047358072e-05,
+      "loss": 2.308,
       "step": 5869
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 38.02273178100586,
-      "learning_rate": 8.23494036283452e-06,
-      "loss": 1.6496,
+      "epoch": 0.74,
+      "grad_norm": 6.752317428588867,
+      "learning_rate": 1.5089319332301386e-05,
+      "loss": 1.8953,
       "step": 5870
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 27.261823654174805,
-      "learning_rate": 8.232935752230131e-06,
-      "loss": 2.1812,
+      "epoch": 0.74,
+      "grad_norm": 10.293950080871582,
+      "learning_rate": 1.50884826172447e-05,
+      "loss": 1.0804,
       "step": 5871
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 10.888236999511719,
-      "learning_rate": 8.23093114162574e-06,
-      "loss": 1.3974,
+      "epoch": 0.74,
+      "grad_norm": 16.059457778930664,
+      "learning_rate": 1.508764590218801e-05,
+      "loss": 3.0737,
       "step": 5872
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 33.661293029785156,
-      "learning_rate": 8.22892653102135e-06,
-      "loss": 2.1975,
+      "epoch": 0.74,
+      "grad_norm": 13.257723808288574,
+      "learning_rate": 1.5086809187131324e-05,
+      "loss": 1.2577,
       "step": 5873
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 38.529293060302734,
-      "learning_rate": 8.22692192041696e-06,
-      "loss": 1.8204,
+      "epoch": 0.74,
+      "grad_norm": 7.22626256942749,
+      "learning_rate": 1.5085972472074637e-05,
+      "loss": 1.2732,
       "step": 5874
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 28.78758430480957,
-      "learning_rate": 8.22491730981257e-06,
-      "loss": 2.9932,
+      "epoch": 0.74,
+      "grad_norm": 16.098403930664062,
+      "learning_rate": 1.5085135757017948e-05,
+      "loss": 2.8811,
       "step": 5875
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 10.32004165649414,
-      "learning_rate": 8.22291269920818e-06,
-      "loss": 2.0646,
+      "epoch": 0.74,
+      "grad_norm": 11.780889511108398,
+      "learning_rate": 1.5084299041961261e-05,
+      "loss": 0.93,
       "step": 5876
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 18.924707412719727,
-      "learning_rate": 8.22090808860379e-06,
-      "loss": 2.3191,
+      "epoch": 0.74,
+      "grad_norm": 14.249544143676758,
+      "learning_rate": 1.5083462326904575e-05,
+      "loss": 0.5134,
       "step": 5877
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 14.681024551391602,
-      "learning_rate": 8.2189034779994e-06,
-      "loss": 1.1271,
+      "epoch": 0.74,
+      "grad_norm": 12.100791931152344,
+      "learning_rate": 1.5082625611847887e-05,
+      "loss": 1.4711,
       "step": 5878
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 13.027987480163574,
-      "learning_rate": 8.216898867395008e-06,
-      "loss": 1.6397,
+      "epoch": 0.74,
+      "grad_norm": 7.681851387023926,
+      "learning_rate": 1.5081788896791199e-05,
+      "loss": 0.7961,
       "step": 5879
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 37.9376335144043,
-      "learning_rate": 8.214894256790618e-06,
-      "loss": 1.7215,
-      "step": 5880
-    },
-    {
-      "epoch": 1.77,
-      "eval_loss": 0.19819287955760956,
-      "eval_runtime": 43.4174,
-      "eval_samples_per_second": 34.065,
-      "eval_steps_per_second": 34.065,
+      "epoch": 0.74,
+      "grad_norm": 16.879785537719727,
+      "learning_rate": 1.5080952181734512e-05,
+      "loss": 2.2265,
       "step": 5880
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 37.3797492980957,
-      "learning_rate": 8.21288964618623e-06,
-      "loss": 2.3832,
+      "epoch": 0.74,
+      "grad_norm": 4.811009883880615,
+      "learning_rate": 1.5080115466677824e-05,
+      "loss": 1.4758,
       "step": 5881
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 20.478801727294922,
-      "learning_rate": 8.210885035581838e-06,
-      "loss": 1.8481,
+      "epoch": 0.74,
+      "grad_norm": 8.104880332946777,
+      "learning_rate": 1.5079278751621136e-05,
+      "loss": 0.3154,
       "step": 5882
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 22.586971282958984,
-      "learning_rate": 8.208880424977449e-06,
-      "loss": 1.8847,
+      "epoch": 0.74,
+      "grad_norm": 12.702399253845215,
+      "learning_rate": 1.5078442036564448e-05,
+      "loss": 1.6427,
       "step": 5883
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 25.06772232055664,
-      "learning_rate": 8.206875814373059e-06,
-      "loss": 2.011,
+      "epoch": 0.74,
+      "grad_norm": 12.479487419128418,
+      "learning_rate": 1.5077605321507762e-05,
+      "loss": 1.8192,
       "step": 5884
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 13.59852123260498,
-      "learning_rate": 8.204871203768669e-06,
-      "loss": 1.5316,
+      "epoch": 0.74,
+      "grad_norm": 14.454097747802734,
+      "learning_rate": 1.5076768606451076e-05,
+      "loss": 0.5032,
       "step": 5885
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 15.501117706298828,
-      "learning_rate": 8.202866593164279e-06,
-      "loss": 1.5476,
+      "epoch": 0.74,
+      "grad_norm": 16.55340003967285,
+      "learning_rate": 1.5075931891394386e-05,
+      "loss": 1.97,
       "step": 5886
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 15.47221851348877,
-      "learning_rate": 8.200861982559889e-06,
-      "loss": 1.5741,
+      "epoch": 0.74,
+      "grad_norm": 7.660234451293945,
+      "learning_rate": 1.50750951763377e-05,
+      "loss": 1.47,
       "step": 5887
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 25.984909057617188,
-      "learning_rate": 8.198857371955499e-06,
-      "loss": 1.8725,
+      "epoch": 0.74,
+      "grad_norm": 20.445825576782227,
+      "learning_rate": 1.5074258461281013e-05,
+      "loss": 1.4376,
       "step": 5888
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 23.49461555480957,
-      "learning_rate": 8.196852761351109e-06,
-      "loss": 2.3657,
+      "epoch": 0.74,
+      "grad_norm": 18.65691375732422,
+      "learning_rate": 1.5073421746224323e-05,
+      "loss": 2.0051,
       "step": 5889
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 12.506348609924316,
-      "learning_rate": 8.194848150746717e-06,
-      "loss": 1.4631,
+      "epoch": 0.74,
+      "grad_norm": 10.39387321472168,
+      "learning_rate": 1.5072585031167637e-05,
+      "loss": 0.6718,
       "step": 5890
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 22.092987060546875,
-      "learning_rate": 8.192843540142327e-06,
-      "loss": 1.3425,
+      "epoch": 0.74,
+      "grad_norm": 9.236499786376953,
+      "learning_rate": 1.507174831611095e-05,
+      "loss": 3.0904,
       "step": 5891
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 24.98388671875,
-      "learning_rate": 8.190838929537938e-06,
-      "loss": 2.064,
+      "epoch": 0.74,
+      "grad_norm": 18.530906677246094,
+      "learning_rate": 1.5070911601054261e-05,
+      "loss": 3.6638,
       "step": 5892
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 15.064332962036133,
-      "learning_rate": 8.188834318933548e-06,
-      "loss": 1.257,
+      "epoch": 0.74,
+      "grad_norm": 7.476229667663574,
+      "learning_rate": 1.5070074885997575e-05,
+      "loss": 2.8985,
       "step": 5893
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 54.17677307128906,
-      "learning_rate": 8.186829708329158e-06,
-      "loss": 2.009,
+      "epoch": 0.74,
+      "grad_norm": 7.896212100982666,
+      "learning_rate": 1.5069238170940888e-05,
+      "loss": 0.7143,
       "step": 5894
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 18.08331298828125,
-      "learning_rate": 8.184825097724768e-06,
-      "loss": 1.2969,
+      "epoch": 0.74,
+      "grad_norm": 12.695160865783691,
+      "learning_rate": 1.50684014558842e-05,
+      "loss": 2.2646,
       "step": 5895
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 32.95471954345703,
-      "learning_rate": 8.182820487120378e-06,
-      "loss": 1.2109,
+      "epoch": 0.74,
+      "grad_norm": 9.66424560546875,
+      "learning_rate": 1.5067564740827512e-05,
+      "loss": 1.8797,
       "step": 5896
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 11.22049331665039,
-      "learning_rate": 8.180815876515988e-06,
-      "loss": 1.2192,
+      "epoch": 0.74,
+      "grad_norm": 15.434563636779785,
+      "learning_rate": 1.5066728025770824e-05,
+      "loss": 2.3841,
       "step": 5897
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 14.255977630615234,
-      "learning_rate": 8.178811265911596e-06,
-      "loss": 1.887,
+      "epoch": 0.74,
+      "grad_norm": 16.04581642150879,
+      "learning_rate": 1.5065891310714138e-05,
+      "loss": 2.2305,
       "step": 5898
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 12.499604225158691,
-      "learning_rate": 8.176806655307206e-06,
-      "loss": 1.3196,
+      "epoch": 0.74,
+      "grad_norm": 14.663934707641602,
+      "learning_rate": 1.506505459565745e-05,
+      "loss": 2.8839,
       "step": 5899
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 15.631196975708008,
-      "learning_rate": 8.174802044702818e-06,
-      "loss": 1.3063,
+      "epoch": 0.74,
+      "grad_norm": 10.927321434020996,
+      "learning_rate": 1.5064217880600762e-05,
+      "loss": 0.4926,
       "step": 5900
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 13.534256935119629,
-      "learning_rate": 8.172797434098427e-06,
-      "loss": 1.6275,
+      "epoch": 0.74,
+      "grad_norm": 9.020415306091309,
+      "learning_rate": 1.5063381165544075e-05,
+      "loss": 2.0736,
       "step": 5901
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 10.233455657958984,
-      "learning_rate": 8.170792823494037e-06,
-      "loss": 1.1873,
+      "epoch": 0.74,
+      "grad_norm": 8.869327545166016,
+      "learning_rate": 1.5062544450487389e-05,
+      "loss": 1.334,
       "step": 5902
     },
     {
-      "epoch": 1.77,
-      "grad_norm": 11.301042556762695,
-      "learning_rate": 8.168788212889647e-06,
-      "loss": 1.2987,
+      "epoch": 0.74,
+      "grad_norm": 12.613329887390137,
+      "learning_rate": 1.50617077354307e-05,
+      "loss": 1.3902,
       "step": 5903
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 46.812076568603516,
-      "learning_rate": 8.166783602285257e-06,
-      "loss": 1.7821,
+      "epoch": 0.74,
+      "grad_norm": 3.1661107540130615,
+      "learning_rate": 1.5060871020374013e-05,
+      "loss": 0.0721,
       "step": 5904
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 34.08024597167969,
-      "learning_rate": 8.164778991680867e-06,
-      "loss": 1.3355,
+      "epoch": 0.74,
+      "grad_norm": 4.516031265258789,
+      "learning_rate": 1.5060034305317327e-05,
+      "loss": 0.3199,
       "step": 5905
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 12.29112720489502,
-      "learning_rate": 8.162774381076477e-06,
-      "loss": 1.3053,
+      "epoch": 0.74,
+      "grad_norm": 35.51274490356445,
+      "learning_rate": 1.5059197590260637e-05,
+      "loss": 1.3607,
       "step": 5906
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 16.215862274169922,
-      "learning_rate": 8.160769770472087e-06,
-      "loss": 1.6378,
+      "epoch": 0.74,
+      "grad_norm": 12.607372283935547,
+      "learning_rate": 1.505836087520395e-05,
+      "loss": 1.5191,
       "step": 5907
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 33.6343994140625,
-      "learning_rate": 8.158765159867697e-06,
-      "loss": 1.9444,
+      "epoch": 0.74,
+      "grad_norm": 16.76181411743164,
+      "learning_rate": 1.5057524160147264e-05,
+      "loss": 2.4684,
       "step": 5908
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 27.786399841308594,
-      "learning_rate": 8.156760549263306e-06,
-      "loss": 1.1768,
+      "epoch": 0.74,
+      "grad_norm": 17.170143127441406,
+      "learning_rate": 1.5056687445090576e-05,
+      "loss": 1.527,
       "step": 5909
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 22.714481353759766,
-      "learning_rate": 8.154755938658916e-06,
-      "loss": 1.5109,
+      "epoch": 0.74,
+      "grad_norm": 16.721603393554688,
+      "learning_rate": 1.5055850730033888e-05,
+      "loss": 1.5672,
       "step": 5910
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 14.203557968139648,
-      "learning_rate": 8.152751328054526e-06,
-      "loss": 1.2664,
+      "epoch": 0.74,
+      "grad_norm": 18.294769287109375,
+      "learning_rate": 1.50550140149772e-05,
+      "loss": 1.8744,
       "step": 5911
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 25.13650894165039,
-      "learning_rate": 8.150746717450136e-06,
-      "loss": 1.6833,
+      "epoch": 0.74,
+      "grad_norm": 12.959049224853516,
+      "learning_rate": 1.5054177299920514e-05,
+      "loss": 3.0204,
       "step": 5912
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 8.750547409057617,
-      "learning_rate": 8.148742106845746e-06,
-      "loss": 0.7758,
+      "epoch": 0.74,
+      "grad_norm": 35.87045669555664,
+      "learning_rate": 1.5053340584863826e-05,
+      "loss": 2.0355,
       "step": 5913
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 13.47951889038086,
-      "learning_rate": 8.146737496241356e-06,
-      "loss": 0.9706,
+      "epoch": 0.74,
+      "grad_norm": 27.6667423248291,
+      "learning_rate": 1.5052503869807138e-05,
+      "loss": 3.6111,
       "step": 5914
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 25.456212997436523,
-      "learning_rate": 8.144732885636966e-06,
-      "loss": 2.0665,
+      "epoch": 0.74,
+      "grad_norm": 10.446527481079102,
+      "learning_rate": 1.5051667154750451e-05,
+      "loss": 1.8251,
       "step": 5915
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 26.591684341430664,
-      "learning_rate": 8.142728275032575e-06,
-      "loss": 2.694,
+      "epoch": 0.74,
+      "grad_norm": 23.50820541381836,
+      "learning_rate": 1.5050830439693765e-05,
+      "loss": 2.3433,
       "step": 5916
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 141.93113708496094,
-      "learning_rate": 8.140723664428185e-06,
-      "loss": 2.0094,
+      "epoch": 0.74,
+      "grad_norm": 12.742378234863281,
+      "learning_rate": 1.5049993724637075e-05,
+      "loss": 1.8785,
       "step": 5917
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 14.334453582763672,
-      "learning_rate": 8.138719053823796e-06,
-      "loss": 1.156,
+      "epoch": 0.74,
+      "grad_norm": 9.84451675415039,
+      "learning_rate": 1.5049157009580389e-05,
+      "loss": 0.8814,
       "step": 5918
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 31.279136657714844,
-      "learning_rate": 8.136714443219406e-06,
-      "loss": 1.2964,
+      "epoch": 0.74,
+      "grad_norm": 16.8998966217041,
+      "learning_rate": 1.5048320294523702e-05,
+      "loss": 1.2168,
       "step": 5919
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 13.835858345031738,
-      "learning_rate": 8.134709832615015e-06,
-      "loss": 0.992,
+      "epoch": 0.74,
+      "grad_norm": 43.36582946777344,
+      "learning_rate": 1.5047483579467013e-05,
+      "loss": 1.2332,
       "step": 5920
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 20.324710845947266,
-      "learning_rate": 8.132705222010625e-06,
-      "loss": 2.0425,
+      "epoch": 0.74,
+      "grad_norm": 10.693442344665527,
+      "learning_rate": 1.5046646864410326e-05,
+      "loss": 1.291,
       "step": 5921
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 9.181856155395508,
-      "learning_rate": 8.130700611406235e-06,
-      "loss": 1.2267,
+      "epoch": 0.74,
+      "grad_norm": 11.892716407775879,
+      "learning_rate": 1.5045810149353638e-05,
+      "loss": 1.3415,
       "step": 5922
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 12.382397651672363,
-      "learning_rate": 8.128696000801845e-06,
-      "loss": 1.4096,
+      "epoch": 0.74,
+      "grad_norm": 9.612030029296875,
+      "learning_rate": 1.5044973434296952e-05,
+      "loss": 2.065,
       "step": 5923
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 19.25230598449707,
-      "learning_rate": 8.126691390197455e-06,
-      "loss": 1.86,
+      "epoch": 0.74,
+      "grad_norm": 462.9150695800781,
+      "learning_rate": 1.5044136719240264e-05,
+      "loss": 3.5542,
       "step": 5924
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 9.224576950073242,
-      "learning_rate": 8.124686779593065e-06,
-      "loss": 1.291,
+      "epoch": 0.74,
+      "grad_norm": 34.43170166015625,
+      "learning_rate": 1.5043300004183576e-05,
+      "loss": 2.6402,
       "step": 5925
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 26.03525161743164,
-      "learning_rate": 8.122682168988675e-06,
-      "loss": 2.0772,
+      "epoch": 0.74,
+      "grad_norm": 8.972990989685059,
+      "learning_rate": 1.504246328912689e-05,
+      "loss": 1.6312,
       "step": 5926
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 12.974356651306152,
-      "learning_rate": 8.120677558384284e-06,
-      "loss": 1.524,
+      "epoch": 0.74,
+      "grad_norm": 19.01601791381836,
+      "learning_rate": 1.50416265740702e-05,
+      "loss": 3.8481,
       "step": 5927
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 9.397273063659668,
-      "learning_rate": 8.118672947779894e-06,
-      "loss": 1.2373,
+      "epoch": 0.74,
+      "grad_norm": 7.473775386810303,
+      "learning_rate": 1.5040789859013513e-05,
+      "loss": 0.6885,
       "step": 5928
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 12.543724060058594,
-      "learning_rate": 8.116668337175504e-06,
-      "loss": 1.4601,
+      "epoch": 0.74,
+      "grad_norm": 5.208780765533447,
+      "learning_rate": 1.5039953143956827e-05,
+      "loss": 0.7665,
       "step": 5929
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 10.495887756347656,
-      "learning_rate": 8.114663726571114e-06,
-      "loss": 1.3694,
+      "epoch": 0.74,
+      "grad_norm": 23.150463104248047,
+      "learning_rate": 1.503911642890014e-05,
+      "loss": 1.9791,
       "step": 5930
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 12.478328704833984,
-      "learning_rate": 8.112659115966724e-06,
-      "loss": 1.8051,
+      "epoch": 0.74,
+      "grad_norm": 14.49713134765625,
+      "learning_rate": 1.5038279713843451e-05,
+      "loss": 1.5597,
       "step": 5931
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 18.770479202270508,
-      "learning_rate": 8.110654505362334e-06,
-      "loss": 1.3905,
+      "epoch": 0.74,
+      "grad_norm": 6.02341365814209,
+      "learning_rate": 1.5037442998786765e-05,
+      "loss": 0.8754,
       "step": 5932
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 14.52056884765625,
-      "learning_rate": 8.108649894757944e-06,
-      "loss": 1.4971,
+      "epoch": 0.74,
+      "grad_norm": 12.140684127807617,
+      "learning_rate": 1.5036606283730078e-05,
+      "loss": 1.5937,
       "step": 5933
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 25.12417221069336,
-      "learning_rate": 8.106645284153554e-06,
-      "loss": 1.9117,
+      "epoch": 0.74,
+      "grad_norm": 19.201501846313477,
+      "learning_rate": 1.5035769568673388e-05,
+      "loss": 1.6273,
       "step": 5934
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 40.989715576171875,
-      "learning_rate": 8.104640673549163e-06,
-      "loss": 2.4517,
+      "epoch": 0.74,
+      "grad_norm": 10.110654830932617,
+      "learning_rate": 1.5034932853616702e-05,
+      "loss": 1.313,
       "step": 5935
     },
     {
-      "epoch": 1.78,
-      "grad_norm": 7.442935943603516,
-      "learning_rate": 8.102636062944773e-06,
-      "loss": 1.0286,
+      "epoch": 0.74,
+      "grad_norm": 18.283098220825195,
+      "learning_rate": 1.5034096138560014e-05,
+      "loss": 1.1007,
       "step": 5936
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 13.013799667358398,
-      "learning_rate": 8.100631452340385e-06,
-      "loss": 1.1536,
+      "epoch": 0.75,
+      "grad_norm": 10.114410400390625,
+      "learning_rate": 1.5033259423503328e-05,
+      "loss": 0.6061,
       "step": 5937
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 10.014057159423828,
-      "learning_rate": 8.098626841735993e-06,
-      "loss": 1.193,
+      "epoch": 0.75,
+      "grad_norm": 3.9301300048828125,
+      "learning_rate": 1.503242270844664e-05,
+      "loss": 1.2018,
       "step": 5938
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 24.215864181518555,
-      "learning_rate": 8.096622231131603e-06,
-      "loss": 1.3133,
+      "epoch": 0.75,
+      "grad_norm": 30.078611373901367,
+      "learning_rate": 1.5031585993389952e-05,
+      "loss": 2.5315,
       "step": 5939
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 14.128569602966309,
-      "learning_rate": 8.094617620527213e-06,
-      "loss": 1.0683,
+      "epoch": 0.75,
+      "grad_norm": 2.8266096115112305,
+      "learning_rate": 1.5030749278333265e-05,
+      "loss": 0.1006,
       "step": 5940
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 18.90635108947754,
-      "learning_rate": 8.092613009922823e-06,
-      "loss": 1.3165,
+      "epoch": 0.75,
+      "grad_norm": 7.4051995277404785,
+      "learning_rate": 1.5029912563276576e-05,
+      "loss": 0.9698,
       "step": 5941
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 7.926889896392822,
-      "learning_rate": 8.090608399318433e-06,
-      "loss": 1.3542,
+      "epoch": 0.75,
+      "grad_norm": 23.7181396484375,
+      "learning_rate": 1.502907584821989e-05,
+      "loss": 3.3637,
       "step": 5942
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 15.462095260620117,
-      "learning_rate": 8.088603788714043e-06,
-      "loss": 1.7081,
+      "epoch": 0.75,
+      "grad_norm": 15.384878158569336,
+      "learning_rate": 1.5028239133163203e-05,
+      "loss": 0.8438,
       "step": 5943
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 13.095417976379395,
-      "learning_rate": 8.086599178109653e-06,
-      "loss": 1.3562,
+      "epoch": 0.75,
+      "grad_norm": 12.596027374267578,
+      "learning_rate": 1.5027402418106516e-05,
+      "loss": 1.9872,
       "step": 5944
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 16.457101821899414,
-      "learning_rate": 8.084594567505264e-06,
-      "loss": 1.1891,
+      "epoch": 0.75,
+      "grad_norm": 10.537801742553711,
+      "learning_rate": 1.5026565703049827e-05,
+      "loss": 2.608,
       "step": 5945
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 9.37852668762207,
-      "learning_rate": 8.082589956900872e-06,
-      "loss": 1.3149,
+      "epoch": 0.75,
+      "grad_norm": 13.232535362243652,
+      "learning_rate": 1.502572898799314e-05,
+      "loss": 1.8912,
       "step": 5946
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 10.058578491210938,
-      "learning_rate": 8.080585346296482e-06,
-      "loss": 1.2129,
+      "epoch": 0.75,
+      "grad_norm": 80.67562866210938,
+      "learning_rate": 1.5024892272936454e-05,
+      "loss": 1.1722,
       "step": 5947
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 15.420865058898926,
-      "learning_rate": 8.078580735692092e-06,
-      "loss": 2.2392,
+      "epoch": 0.75,
+      "grad_norm": 12.835854530334473,
+      "learning_rate": 1.5024055557879764e-05,
+      "loss": 1.1671,
       "step": 5948
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 17.6695556640625,
-      "learning_rate": 8.076576125087702e-06,
-      "loss": 1.1624,
+      "epoch": 0.75,
+      "grad_norm": 39.22283172607422,
+      "learning_rate": 1.5023218842823078e-05,
+      "loss": 1.5886,
       "step": 5949
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 12.30250072479248,
-      "learning_rate": 8.074571514483312e-06,
-      "loss": 1.4646,
+      "epoch": 0.75,
+      "grad_norm": 5.5500922203063965,
+      "learning_rate": 1.502238212776639e-05,
+      "loss": 0.7196,
       "step": 5950
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 26.107072830200195,
-      "learning_rate": 8.072566903878922e-06,
-      "loss": 1.4681,
+      "epoch": 0.75,
+      "grad_norm": 7.820392608642578,
+      "learning_rate": 1.5021545412709704e-05,
+      "loss": 0.0653,
       "step": 5951
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 16.788949966430664,
-      "learning_rate": 8.070562293274532e-06,
-      "loss": 1.3324,
+      "epoch": 0.75,
+      "grad_norm": 14.047516822814941,
+      "learning_rate": 1.5020708697653016e-05,
+      "loss": 0.8211,
       "step": 5952
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 43.306427001953125,
-      "learning_rate": 8.068557682670143e-06,
-      "loss": 1.6477,
+      "epoch": 0.75,
+      "grad_norm": 16.972585678100586,
+      "learning_rate": 1.5019871982596327e-05,
+      "loss": 2.8933,
       "step": 5953
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 14.007320404052734,
-      "learning_rate": 8.066553072065751e-06,
-      "loss": 0.9753,
+      "epoch": 0.75,
+      "grad_norm": 2.9097461700439453,
+      "learning_rate": 1.5019035267539641e-05,
+      "loss": 0.1308,
       "step": 5954
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 9.967177391052246,
-      "learning_rate": 8.064548461461363e-06,
-      "loss": 0.8905,
+      "epoch": 0.75,
+      "grad_norm": 42.2338981628418,
+      "learning_rate": 1.5018198552482951e-05,
+      "loss": 2.3033,
       "step": 5955
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 63.55514144897461,
-      "learning_rate": 8.062543850856973e-06,
-      "loss": 2.9485,
+      "epoch": 0.75,
+      "grad_norm": 4.907975673675537,
+      "learning_rate": 1.5017361837426265e-05,
+      "loss": 2.3643,
       "step": 5956
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 18.12586212158203,
-      "learning_rate": 8.060539240252581e-06,
-      "loss": 1.9415,
+      "epoch": 0.75,
+      "grad_norm": 28.78272819519043,
+      "learning_rate": 1.5016525122369579e-05,
+      "loss": 2.56,
       "step": 5957
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 9.148773193359375,
-      "learning_rate": 8.058534629648191e-06,
-      "loss": 1.476,
+      "epoch": 0.75,
+      "grad_norm": 6.502068996429443,
+      "learning_rate": 1.5015688407312892e-05,
+      "loss": 0.5838,
       "step": 5958
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 10.708128929138184,
-      "learning_rate": 8.056530019043801e-06,
-      "loss": 1.3661,
+      "epoch": 0.75,
+      "grad_norm": 24.910249710083008,
+      "learning_rate": 1.5014851692256203e-05,
+      "loss": 1.3385,
       "step": 5959
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 20.42496681213379,
-      "learning_rate": 8.054525408439411e-06,
-      "loss": 1.4746,
+      "epoch": 0.75,
+      "grad_norm": 18.505050659179688,
+      "learning_rate": 1.5014014977199516e-05,
+      "loss": 1.3337,
       "step": 5960
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 15.104199409484863,
-      "learning_rate": 8.052520797835022e-06,
-      "loss": 1.8083,
+      "epoch": 0.75,
+      "grad_norm": 13.536693572998047,
+      "learning_rate": 1.5013178262142828e-05,
+      "loss": 3.3743,
       "step": 5961
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 20.05451011657715,
-      "learning_rate": 8.050516187230632e-06,
-      "loss": 1.6234,
+      "epoch": 0.75,
+      "grad_norm": 9.997637748718262,
+      "learning_rate": 1.501234154708614e-05,
+      "loss": 1.1199,
       "step": 5962
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 23.32365608215332,
-      "learning_rate": 8.048511576626242e-06,
-      "loss": 1.5622,
+      "epoch": 0.75,
+      "grad_norm": 10.148368835449219,
+      "learning_rate": 1.5011504832029454e-05,
+      "loss": 0.7278,
       "step": 5963
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 17.466800689697266,
-      "learning_rate": 8.04650696602185e-06,
-      "loss": 0.6051,
+      "epoch": 0.75,
+      "grad_norm": 13.460878372192383,
+      "learning_rate": 1.5010668116972766e-05,
+      "loss": 2.8044,
       "step": 5964
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 26.45103645324707,
-      "learning_rate": 8.04450235541746e-06,
-      "loss": 1.5636,
+      "epoch": 0.75,
+      "grad_norm": 23.029918670654297,
+      "learning_rate": 1.500983140191608e-05,
+      "loss": 2.7466,
       "step": 5965
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 25.677671432495117,
-      "learning_rate": 8.04249774481307e-06,
-      "loss": 1.1426,
+      "epoch": 0.75,
+      "grad_norm": 19.22269058227539,
+      "learning_rate": 1.5008994686859391e-05,
+      "loss": 3.4052,
       "step": 5966
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 17.478527069091797,
-      "learning_rate": 8.04049313420868e-06,
-      "loss": 0.8924,
+      "epoch": 0.75,
+      "grad_norm": 5.88612699508667,
+      "learning_rate": 1.5008157971802703e-05,
+      "loss": 0.1444,
       "step": 5967
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 15.59758472442627,
-      "learning_rate": 8.03848852360429e-06,
-      "loss": 1.4224,
+      "epoch": 0.75,
+      "grad_norm": 24.3934268951416,
+      "learning_rate": 1.5007321256746017e-05,
+      "loss": 2.4715,
       "step": 5968
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 20.508508682250977,
-      "learning_rate": 8.0364839129999e-06,
-      "loss": 2.0134,
+      "epoch": 0.75,
+      "grad_norm": 48.39157485961914,
+      "learning_rate": 1.5006484541689327e-05,
+      "loss": 0.5484,
       "step": 5969
     },
     {
-      "epoch": 1.79,
-      "grad_norm": 14.481962203979492,
-      "learning_rate": 8.03447930239551e-06,
-      "loss": 1.176,
+      "epoch": 0.75,
+      "grad_norm": 9.383609771728516,
+      "learning_rate": 1.5005647826632641e-05,
+      "loss": 0.8422,
       "step": 5970
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 17.673538208007812,
-      "learning_rate": 8.03247469179112e-06,
-      "loss": 1.5274,
+      "epoch": 0.75,
+      "grad_norm": 10.745577812194824,
+      "learning_rate": 1.5004811111575955e-05,
+      "loss": 1.2341,
       "step": 5971
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 16.05086898803711,
-      "learning_rate": 8.030470081186729e-06,
-      "loss": 1.2769,
+      "epoch": 0.75,
+      "grad_norm": 22.75591468811035,
+      "learning_rate": 1.5003974396519268e-05,
+      "loss": 2.6592,
       "step": 5972
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 26.023082733154297,
-      "learning_rate": 8.028465470582339e-06,
-      "loss": 0.8421,
+      "epoch": 0.75,
+      "grad_norm": 17.623111724853516,
+      "learning_rate": 1.5003137681462578e-05,
+      "loss": 1.9956,
       "step": 5973
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 19.35530662536621,
-      "learning_rate": 8.026460859977951e-06,
-      "loss": 2.123,
+      "epoch": 0.75,
+      "grad_norm": 7.998566150665283,
+      "learning_rate": 1.5002300966405892e-05,
+      "loss": 0.7073,
       "step": 5974
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 45.65658950805664,
-      "learning_rate": 8.02445624937356e-06,
-      "loss": 2.4255,
+      "epoch": 0.75,
+      "grad_norm": 5.96798038482666,
+      "learning_rate": 1.5001464251349204e-05,
+      "loss": 0.475,
       "step": 5975
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 50.98988342285156,
-      "learning_rate": 8.02245163876917e-06,
-      "loss": 2.1818,
+      "epoch": 0.75,
+      "grad_norm": 7.934585094451904,
+      "learning_rate": 1.5000627536292516e-05,
+      "loss": 0.9308,
       "step": 5976
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 25.256816864013672,
-      "learning_rate": 8.02044702816478e-06,
-      "loss": 2.182,
+      "epoch": 0.75,
+      "grad_norm": 4.579667568206787,
+      "learning_rate": 1.499979082123583e-05,
+      "loss": 0.6803,
       "step": 5977
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 18.960912704467773,
-      "learning_rate": 8.01844241756039e-06,
-      "loss": 1.4124,
+      "epoch": 0.75,
+      "grad_norm": 38.206138610839844,
+      "learning_rate": 1.4998954106179142e-05,
+      "loss": 0.7277,
       "step": 5978
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 20.172710418701172,
-      "learning_rate": 8.016437806956e-06,
-      "loss": 0.64,
+      "epoch": 0.75,
+      "grad_norm": 4.93873929977417,
+      "learning_rate": 1.4998117391122455e-05,
+      "loss": 0.3812,
       "step": 5979
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 16.025949478149414,
-      "learning_rate": 8.01443319635161e-06,
-      "loss": 1.1168,
+      "epoch": 0.75,
+      "grad_norm": 19.105716705322266,
+      "learning_rate": 1.4997280676065766e-05,
+      "loss": 0.8092,
       "step": 5980
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 19.96694564819336,
-      "learning_rate": 8.01242858574722e-06,
-      "loss": 0.9764,
+      "epoch": 0.75,
+      "grad_norm": 9.995222091674805,
+      "learning_rate": 1.4996443961009079e-05,
+      "loss": 1.533,
       "step": 5981
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 28.706392288208008,
-      "learning_rate": 8.01042397514283e-06,
-      "loss": 1.6412,
+      "epoch": 0.75,
+      "grad_norm": 8.4161958694458,
+      "learning_rate": 1.4995607245952393e-05,
+      "loss": 0.6504,
       "step": 5982
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 9.621232032775879,
-      "learning_rate": 8.008419364538438e-06,
-      "loss": 0.9458,
+      "epoch": 0.75,
+      "grad_norm": 43.21015548706055,
+      "learning_rate": 1.4994770530895703e-05,
+      "loss": 1.8054,
       "step": 5983
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 50.64955520629883,
-      "learning_rate": 8.006414753934048e-06,
-      "loss": 1.558,
+      "epoch": 0.75,
+      "grad_norm": 16.76747703552246,
+      "learning_rate": 1.4993933815839017e-05,
+      "loss": 3.7512,
       "step": 5984
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 54.46828842163086,
-      "learning_rate": 8.004410143329658e-06,
-      "loss": 2.4981,
+      "epoch": 0.75,
+      "grad_norm": 9.055212020874023,
+      "learning_rate": 1.499309710078233e-05,
+      "loss": 1.5094,
       "step": 5985
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 41.00409698486328,
-      "learning_rate": 8.002405532725269e-06,
-      "loss": 1.5329,
+      "epoch": 0.75,
+      "grad_norm": 48.513824462890625,
+      "learning_rate": 1.4992260385725644e-05,
+      "loss": 2.6326,
       "step": 5986
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 14.016818046569824,
-      "learning_rate": 8.000400922120879e-06,
-      "loss": 1.4779,
+      "epoch": 0.75,
+      "grad_norm": 10.663233757019043,
+      "learning_rate": 1.4991423670668954e-05,
+      "loss": 1.4526,
       "step": 5987
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 46.308258056640625,
-      "learning_rate": 7.998396311516489e-06,
-      "loss": 1.1857,
+      "epoch": 0.75,
+      "grad_norm": 14.359576225280762,
+      "learning_rate": 1.4990586955612268e-05,
+      "loss": 2.2563,
       "step": 5988
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 10.816703796386719,
-      "learning_rate": 7.996391700912099e-06,
-      "loss": 1.2238,
+      "epoch": 0.75,
+      "grad_norm": 21.671953201293945,
+      "learning_rate": 1.498975024055558e-05,
+      "loss": 1.9075,
       "step": 5989
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 14.56143856048584,
-      "learning_rate": 7.994387090307709e-06,
-      "loss": 1.5197,
+      "epoch": 0.75,
+      "grad_norm": 5.257259845733643,
+      "learning_rate": 1.4988913525498892e-05,
+      "loss": 1.7287,
       "step": 5990
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 13.148183822631836,
-      "learning_rate": 7.992382479703317e-06,
-      "loss": 1.0046,
+      "epoch": 0.75,
+      "grad_norm": 22.53580665588379,
+      "learning_rate": 1.4988076810442205e-05,
+      "loss": 0.7538,
       "step": 5991
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 54.134552001953125,
-      "learning_rate": 7.990377869098929e-06,
-      "loss": 1.8494,
+      "epoch": 0.75,
+      "grad_norm": 24.42538833618164,
+      "learning_rate": 1.4987240095385517e-05,
+      "loss": 2.8615,
       "step": 5992
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 12.82621955871582,
-      "learning_rate": 7.988373258494539e-06,
-      "loss": 1.4167,
+      "epoch": 0.75,
+      "grad_norm": 7.242316722869873,
+      "learning_rate": 1.4986403380328831e-05,
+      "loss": 2.4656,
       "step": 5993
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 14.938492774963379,
-      "learning_rate": 7.986368647890148e-06,
-      "loss": 0.6812,
+      "epoch": 0.75,
+      "grad_norm": 15.731229782104492,
+      "learning_rate": 1.4985566665272141e-05,
+      "loss": 1.8357,
       "step": 5994
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 39.79289627075195,
-      "learning_rate": 7.984364037285758e-06,
-      "loss": 4.1951,
+      "epoch": 0.75,
+      "grad_norm": 17.896705627441406,
+      "learning_rate": 1.4984729950215455e-05,
+      "loss": 2.1809,
       "step": 5995
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 12.317481994628906,
-      "learning_rate": 7.982359426681368e-06,
-      "loss": 1.0172,
+      "epoch": 0.75,
+      "grad_norm": 3.5809171199798584,
+      "learning_rate": 1.4983893235158769e-05,
+      "loss": 0.3103,
       "step": 5996
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 28.569238662719727,
-      "learning_rate": 7.980354816076978e-06,
-      "loss": 2.539,
+      "epoch": 0.75,
+      "grad_norm": 11.365817070007324,
+      "learning_rate": 1.4983056520102079e-05,
+      "loss": 0.815,
       "step": 5997
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 28.877849578857422,
-      "learning_rate": 7.978350205472588e-06,
-      "loss": 1.4473,
+      "epoch": 0.75,
+      "grad_norm": 29.889799118041992,
+      "learning_rate": 1.4982219805045393e-05,
+      "loss": 2.3377,
       "step": 5998
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 42.884979248046875,
-      "learning_rate": 7.976345594868198e-06,
-      "loss": 0.6649,
+      "epoch": 0.75,
+      "grad_norm": 100.96052551269531,
+      "learning_rate": 1.4981383089988706e-05,
+      "loss": 2.4042,
       "step": 5999
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 8.961115837097168,
-      "learning_rate": 7.974340984263808e-06,
-      "loss": 1.1251,
+      "epoch": 0.75,
+      "grad_norm": 27.039257049560547,
+      "learning_rate": 1.498054637493202e-05,
+      "loss": 3.018,
       "step": 6000
     },
     {
-      "epoch": 1.8,
-      "eval_loss": 0.19368119537830353,
-      "eval_runtime": 43.5079,
-      "eval_samples_per_second": 33.994,
-      "eval_steps_per_second": 33.994,
+      "epoch": 0.75,
+      "eval_loss": 0.11360074579715729,
+      "eval_runtime": 95.1759,
+      "eval_samples_per_second": 37.215,
+      "eval_steps_per_second": 37.215,
       "step": 6000
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 72.2662124633789,
-      "learning_rate": 7.972336373659416e-06,
-      "loss": 2.015,
+      "epoch": 0.75,
+      "grad_norm": 4.2123284339904785,
+      "learning_rate": 1.497970965987533e-05,
+      "loss": 0.4822,
       "step": 6001
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 16.20361328125,
-      "learning_rate": 7.970331763055027e-06,
-      "loss": 1.2574,
+      "epoch": 0.75,
+      "grad_norm": 20.182884216308594,
+      "learning_rate": 1.4978872944818644e-05,
+      "loss": 3.4223,
       "step": 6002
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 20.717201232910156,
-      "learning_rate": 7.968327152450637e-06,
-      "loss": 1.5776,
+      "epoch": 0.75,
+      "grad_norm": 7.4854559898376465,
+      "learning_rate": 1.4978036229761956e-05,
+      "loss": 1.7623,
       "step": 6003
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 23.648340225219727,
-      "learning_rate": 7.966322541846248e-06,
-      "loss": 1.8967,
+      "epoch": 0.75,
+      "grad_norm": 10.160518646240234,
+      "learning_rate": 1.4977199514705268e-05,
+      "loss": 0.7109,
       "step": 6004
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 28.407695770263672,
-      "learning_rate": 7.964317931241857e-06,
-      "loss": 2.1085,
+      "epoch": 0.75,
+      "grad_norm": 7.039008140563965,
+      "learning_rate": 1.4976362799648581e-05,
+      "loss": 2.4178,
       "step": 6005
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 13.047331809997559,
-      "learning_rate": 7.962313320637467e-06,
-      "loss": 1.9436,
+      "epoch": 0.75,
+      "grad_norm": 16.30356216430664,
+      "learning_rate": 1.4975526084591893e-05,
+      "loss": 2.6283,
       "step": 6006
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 12.131548881530762,
-      "learning_rate": 7.960308710033077e-06,
-      "loss": 1.3458,
+      "epoch": 0.75,
+      "grad_norm": 7.128363609313965,
+      "learning_rate": 1.4974689369535207e-05,
+      "loss": 1.4611,
       "step": 6007
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 15.92016315460205,
-      "learning_rate": 7.958304099428687e-06,
-      "loss": 0.8177,
+      "epoch": 0.75,
+      "grad_norm": 4.830732822418213,
+      "learning_rate": 1.4973852654478517e-05,
+      "loss": 0.7371,
       "step": 6008
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 9.055414199829102,
-      "learning_rate": 7.956299488824295e-06,
-      "loss": 1.3297,
+      "epoch": 0.75,
+      "grad_norm": 16.139087677001953,
+      "learning_rate": 1.497301593942183e-05,
+      "loss": 2.0725,
       "step": 6009
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 13.94730281829834,
-      "learning_rate": 7.954294878219907e-06,
-      "loss": 1.4445,
+      "epoch": 0.75,
+      "grad_norm": 16.47942352294922,
+      "learning_rate": 1.4972179224365144e-05,
+      "loss": 1.6814,
       "step": 6010
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 17.391616821289062,
-      "learning_rate": 7.952290267615517e-06,
-      "loss": 0.9038,
+      "epoch": 0.75,
+      "grad_norm": 17.696067810058594,
+      "learning_rate": 1.4971342509308455e-05,
+      "loss": 2.2306,
       "step": 6011
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 198.21731567382812,
-      "learning_rate": 7.950285657011126e-06,
-      "loss": 2.3957,
+      "epoch": 0.75,
+      "grad_norm": 13.801817893981934,
+      "learning_rate": 1.4970505794251768e-05,
+      "loss": 1.6959,
       "step": 6012
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 12.243670463562012,
-      "learning_rate": 7.948281046406736e-06,
-      "loss": 1.4155,
+      "epoch": 0.75,
+      "grad_norm": 21.30613899230957,
+      "learning_rate": 1.4969669079195082e-05,
+      "loss": 3.468,
       "step": 6013
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 13.951066017150879,
-      "learning_rate": 7.946276435802346e-06,
-      "loss": 1.4097,
+      "epoch": 0.75,
+      "grad_norm": 45.791160583496094,
+      "learning_rate": 1.4968832364138394e-05,
+      "loss": 3.4723,
       "step": 6014
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 24.333097457885742,
-      "learning_rate": 7.944271825197956e-06,
-      "loss": 2.4542,
+      "epoch": 0.75,
+      "grad_norm": 9.92892074584961,
+      "learning_rate": 1.4967995649081706e-05,
+      "loss": 2.3938,
       "step": 6015
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 17.491724014282227,
-      "learning_rate": 7.942267214593566e-06,
-      "loss": 1.649,
+      "epoch": 0.75,
+      "grad_norm": 22.17215919494629,
+      "learning_rate": 1.496715893402502e-05,
+      "loss": 1.9325,
       "step": 6016
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 11.039308547973633,
-      "learning_rate": 7.940262603989176e-06,
-      "loss": 1.1686,
+      "epoch": 0.76,
+      "grad_norm": 59.155479431152344,
+      "learning_rate": 1.4966322218968332e-05,
+      "loss": 3.5734,
       "step": 6017
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 18.201019287109375,
-      "learning_rate": 7.938257993384786e-06,
-      "loss": 1.5321,
+      "epoch": 0.76,
+      "grad_norm": 16.35432243347168,
+      "learning_rate": 1.4965485503911644e-05,
+      "loss": 2.029,
       "step": 6018
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 29.47193717956543,
-      "learning_rate": 7.936253382780396e-06,
-      "loss": 1.5978,
+      "epoch": 0.76,
+      "grad_norm": 12.570013999938965,
+      "learning_rate": 1.4964648788854957e-05,
+      "loss": 1.008,
       "step": 6019
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 24.67177391052246,
-      "learning_rate": 7.934248772176005e-06,
-      "loss": 2.0138,
+      "epoch": 0.76,
+      "grad_norm": 34.56005096435547,
+      "learning_rate": 1.4963812073798269e-05,
+      "loss": 1.8091,
       "step": 6020
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 25.735088348388672,
-      "learning_rate": 7.932244161571615e-06,
-      "loss": 3.0204,
+      "epoch": 0.76,
+      "grad_norm": 33.60686111450195,
+      "learning_rate": 1.4962975358741583e-05,
+      "loss": 1.8547,
       "step": 6021
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 63.19715118408203,
-      "learning_rate": 7.930239550967225e-06,
-      "loss": 1.5556,
+      "epoch": 0.76,
+      "grad_norm": 17.62992286682129,
+      "learning_rate": 1.4962138643684893e-05,
+      "loss": 0.9139,
       "step": 6022
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 12.563130378723145,
-      "learning_rate": 7.928234940362835e-06,
-      "loss": 1.3196,
+      "epoch": 0.76,
+      "grad_norm": 10.80237865447998,
+      "learning_rate": 1.4961301928628207e-05,
+      "loss": 1.0544,
       "step": 6023
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 21.176742553710938,
-      "learning_rate": 7.926230329758445e-06,
-      "loss": 2.416,
+      "epoch": 0.76,
+      "grad_norm": 13.512557983398438,
+      "learning_rate": 1.496046521357152e-05,
+      "loss": 1.1676,
       "step": 6024
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 14.44190502166748,
-      "learning_rate": 7.924225719154055e-06,
-      "loss": 1.14,
+      "epoch": 0.76,
+      "grad_norm": 8.398104667663574,
+      "learning_rate": 1.495962849851483e-05,
+      "loss": 2.2205,
       "step": 6025
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 10.227465629577637,
-      "learning_rate": 7.922221108549665e-06,
-      "loss": 0.9925,
+      "epoch": 0.76,
+      "grad_norm": 24.02195167541504,
+      "learning_rate": 1.4958791783458144e-05,
+      "loss": 4.6731,
       "step": 6026
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 16.822582244873047,
-      "learning_rate": 7.920216497945275e-06,
-      "loss": 1.413,
+      "epoch": 0.76,
+      "grad_norm": 34.829097747802734,
+      "learning_rate": 1.4957955068401458e-05,
+      "loss": 2.6504,
       "step": 6027
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 23.78660774230957,
-      "learning_rate": 7.918211887340884e-06,
-      "loss": 2.2229,
+      "epoch": 0.76,
+      "grad_norm": 21.19928741455078,
+      "learning_rate": 1.495711835334477e-05,
+      "loss": 2.3643,
       "step": 6028
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 22.946805953979492,
-      "learning_rate": 7.916207276736495e-06,
-      "loss": 1.9366,
+      "epoch": 0.76,
+      "grad_norm": 11.841145515441895,
+      "learning_rate": 1.4956281638288082e-05,
+      "loss": 0.8766,
       "step": 6029
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 11.842130661010742,
-      "learning_rate": 7.914202666132105e-06,
-      "loss": 2.1797,
+      "epoch": 0.76,
+      "grad_norm": 5.688900947570801,
+      "learning_rate": 1.4955444923231395e-05,
+      "loss": 1.7931,
       "step": 6030
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 11.936902046203613,
-      "learning_rate": 7.912198055527714e-06,
-      "loss": 0.9288,
+      "epoch": 0.76,
+      "grad_norm": 12.185578346252441,
+      "learning_rate": 1.4954608208174707e-05,
+      "loss": 1.2761,
       "step": 6031
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 12.97545337677002,
-      "learning_rate": 7.910193444923324e-06,
-      "loss": 0.8792,
+      "epoch": 0.76,
+      "grad_norm": 19.21368980407715,
+      "learning_rate": 1.495377149311802e-05,
+      "loss": 1.5665,
       "step": 6032
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 22.72028160095215,
-      "learning_rate": 7.908188834318934e-06,
-      "loss": 1.8991,
+      "epoch": 0.76,
+      "grad_norm": 11.331363677978516,
+      "learning_rate": 1.4952934778061331e-05,
+      "loss": 2.9063,
       "step": 6033
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 68.23728942871094,
-      "learning_rate": 7.906184223714544e-06,
-      "loss": 2.0705,
+      "epoch": 0.76,
+      "grad_norm": 16.766036987304688,
+      "learning_rate": 1.4952098063004645e-05,
+      "loss": 2.0617,
       "step": 6034
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 17.14826202392578,
-      "learning_rate": 7.904179613110154e-06,
-      "loss": 1.9903,
+      "epoch": 0.76,
+      "grad_norm": 18.78944206237793,
+      "learning_rate": 1.4951261347947959e-05,
+      "loss": 1.3033,
       "step": 6035
     },
     {
-      "epoch": 1.81,
-      "grad_norm": 33.424381256103516,
-      "learning_rate": 7.902175002505764e-06,
-      "loss": 1.5851,
+      "epoch": 0.76,
+      "grad_norm": 19.58803939819336,
+      "learning_rate": 1.4950424632891269e-05,
+      "loss": 2.1772,
       "step": 6036
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 21.694814682006836,
-      "learning_rate": 7.900170391901374e-06,
-      "loss": 2.1811,
+      "epoch": 0.76,
+      "grad_norm": 24.190006256103516,
+      "learning_rate": 1.4949587917834583e-05,
+      "loss": 1.8668,
       "step": 6037
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 11.936399459838867,
-      "learning_rate": 7.898165781296984e-06,
-      "loss": 1.4652,
+      "epoch": 0.76,
+      "grad_norm": 14.641880989074707,
+      "learning_rate": 1.4948751202777896e-05,
+      "loss": 2.1728,
       "step": 6038
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 10.102071762084961,
-      "learning_rate": 7.896161170692593e-06,
-      "loss": 1.9421,
+      "epoch": 0.76,
+      "grad_norm": 20.583711624145508,
+      "learning_rate": 1.4947914487721206e-05,
+      "loss": 2.317,
       "step": 6039
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 32.03966522216797,
-      "learning_rate": 7.894156560088203e-06,
-      "loss": 1.4121,
+      "epoch": 0.76,
+      "grad_norm": 16.76109504699707,
+      "learning_rate": 1.494707777266452e-05,
+      "loss": 1.2081,
       "step": 6040
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 31.259422302246094,
-      "learning_rate": 7.892151949483815e-06,
-      "loss": 1.255,
+      "epoch": 0.76,
+      "grad_norm": 61.240535736083984,
+      "learning_rate": 1.4946241057607834e-05,
+      "loss": 2.282,
       "step": 6041
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 11.89094352722168,
-      "learning_rate": 7.890147338879423e-06,
-      "loss": 1.1638,
+      "epoch": 0.76,
+      "grad_norm": 13.91823673248291,
+      "learning_rate": 1.4945404342551146e-05,
+      "loss": 3.5434,
       "step": 6042
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 16.84495735168457,
-      "learning_rate": 7.888142728275033e-06,
-      "loss": 1.4586,
+      "epoch": 0.76,
+      "grad_norm": 7.848240852355957,
+      "learning_rate": 1.4944567627494458e-05,
+      "loss": 2.3143,
       "step": 6043
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 9.196840286254883,
-      "learning_rate": 7.886138117670643e-06,
-      "loss": 2.2115,
+      "epoch": 0.76,
+      "grad_norm": 4.802414894104004,
+      "learning_rate": 1.4943730912437771e-05,
+      "loss": 0.6465,
       "step": 6044
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 20.001394271850586,
-      "learning_rate": 7.884133507066253e-06,
-      "loss": 1.0957,
+      "epoch": 0.76,
+      "grad_norm": 17.290687561035156,
+      "learning_rate": 1.4942894197381083e-05,
+      "loss": 1.7776,
       "step": 6045
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 30.092775344848633,
-      "learning_rate": 7.882128896461862e-06,
-      "loss": 1.6151,
+      "epoch": 0.76,
+      "grad_norm": 22.302507400512695,
+      "learning_rate": 1.4942057482324395e-05,
+      "loss": 0.5936,
       "step": 6046
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 10.02807331085205,
-      "learning_rate": 7.880124285857474e-06,
-      "loss": 0.872,
+      "epoch": 0.76,
+      "grad_norm": 5.213108062744141,
+      "learning_rate": 1.4941220767267707e-05,
+      "loss": 0.6772,
       "step": 6047
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 21.552358627319336,
-      "learning_rate": 7.878119675253084e-06,
-      "loss": 1.4557,
+      "epoch": 0.76,
+      "grad_norm": 12.678808212280273,
+      "learning_rate": 1.494038405221102e-05,
+      "loss": 1.4635,
       "step": 6048
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 21.00843620300293,
-      "learning_rate": 7.876115064648692e-06,
-      "loss": 1.379,
+      "epoch": 0.76,
+      "grad_norm": 7.262757301330566,
+      "learning_rate": 1.4939547337154334e-05,
+      "loss": 3.6633,
       "step": 6049
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 16.021032333374023,
-      "learning_rate": 7.874110454044302e-06,
-      "loss": 1.8451,
+      "epoch": 0.76,
+      "grad_norm": 10.369536399841309,
+      "learning_rate": 1.4938710622097645e-05,
+      "loss": 0.4444,
       "step": 6050
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 14.853699684143066,
-      "learning_rate": 7.872105843439912e-06,
-      "loss": 1.125,
+      "epoch": 0.76,
+      "grad_norm": 14.876389503479004,
+      "learning_rate": 1.4937873907040958e-05,
+      "loss": 1.0291,
       "step": 6051
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 20.046295166015625,
-      "learning_rate": 7.870101232835522e-06,
-      "loss": 1.9366,
+      "epoch": 0.76,
+      "grad_norm": 8.755902290344238,
+      "learning_rate": 1.4937037191984272e-05,
+      "loss": 1.6724,
       "step": 6052
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 17.23594856262207,
-      "learning_rate": 7.868096622231132e-06,
-      "loss": 1.6949,
+      "epoch": 0.76,
+      "grad_norm": 15.990140914916992,
+      "learning_rate": 1.4936200476927582e-05,
+      "loss": 2.0313,
       "step": 6053
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 16.25902557373047,
-      "learning_rate": 7.866092011626742e-06,
-      "loss": 1.9581,
+      "epoch": 0.76,
+      "grad_norm": 4.509998321533203,
+      "learning_rate": 1.4935363761870896e-05,
+      "loss": 0.436,
       "step": 6054
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 25.339801788330078,
-      "learning_rate": 7.864087401022353e-06,
-      "loss": 1.4696,
+      "epoch": 0.76,
+      "grad_norm": 14.44178295135498,
+      "learning_rate": 1.493452704681421e-05,
+      "loss": 2.0875,
       "step": 6055
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 45.478515625,
-      "learning_rate": 7.862082790417963e-06,
-      "loss": 2.0867,
+      "epoch": 0.76,
+      "grad_norm": 13.471675872802734,
+      "learning_rate": 1.4933690331757521e-05,
+      "loss": 0.8051,
       "step": 6056
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 19.254344940185547,
-      "learning_rate": 7.860078179813571e-06,
-      "loss": 1.6101,
+      "epoch": 0.76,
+      "grad_norm": 23.050397872924805,
+      "learning_rate": 1.4932853616700833e-05,
+      "loss": 1.3614,
       "step": 6057
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 23.987213134765625,
-      "learning_rate": 7.858073569209181e-06,
-      "loss": 1.1053,
+      "epoch": 0.76,
+      "grad_norm": 7.089566707611084,
+      "learning_rate": 1.4932016901644147e-05,
+      "loss": 0.6359,
       "step": 6058
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 33.420997619628906,
-      "learning_rate": 7.856068958604791e-06,
-      "loss": 1.3957,
+      "epoch": 0.76,
+      "grad_norm": 18.266427993774414,
+      "learning_rate": 1.4931180186587459e-05,
+      "loss": 0.9629,
       "step": 6059
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 53.56259536743164,
-      "learning_rate": 7.854064348000401e-06,
-      "loss": 2.8107,
+      "epoch": 0.76,
+      "grad_norm": 14.109417915344238,
+      "learning_rate": 1.4930343471530771e-05,
+      "loss": 1.1746,
       "step": 6060
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 14.887017250061035,
-      "learning_rate": 7.852059737396011e-06,
-      "loss": 2.4886,
+      "epoch": 0.76,
+      "grad_norm": 9.30927848815918,
+      "learning_rate": 1.4929506756474083e-05,
+      "loss": 1.5184,
       "step": 6061
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 15.285018920898438,
-      "learning_rate": 7.850055126791621e-06,
-      "loss": 1.1363,
+      "epoch": 0.76,
+      "grad_norm": 7.900919437408447,
+      "learning_rate": 1.4928670041417397e-05,
+      "loss": 0.5081,
       "step": 6062
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 16.956491470336914,
-      "learning_rate": 7.848050516187231e-06,
-      "loss": 1.4113,
+      "epoch": 0.76,
+      "grad_norm": 14.977768898010254,
+      "learning_rate": 1.492783332636071e-05,
+      "loss": 1.6299,
       "step": 6063
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 22.227479934692383,
-      "learning_rate": 7.846045905582842e-06,
-      "loss": 1.2002,
+      "epoch": 0.76,
+      "grad_norm": 18.095666885375977,
+      "learning_rate": 1.492699661130402e-05,
+      "loss": 4.2114,
       "step": 6064
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 44.9412727355957,
-      "learning_rate": 7.84404129497845e-06,
-      "loss": 1.583,
+      "epoch": 0.76,
+      "grad_norm": 25.11492156982422,
+      "learning_rate": 1.4926159896247334e-05,
+      "loss": 2.605,
       "step": 6065
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 37.56648254394531,
-      "learning_rate": 7.842036684374062e-06,
-      "loss": 2.022,
+      "epoch": 0.76,
+      "grad_norm": 44.8865966796875,
+      "learning_rate": 1.4925323181190648e-05,
+      "loss": 2.4028,
       "step": 6066
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 15.846146583557129,
-      "learning_rate": 7.840032073769672e-06,
-      "loss": 0.9523,
+      "epoch": 0.76,
+      "grad_norm": 8.544513702392578,
+      "learning_rate": 1.4924486466133958e-05,
+      "loss": 1.1757,
       "step": 6067
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 18.317182540893555,
-      "learning_rate": 7.83802746316528e-06,
-      "loss": 1.5857,
+      "epoch": 0.76,
+      "grad_norm": 14.130455017089844,
+      "learning_rate": 1.4923649751077272e-05,
+      "loss": 1.6795,
       "step": 6068
     },
     {
-      "epoch": 1.82,
-      "grad_norm": 14.531113624572754,
-      "learning_rate": 7.83602285256089e-06,
-      "loss": 1.2722,
+      "epoch": 0.76,
+      "grad_norm": 107.8146743774414,
+      "learning_rate": 1.4922813036020585e-05,
+      "loss": 1.8548,
       "step": 6069
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 9.223918914794922,
-      "learning_rate": 7.8340182419565e-06,
-      "loss": 1.6124,
+      "epoch": 0.76,
+      "grad_norm": 18.72151756286621,
+      "learning_rate": 1.4921976320963897e-05,
+      "loss": 1.1193,
       "step": 6070
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 12.137374877929688,
-      "learning_rate": 7.83201363135211e-06,
-      "loss": 1.0094,
+      "epoch": 0.76,
+      "grad_norm": 28.117448806762695,
+      "learning_rate": 1.492113960590721e-05,
+      "loss": 1.3729,
       "step": 6071
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 8.65196418762207,
-      "learning_rate": 7.83000902074772e-06,
-      "loss": 0.9413,
+      "epoch": 0.76,
+      "grad_norm": 7.518640995025635,
+      "learning_rate": 1.4920302890850521e-05,
+      "loss": 0.8463,
       "step": 6072
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 16.220958709716797,
-      "learning_rate": 7.82800441014333e-06,
-      "loss": 2.1056,
+      "epoch": 0.76,
+      "grad_norm": 15.60339641571045,
+      "learning_rate": 1.4919466175793835e-05,
+      "loss": 1.315,
       "step": 6073
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 19.515766143798828,
-      "learning_rate": 7.82599979953894e-06,
-      "loss": 1.763,
+      "epoch": 0.76,
+      "grad_norm": 19.411062240600586,
+      "learning_rate": 1.4918629460737147e-05,
+      "loss": 2.0167,
       "step": 6074
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 11.028536796569824,
-      "learning_rate": 7.82399518893455e-06,
-      "loss": 1.577,
+      "epoch": 0.76,
+      "grad_norm": 9.9929780960083,
+      "learning_rate": 1.4917792745680459e-05,
+      "loss": 0.8052,
       "step": 6075
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 14.080721855163574,
-      "learning_rate": 7.82199057833016e-06,
-      "loss": 1.2928,
+      "epoch": 0.76,
+      "grad_norm": 24.951963424682617,
+      "learning_rate": 1.4916956030623772e-05,
+      "loss": 1.8592,
       "step": 6076
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 22.41744041442871,
-      "learning_rate": 7.81998596772577e-06,
-      "loss": 1.6717,
+      "epoch": 0.76,
+      "grad_norm": 12.254054069519043,
+      "learning_rate": 1.4916119315567086e-05,
+      "loss": 0.5268,
       "step": 6077
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 29.117725372314453,
-      "learning_rate": 7.817981357121381e-06,
-      "loss": 2.0723,
+      "epoch": 0.76,
+      "grad_norm": 24.44095802307129,
+      "learning_rate": 1.4915282600510396e-05,
+      "loss": 3.7504,
       "step": 6078
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 10.062914848327637,
-      "learning_rate": 7.81597674651699e-06,
-      "loss": 0.9723,
+      "epoch": 0.76,
+      "grad_norm": 16.46070098876953,
+      "learning_rate": 1.491444588545371e-05,
+      "loss": 2.3923,
       "step": 6079
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 19.473957061767578,
-      "learning_rate": 7.8139721359126e-06,
-      "loss": 1.6513,
+      "epoch": 0.76,
+      "grad_norm": 22.783987045288086,
+      "learning_rate": 1.4913609170397024e-05,
+      "loss": 1.3608,
       "step": 6080
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 13.241780281066895,
-      "learning_rate": 7.81196752530821e-06,
-      "loss": 1.4396,
+      "epoch": 0.76,
+      "grad_norm": 13.54172420501709,
+      "learning_rate": 1.4912772455340334e-05,
+      "loss": 1.4165,
       "step": 6081
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 50.16336441040039,
-      "learning_rate": 7.80996291470382e-06,
-      "loss": 1.4343,
+      "epoch": 0.76,
+      "grad_norm": 14.175755500793457,
+      "learning_rate": 1.4911935740283648e-05,
+      "loss": 1.288,
       "step": 6082
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 8.716567039489746,
-      "learning_rate": 7.807958304099428e-06,
-      "loss": 0.7952,
+      "epoch": 0.76,
+      "grad_norm": 36.6352653503418,
+      "learning_rate": 1.4911099025226961e-05,
+      "loss": 1.3954,
       "step": 6083
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 17.933069229125977,
-      "learning_rate": 7.80595369349504e-06,
-      "loss": 2.706,
+      "epoch": 0.76,
+      "grad_norm": 6.960003852844238,
+      "learning_rate": 1.4910262310170273e-05,
+      "loss": 1.4535,
       "step": 6084
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 30.77072525024414,
-      "learning_rate": 7.80394908289065e-06,
-      "loss": 1.9985,
+      "epoch": 0.76,
+      "grad_norm": 5.8838067054748535,
+      "learning_rate": 1.4909425595113585e-05,
+      "loss": 0.5222,
       "step": 6085
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 13.682772636413574,
-      "learning_rate": 7.801944472286258e-06,
-      "loss": 0.9191,
+      "epoch": 0.76,
+      "grad_norm": 29.456592559814453,
+      "learning_rate": 1.4908588880056897e-05,
+      "loss": 1.6292,
       "step": 6086
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 17.457841873168945,
-      "learning_rate": 7.799939861681868e-06,
-      "loss": 1.518,
+      "epoch": 0.76,
+      "grad_norm": 13.978496551513672,
+      "learning_rate": 1.490775216500021e-05,
+      "loss": 1.9543,
       "step": 6087
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 41.92538833618164,
-      "learning_rate": 7.797935251077479e-06,
-      "loss": 2.1136,
+      "epoch": 0.76,
+      "grad_norm": 9.037714004516602,
+      "learning_rate": 1.4906915449943523e-05,
+      "loss": 0.9129,
       "step": 6088
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 22.720664978027344,
-      "learning_rate": 7.795930640473089e-06,
-      "loss": 1.8965,
+      "epoch": 0.76,
+      "grad_norm": 13.83279800415039,
+      "learning_rate": 1.4906078734886835e-05,
+      "loss": 1.8087,
       "step": 6089
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 13.168417930603027,
-      "learning_rate": 7.793926029868699e-06,
-      "loss": 1.3237,
+      "epoch": 0.76,
+      "grad_norm": 19.00337028503418,
+      "learning_rate": 1.4905242019830148e-05,
+      "loss": 2.4818,
       "step": 6090
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 22.606231689453125,
-      "learning_rate": 7.791921419264309e-06,
-      "loss": 1.6433,
+      "epoch": 0.76,
+      "grad_norm": 20.148292541503906,
+      "learning_rate": 1.4904405304773462e-05,
+      "loss": 2.7005,
       "step": 6091
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 22.04924964904785,
-      "learning_rate": 7.789916808659919e-06,
-      "loss": 1.9051,
+      "epoch": 0.76,
+      "grad_norm": 9.858909606933594,
+      "learning_rate": 1.4903568589716772e-05,
+      "loss": 1.2669,
       "step": 6092
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 12.943755149841309,
-      "learning_rate": 7.787912198055529e-06,
-      "loss": 1.3566,
+      "epoch": 0.76,
+      "grad_norm": 33.501895904541016,
+      "learning_rate": 1.4902731874660086e-05,
+      "loss": 2.5563,
       "step": 6093
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 8.388080596923828,
-      "learning_rate": 7.785907587451137e-06,
-      "loss": 1.0533,
+      "epoch": 0.76,
+      "grad_norm": 15.421066284179688,
+      "learning_rate": 1.49018951596034e-05,
+      "loss": 1.3921,
       "step": 6094
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 39.58984375,
-      "learning_rate": 7.783902976846747e-06,
-      "loss": 2.1753,
+      "epoch": 0.76,
+      "grad_norm": 20.976089477539062,
+      "learning_rate": 1.490105844454671e-05,
+      "loss": 1.3577,
       "step": 6095
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 17.842021942138672,
-      "learning_rate": 7.781898366242357e-06,
-      "loss": 0.9146,
+      "epoch": 0.77,
+      "grad_norm": 10.768942832946777,
+      "learning_rate": 1.4900221729490023e-05,
+      "loss": 0.8958,
       "step": 6096
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 12.273473739624023,
-      "learning_rate": 7.779893755637968e-06,
-      "loss": 1.0507,
+      "epoch": 0.77,
+      "grad_norm": 6.768867015838623,
+      "learning_rate": 1.4899385014433337e-05,
+      "loss": 2.9632,
       "step": 6097
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 34.619178771972656,
-      "learning_rate": 7.777889145033578e-06,
-      "loss": 2.2011,
+      "epoch": 0.77,
+      "grad_norm": 4.851020336151123,
+      "learning_rate": 1.4898548299376649e-05,
+      "loss": 1.1444,
       "step": 6098
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 87.12923431396484,
-      "learning_rate": 7.775884534429188e-06,
-      "loss": 1.4831,
+      "epoch": 0.77,
+      "grad_norm": 21.36313819885254,
+      "learning_rate": 1.4897711584319961e-05,
+      "loss": 1.6182,
       "step": 6099
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 9.995388984680176,
-      "learning_rate": 7.773879923824798e-06,
-      "loss": 1.2102,
+      "epoch": 0.77,
+      "grad_norm": 10.490818977355957,
+      "learning_rate": 1.4896874869263273e-05,
+      "loss": 1.1183,
       "step": 6100
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 13.057961463928223,
-      "learning_rate": 7.771875313220408e-06,
-      "loss": 1.3872,
+      "epoch": 0.77,
+      "grad_norm": 9.165608406066895,
+      "learning_rate": 1.4896038154206587e-05,
+      "loss": 1.6784,
       "step": 6101
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 8.191656112670898,
-      "learning_rate": 7.769870702616016e-06,
-      "loss": 1.3387,
+      "epoch": 0.77,
+      "grad_norm": 22.0075740814209,
+      "learning_rate": 1.4895201439149899e-05,
+      "loss": 1.4024,
       "step": 6102
     },
     {
-      "epoch": 1.83,
-      "grad_norm": 25.005573272705078,
-      "learning_rate": 7.767866092011628e-06,
-      "loss": 1.393,
+      "epoch": 0.77,
+      "grad_norm": 21.335983276367188,
+      "learning_rate": 1.489436472409321e-05,
+      "loss": 2.9008,
       "step": 6103
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 11.448406219482422,
-      "learning_rate": 7.765861481407238e-06,
-      "loss": 1.2107,
+      "epoch": 0.77,
+      "grad_norm": 11.496278762817383,
+      "learning_rate": 1.4893528009036524e-05,
+      "loss": 2.6356,
       "step": 6104
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 10.444814682006836,
-      "learning_rate": 7.763856870802847e-06,
-      "loss": 1.1085,
+      "epoch": 0.77,
+      "grad_norm": 10.158914566040039,
+      "learning_rate": 1.4892691293979838e-05,
+      "loss": 3.4004,
       "step": 6105
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 22.36345863342285,
-      "learning_rate": 7.761852260198457e-06,
-      "loss": 1.4649,
+      "epoch": 0.77,
+      "grad_norm": 23.894088745117188,
+      "learning_rate": 1.4891854578923148e-05,
+      "loss": 2.5214,
       "step": 6106
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 12.638916015625,
-      "learning_rate": 7.759847649594067e-06,
-      "loss": 1.0798,
+      "epoch": 0.77,
+      "grad_norm": 5.749575614929199,
+      "learning_rate": 1.4891017863866462e-05,
+      "loss": 0.5022,
       "step": 6107
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 30.909666061401367,
-      "learning_rate": 7.757843038989677e-06,
-      "loss": 1.3542,
+      "epoch": 0.77,
+      "grad_norm": 17.750463485717773,
+      "learning_rate": 1.4890181148809775e-05,
+      "loss": 2.0954,
       "step": 6108
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 21.984508514404297,
-      "learning_rate": 7.755838428385287e-06,
-      "loss": 1.5787,
+      "epoch": 0.77,
+      "grad_norm": 9.65576457977295,
+      "learning_rate": 1.4889344433753086e-05,
+      "loss": 1.0989,
       "step": 6109
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 28.869102478027344,
-      "learning_rate": 7.753833817780897e-06,
-      "loss": 1.8047,
+      "epoch": 0.77,
+      "grad_norm": 16.375839233398438,
+      "learning_rate": 1.48885077186964e-05,
+      "loss": 1.2684,
       "step": 6110
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 8.794825553894043,
-      "learning_rate": 7.751829207176507e-06,
-      "loss": 0.4721,
+      "epoch": 0.77,
+      "grad_norm": 34.77222442626953,
+      "learning_rate": 1.4887671003639713e-05,
+      "loss": 2.8609,
       "step": 6111
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 20.192941665649414,
-      "learning_rate": 7.749824596572117e-06,
-      "loss": 1.8928,
+      "epoch": 0.77,
+      "grad_norm": 11.829339981079102,
+      "learning_rate": 1.4886834288583025e-05,
+      "loss": 1.7961,
       "step": 6112
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 8.28946304321289,
-      "learning_rate": 7.747819985967726e-06,
-      "loss": 0.8887,
+      "epoch": 0.77,
+      "grad_norm": 21.34874153137207,
+      "learning_rate": 1.4885997573526337e-05,
+      "loss": 2.5703,
       "step": 6113
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 7.44697380065918,
-      "learning_rate": 7.745815375363336e-06,
-      "loss": 1.0449,
+      "epoch": 0.77,
+      "grad_norm": 14.126192092895508,
+      "learning_rate": 1.4885160858469649e-05,
+      "loss": 2.9025,
       "step": 6114
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 47.45841598510742,
-      "learning_rate": 7.743810764758947e-06,
-      "loss": 1.8076,
+      "epoch": 0.77,
+      "grad_norm": 15.445282936096191,
+      "learning_rate": 1.4884324143412962e-05,
+      "loss": 1.7945,
       "step": 6115
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 22.896045684814453,
-      "learning_rate": 7.741806154154556e-06,
-      "loss": 1.3109,
+      "epoch": 0.77,
+      "grad_norm": 10.755892753601074,
+      "learning_rate": 1.4883487428356274e-05,
+      "loss": 1.4015,
       "step": 6116
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 51.84198760986328,
-      "learning_rate": 7.739801543550166e-06,
-      "loss": 1.9647,
+      "epoch": 0.77,
+      "grad_norm": 7.058221817016602,
+      "learning_rate": 1.4882650713299586e-05,
+      "loss": 1.9887,
       "step": 6117
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 16.364362716674805,
-      "learning_rate": 7.737796932945776e-06,
-      "loss": 1.6399,
+      "epoch": 0.77,
+      "grad_norm": 21.246171951293945,
+      "learning_rate": 1.48818139982429e-05,
+      "loss": 3.261,
       "step": 6118
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 14.368910789489746,
-      "learning_rate": 7.735792322341386e-06,
-      "loss": 1.0631,
+      "epoch": 0.77,
+      "grad_norm": 13.608878135681152,
+      "learning_rate": 1.4880977283186214e-05,
+      "loss": 2.2239,
       "step": 6119
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 11.560124397277832,
-      "learning_rate": 7.733787711736994e-06,
-      "loss": 1.6401,
-      "step": 6120
-    },
-    {
-      "epoch": 1.84,
-      "eval_loss": 0.1900317370891571,
-      "eval_runtime": 43.601,
-      "eval_samples_per_second": 33.921,
-      "eval_steps_per_second": 33.921,
+      "epoch": 0.77,
+      "grad_norm": 20.881624221801758,
+      "learning_rate": 1.4880140568129524e-05,
+      "loss": 1.8091,
       "step": 6120
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 11.354382514953613,
-      "learning_rate": 7.731783101132606e-06,
-      "loss": 1.3231,
+      "epoch": 0.77,
+      "grad_norm": 7.887094497680664,
+      "learning_rate": 1.4879303853072838e-05,
+      "loss": 0.7012,
       "step": 6121
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 8.842697143554688,
-      "learning_rate": 7.729778490528216e-06,
-      "loss": 0.5822,
+      "epoch": 0.77,
+      "grad_norm": 19.80340003967285,
+      "learning_rate": 1.4878467138016151e-05,
+      "loss": 1.027,
       "step": 6122
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 44.98017501831055,
-      "learning_rate": 7.727773879923826e-06,
-      "loss": 1.6967,
+      "epoch": 0.77,
+      "grad_norm": 22.05449104309082,
+      "learning_rate": 1.4877630422959461e-05,
+      "loss": 0.8,
       "step": 6123
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 29.9891414642334,
-      "learning_rate": 7.725769269319435e-06,
-      "loss": 1.5777,
+      "epoch": 0.77,
+      "grad_norm": 8.357983589172363,
+      "learning_rate": 1.4876793707902775e-05,
+      "loss": 0.6011,
       "step": 6124
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 25.591136932373047,
-      "learning_rate": 7.723764658715045e-06,
-      "loss": 0.9665,
+      "epoch": 0.77,
+      "grad_norm": 8.36915397644043,
+      "learning_rate": 1.4875956992846087e-05,
+      "loss": 0.7543,
       "step": 6125
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 13.819255828857422,
-      "learning_rate": 7.721760048110655e-06,
-      "loss": 2.2739,
+      "epoch": 0.77,
+      "grad_norm": 31.40436553955078,
+      "learning_rate": 1.48751202777894e-05,
+      "loss": 2.5164,
       "step": 6126
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 28.475740432739258,
-      "learning_rate": 7.719755437506265e-06,
-      "loss": 2.247,
+      "epoch": 0.77,
+      "grad_norm": 15.912656784057617,
+      "learning_rate": 1.4874283562732713e-05,
+      "loss": 1.6534,
       "step": 6127
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 12.238512992858887,
-      "learning_rate": 7.717750826901875e-06,
-      "loss": 0.8953,
+      "epoch": 0.77,
+      "grad_norm": 7.357669353485107,
+      "learning_rate": 1.4873446847676025e-05,
+      "loss": 2.3133,
       "step": 6128
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 35.80490493774414,
-      "learning_rate": 7.715746216297485e-06,
-      "loss": 3.4695,
+      "epoch": 0.77,
+      "grad_norm": 7.6885881423950195,
+      "learning_rate": 1.4872610132619338e-05,
+      "loss": 0.8044,
       "step": 6129
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 7.781877040863037,
-      "learning_rate": 7.713741605693095e-06,
-      "loss": 1.8642,
+      "epoch": 0.77,
+      "grad_norm": 11.81246280670166,
+      "learning_rate": 1.487177341756265e-05,
+      "loss": 0.8979,
       "step": 6130
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 13.301283836364746,
-      "learning_rate": 7.711736995088704e-06,
-      "loss": 1.3696,
+      "epoch": 0.77,
+      "grad_norm": 5.839165210723877,
+      "learning_rate": 1.4870936702505962e-05,
+      "loss": 0.5538,
       "step": 6131
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 10.512940406799316,
-      "learning_rate": 7.709732384484314e-06,
-      "loss": 0.6859,
+      "epoch": 0.77,
+      "grad_norm": 9.871504783630371,
+      "learning_rate": 1.4870099987449276e-05,
+      "loss": 0.5526,
       "step": 6132
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 25.800260543823242,
-      "learning_rate": 7.707727773879926e-06,
-      "loss": 1.9089,
+      "epoch": 0.77,
+      "grad_norm": 27.299074172973633,
+      "learning_rate": 1.486926327239259e-05,
+      "loss": 2.091,
       "step": 6133
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 15.625998497009277,
-      "learning_rate": 7.705723163275534e-06,
-      "loss": 1.8234,
+      "epoch": 0.77,
+      "grad_norm": 60.671478271484375,
+      "learning_rate": 1.48684265573359e-05,
+      "loss": 3.327,
       "step": 6134
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 42.901615142822266,
-      "learning_rate": 7.703718552671144e-06,
-      "loss": 2.0355,
+      "epoch": 0.77,
+      "grad_norm": 9.930606842041016,
+      "learning_rate": 1.4867589842279213e-05,
+      "loss": 1.24,
       "step": 6135
     },
     {
-      "epoch": 1.84,
-      "grad_norm": 15.231833457946777,
-      "learning_rate": 7.701713942066754e-06,
-      "loss": 1.1297,
+      "epoch": 0.77,
+      "grad_norm": 23.737319946289062,
+      "learning_rate": 1.4866753127222527e-05,
+      "loss": 1.906,
       "step": 6136
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 21.943153381347656,
-      "learning_rate": 7.699709331462364e-06,
-      "loss": 1.2517,
+      "epoch": 0.77,
+      "grad_norm": 15.565106391906738,
+      "learning_rate": 1.4865916412165837e-05,
+      "loss": 1.4209,
       "step": 6137
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 22.251258850097656,
-      "learning_rate": 7.697704720857974e-06,
-      "loss": 1.5025,
+      "epoch": 0.77,
+      "grad_norm": 14.134838104248047,
+      "learning_rate": 1.4865079697109151e-05,
+      "loss": 1.0566,
       "step": 6138
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 15.252081871032715,
-      "learning_rate": 7.695700110253583e-06,
-      "loss": 1.0183,
+      "epoch": 0.77,
+      "grad_norm": 29.05521583557129,
+      "learning_rate": 1.4864242982052463e-05,
+      "loss": 1.4515,
       "step": 6139
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 26.634531021118164,
-      "learning_rate": 7.693695499649194e-06,
-      "loss": 1.5842,
+      "epoch": 0.77,
+      "grad_norm": 23.578039169311523,
+      "learning_rate": 1.4863406266995777e-05,
+      "loss": 1.5878,
       "step": 6140
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 29.84535789489746,
-      "learning_rate": 7.691690889044805e-06,
-      "loss": 1.862,
+      "epoch": 0.77,
+      "grad_norm": 31.573808670043945,
+      "learning_rate": 1.4862569551939088e-05,
+      "loss": 3.083,
       "step": 6141
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 24.699621200561523,
-      "learning_rate": 7.689686278440413e-06,
-      "loss": 1.8952,
+      "epoch": 0.77,
+      "grad_norm": 14.645096778869629,
+      "learning_rate": 1.48617328368824e-05,
+      "loss": 1.0865,
       "step": 6142
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 26.514081954956055,
-      "learning_rate": 7.687681667836023e-06,
-      "loss": 1.948,
+      "epoch": 0.77,
+      "grad_norm": 22.972885131835938,
+      "learning_rate": 1.4860896121825714e-05,
+      "loss": 1.9469,
       "step": 6143
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 11.536233901977539,
-      "learning_rate": 7.685677057231633e-06,
-      "loss": 2.2692,
+      "epoch": 0.77,
+      "grad_norm": 6.411251544952393,
+      "learning_rate": 1.4860059406769024e-05,
+      "loss": 0.3834,
       "step": 6144
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 30.69923973083496,
-      "learning_rate": 7.683672446627243e-06,
-      "loss": 1.2475,
+      "epoch": 0.77,
+      "grad_norm": 21.848386764526367,
+      "learning_rate": 1.4859222691712338e-05,
+      "loss": 1.6994,
       "step": 6145
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 19.856365203857422,
-      "learning_rate": 7.681667836022853e-06,
-      "loss": 1.361,
+      "epoch": 0.77,
+      "grad_norm": 57.35751724243164,
+      "learning_rate": 1.4858385976655652e-05,
+      "loss": 2.9288,
       "step": 6146
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 10.169954299926758,
-      "learning_rate": 7.679663225418463e-06,
-      "loss": 1.3548,
+      "epoch": 0.77,
+      "grad_norm": 10.219704627990723,
+      "learning_rate": 1.4857549261598965e-05,
+      "loss": 0.3866,
       "step": 6147
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 36.704524993896484,
-      "learning_rate": 7.677658614814073e-06,
-      "loss": 1.8048,
+      "epoch": 0.77,
+      "grad_norm": 17.33302879333496,
+      "learning_rate": 1.4856712546542276e-05,
+      "loss": 2.0465,
       "step": 6148
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 14.586971282958984,
-      "learning_rate": 7.675654004209683e-06,
-      "loss": 1.6924,
+      "epoch": 0.77,
+      "grad_norm": 6.763572692871094,
+      "learning_rate": 1.485587583148559e-05,
+      "loss": 0.2575,
       "step": 6149
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 31.322877883911133,
-      "learning_rate": 7.673649393605292e-06,
-      "loss": 2.1292,
+      "epoch": 0.77,
+      "grad_norm": 8.814136505126953,
+      "learning_rate": 1.4855039116428903e-05,
+      "loss": 2.5439,
       "step": 6150
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 13.903470993041992,
-      "learning_rate": 7.671644783000902e-06,
-      "loss": 1.5199,
+      "epoch": 0.77,
+      "grad_norm": 16.717769622802734,
+      "learning_rate": 1.4854202401372213e-05,
+      "loss": 1.1557,
       "step": 6151
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 206.5603485107422,
-      "learning_rate": 7.669640172396514e-06,
-      "loss": 2.1998,
+      "epoch": 0.77,
+      "grad_norm": 10.755870819091797,
+      "learning_rate": 1.4853365686315527e-05,
+      "loss": 1.4779,
       "step": 6152
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 42.4091682434082,
-      "learning_rate": 7.667635561792122e-06,
-      "loss": 1.8714,
+      "epoch": 0.77,
+      "grad_norm": 79.06465911865234,
+      "learning_rate": 1.4852528971258839e-05,
+      "loss": 1.947,
       "step": 6153
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 10.952982902526855,
-      "learning_rate": 7.665630951187732e-06,
-      "loss": 1.7223,
+      "epoch": 0.77,
+      "grad_norm": 13.590725898742676,
+      "learning_rate": 1.4851692256202152e-05,
+      "loss": 2.8569,
       "step": 6154
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 12.478805541992188,
-      "learning_rate": 7.663626340583342e-06,
-      "loss": 0.9292,
+      "epoch": 0.77,
+      "grad_norm": 7.518975734710693,
+      "learning_rate": 1.4850855541145464e-05,
+      "loss": 1.182,
       "step": 6155
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 45.16120910644531,
-      "learning_rate": 7.661621729978952e-06,
-      "loss": 1.1461,
+      "epoch": 0.77,
+      "grad_norm": 8.839305877685547,
+      "learning_rate": 1.4850018826088776e-05,
+      "loss": 1.5005,
       "step": 6156
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 11.00290298461914,
-      "learning_rate": 7.659617119374562e-06,
-      "loss": 1.3505,
+      "epoch": 0.77,
+      "grad_norm": 10.575740814208984,
+      "learning_rate": 1.484918211103209e-05,
+      "loss": 0.9277,
       "step": 6157
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 12.987632751464844,
-      "learning_rate": 7.657612508770173e-06,
-      "loss": 1.1632,
+      "epoch": 0.77,
+      "grad_norm": 11.546903610229492,
+      "learning_rate": 1.48483453959754e-05,
+      "loss": 1.5759,
       "step": 6158
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 23.716419219970703,
-      "learning_rate": 7.655607898165783e-06,
-      "loss": 1.9068,
+      "epoch": 0.77,
+      "grad_norm": 310.0779113769531,
+      "learning_rate": 1.4847508680918714e-05,
+      "loss": 0.823,
       "step": 6159
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 30.917993545532227,
-      "learning_rate": 7.653603287561393e-06,
-      "loss": 1.9522,
+      "epoch": 0.77,
+      "grad_norm": 16.000261306762695,
+      "learning_rate": 1.4846671965862027e-05,
+      "loss": 1.8281,
       "step": 6160
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 9.036660194396973,
-      "learning_rate": 7.651598676957001e-06,
-      "loss": 1.1854,
+      "epoch": 0.77,
+      "grad_norm": 13.16162395477295,
+      "learning_rate": 1.4845835250805341e-05,
+      "loss": 1.8281,
       "step": 6161
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 53.36625289916992,
-      "learning_rate": 7.649594066352611e-06,
-      "loss": 1.6606,
+      "epoch": 0.77,
+      "grad_norm": 37.1568603515625,
+      "learning_rate": 1.4844998535748651e-05,
+      "loss": 2.4353,
       "step": 6162
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 23.467453002929688,
-      "learning_rate": 7.647589455748221e-06,
-      "loss": 1.7637,
+      "epoch": 0.77,
+      "grad_norm": 7.650110721588135,
+      "learning_rate": 1.4844161820691965e-05,
+      "loss": 2.1214,
       "step": 6163
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 16.767295837402344,
-      "learning_rate": 7.645584845143831e-06,
-      "loss": 1.3892,
+      "epoch": 0.77,
+      "grad_norm": 19.80661964416504,
+      "learning_rate": 1.4843325105635279e-05,
+      "loss": 1.6672,
       "step": 6164
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 21.95707893371582,
-      "learning_rate": 7.643580234539441e-06,
-      "loss": 2.415,
+      "epoch": 0.77,
+      "grad_norm": 14.368863105773926,
+      "learning_rate": 1.4842488390578589e-05,
+      "loss": 1.7552,
       "step": 6165
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 14.439309120178223,
-      "learning_rate": 7.641575623935052e-06,
-      "loss": 1.575,
+      "epoch": 0.77,
+      "grad_norm": 14.653712272644043,
+      "learning_rate": 1.4841651675521903e-05,
+      "loss": 0.6312,
       "step": 6166
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 8.983049392700195,
-      "learning_rate": 7.639571013330662e-06,
-      "loss": 1.2169,
+      "epoch": 0.77,
+      "grad_norm": 19.684627532958984,
+      "learning_rate": 1.4840814960465215e-05,
+      "loss": 0.2277,
       "step": 6167
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 15.929500579833984,
-      "learning_rate": 7.63756640272627e-06,
-      "loss": 1.3968,
+      "epoch": 0.77,
+      "grad_norm": 11.68897819519043,
+      "learning_rate": 1.4839978245408528e-05,
+      "loss": 0.9268,
       "step": 6168
     },
     {
-      "epoch": 1.85,
-      "grad_norm": 16.895946502685547,
-      "learning_rate": 7.63556179212188e-06,
-      "loss": 1.773,
+      "epoch": 0.77,
+      "grad_norm": 13.049945831298828,
+      "learning_rate": 1.483914153035184e-05,
+      "loss": 0.775,
       "step": 6169
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 11.67673110961914,
-      "learning_rate": 7.633557181517492e-06,
-      "loss": 0.8359,
+      "epoch": 0.77,
+      "grad_norm": 23.58036231994629,
+      "learning_rate": 1.4838304815295152e-05,
+      "loss": 1.8576,
       "step": 6170
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 12.90839958190918,
-      "learning_rate": 7.6315525709131e-06,
-      "loss": 1.2435,
+      "epoch": 0.77,
+      "grad_norm": 4.773311138153076,
+      "learning_rate": 1.4837468100238466e-05,
+      "loss": 1.0645,
       "step": 6171
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 50.556400299072266,
-      "learning_rate": 7.62954796030871e-06,
-      "loss": 3.8311,
+      "epoch": 0.77,
+      "grad_norm": 20.70108985900879,
+      "learning_rate": 1.4836631385181776e-05,
+      "loss": 3.5102,
       "step": 6172
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 9.137083053588867,
-      "learning_rate": 7.6275433497043204e-06,
-      "loss": 0.9147,
+      "epoch": 0.77,
+      "grad_norm": 8.755514144897461,
+      "learning_rate": 1.483579467012509e-05,
+      "loss": 1.0022,
       "step": 6173
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 39.34935760498047,
-      "learning_rate": 7.6255387390999305e-06,
-      "loss": 2.6975,
+      "epoch": 0.77,
+      "grad_norm": 17.600788116455078,
+      "learning_rate": 1.4834957955068403e-05,
+      "loss": 0.7465,
       "step": 6174
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 14.056702613830566,
-      "learning_rate": 7.62353412849554e-06,
-      "loss": 1.3314,
+      "epoch": 0.77,
+      "grad_norm": 21.20954132080078,
+      "learning_rate": 1.4834121240011717e-05,
+      "loss": 2.1938,
       "step": 6175
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 14.793180465698242,
-      "learning_rate": 7.621529517891151e-06,
-      "loss": 1.8193,
+      "epoch": 0.78,
+      "grad_norm": 9.03451156616211,
+      "learning_rate": 1.4833284524955027e-05,
+      "loss": 1.047,
       "step": 6176
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 6.99417781829834,
-      "learning_rate": 7.619524907286761e-06,
-      "loss": 0.9592,
+      "epoch": 0.78,
+      "grad_norm": 10.841198921203613,
+      "learning_rate": 1.4832447809898341e-05,
+      "loss": 2.5945,
       "step": 6177
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 8.927657127380371,
-      "learning_rate": 7.61752029668237e-06,
-      "loss": 0.7251,
+      "epoch": 0.78,
+      "grad_norm": 8.859786033630371,
+      "learning_rate": 1.4831611094841653e-05,
+      "loss": 1.9457,
       "step": 6178
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 22.89257049560547,
-      "learning_rate": 7.61551568607798e-06,
-      "loss": 2.0944,
+      "epoch": 0.78,
+      "grad_norm": 3.40340518951416,
+      "learning_rate": 1.4830774379784965e-05,
+      "loss": 0.1393,
       "step": 6179
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 33.927913665771484,
-      "learning_rate": 7.613511075473589e-06,
-      "loss": 2.6198,
+      "epoch": 0.78,
+      "grad_norm": 24.379648208618164,
+      "learning_rate": 1.4829937664728278e-05,
+      "loss": 1.8299,
       "step": 6180
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 18.200057983398438,
-      "learning_rate": 7.611506464869199e-06,
-      "loss": 1.3679,
+      "epoch": 0.78,
+      "grad_norm": 10.48060131072998,
+      "learning_rate": 1.482910094967159e-05,
+      "loss": 2.0846,
       "step": 6181
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 17.662506103515625,
-      "learning_rate": 7.6095018542648095e-06,
-      "loss": 2.0992,
+      "epoch": 0.78,
+      "grad_norm": 48.35779571533203,
+      "learning_rate": 1.4828264234614904e-05,
+      "loss": 3.2575,
       "step": 6182
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 11.235939979553223,
-      "learning_rate": 7.60749724366042e-06,
-      "loss": 1.4116,
+      "epoch": 0.78,
+      "grad_norm": 9.917369842529297,
+      "learning_rate": 1.4827427519558214e-05,
+      "loss": 0.4835,
       "step": 6183
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 19.397539138793945,
-      "learning_rate": 7.60549263305603e-06,
-      "loss": 1.3618,
+      "epoch": 0.78,
+      "grad_norm": 4.46663236618042,
+      "learning_rate": 1.4826590804501528e-05,
+      "loss": 0.9349,
       "step": 6184
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 10.240893363952637,
-      "learning_rate": 7.60348802245164e-06,
-      "loss": 1.0369,
+      "epoch": 0.78,
+      "grad_norm": 11.6162691116333,
+      "learning_rate": 1.4825754089444842e-05,
+      "loss": 1.7973,
       "step": 6185
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 14.981084823608398,
-      "learning_rate": 7.601483411847249e-06,
-      "loss": 0.8672,
+      "epoch": 0.78,
+      "grad_norm": 26.446144104003906,
+      "learning_rate": 1.4824917374388152e-05,
+      "loss": 1.2248,
       "step": 6186
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 12.943142890930176,
-      "learning_rate": 7.599478801242859e-06,
-      "loss": 1.19,
+      "epoch": 0.78,
+      "grad_norm": 6.8402838706970215,
+      "learning_rate": 1.4824080659331466e-05,
+      "loss": 0.6867,
       "step": 6187
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 15.80017375946045,
-      "learning_rate": 7.597474190638468e-06,
-      "loss": 1.2749,
+      "epoch": 0.78,
+      "grad_norm": 13.780156135559082,
+      "learning_rate": 1.4823243944274779e-05,
+      "loss": 2.4299,
       "step": 6188
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 15.71901798248291,
-      "learning_rate": 7.595469580034079e-06,
-      "loss": 1.0749,
+      "epoch": 0.78,
+      "grad_norm": 7.482303142547607,
+      "learning_rate": 1.4822407229218093e-05,
+      "loss": 0.7451,
       "step": 6189
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 13.598134994506836,
-      "learning_rate": 7.593464969429689e-06,
-      "loss": 1.2334,
+      "epoch": 0.78,
+      "grad_norm": 10.03032398223877,
+      "learning_rate": 1.4821570514161403e-05,
+      "loss": 0.628,
       "step": 6190
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 13.935446739196777,
-      "learning_rate": 7.5914603588252986e-06,
-      "loss": 1.2212,
+      "epoch": 0.78,
+      "grad_norm": 20.40422248840332,
+      "learning_rate": 1.4820733799104717e-05,
+      "loss": 1.8446,
       "step": 6191
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 28.57442855834961,
-      "learning_rate": 7.589455748220909e-06,
-      "loss": 2.0672,
+      "epoch": 0.78,
+      "grad_norm": 10.26680850982666,
+      "learning_rate": 1.4819897084048029e-05,
+      "loss": 2.0607,
       "step": 6192
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 35.023197174072266,
-      "learning_rate": 7.587451137616518e-06,
-      "loss": 1.8612,
+      "epoch": 0.78,
+      "grad_norm": 16.696704864501953,
+      "learning_rate": 1.481906036899134e-05,
+      "loss": 2.4927,
       "step": 6193
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 11.907183647155762,
-      "learning_rate": 7.585446527012128e-06,
-      "loss": 1.6789,
+      "epoch": 0.78,
+      "grad_norm": 21.497238159179688,
+      "learning_rate": 1.4818223653934654e-05,
+      "loss": 1.3641,
       "step": 6194
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 18.430017471313477,
-      "learning_rate": 7.583441916407739e-06,
-      "loss": 1.3619,
+      "epoch": 0.78,
+      "grad_norm": 5.724583148956299,
+      "learning_rate": 1.4817386938877966e-05,
+      "loss": 0.6211,
       "step": 6195
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 29.030160903930664,
-      "learning_rate": 7.581437305803349e-06,
-      "loss": 1.116,
+      "epoch": 0.78,
+      "grad_norm": 14.418787956237793,
+      "learning_rate": 1.481655022382128e-05,
+      "loss": 1.2559,
       "step": 6196
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 18.958419799804688,
-      "learning_rate": 7.579432695198958e-06,
-      "loss": 1.6912,
+      "epoch": 0.78,
+      "grad_norm": 18.260189056396484,
+      "learning_rate": 1.481571350876459e-05,
+      "loss": 1.4076,
       "step": 6197
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 19.51006317138672,
-      "learning_rate": 7.577428084594568e-06,
-      "loss": 1.6718,
+      "epoch": 0.78,
+      "grad_norm": 16.18914794921875,
+      "learning_rate": 1.4814876793707904e-05,
+      "loss": 1.6051,
       "step": 6198
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 13.113887786865234,
-      "learning_rate": 7.5754234739901775e-06,
-      "loss": 1.2158,
+      "epoch": 0.78,
+      "grad_norm": 11.067285537719727,
+      "learning_rate": 1.4814040078651217e-05,
+      "loss": 1.6897,
       "step": 6199
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 13.66335678100586,
-      "learning_rate": 7.573418863385788e-06,
-      "loss": 0.8829,
+      "epoch": 0.78,
+      "grad_norm": 13.584056854248047,
+      "learning_rate": 1.4813203363594528e-05,
+      "loss": 2.4539,
       "step": 6200
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 12.062017440795898,
-      "learning_rate": 7.5714142527813986e-06,
-      "loss": 1.7451,
+      "epoch": 0.78,
+      "grad_norm": 18.683652877807617,
+      "learning_rate": 1.4812366648537841e-05,
+      "loss": 1.0933,
       "step": 6201
     },
     {
-      "epoch": 1.86,
-      "grad_norm": 12.591232299804688,
-      "learning_rate": 7.569409642177008e-06,
-      "loss": 1.4194,
+      "epoch": 0.78,
+      "grad_norm": 60.37272262573242,
+      "learning_rate": 1.4811529933481155e-05,
+      "loss": 2.5123,
       "step": 6202
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 9.007003784179688,
-      "learning_rate": 7.567405031572618e-06,
-      "loss": 0.6593,
+      "epoch": 0.78,
+      "grad_norm": 9.876708030700684,
+      "learning_rate": 1.4810693218424469e-05,
+      "loss": 0.7678,
       "step": 6203
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 15.050545692443848,
-      "learning_rate": 7.565400420968227e-06,
-      "loss": 1.2698,
+      "epoch": 0.78,
+      "grad_norm": 17.32562255859375,
+      "learning_rate": 1.4809856503367779e-05,
+      "loss": 0.7407,
       "step": 6204
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 53.996299743652344,
-      "learning_rate": 7.563395810363837e-06,
-      "loss": 2.5506,
+      "epoch": 0.78,
+      "grad_norm": 28.528152465820312,
+      "learning_rate": 1.4809019788311093e-05,
+      "loss": 1.6875,
       "step": 6205
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 16.08456039428711,
-      "learning_rate": 7.561391199759447e-06,
-      "loss": 0.7814,
+      "epoch": 0.78,
+      "grad_norm": 30.383686065673828,
+      "learning_rate": 1.4808183073254404e-05,
+      "loss": 2.2793,
       "step": 6206
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 19.180896759033203,
-      "learning_rate": 7.559386589155057e-06,
-      "loss": 2.2722,
+      "epoch": 0.78,
+      "grad_norm": 13.102036476135254,
+      "learning_rate": 1.4807346358197716e-05,
+      "loss": 1.1206,
       "step": 6207
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 12.53034782409668,
-      "learning_rate": 7.5573819785506675e-06,
-      "loss": 1.113,
+      "epoch": 0.78,
+      "grad_norm": 16.857431411743164,
+      "learning_rate": 1.480650964314103e-05,
+      "loss": 2.5142,
       "step": 6208
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 12.597771644592285,
-      "learning_rate": 7.5553773679462775e-06,
-      "loss": 1.7598,
+      "epoch": 0.78,
+      "grad_norm": 9.684059143066406,
+      "learning_rate": 1.4805672928084342e-05,
+      "loss": 1.1404,
       "step": 6209
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 8.483002662658691,
-      "learning_rate": 7.553372757341887e-06,
-      "loss": 0.8353,
+      "epoch": 0.78,
+      "grad_norm": 10.59528923034668,
+      "learning_rate": 1.4804836213027656e-05,
+      "loss": 1.5664,
       "step": 6210
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 9.132856369018555,
-      "learning_rate": 7.551368146737497e-06,
-      "loss": 1.0596,
+      "epoch": 0.78,
+      "grad_norm": 14.23654556274414,
+      "learning_rate": 1.4803999497970966e-05,
+      "loss": 1.0221,
       "step": 6211
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 12.06258773803711,
-      "learning_rate": 7.549363536133106e-06,
-      "loss": 1.3816,
+      "epoch": 0.78,
+      "grad_norm": 7.1277923583984375,
+      "learning_rate": 1.480316278291428e-05,
+      "loss": 1.7848,
       "step": 6212
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 12.495207786560059,
-      "learning_rate": 7.547358925528717e-06,
-      "loss": 2.0172,
+      "epoch": 0.78,
+      "grad_norm": 10.347265243530273,
+      "learning_rate": 1.4802326067857593e-05,
+      "loss": 0.6381,
       "step": 6213
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 15.568368911743164,
-      "learning_rate": 7.545354314924327e-06,
-      "loss": 2.3213,
+      "epoch": 0.78,
+      "grad_norm": 15.681981086730957,
+      "learning_rate": 1.4801489352800904e-05,
+      "loss": 1.7937,
       "step": 6214
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 57.913516998291016,
-      "learning_rate": 7.543349704319936e-06,
-      "loss": 1.5673,
+      "epoch": 0.78,
+      "grad_norm": 12.805305480957031,
+      "learning_rate": 1.4800652637744217e-05,
+      "loss": 2.8912,
       "step": 6215
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 11.138136863708496,
-      "learning_rate": 7.5413450937155464e-06,
-      "loss": 1.1155,
+      "epoch": 0.78,
+      "grad_norm": 11.391907691955566,
+      "learning_rate": 1.479981592268753e-05,
+      "loss": 2.3061,
       "step": 6216
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 36.35742950439453,
-      "learning_rate": 7.539340483111156e-06,
-      "loss": 0.877,
+      "epoch": 0.78,
+      "grad_norm": 13.441726684570312,
+      "learning_rate": 1.4798979207630843e-05,
+      "loss": 1.463,
       "step": 6217
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 11.55097484588623,
-      "learning_rate": 7.537335872506766e-06,
-      "loss": 0.539,
+      "epoch": 0.78,
+      "grad_norm": 11.16281509399414,
+      "learning_rate": 1.4798142492574155e-05,
+      "loss": 1.4862,
       "step": 6218
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 18.280834197998047,
-      "learning_rate": 7.535331261902377e-06,
-      "loss": 1.5513,
+      "epoch": 0.78,
+      "grad_norm": 9.24189567565918,
+      "learning_rate": 1.4797305777517468e-05,
+      "loss": 1.9563,
       "step": 6219
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 30.530216217041016,
-      "learning_rate": 7.533326651297986e-06,
-      "loss": 1.9069,
+      "epoch": 0.78,
+      "grad_norm": 39.83635330200195,
+      "learning_rate": 1.479646906246078e-05,
+      "loss": 2.1532,
       "step": 6220
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 28.866329193115234,
-      "learning_rate": 7.531322040693596e-06,
-      "loss": 1.4386,
+      "epoch": 0.78,
+      "grad_norm": 21.487403869628906,
+      "learning_rate": 1.4795632347404092e-05,
+      "loss": 1.46,
       "step": 6221
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 22.6090145111084,
-      "learning_rate": 7.529317430089206e-06,
-      "loss": 2.3103,
+      "epoch": 0.78,
+      "grad_norm": 16.41718292236328,
+      "learning_rate": 1.4794795632347406e-05,
+      "loss": 2.1983,
       "step": 6222
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 9.455753326416016,
-      "learning_rate": 7.527312819484815e-06,
-      "loss": 1.0053,
+      "epoch": 0.78,
+      "grad_norm": 30.581064224243164,
+      "learning_rate": 1.4793958917290718e-05,
+      "loss": 1.1738,
       "step": 6223
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 47.22819519042969,
-      "learning_rate": 7.525308208880425e-06,
-      "loss": 2.0746,
+      "epoch": 0.78,
+      "grad_norm": 11.414182662963867,
+      "learning_rate": 1.4793122202234032e-05,
+      "loss": 1.7153,
       "step": 6224
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 16.05484962463379,
-      "learning_rate": 7.523303598276035e-06,
-      "loss": 1.4145,
+      "epoch": 0.78,
+      "grad_norm": 12.448118209838867,
+      "learning_rate": 1.4792285487177342e-05,
+      "loss": 1.2993,
       "step": 6225
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 23.540828704833984,
-      "learning_rate": 7.521298987671646e-06,
-      "loss": 1.1681,
+      "epoch": 0.78,
+      "grad_norm": 7.1149373054504395,
+      "learning_rate": 1.4791448772120655e-05,
+      "loss": 0.6568,
       "step": 6226
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 14.86224365234375,
-      "learning_rate": 7.519294377067256e-06,
-      "loss": 1.6844,
+      "epoch": 0.78,
+      "grad_norm": 6.569999694824219,
+      "learning_rate": 1.4790612057063969e-05,
+      "loss": 0.5046,
       "step": 6227
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 76.27838134765625,
-      "learning_rate": 7.517289766462865e-06,
-      "loss": 1.9908,
+      "epoch": 0.78,
+      "grad_norm": 14.83082103729248,
+      "learning_rate": 1.478977534200728e-05,
+      "loss": 1.4867,
       "step": 6228
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 18.715179443359375,
-      "learning_rate": 7.515285155858475e-06,
-      "loss": 1.2129,
+      "epoch": 0.78,
+      "grad_norm": 19.81695556640625,
+      "learning_rate": 1.4788938626950593e-05,
+      "loss": 1.2355,
       "step": 6229
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 16.809297561645508,
-      "learning_rate": 7.513280545254085e-06,
-      "loss": 1.1011,
+      "epoch": 0.78,
+      "grad_norm": 7.853550434112549,
+      "learning_rate": 1.4788101911893907e-05,
+      "loss": 0.8718,
       "step": 6230
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 8.646817207336426,
-      "learning_rate": 7.511275934649694e-06,
-      "loss": 1.5495,
+      "epoch": 0.78,
+      "grad_norm": 16.03679656982422,
+      "learning_rate": 1.4787265196837219e-05,
+      "loss": 2.5044,
       "step": 6231
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 35.9256591796875,
-      "learning_rate": 7.509271324045305e-06,
-      "loss": 1.7407,
+      "epoch": 0.78,
+      "grad_norm": 10.5372896194458,
+      "learning_rate": 1.478642848178053e-05,
+      "loss": 1.4197,
       "step": 6232
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 8.104480743408203,
-      "learning_rate": 7.507266713440915e-06,
-      "loss": 0.5633,
+      "epoch": 0.78,
+      "grad_norm": 6.5133466720581055,
+      "learning_rate": 1.4785591766723844e-05,
+      "loss": 1.7093,
       "step": 6233
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 10.297518730163574,
-      "learning_rate": 7.5052621028365246e-06,
-      "loss": 1.2832,
+      "epoch": 0.78,
+      "grad_norm": 11.621196746826172,
+      "learning_rate": 1.4784755051667156e-05,
+      "loss": 1.5063,
       "step": 6234
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 15.735064506530762,
-      "learning_rate": 7.503257492232135e-06,
-      "loss": 1.4141,
+      "epoch": 0.78,
+      "grad_norm": 31.792678833007812,
+      "learning_rate": 1.4783918336610468e-05,
+      "loss": 1.9922,
       "step": 6235
     },
     {
-      "epoch": 1.87,
-      "grad_norm": 12.001105308532715,
-      "learning_rate": 7.501252881627744e-06,
-      "loss": 1.4383,
+      "epoch": 0.78,
+      "grad_norm": 23.694442749023438,
+      "learning_rate": 1.478308162155378e-05,
+      "loss": 1.438,
       "step": 6236
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 15.454205513000488,
-      "learning_rate": 7.499248271023354e-06,
-      "loss": 1.0,
+      "epoch": 0.78,
+      "grad_norm": 3.3710951805114746,
+      "learning_rate": 1.4782244906497094e-05,
+      "loss": 0.1078,
       "step": 6237
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 19.108205795288086,
-      "learning_rate": 7.497243660418965e-06,
-      "loss": 1.6183,
+      "epoch": 0.78,
+      "grad_norm": 11.370076179504395,
+      "learning_rate": 1.4781408191440407e-05,
+      "loss": 0.801,
       "step": 6238
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 38.10220718383789,
-      "learning_rate": 7.495239049814574e-06,
-      "loss": 2.5108,
+      "epoch": 0.78,
+      "grad_norm": 33.836185455322266,
+      "learning_rate": 1.4780571476383718e-05,
+      "loss": 1.8408,
       "step": 6239
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 10.768503189086914,
-      "learning_rate": 7.493234439210184e-06,
-      "loss": 1.4355,
-      "step": 6240
-    },
-    {
-      "epoch": 1.88,
-      "eval_loss": 0.19893887639045715,
-      "eval_runtime": 43.6319,
-      "eval_samples_per_second": 33.897,
-      "eval_steps_per_second": 33.897,
+      "epoch": 0.78,
+      "grad_norm": 6.308077812194824,
+      "learning_rate": 1.4779734761327031e-05,
+      "loss": 0.2894,
       "step": 6240
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 7.937902450561523,
-      "learning_rate": 7.4912298286057935e-06,
-      "loss": 0.9729,
+      "epoch": 0.78,
+      "grad_norm": 7.155177593231201,
+      "learning_rate": 1.4778898046270345e-05,
+      "loss": 0.2025,
       "step": 6241
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 10.797032356262207,
-      "learning_rate": 7.4892252180014035e-06,
-      "loss": 0.9792,
+      "epoch": 0.78,
+      "grad_norm": 29.36728858947754,
+      "learning_rate": 1.4778061331213655e-05,
+      "loss": 1.8414,
       "step": 6242
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 8.380309104919434,
-      "learning_rate": 7.487220607397014e-06,
-      "loss": 0.7776,
+      "epoch": 0.78,
+      "grad_norm": 24.564516067504883,
+      "learning_rate": 1.4777224616156969e-05,
+      "loss": 2.6769,
       "step": 6243
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 10.550333023071289,
-      "learning_rate": 7.485215996792624e-06,
-      "loss": 1.1951,
+      "epoch": 0.78,
+      "grad_norm": 18.393569946289062,
+      "learning_rate": 1.4776387901100282e-05,
+      "loss": 2.8948,
       "step": 6244
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 16.15055274963379,
-      "learning_rate": 7.483211386188234e-06,
-      "loss": 0.9995,
+      "epoch": 0.78,
+      "grad_norm": 8.56533432006836,
+      "learning_rate": 1.4775551186043593e-05,
+      "loss": 2.4428,
       "step": 6245
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 18.002609252929688,
-      "learning_rate": 7.481206775583844e-06,
-      "loss": 1.6159,
+      "epoch": 0.78,
+      "grad_norm": 83.42866516113281,
+      "learning_rate": 1.4774714470986906e-05,
+      "loss": 2.374,
       "step": 6246
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 19.96839714050293,
-      "learning_rate": 7.479202164979453e-06,
-      "loss": 1.9925,
+      "epoch": 0.78,
+      "grad_norm": 40.70931625366211,
+      "learning_rate": 1.477387775593022e-05,
+      "loss": 1.0785,
       "step": 6247
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 22.932233810424805,
-      "learning_rate": 7.477197554375063e-06,
-      "loss": 1.5709,
+      "epoch": 0.78,
+      "grad_norm": 11.857887268066406,
+      "learning_rate": 1.4773041040873532e-05,
+      "loss": 1.0883,
       "step": 6248
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 10.494919776916504,
-      "learning_rate": 7.4751929437706724e-06,
-      "loss": 1.1168,
+      "epoch": 0.78,
+      "grad_norm": 11.886443138122559,
+      "learning_rate": 1.4772204325816844e-05,
+      "loss": 1.2131,
       "step": 6249
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 37.70262145996094,
-      "learning_rate": 7.473188333166283e-06,
-      "loss": 1.6439,
+      "epoch": 0.78,
+      "grad_norm": 20.011035919189453,
+      "learning_rate": 1.4771367610760156e-05,
+      "loss": 1.8072,
       "step": 6250
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 14.26046085357666,
-      "learning_rate": 7.4711837225618935e-06,
-      "loss": 1.4012,
+      "epoch": 0.78,
+      "grad_norm": 10.199731826782227,
+      "learning_rate": 1.477053089570347e-05,
+      "loss": 0.7874,
       "step": 6251
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 19.355712890625,
-      "learning_rate": 7.469179111957503e-06,
-      "loss": 1.3238,
+      "epoch": 0.78,
+      "grad_norm": 6.612044334411621,
+      "learning_rate": 1.4769694180646782e-05,
+      "loss": 0.6469,
       "step": 6252
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 43.30799865722656,
-      "learning_rate": 7.467174501353113e-06,
-      "loss": 2.3852,
+      "epoch": 0.78,
+      "grad_norm": 10.592775344848633,
+      "learning_rate": 1.4768857465590093e-05,
+      "loss": 0.3784,
       "step": 6253
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 17.540157318115234,
-      "learning_rate": 7.465169890748723e-06,
-      "loss": 0.9667,
+      "epoch": 0.78,
+      "grad_norm": 10.249109268188477,
+      "learning_rate": 1.4768020750533407e-05,
+      "loss": 1.2488,
       "step": 6254
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 14.730655670166016,
-      "learning_rate": 7.463165280144332e-06,
-      "loss": 1.454,
+      "epoch": 0.78,
+      "grad_norm": 22.15861701965332,
+      "learning_rate": 1.476718403547672e-05,
+      "loss": 1.2772,
       "step": 6255
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 11.64047908782959,
-      "learning_rate": 7.461160669539943e-06,
-      "loss": 1.209,
+      "epoch": 0.79,
+      "grad_norm": 15.571314811706543,
+      "learning_rate": 1.4766347320420031e-05,
+      "loss": 1.6813,
       "step": 6256
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 8.219613075256348,
-      "learning_rate": 7.459156058935553e-06,
-      "loss": 1.3621,
+      "epoch": 0.79,
+      "grad_norm": 10.145541191101074,
+      "learning_rate": 1.4765510605363345e-05,
+      "loss": 2.0737,
       "step": 6257
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 15.764402389526367,
-      "learning_rate": 7.457151448331162e-06,
-      "loss": 1.7374,
+      "epoch": 0.79,
+      "grad_norm": 11.789886474609375,
+      "learning_rate": 1.4764673890306658e-05,
+      "loss": 1.0396,
       "step": 6258
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 17.738845825195312,
-      "learning_rate": 7.4551468377267724e-06,
-      "loss": 1.5129,
+      "epoch": 0.79,
+      "grad_norm": 29.804969787597656,
+      "learning_rate": 1.4763837175249969e-05,
+      "loss": 3.3272,
       "step": 6259
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 18.939716339111328,
-      "learning_rate": 7.453142227122382e-06,
-      "loss": 1.6715,
+      "epoch": 0.79,
+      "grad_norm": 16.483253479003906,
+      "learning_rate": 1.4763000460193282e-05,
+      "loss": 1.5897,
       "step": 6260
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 25.509105682373047,
-      "learning_rate": 7.451137616517992e-06,
-      "loss": 1.6497,
+      "epoch": 0.79,
+      "grad_norm": 11.531888008117676,
+      "learning_rate": 1.4762163745136596e-05,
+      "loss": 2.6644,
       "step": 6261
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 30.341272354125977,
-      "learning_rate": 7.449133005913603e-06,
-      "loss": 2.2224,
+      "epoch": 0.79,
+      "grad_norm": 6.730110168457031,
+      "learning_rate": 1.4761327030079908e-05,
+      "loss": 0.5048,
       "step": 6262
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 29.086957931518555,
-      "learning_rate": 7.447128395309212e-06,
-      "loss": 2.018,
+      "epoch": 0.79,
+      "grad_norm": 36.700557708740234,
+      "learning_rate": 1.476049031502322e-05,
+      "loss": 3.3426,
       "step": 6263
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 13.672849655151367,
-      "learning_rate": 7.445123784704822e-06,
-      "loss": 1.4355,
+      "epoch": 0.79,
+      "grad_norm": 25.597129821777344,
+      "learning_rate": 1.4759653599966532e-05,
+      "loss": 2.0085,
       "step": 6264
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 27.003616333007812,
-      "learning_rate": 7.443119174100431e-06,
-      "loss": 1.4442,
+      "epoch": 0.79,
+      "grad_norm": 19.021533966064453,
+      "learning_rate": 1.4758816884909845e-05,
+      "loss": 1.959,
       "step": 6265
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 19.808364868164062,
-      "learning_rate": 7.441114563496041e-06,
-      "loss": 1.2619,
+      "epoch": 0.79,
+      "grad_norm": 169.17330932617188,
+      "learning_rate": 1.4757980169853157e-05,
+      "loss": 1.5518,
       "step": 6266
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 17.011503219604492,
-      "learning_rate": 7.439109952891651e-06,
-      "loss": 1.2087,
+      "epoch": 0.79,
+      "grad_norm": 27.448816299438477,
+      "learning_rate": 1.475714345479647e-05,
+      "loss": 2.5831,
       "step": 6267
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 16.227067947387695,
-      "learning_rate": 7.437105342287261e-06,
-      "loss": 2.4985,
+      "epoch": 0.79,
+      "grad_norm": 8.960739135742188,
+      "learning_rate": 1.4756306739739783e-05,
+      "loss": 1.958,
       "step": 6268
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 15.978217124938965,
-      "learning_rate": 7.435100731682872e-06,
-      "loss": 1.3265,
+      "epoch": 0.79,
+      "grad_norm": 11.165156364440918,
+      "learning_rate": 1.4755470024683097e-05,
+      "loss": 1.4424,
       "step": 6269
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 16.302669525146484,
-      "learning_rate": 7.433096121078482e-06,
-      "loss": 0.8265,
+      "epoch": 0.79,
+      "grad_norm": 9.504926681518555,
+      "learning_rate": 1.4754633309626407e-05,
+      "loss": 0.9984,
       "step": 6270
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 20.792295455932617,
-      "learning_rate": 7.431091510474091e-06,
-      "loss": 1.4503,
+      "epoch": 0.79,
+      "grad_norm": 10.795631408691406,
+      "learning_rate": 1.475379659456972e-05,
+      "loss": 0.886,
       "step": 6271
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 9.446883201599121,
-      "learning_rate": 7.429086899869701e-06,
-      "loss": 1.8997,
+      "epoch": 0.79,
+      "grad_norm": 9.808711051940918,
+      "learning_rate": 1.4752959879513034e-05,
+      "loss": 1.8094,
       "step": 6272
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 24.59245491027832,
-      "learning_rate": 7.42708228926531e-06,
-      "loss": 1.1568,
+      "epoch": 0.79,
+      "grad_norm": 13.138855934143066,
+      "learning_rate": 1.4752123164456344e-05,
+      "loss": 2.1202,
       "step": 6273
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 13.666716575622559,
-      "learning_rate": 7.42507767866092e-06,
-      "loss": 0.9083,
+      "epoch": 0.79,
+      "grad_norm": 4.985891819000244,
+      "learning_rate": 1.4751286449399658e-05,
+      "loss": 0.0602,
       "step": 6274
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 17.118921279907227,
-      "learning_rate": 7.423073068056531e-06,
-      "loss": 1.5059,
+      "epoch": 0.79,
+      "grad_norm": 10.045243263244629,
+      "learning_rate": 1.4750449734342972e-05,
+      "loss": 1.6623,
       "step": 6275
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 56.16904067993164,
-      "learning_rate": 7.4210684574521405e-06,
-      "loss": 3.6439,
+      "epoch": 0.79,
+      "grad_norm": 13.370321273803711,
+      "learning_rate": 1.4749613019286284e-05,
+      "loss": 2.2939,
       "step": 6276
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 13.676959037780762,
-      "learning_rate": 7.4190638468477506e-06,
-      "loss": 1.6671,
+      "epoch": 0.79,
+      "grad_norm": 12.222705841064453,
+      "learning_rate": 1.4748776304229596e-05,
+      "loss": 1.8068,
       "step": 6277
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 10.493898391723633,
-      "learning_rate": 7.41705923624336e-06,
-      "loss": 1.4543,
+      "epoch": 0.79,
+      "grad_norm": 74.30975341796875,
+      "learning_rate": 1.4747939589172908e-05,
+      "loss": 4.0831,
       "step": 6278
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 22.156248092651367,
-      "learning_rate": 7.41505462563897e-06,
-      "loss": 1.5765,
+      "epoch": 0.79,
+      "grad_norm": 57.3051872253418,
+      "learning_rate": 1.4747102874116221e-05,
+      "loss": 2.1201,
       "step": 6279
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 15.999430656433105,
-      "learning_rate": 7.41305001503458e-06,
-      "loss": 1.7229,
+      "epoch": 0.79,
+      "grad_norm": 21.052371978759766,
+      "learning_rate": 1.4746266159059533e-05,
+      "loss": 2.073,
       "step": 6280
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 10.846489906311035,
-      "learning_rate": 7.411045404430191e-06,
-      "loss": 1.9086,
+      "epoch": 0.79,
+      "grad_norm": 29.350799560546875,
+      "learning_rate": 1.4745429444002845e-05,
+      "loss": 1.2385,
       "step": 6281
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 17.6227970123291,
-      "learning_rate": 7.4090407938258e-06,
-      "loss": 1.2638,
+      "epoch": 0.79,
+      "grad_norm": 6.653513431549072,
+      "learning_rate": 1.4744592728946159e-05,
+      "loss": 0.3835,
       "step": 6282
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 16.64601707458496,
-      "learning_rate": 7.40703618322141e-06,
-      "loss": 1.2801,
+      "epoch": 0.79,
+      "grad_norm": 13.995912551879883,
+      "learning_rate": 1.4743756013889472e-05,
+      "loss": 1.3589,
       "step": 6283
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 18.028026580810547,
-      "learning_rate": 7.4050315726170195e-06,
-      "loss": 1.4945,
+      "epoch": 0.79,
+      "grad_norm": 8.13442325592041,
+      "learning_rate": 1.4742919298832783e-05,
+      "loss": 2.8616,
       "step": 6284
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 35.858795166015625,
-      "learning_rate": 7.4030269620126295e-06,
-      "loss": 1.969,
+      "epoch": 0.79,
+      "grad_norm": 12.456286430358887,
+      "learning_rate": 1.4742082583776096e-05,
+      "loss": 1.8573,
       "step": 6285
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 9.260512351989746,
-      "learning_rate": 7.401022351408239e-06,
-      "loss": 1.0171,
+      "epoch": 0.79,
+      "grad_norm": 9.940149307250977,
+      "learning_rate": 1.474124586871941e-05,
+      "loss": 3.2977,
       "step": 6286
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 28.248775482177734,
-      "learning_rate": 7.39901774080385e-06,
-      "loss": 1.2457,
+      "epoch": 0.79,
+      "grad_norm": 48.481712341308594,
+      "learning_rate": 1.474040915366272e-05,
+      "loss": 4.0217,
       "step": 6287
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 23.339683532714844,
-      "learning_rate": 7.39701313019946e-06,
-      "loss": 1.8959,
+      "epoch": 0.79,
+      "grad_norm": 5.557957172393799,
+      "learning_rate": 1.4739572438606034e-05,
+      "loss": 0.6419,
       "step": 6288
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 16.68250274658203,
-      "learning_rate": 7.395008519595069e-06,
-      "loss": 1.1061,
+      "epoch": 0.79,
+      "grad_norm": 12.64964485168457,
+      "learning_rate": 1.4738735723549346e-05,
+      "loss": 0.9878,
       "step": 6289
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 22.26504898071289,
-      "learning_rate": 7.393003908990679e-06,
-      "loss": 1.7896,
+      "epoch": 0.79,
+      "grad_norm": 28.808948516845703,
+      "learning_rate": 1.473789900849266e-05,
+      "loss": 1.926,
       "step": 6290
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 16.31560707092285,
-      "learning_rate": 7.390999298386289e-06,
-      "loss": 0.9241,
+      "epoch": 0.79,
+      "grad_norm": 8.438365936279297,
+      "learning_rate": 1.4737062293435971e-05,
+      "loss": 1.3627,
       "step": 6291
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 132.97479248046875,
-      "learning_rate": 7.3889946877818984e-06,
-      "loss": 2.0392,
+      "epoch": 0.79,
+      "grad_norm": 46.933963775634766,
+      "learning_rate": 1.4736225578379283e-05,
+      "loss": 3.0685,
       "step": 6292
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 26.72028350830078,
-      "learning_rate": 7.386990077177509e-06,
-      "loss": 1.2421,
+      "epoch": 0.79,
+      "grad_norm": 8.370945930480957,
+      "learning_rate": 1.4735388863322597e-05,
+      "loss": 1.3335,
       "step": 6293
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 7.662388324737549,
-      "learning_rate": 7.3849854665731195e-06,
-      "loss": 1.324,
+      "epoch": 0.79,
+      "grad_norm": 10.044309616088867,
+      "learning_rate": 1.4734552148265907e-05,
+      "loss": 1.9237,
       "step": 6294
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 17.365476608276367,
-      "learning_rate": 7.382980855968729e-06,
-      "loss": 1.5321,
+      "epoch": 0.79,
+      "grad_norm": 12.739347457885742,
+      "learning_rate": 1.4733715433209221e-05,
+      "loss": 0.7081,
       "step": 6295
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 12.46912956237793,
-      "learning_rate": 7.380976245364339e-06,
-      "loss": 1.4461,
+      "epoch": 0.79,
+      "grad_norm": 9.426420211791992,
+      "learning_rate": 1.4732878718152535e-05,
+      "loss": 1.0961,
       "step": 6296
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 11.624770164489746,
-      "learning_rate": 7.378971634759948e-06,
-      "loss": 1.7608,
+      "epoch": 0.79,
+      "grad_norm": 9.006810188293457,
+      "learning_rate": 1.4732042003095848e-05,
+      "loss": 1.2268,
       "step": 6297
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 18.62898826599121,
-      "learning_rate": 7.376967024155558e-06,
-      "loss": 1.0102,
+      "epoch": 0.79,
+      "grad_norm": 30.43055534362793,
+      "learning_rate": 1.4731205288039159e-05,
+      "loss": 1.4886,
       "step": 6298
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 7.544929504394531,
-      "learning_rate": 7.374962413551169e-06,
-      "loss": 0.8058,
+      "epoch": 0.79,
+      "grad_norm": 21.447704315185547,
+      "learning_rate": 1.4730368572982472e-05,
+      "loss": 2.3547,
       "step": 6299
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 8.430646896362305,
-      "learning_rate": 7.372957802946778e-06,
-      "loss": 1.2943,
+      "epoch": 0.79,
+      "grad_norm": 7.026731967926025,
+      "learning_rate": 1.4729531857925786e-05,
+      "loss": 0.6442,
       "step": 6300
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 16.084999084472656,
-      "learning_rate": 7.370953192342388e-06,
-      "loss": 1.6322,
+      "epoch": 0.79,
+      "grad_norm": 4.663534641265869,
+      "learning_rate": 1.4728695142869096e-05,
+      "loss": 0.3037,
       "step": 6301
     },
     {
-      "epoch": 1.89,
-      "grad_norm": 26.333581924438477,
-      "learning_rate": 7.368948581737998e-06,
-      "loss": 1.3346,
+      "epoch": 0.79,
+      "grad_norm": 25.217287063598633,
+      "learning_rate": 1.472785842781241e-05,
+      "loss": 1.9375,
       "step": 6302
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 12.999780654907227,
-      "learning_rate": 7.366943971133608e-06,
-      "loss": 1.312,
+      "epoch": 0.79,
+      "grad_norm": 3.978163719177246,
+      "learning_rate": 1.4727021712755722e-05,
+      "loss": 0.4616,
       "step": 6303
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 20.878276824951172,
-      "learning_rate": 7.364939360529218e-06,
-      "loss": 1.235,
+      "epoch": 0.79,
+      "grad_norm": 13.031975746154785,
+      "learning_rate": 1.4726184997699035e-05,
+      "loss": 1.4241,
       "step": 6304
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 39.3919677734375,
-      "learning_rate": 7.362934749924827e-06,
-      "loss": 1.3573,
+      "epoch": 0.79,
+      "grad_norm": 33.387062072753906,
+      "learning_rate": 1.4725348282642347e-05,
+      "loss": 2.9831,
       "step": 6305
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 42.52452087402344,
-      "learning_rate": 7.360930139320438e-06,
-      "loss": 2.1336,
+      "epoch": 0.79,
+      "grad_norm": 11.21850872039795,
+      "learning_rate": 1.472451156758566e-05,
+      "loss": 2.8881,
       "step": 6306
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 10.164920806884766,
-      "learning_rate": 7.358925528716048e-06,
-      "loss": 1.1969,
+      "epoch": 0.79,
+      "grad_norm": 22.746912002563477,
+      "learning_rate": 1.4723674852528973e-05,
+      "loss": 2.044,
       "step": 6307
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 14.353740692138672,
-      "learning_rate": 7.356920918111657e-06,
-      "loss": 1.6327,
+      "epoch": 0.79,
+      "grad_norm": 14.89169692993164,
+      "learning_rate": 1.4722838137472283e-05,
+      "loss": 1.5603,
       "step": 6308
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 11.760319709777832,
-      "learning_rate": 7.354916307507267e-06,
-      "loss": 1.8274,
+      "epoch": 0.79,
+      "grad_norm": 9.180404663085938,
+      "learning_rate": 1.4722001422415597e-05,
+      "loss": 3.4738,
       "step": 6309
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 65.9858169555664,
-      "learning_rate": 7.3529116969028766e-06,
-      "loss": 1.0124,
+      "epoch": 0.79,
+      "grad_norm": 5.776002407073975,
+      "learning_rate": 1.472116470735891e-05,
+      "loss": 0.4565,
       "step": 6310
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 13.734026908874512,
-      "learning_rate": 7.350907086298487e-06,
-      "loss": 1.2509,
+      "epoch": 0.79,
+      "grad_norm": 9.058882713317871,
+      "learning_rate": 1.4720327992302224e-05,
+      "loss": 0.7662,
       "step": 6311
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 21.561870574951172,
-      "learning_rate": 7.348902475694098e-06,
-      "loss": 1.3681,
+      "epoch": 0.79,
+      "grad_norm": 26.460262298583984,
+      "learning_rate": 1.4719491277245534e-05,
+      "loss": 2.7587,
       "step": 6312
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 51.21876525878906,
-      "learning_rate": 7.346897865089707e-06,
-      "loss": 2.3598,
+      "epoch": 0.79,
+      "grad_norm": 13.0591402053833,
+      "learning_rate": 1.4718654562188848e-05,
+      "loss": 0.6505,
       "step": 6313
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 35.55440139770508,
-      "learning_rate": 7.344893254485317e-06,
-      "loss": 1.6387,
+      "epoch": 0.79,
+      "grad_norm": 13.908039093017578,
+      "learning_rate": 1.4717817847132162e-05,
+      "loss": 0.9457,
       "step": 6314
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 9.630316734313965,
-      "learning_rate": 7.342888643880927e-06,
-      "loss": 1.0267,
+      "epoch": 0.79,
+      "grad_norm": 18.12594223022461,
+      "learning_rate": 1.4716981132075472e-05,
+      "loss": 1.9944,
       "step": 6315
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 17.24791717529297,
-      "learning_rate": 7.340884033276536e-06,
-      "loss": 2.4284,
+      "epoch": 0.79,
+      "grad_norm": 7.108876705169678,
+      "learning_rate": 1.4716144417018786e-05,
+      "loss": 0.989,
       "step": 6316
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 12.127941131591797,
-      "learning_rate": 7.338879422672146e-06,
-      "loss": 1.3257,
+      "epoch": 0.79,
+      "grad_norm": 8.037245750427246,
+      "learning_rate": 1.4715307701962098e-05,
+      "loss": 0.368,
       "step": 6317
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 11.26683235168457,
-      "learning_rate": 7.336874812067757e-06,
-      "loss": 0.946,
+      "epoch": 0.79,
+      "grad_norm": 20.03801155090332,
+      "learning_rate": 1.4714470986905411e-05,
+      "loss": 2.1021,
       "step": 6318
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 7.805086612701416,
-      "learning_rate": 7.3348702014633665e-06,
-      "loss": 0.6828,
+      "epoch": 0.79,
+      "grad_norm": 10.88740348815918,
+      "learning_rate": 1.4713634271848723e-05,
+      "loss": 0.5601,
       "step": 6319
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 124.16524505615234,
-      "learning_rate": 7.3328655908589766e-06,
-      "loss": 2.6128,
+      "epoch": 0.79,
+      "grad_norm": 13.821710586547852,
+      "learning_rate": 1.4712797556792035e-05,
+      "loss": 2.988,
       "step": 6320
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 26.658418655395508,
-      "learning_rate": 7.330860980254586e-06,
-      "loss": 1.359,
+      "epoch": 0.79,
+      "grad_norm": 25.372257232666016,
+      "learning_rate": 1.4711960841735349e-05,
+      "loss": 2.3201,
       "step": 6321
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 63.66525650024414,
-      "learning_rate": 7.328856369650196e-06,
-      "loss": 2.1401,
+      "epoch": 0.79,
+      "grad_norm": 18.005115509033203,
+      "learning_rate": 1.4711124126678659e-05,
+      "loss": 2.3348,
       "step": 6322
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 19.522897720336914,
-      "learning_rate": 7.326851759045805e-06,
-      "loss": 1.6017,
+      "epoch": 0.79,
+      "grad_norm": 11.153677940368652,
+      "learning_rate": 1.4710287411621973e-05,
+      "loss": 1.6604,
       "step": 6323
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 18.54653549194336,
-      "learning_rate": 7.324847148441416e-06,
-      "loss": 1.902,
+      "epoch": 0.79,
+      "grad_norm": 7.8768510818481445,
+      "learning_rate": 1.4709450696565286e-05,
+      "loss": 0.9796,
       "step": 6324
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 40.539337158203125,
-      "learning_rate": 7.322842537837026e-06,
-      "loss": 1.6362,
+      "epoch": 0.79,
+      "grad_norm": 7.6989874839782715,
+      "learning_rate": 1.47086139815086e-05,
+      "loss": 0.8554,
       "step": 6325
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 24.200485229492188,
-      "learning_rate": 7.320837927232635e-06,
-      "loss": 2.5706,
+      "epoch": 0.79,
+      "grad_norm": 16.05092430114746,
+      "learning_rate": 1.470777726645191e-05,
+      "loss": 0.9129,
       "step": 6326
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 38.600364685058594,
-      "learning_rate": 7.3188333166282455e-06,
-      "loss": 2.5773,
+      "epoch": 0.79,
+      "grad_norm": 20.265979766845703,
+      "learning_rate": 1.4706940551395224e-05,
+      "loss": 1.0729,
       "step": 6327
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 13.316827774047852,
-      "learning_rate": 7.3168287060238555e-06,
-      "loss": 1.6111,
+      "epoch": 0.79,
+      "grad_norm": 17.98265266418457,
+      "learning_rate": 1.4706103836338536e-05,
+      "loss": 1.3576,
       "step": 6328
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 20.68040657043457,
-      "learning_rate": 7.314824095419465e-06,
-      "loss": 1.8822,
+      "epoch": 0.79,
+      "grad_norm": 10.445375442504883,
+      "learning_rate": 1.4705267121281848e-05,
+      "loss": 1.1647,
       "step": 6329
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 20.918405532836914,
-      "learning_rate": 7.312819484815076e-06,
-      "loss": 0.873,
+      "epoch": 0.79,
+      "grad_norm": 21.013591766357422,
+      "learning_rate": 1.4704430406225161e-05,
+      "loss": 1.1448,
       "step": 6330
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 19.907466888427734,
-      "learning_rate": 7.310814874210686e-06,
-      "loss": 1.2545,
+      "epoch": 0.79,
+      "grad_norm": 17.206157684326172,
+      "learning_rate": 1.4703593691168473e-05,
+      "loss": 2.1738,
       "step": 6331
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 16.205663681030273,
-      "learning_rate": 7.308810263606295e-06,
-      "loss": 1.2967,
+      "epoch": 0.79,
+      "grad_norm": 15.27292251586914,
+      "learning_rate": 1.4702756976111787e-05,
+      "loss": 2.8595,
       "step": 6332
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 13.830738067626953,
-      "learning_rate": 7.306805653001905e-06,
-      "loss": 0.9852,
+      "epoch": 0.79,
+      "grad_norm": 14.216376304626465,
+      "learning_rate": 1.4701920261055099e-05,
+      "loss": 1.8285,
       "step": 6333
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 12.994755744934082,
-      "learning_rate": 7.304801042397514e-06,
-      "loss": 1.0101,
+      "epoch": 0.79,
+      "grad_norm": 71.95661163330078,
+      "learning_rate": 1.4701083545998411e-05,
+      "loss": 3.5818,
       "step": 6334
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 10.922076225280762,
-      "learning_rate": 7.3027964317931244e-06,
-      "loss": 1.1132,
+      "epoch": 0.8,
+      "grad_norm": 8.187115669250488,
+      "learning_rate": 1.4700246830941725e-05,
+      "loss": 2.0697,
       "step": 6335
     },
     {
-      "epoch": 1.9,
-      "grad_norm": 14.2009859085083,
-      "learning_rate": 7.300791821188735e-06,
-      "loss": 1.1271,
+      "epoch": 0.8,
+      "grad_norm": 11.393028259277344,
+      "learning_rate": 1.4699410115885035e-05,
+      "loss": 0.9834,
       "step": 6336
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 13.78223705291748,
-      "learning_rate": 7.298787210584345e-06,
-      "loss": 1.4629,
+      "epoch": 0.8,
+      "grad_norm": 25.30308723449707,
+      "learning_rate": 1.4698573400828349e-05,
+      "loss": 4.2712,
       "step": 6337
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 47.76295852661133,
-      "learning_rate": 7.296782599979955e-06,
-      "loss": 1.6983,
+      "epoch": 0.8,
+      "grad_norm": 18.839860916137695,
+      "learning_rate": 1.4697736685771662e-05,
+      "loss": 2.2294,
       "step": 6338
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 26.619237899780273,
-      "learning_rate": 7.294777989375565e-06,
-      "loss": 2.0114,
+      "epoch": 0.8,
+      "grad_norm": 37.84192657470703,
+      "learning_rate": 1.4696899970714976e-05,
+      "loss": 1.652,
       "step": 6339
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 10.558842658996582,
-      "learning_rate": 7.292773378771174e-06,
-      "loss": 0.7375,
+      "epoch": 0.8,
+      "grad_norm": 11.718338966369629,
+      "learning_rate": 1.4696063255658286e-05,
+      "loss": 1.5113,
       "step": 6340
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 11.236355781555176,
-      "learning_rate": 7.290768768166784e-06,
-      "loss": 1.406,
+      "epoch": 0.8,
+      "grad_norm": 3.9569454193115234,
+      "learning_rate": 1.46952265406016e-05,
+      "loss": 0.1657,
       "step": 6341
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 10.12108325958252,
-      "learning_rate": 7.288764157562395e-06,
-      "loss": 1.4366,
+      "epoch": 0.8,
+      "grad_norm": 40.034698486328125,
+      "learning_rate": 1.4694389825544912e-05,
+      "loss": 1.7666,
       "step": 6342
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 23.887067794799805,
-      "learning_rate": 7.286759546958004e-06,
-      "loss": 1.2171,
+      "epoch": 0.8,
+      "grad_norm": 10.887019157409668,
+      "learning_rate": 1.4693553110488224e-05,
+      "loss": 0.9377,
       "step": 6343
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 11.277857780456543,
-      "learning_rate": 7.284754936353614e-06,
-      "loss": 1.1178,
+      "epoch": 0.8,
+      "grad_norm": 13.77903938293457,
+      "learning_rate": 1.4692716395431537e-05,
+      "loss": 0.883,
       "step": 6344
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 8.617938041687012,
-      "learning_rate": 7.282750325749224e-06,
-      "loss": 0.9352,
+      "epoch": 0.8,
+      "grad_norm": 10.93370532989502,
+      "learning_rate": 1.469187968037485e-05,
+      "loss": 1.1001,
       "step": 6345
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 7.352804660797119,
-      "learning_rate": 7.280745715144834e-06,
-      "loss": 1.1608,
+      "epoch": 0.8,
+      "grad_norm": 18.387582778930664,
+      "learning_rate": 1.4691042965318163e-05,
+      "loss": 2.2495,
       "step": 6346
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 19.099529266357422,
-      "learning_rate": 7.278741104540443e-06,
-      "loss": 1.3358,
+      "epoch": 0.8,
+      "grad_norm": 43.04143142700195,
+      "learning_rate": 1.4690206250261473e-05,
+      "loss": 1.8831,
       "step": 6347
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 7.905605792999268,
-      "learning_rate": 7.276736493936053e-06,
-      "loss": 1.8386,
+      "epoch": 0.8,
+      "grad_norm": 6.625004291534424,
+      "learning_rate": 1.4689369535204787e-05,
+      "loss": 0.6768,
       "step": 6348
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 23.328462600708008,
-      "learning_rate": 7.274731883331664e-06,
-      "loss": 1.5516,
+      "epoch": 0.8,
+      "grad_norm": 7.089539051055908,
+      "learning_rate": 1.46885328201481e-05,
+      "loss": 0.7392,
       "step": 6349
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 55.610599517822266,
-      "learning_rate": 7.272727272727273e-06,
-      "loss": 1.2527,
+      "epoch": 0.8,
+      "grad_norm": 35.005977630615234,
+      "learning_rate": 1.468769610509141e-05,
+      "loss": 3.1561,
       "step": 6350
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 14.633689880371094,
-      "learning_rate": 7.270722662122883e-06,
-      "loss": 1.4997,
+      "epoch": 0.8,
+      "grad_norm": 10.768105506896973,
+      "learning_rate": 1.4686859390034724e-05,
+      "loss": 2.7292,
       "step": 6351
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 15.966309547424316,
-      "learning_rate": 7.268718051518493e-06,
-      "loss": 1.4803,
+      "epoch": 0.8,
+      "grad_norm": 12.711040496826172,
+      "learning_rate": 1.4686022674978038e-05,
+      "loss": 1.4311,
       "step": 6352
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 15.084650039672852,
-      "learning_rate": 7.2667134409141026e-06,
-      "loss": 0.8466,
+      "epoch": 0.8,
+      "grad_norm": 9.029401779174805,
+      "learning_rate": 1.4685185959921352e-05,
+      "loss": 1.0504,
       "step": 6353
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 24.249235153198242,
-      "learning_rate": 7.264708830309713e-06,
-      "loss": 1.2289,
+      "epoch": 0.8,
+      "grad_norm": 12.517583847045898,
+      "learning_rate": 1.4684349244864662e-05,
+      "loss": 2.1899,
       "step": 6354
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 30.721338272094727,
-      "learning_rate": 7.262704219705324e-06,
-      "loss": 1.5891,
+      "epoch": 0.8,
+      "grad_norm": 18.929569244384766,
+      "learning_rate": 1.4683512529807976e-05,
+      "loss": 4.0764,
       "step": 6355
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 16.246719360351562,
-      "learning_rate": 7.260699609100933e-06,
-      "loss": 1.4035,
+      "epoch": 0.8,
+      "grad_norm": 15.242326736450195,
+      "learning_rate": 1.4682675814751287e-05,
+      "loss": 1.8787,
       "step": 6356
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 21.17243194580078,
-      "learning_rate": 7.258694998496543e-06,
-      "loss": 1.8486,
+      "epoch": 0.8,
+      "grad_norm": 19.440195083618164,
+      "learning_rate": 1.46818390996946e-05,
+      "loss": 1.7888,
       "step": 6357
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 16.293087005615234,
-      "learning_rate": 7.256690387892152e-06,
-      "loss": 0.9408,
+      "epoch": 0.8,
+      "grad_norm": 16.57932472229004,
+      "learning_rate": 1.4681002384637913e-05,
+      "loss": 1.6356,
       "step": 6358
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 96.68749237060547,
-      "learning_rate": 7.254685777287762e-06,
-      "loss": 2.8577,
+      "epoch": 0.8,
+      "grad_norm": 18.87995719909668,
+      "learning_rate": 1.4680165669581225e-05,
+      "loss": 2.8658,
       "step": 6359
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 17.761423110961914,
-      "learning_rate": 7.2526811666833715e-06,
-      "loss": 2.8659,
-      "step": 6360
-    },
-    {
-      "epoch": 1.91,
-      "eval_loss": 0.17964056134223938,
-      "eval_runtime": 43.926,
-      "eval_samples_per_second": 33.67,
-      "eval_steps_per_second": 33.67,
+      "epoch": 0.8,
+      "grad_norm": 16.52043914794922,
+      "learning_rate": 1.4679328954524539e-05,
+      "loss": 1.8064,
       "step": 6360
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 23.259288787841797,
-      "learning_rate": 7.250676556078982e-06,
-      "loss": 1.1937,
+      "epoch": 0.8,
+      "grad_norm": 10.4945068359375,
+      "learning_rate": 1.4678492239467849e-05,
+      "loss": 1.7405,
       "step": 6361
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 28.89937400817871,
-      "learning_rate": 7.2486719454745925e-06,
-      "loss": 1.9101,
+      "epoch": 0.8,
+      "grad_norm": 36.318603515625,
+      "learning_rate": 1.4677655524411163e-05,
+      "loss": 2.7147,
       "step": 6362
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 13.6145601272583,
-      "learning_rate": 7.246667334870202e-06,
-      "loss": 1.4405,
+      "epoch": 0.8,
+      "grad_norm": 10.358726501464844,
+      "learning_rate": 1.4676818809354476e-05,
+      "loss": 0.9465,
       "step": 6363
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 28.463932037353516,
-      "learning_rate": 7.244662724265812e-06,
-      "loss": 1.7418,
+      "epoch": 0.8,
+      "grad_norm": 11.27572250366211,
+      "learning_rate": 1.4675982094297787e-05,
+      "loss": 3.3759,
       "step": 6364
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 29.748435974121094,
-      "learning_rate": 7.242658113661422e-06,
-      "loss": 1.7554,
+      "epoch": 0.8,
+      "grad_norm": 8.469902038574219,
+      "learning_rate": 1.46751453792411e-05,
+      "loss": 0.8402,
       "step": 6365
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 10.47834300994873,
-      "learning_rate": 7.240653503057031e-06,
-      "loss": 0.9418,
+      "epoch": 0.8,
+      "grad_norm": 17.69377899169922,
+      "learning_rate": 1.4674308664184414e-05,
+      "loss": 2.8532,
       "step": 6366
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 66.0386962890625,
-      "learning_rate": 7.238648892452642e-06,
-      "loss": 1.4975,
+      "epoch": 0.8,
+      "grad_norm": 10.347311019897461,
+      "learning_rate": 1.4673471949127727e-05,
+      "loss": 1.5435,
       "step": 6367
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 17.965469360351562,
-      "learning_rate": 7.236644281848252e-06,
-      "loss": 1.2229,
+      "epoch": 0.8,
+      "grad_norm": 10.75600814819336,
+      "learning_rate": 1.4672635234071038e-05,
+      "loss": 0.7196,
       "step": 6368
     },
     {
-      "epoch": 1.91,
-      "grad_norm": 12.599654197692871,
-      "learning_rate": 7.234639671243861e-06,
-      "loss": 1.606,
+      "epoch": 0.8,
+      "grad_norm": 76.24691009521484,
+      "learning_rate": 1.4671798519014351e-05,
+      "loss": 2.5328,
       "step": 6369
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 17.390832901000977,
-      "learning_rate": 7.2326350606394715e-06,
-      "loss": 1.387,
+      "epoch": 0.8,
+      "grad_norm": 32.28241729736328,
+      "learning_rate": 1.4670961803957663e-05,
+      "loss": 1.665,
       "step": 6370
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 15.174606323242188,
-      "learning_rate": 7.230630450035081e-06,
-      "loss": 1.3947,
+      "epoch": 0.8,
+      "grad_norm": 12.82848072052002,
+      "learning_rate": 1.4670125088900975e-05,
+      "loss": 0.4803,
       "step": 6371
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 16.567323684692383,
-      "learning_rate": 7.228625839430691e-06,
-      "loss": 1.5866,
+      "epoch": 0.8,
+      "grad_norm": 8.32317066192627,
+      "learning_rate": 1.4669288373844289e-05,
+      "loss": 0.5588,
       "step": 6372
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 33.17914962768555,
-      "learning_rate": 7.226621228826302e-06,
-      "loss": 1.8752,
+      "epoch": 0.8,
+      "grad_norm": 18.013607025146484,
+      "learning_rate": 1.4668451658787601e-05,
+      "loss": 0.9563,
       "step": 6373
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 18.26946258544922,
-      "learning_rate": 7.224616618221911e-06,
-      "loss": 1.1493,
+      "epoch": 0.8,
+      "grad_norm": 10.781624794006348,
+      "learning_rate": 1.4667614943730915e-05,
+      "loss": 1.842,
       "step": 6374
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 26.841148376464844,
-      "learning_rate": 7.222612007617521e-06,
-      "loss": 1.86,
+      "epoch": 0.8,
+      "grad_norm": 6.22960901260376,
+      "learning_rate": 1.4666778228674225e-05,
+      "loss": 1.6661,
       "step": 6375
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 15.99163818359375,
-      "learning_rate": 7.220607397013131e-06,
-      "loss": 1.7793,
+      "epoch": 0.8,
+      "grad_norm": 9.189697265625,
+      "learning_rate": 1.4665941513617538e-05,
+      "loss": 2.0607,
       "step": 6376
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 85.62297821044922,
-      "learning_rate": 7.21860278640874e-06,
-      "loss": 2.2498,
+      "epoch": 0.8,
+      "grad_norm": 51.77899932861328,
+      "learning_rate": 1.4665104798560852e-05,
+      "loss": 1.2212,
       "step": 6377
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 14.444759368896484,
-      "learning_rate": 7.2165981758043504e-06,
-      "loss": 1.0914,
+      "epoch": 0.8,
+      "grad_norm": 14.942885398864746,
+      "learning_rate": 1.4664268083504162e-05,
+      "loss": 2.9491,
       "step": 6378
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 14.665372848510742,
-      "learning_rate": 7.214593565199961e-06,
-      "loss": 2.3208,
+      "epoch": 0.8,
+      "grad_norm": 6.897355079650879,
+      "learning_rate": 1.4663431368447476e-05,
+      "loss": 0.2136,
       "step": 6379
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 6.8967604637146,
-      "learning_rate": 7.212588954595571e-06,
-      "loss": 1.0709,
+      "epoch": 0.8,
+      "grad_norm": 6.80172872543335,
+      "learning_rate": 1.466259465339079e-05,
+      "loss": 1.2004,
       "step": 6380
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 27.484949111938477,
-      "learning_rate": 7.210584343991181e-06,
-      "loss": 1.6652,
+      "epoch": 0.8,
+      "grad_norm": 18.366886138916016,
+      "learning_rate": 1.4661757938334102e-05,
+      "loss": 1.1385,
       "step": 6381
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 36.965274810791016,
-      "learning_rate": 7.20857973338679e-06,
-      "loss": 2.1163,
+      "epoch": 0.8,
+      "grad_norm": 38.32312774658203,
+      "learning_rate": 1.4660921223277414e-05,
+      "loss": 2.4598,
       "step": 6382
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 16.792247772216797,
-      "learning_rate": 7.2065751227824e-06,
-      "loss": 1.4545,
+      "epoch": 0.8,
+      "grad_norm": 10.939009666442871,
+      "learning_rate": 1.4660084508220727e-05,
+      "loss": 1.1789,
       "step": 6383
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 18.809303283691406,
-      "learning_rate": 7.204570512178009e-06,
-      "loss": 1.9265,
+      "epoch": 0.8,
+      "grad_norm": 19.45528221130371,
+      "learning_rate": 1.465924779316404e-05,
+      "loss": 0.8006,
       "step": 6384
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 23.61911964416504,
-      "learning_rate": 7.20256590157362e-06,
-      "loss": 1.4629,
+      "epoch": 0.8,
+      "grad_norm": 29.048940658569336,
+      "learning_rate": 1.4658411078107351e-05,
+      "loss": 2.3528,
       "step": 6385
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 15.787665367126465,
-      "learning_rate": 7.20056129096923e-06,
-      "loss": 1.5397,
+      "epoch": 0.8,
+      "grad_norm": 20.245954513549805,
+      "learning_rate": 1.4657574363050665e-05,
+      "loss": 2.627,
       "step": 6386
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 25.38530158996582,
-      "learning_rate": 7.1985566803648395e-06,
-      "loss": 1.9278,
+      "epoch": 0.8,
+      "grad_norm": 14.931648254394531,
+      "learning_rate": 1.4656737647993977e-05,
+      "loss": 1.7569,
       "step": 6387
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 52.69976806640625,
-      "learning_rate": 7.19655206976045e-06,
-      "loss": 1.0921,
+      "epoch": 0.8,
+      "grad_norm": 11.111387252807617,
+      "learning_rate": 1.465590093293729e-05,
+      "loss": 1.1751,
       "step": 6388
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 12.26889419555664,
-      "learning_rate": 7.19454745915606e-06,
-      "loss": 0.9164,
+      "epoch": 0.8,
+      "grad_norm": 9.465842247009277,
+      "learning_rate": 1.46550642178806e-05,
+      "loss": 1.4811,
       "step": 6389
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 11.600250244140625,
-      "learning_rate": 7.192542848551669e-06,
-      "loss": 0.8845,
+      "epoch": 0.8,
+      "grad_norm": 7.897397041320801,
+      "learning_rate": 1.4654227502823914e-05,
+      "loss": 0.426,
       "step": 6390
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 21.00747299194336,
-      "learning_rate": 7.190538237947279e-06,
-      "loss": 1.6517,
+      "epoch": 0.8,
+      "grad_norm": 6.984261512756348,
+      "learning_rate": 1.4653390787767228e-05,
+      "loss": 0.326,
       "step": 6391
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 34.60345458984375,
-      "learning_rate": 7.18853362734289e-06,
-      "loss": 1.5433,
+      "epoch": 0.8,
+      "grad_norm": 34.08882141113281,
+      "learning_rate": 1.4652554072710538e-05,
+      "loss": 1.84,
       "step": 6392
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 11.045711517333984,
-      "learning_rate": 7.186529016738499e-06,
-      "loss": 1.2821,
+      "epoch": 0.8,
+      "grad_norm": 12.20494270324707,
+      "learning_rate": 1.4651717357653852e-05,
+      "loss": 2.2515,
       "step": 6393
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 15.181732177734375,
-      "learning_rate": 7.184524406134109e-06,
-      "loss": 1.382,
+      "epoch": 0.8,
+      "grad_norm": 11.835962295532227,
+      "learning_rate": 1.4650880642597165e-05,
+      "loss": 0.2851,
       "step": 6394
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 15.159256935119629,
-      "learning_rate": 7.1825197955297185e-06,
-      "loss": 1.1485,
+      "epoch": 0.8,
+      "grad_norm": 22.450462341308594,
+      "learning_rate": 1.4650043927540477e-05,
+      "loss": 1.2688,
       "step": 6395
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 18.364330291748047,
-      "learning_rate": 7.1805151849253286e-06,
-      "loss": 1.5439,
+      "epoch": 0.8,
+      "grad_norm": 11.962854385375977,
+      "learning_rate": 1.464920721248379e-05,
+      "loss": 1.7477,
       "step": 6396
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 25.671911239624023,
-      "learning_rate": 7.178510574320938e-06,
-      "loss": 1.1944,
+      "epoch": 0.8,
+      "grad_norm": 21.87105941772461,
+      "learning_rate": 1.4648370497427103e-05,
+      "loss": 1.4168,
       "step": 6397
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 20.186847686767578,
-      "learning_rate": 7.176505963716549e-06,
-      "loss": 1.7509,
+      "epoch": 0.8,
+      "grad_norm": 17.599401473999023,
+      "learning_rate": 1.4647533782370415e-05,
+      "loss": 1.0415,
       "step": 6398
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 13.25575065612793,
-      "learning_rate": 7.174501353112159e-06,
-      "loss": 0.9348,
+      "epoch": 0.8,
+      "grad_norm": 18.64591407775879,
+      "learning_rate": 1.4646697067313727e-05,
+      "loss": 1.5148,
       "step": 6399
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 16.35478401184082,
-      "learning_rate": 7.172496742507769e-06,
-      "loss": 1.4031,
+      "epoch": 0.8,
+      "grad_norm": 14.95875072479248,
+      "learning_rate": 1.4645860352257039e-05,
+      "loss": 1.3421,
       "step": 6400
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 13.498659133911133,
-      "learning_rate": 7.170492131903378e-06,
-      "loss": 1.4828,
+      "epoch": 0.8,
+      "eval_loss": 0.09361087530851364,
+      "eval_runtime": 100.2017,
+      "eval_samples_per_second": 35.349,
+      "eval_steps_per_second": 35.349,
+      "step": 6400
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.143239974975586,
+      "learning_rate": 1.4645023637200353e-05,
+      "loss": 2.1652,
       "step": 6401
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 88.43494415283203,
-      "learning_rate": 7.168487521298988e-06,
-      "loss": 1.9878,
+      "epoch": 0.8,
+      "grad_norm": 24.113967895507812,
+      "learning_rate": 1.4644186922143666e-05,
+      "loss": 3.1909,
       "step": 6402
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 11.417867660522461,
-      "learning_rate": 7.1664829106945975e-06,
-      "loss": 1.5075,
+      "epoch": 0.8,
+      "grad_norm": 4.8510355949401855,
+      "learning_rate": 1.4643350207086976e-05,
+      "loss": 0.4873,
       "step": 6403
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 12.581781387329102,
-      "learning_rate": 7.164478300090208e-06,
-      "loss": 1.1652,
+      "epoch": 0.8,
+      "grad_norm": 10.595499992370605,
+      "learning_rate": 1.464251349203029e-05,
+      "loss": 2.3981,
       "step": 6404
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 19.474443435668945,
-      "learning_rate": 7.1624736894858185e-06,
-      "loss": 1.7053,
+      "epoch": 0.8,
+      "grad_norm": 15.347944259643555,
+      "learning_rate": 1.4641676776973604e-05,
+      "loss": 1.7141,
       "step": 6405
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 21.434553146362305,
-      "learning_rate": 7.160469078881428e-06,
-      "loss": 2.3571,
+      "epoch": 0.8,
+      "grad_norm": 7.038366317749023,
+      "learning_rate": 1.4640840061916914e-05,
+      "loss": 0.3818,
       "step": 6406
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 17.52594566345215,
-      "learning_rate": 7.158464468277038e-06,
-      "loss": 2.2389,
+      "epoch": 0.8,
+      "grad_norm": 16.829757690429688,
+      "learning_rate": 1.4640003346860228e-05,
+      "loss": 1.2497,
       "step": 6407
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 26.64896583557129,
-      "learning_rate": 7.156459857672647e-06,
-      "loss": 1.8134,
+      "epoch": 0.8,
+      "grad_norm": 18.520206451416016,
+      "learning_rate": 1.4639166631803541e-05,
+      "loss": 1.2516,
       "step": 6408
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 13.084494590759277,
-      "learning_rate": 7.154455247068257e-06,
-      "loss": 1.584,
+      "epoch": 0.8,
+      "grad_norm": 13.922795295715332,
+      "learning_rate": 1.4638329916746853e-05,
+      "loss": 2.1521,
       "step": 6409
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 12.483121871948242,
-      "learning_rate": 7.152450636463868e-06,
-      "loss": 1.8767,
+      "epoch": 0.8,
+      "grad_norm": 19.33915901184082,
+      "learning_rate": 1.4637493201690165e-05,
+      "loss": 1.6892,
       "step": 6410
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 8.219230651855469,
-      "learning_rate": 7.150446025859477e-06,
-      "loss": 1.177,
+      "epoch": 0.8,
+      "grad_norm": 24.729047775268555,
+      "learning_rate": 1.4636656486633479e-05,
+      "loss": 0.8628,
       "step": 6411
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 31.770601272583008,
-      "learning_rate": 7.148441415255087e-06,
-      "loss": 1.9082,
+      "epoch": 0.8,
+      "grad_norm": 22.032670974731445,
+      "learning_rate": 1.4635819771576791e-05,
+      "loss": 1.7879,
       "step": 6412
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 18.06419563293457,
-      "learning_rate": 7.1464368046506975e-06,
-      "loss": 1.2848,
+      "epoch": 0.8,
+      "grad_norm": 15.246326446533203,
+      "learning_rate": 1.4634983056520103e-05,
+      "loss": 2.3433,
       "step": 6413
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 6.39808464050293,
-      "learning_rate": 7.144432194046307e-06,
-      "loss": 0.4952,
+      "epoch": 0.8,
+      "grad_norm": 10.587986946105957,
+      "learning_rate": 1.4634146341463415e-05,
+      "loss": 2.016,
       "step": 6414
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 16.47151756286621,
-      "learning_rate": 7.142427583441917e-06,
-      "loss": 1.4156,
+      "epoch": 0.81,
+      "grad_norm": 15.93343734741211,
+      "learning_rate": 1.4633309626406728e-05,
+      "loss": 2.5226,
       "step": 6415
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 35.63877487182617,
-      "learning_rate": 7.140422972837528e-06,
-      "loss": 1.6205,
+      "epoch": 0.81,
+      "grad_norm": 18.305713653564453,
+      "learning_rate": 1.4632472911350042e-05,
+      "loss": 1.4843,
       "step": 6416
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 16.693328857421875,
-      "learning_rate": 7.138418362233137e-06,
-      "loss": 1.1112,
+      "epoch": 0.81,
+      "grad_norm": 27.7530460357666,
+      "learning_rate": 1.4631636196293352e-05,
+      "loss": 1.9575,
       "step": 6417
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 34.685909271240234,
-      "learning_rate": 7.136413751628747e-06,
-      "loss": 1.704,
+      "epoch": 0.81,
+      "grad_norm": 15.6149263381958,
+      "learning_rate": 1.4630799481236666e-05,
+      "loss": 1.0243,
       "step": 6418
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 17.206161499023438,
-      "learning_rate": 7.134409141024356e-06,
-      "loss": 1.1891,
+      "epoch": 0.81,
+      "grad_norm": 18.455860137939453,
+      "learning_rate": 1.462996276617998e-05,
+      "loss": 2.7726,
       "step": 6419
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 19.83222770690918,
-      "learning_rate": 7.132404530419966e-06,
-      "loss": 2.2121,
+      "epoch": 0.81,
+      "grad_norm": 19.104188919067383,
+      "learning_rate": 1.462912605112329e-05,
+      "loss": 2.0499,
       "step": 6420
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 26.431730270385742,
-      "learning_rate": 7.130399919815576e-06,
-      "loss": 1.7221,
+      "epoch": 0.81,
+      "grad_norm": 12.461531639099121,
+      "learning_rate": 1.4628289336066604e-05,
+      "loss": 0.9269,
       "step": 6421
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 9.490403175354004,
-      "learning_rate": 7.1283953092111865e-06,
-      "loss": 1.1032,
+      "epoch": 0.81,
+      "grad_norm": 10.216887474060059,
+      "learning_rate": 1.4627452621009917e-05,
+      "loss": 1.6536,
       "step": 6422
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 52.717281341552734,
-      "learning_rate": 7.126390698606797e-06,
-      "loss": 1.5446,
+      "epoch": 0.81,
+      "grad_norm": 9.225250244140625,
+      "learning_rate": 1.4626615905953229e-05,
+      "loss": 1.0244,
       "step": 6423
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 14.396468162536621,
-      "learning_rate": 7.124386088002406e-06,
-      "loss": 0.762,
+      "epoch": 0.81,
+      "grad_norm": 15.16700553894043,
+      "learning_rate": 1.4625779190896541e-05,
+      "loss": 0.8063,
       "step": 6424
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 34.430179595947266,
-      "learning_rate": 7.122381477398016e-06,
-      "loss": 2.3349,
+      "epoch": 0.81,
+      "grad_norm": 41.029144287109375,
+      "learning_rate": 1.4624942475839855e-05,
+      "loss": 2.733,
       "step": 6425
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 7.51977014541626,
-      "learning_rate": 7.120376866793626e-06,
-      "loss": 0.1446,
+      "epoch": 0.81,
+      "grad_norm": 10.162013053894043,
+      "learning_rate": 1.4624105760783167e-05,
+      "loss": 1.6527,
       "step": 6426
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 11.881677627563477,
-      "learning_rate": 7.118372256189235e-06,
-      "loss": 1.2101,
+      "epoch": 0.81,
+      "grad_norm": 8.653647422790527,
+      "learning_rate": 1.4623269045726479e-05,
+      "loss": 1.5381,
       "step": 6427
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 17.82876968383789,
-      "learning_rate": 7.116367645584846e-06,
-      "loss": 1.2283,
+      "epoch": 0.81,
+      "grad_norm": 14.731822967529297,
+      "learning_rate": 1.462243233066979e-05,
+      "loss": 0.7498,
       "step": 6428
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 16.561769485473633,
-      "learning_rate": 7.114363034980456e-06,
-      "loss": 2.0483,
+      "epoch": 0.81,
+      "grad_norm": 19.83860206604004,
+      "learning_rate": 1.4621595615613104e-05,
+      "loss": 1.3507,
       "step": 6429
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 35.203155517578125,
-      "learning_rate": 7.1123584243760655e-06,
-      "loss": 1.7524,
+      "epoch": 0.81,
+      "grad_norm": 9.492634773254395,
+      "learning_rate": 1.4620758900556418e-05,
+      "loss": 0.7376,
       "step": 6430
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 24.895706176757812,
-      "learning_rate": 7.110353813771676e-06,
-      "loss": 1.6033,
+      "epoch": 0.81,
+      "grad_norm": 9.63869857788086,
+      "learning_rate": 1.4619922185499728e-05,
+      "loss": 1.3482,
       "step": 6431
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 23.870237350463867,
-      "learning_rate": 7.108349203167285e-06,
-      "loss": 1.4187,
+      "epoch": 0.81,
+      "grad_norm": 8.897042274475098,
+      "learning_rate": 1.4619085470443042e-05,
+      "loss": 1.8334,
       "step": 6432
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 24.825237274169922,
-      "learning_rate": 7.106344592562895e-06,
-      "loss": 1.594,
+      "epoch": 0.81,
+      "grad_norm": 21.0125789642334,
+      "learning_rate": 1.4618248755386355e-05,
+      "loss": 2.6799,
       "step": 6433
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 8.913065910339355,
-      "learning_rate": 7.104339981958505e-06,
-      "loss": 1.0598,
+      "epoch": 0.81,
+      "grad_norm": 12.40052604675293,
+      "learning_rate": 1.4617412040329666e-05,
+      "loss": 2.913,
       "step": 6434
     },
     {
-      "epoch": 1.93,
-      "grad_norm": 41.398555755615234,
-      "learning_rate": 7.102335371354115e-06,
-      "loss": 2.3157,
+      "epoch": 0.81,
+      "grad_norm": 13.130826950073242,
+      "learning_rate": 1.461657532527298e-05,
+      "loss": 1.4907,
       "step": 6435
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 18.028615951538086,
-      "learning_rate": 7.100330760749725e-06,
-      "loss": 2.0904,
+      "epoch": 0.81,
+      "grad_norm": 18.544252395629883,
+      "learning_rate": 1.4615738610216293e-05,
+      "loss": 1.7603,
       "step": 6436
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 11.3342866897583,
-      "learning_rate": 7.098326150145335e-06,
-      "loss": 1.7827,
+      "epoch": 0.81,
+      "grad_norm": 14.421226501464844,
+      "learning_rate": 1.4614901895159605e-05,
+      "loss": 0.8415,
       "step": 6437
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 20.189563751220703,
-      "learning_rate": 7.0963215395409445e-06,
-      "loss": 1.6553,
+      "epoch": 0.81,
+      "grad_norm": 10.523085594177246,
+      "learning_rate": 1.4614065180102917e-05,
+      "loss": 1.099,
       "step": 6438
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 12.127889633178711,
-      "learning_rate": 7.094316928936555e-06,
-      "loss": 1.4503,
+      "epoch": 0.81,
+      "grad_norm": 20.280637741088867,
+      "learning_rate": 1.4613228465046229e-05,
+      "loss": 1.1873,
       "step": 6439
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 37.8315315246582,
-      "learning_rate": 7.092312318332164e-06,
-      "loss": 1.7977,
+      "epoch": 0.81,
+      "grad_norm": 11.694278717041016,
+      "learning_rate": 1.4612391749989543e-05,
+      "loss": 1.9685,
       "step": 6440
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 13.1519775390625,
-      "learning_rate": 7.090307707727775e-06,
-      "loss": 1.6393,
+      "epoch": 0.81,
+      "grad_norm": 19.720291137695312,
+      "learning_rate": 1.4611555034932854e-05,
+      "loss": 2.8196,
       "step": 6441
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 35.46371841430664,
-      "learning_rate": 7.088303097123385e-06,
-      "loss": 2.3399,
+      "epoch": 0.81,
+      "grad_norm": 4.62885856628418,
+      "learning_rate": 1.4610718319876166e-05,
+      "loss": 1.571,
       "step": 6442
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 12.697428703308105,
-      "learning_rate": 7.086298486518994e-06,
-      "loss": 2.5277,
+      "epoch": 0.81,
+      "grad_norm": 21.804670333862305,
+      "learning_rate": 1.460988160481948e-05,
+      "loss": 0.8583,
       "step": 6443
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 26.87334632873535,
-      "learning_rate": 7.084293875914604e-06,
-      "loss": 2.0531,
+      "epoch": 0.81,
+      "grad_norm": 7.430726528167725,
+      "learning_rate": 1.4609044889762794e-05,
+      "loss": 0.3739,
       "step": 6444
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 14.13779354095459,
-      "learning_rate": 7.082289265310213e-06,
-      "loss": 1.8136,
+      "epoch": 0.81,
+      "grad_norm": 7.364584445953369,
+      "learning_rate": 1.4608208174706104e-05,
+      "loss": 2.0177,
       "step": 6445
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 12.27436637878418,
-      "learning_rate": 7.0802846547058235e-06,
-      "loss": 1.2326,
+      "epoch": 0.81,
+      "grad_norm": 7.745721340179443,
+      "learning_rate": 1.4607371459649418e-05,
+      "loss": 1.0005,
       "step": 6446
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 19.918701171875,
-      "learning_rate": 7.078280044101434e-06,
-      "loss": 1.4257,
+      "epoch": 0.81,
+      "grad_norm": 15.221183776855469,
+      "learning_rate": 1.4606534744592731e-05,
+      "loss": 1.7872,
       "step": 6447
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 13.029349327087402,
-      "learning_rate": 7.076275433497044e-06,
-      "loss": 1.7537,
+      "epoch": 0.81,
+      "grad_norm": 11.117356300354004,
+      "learning_rate": 1.4605698029536042e-05,
+      "loss": 0.4948,
       "step": 6448
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 9.116921424865723,
-      "learning_rate": 7.074270822892654e-06,
-      "loss": 0.7983,
+      "epoch": 0.81,
+      "grad_norm": 22.149002075195312,
+      "learning_rate": 1.4604861314479355e-05,
+      "loss": 1.5521,
       "step": 6449
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 30.727689743041992,
-      "learning_rate": 7.072266212288264e-06,
-      "loss": 1.2988,
+      "epoch": 0.81,
+      "grad_norm": 17.480669021606445,
+      "learning_rate": 1.4604024599422669e-05,
+      "loss": 1.7611,
       "step": 6450
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 15.236931800842285,
-      "learning_rate": 7.070261601683873e-06,
-      "loss": 1.4535,
+      "epoch": 0.81,
+      "grad_norm": 10.693264961242676,
+      "learning_rate": 1.460318788436598e-05,
+      "loss": 1.9562,
       "step": 6451
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 38.2266845703125,
-      "learning_rate": 7.068256991079483e-06,
-      "loss": 0.8669,
+      "epoch": 0.81,
+      "grad_norm": 8.48538589477539,
+      "learning_rate": 1.4602351169309293e-05,
+      "loss": 0.6823,
       "step": 6452
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 42.05986785888672,
-      "learning_rate": 7.066252380475094e-06,
-      "loss": 1.7596,
+      "epoch": 0.81,
+      "grad_norm": 13.08189868927002,
+      "learning_rate": 1.4601514454252605e-05,
+      "loss": 2.1066,
       "step": 6453
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 28.586145401000977,
-      "learning_rate": 7.064247769870703e-06,
-      "loss": 1.823,
+      "epoch": 0.81,
+      "grad_norm": 13.800010681152344,
+      "learning_rate": 1.4600677739195918e-05,
+      "loss": 1.4282,
       "step": 6454
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 6.259302616119385,
-      "learning_rate": 7.062243159266313e-06,
-      "loss": 0.3312,
+      "epoch": 0.81,
+      "grad_norm": 15.999839782714844,
+      "learning_rate": 1.459984102413923e-05,
+      "loss": 1.8667,
       "step": 6455
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 8.961036682128906,
-      "learning_rate": 7.060238548661923e-06,
-      "loss": 0.999,
+      "epoch": 0.81,
+      "grad_norm": 20.90586280822754,
+      "learning_rate": 1.4599004309082542e-05,
+      "loss": 2.6342,
       "step": 6456
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 19.617019653320312,
-      "learning_rate": 7.058233938057533e-06,
-      "loss": 1.9528,
+      "epoch": 0.81,
+      "grad_norm": 24.189279556274414,
+      "learning_rate": 1.4598167594025856e-05,
+      "loss": 2.6589,
       "step": 6457
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 30.003917694091797,
-      "learning_rate": 7.056229327453143e-06,
-      "loss": 1.5707,
+      "epoch": 0.81,
+      "grad_norm": 7.959173202514648,
+      "learning_rate": 1.459733087896917e-05,
+      "loss": 1.2131,
       "step": 6458
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 16.13441276550293,
-      "learning_rate": 7.054224716848753e-06,
-      "loss": 2.1332,
+      "epoch": 0.81,
+      "grad_norm": 19.223169326782227,
+      "learning_rate": 1.459649416391248e-05,
+      "loss": 2.325,
       "step": 6459
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 11.956790924072266,
-      "learning_rate": 7.052220106244363e-06,
-      "loss": 1.4419,
+      "epoch": 0.81,
+      "grad_norm": 7.551638126373291,
+      "learning_rate": 1.4595657448855793e-05,
+      "loss": 0.3251,
       "step": 6460
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 17.357301712036133,
-      "learning_rate": 7.050215495639973e-06,
-      "loss": 1.1238,
+      "epoch": 0.81,
+      "grad_norm": 15.9935941696167,
+      "learning_rate": 1.4594820733799107e-05,
+      "loss": 1.999,
       "step": 6461
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 16.655643463134766,
-      "learning_rate": 7.048210885035582e-06,
-      "loss": 2.1496,
+      "epoch": 0.81,
+      "grad_norm": 8.858784675598145,
+      "learning_rate": 1.4593984018742417e-05,
+      "loss": 0.5507,
       "step": 6462
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 17.952760696411133,
-      "learning_rate": 7.046206274431192e-06,
-      "loss": 1.5901,
+      "epoch": 0.81,
+      "grad_norm": 18.052534103393555,
+      "learning_rate": 1.4593147303685731e-05,
+      "loss": 1.3051,
       "step": 6463
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 20.259384155273438,
-      "learning_rate": 7.044201663826802e-06,
-      "loss": 1.2221,
+      "epoch": 0.81,
+      "grad_norm": 19.27977752685547,
+      "learning_rate": 1.4592310588629045e-05,
+      "loss": 1.6754,
       "step": 6464
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 25.712038040161133,
-      "learning_rate": 7.0421970532224125e-06,
-      "loss": 2.141,
+      "epoch": 0.81,
+      "grad_norm": 22.184526443481445,
+      "learning_rate": 1.4591473873572357e-05,
+      "loss": 1.365,
       "step": 6465
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 17.299470901489258,
-      "learning_rate": 7.040192442618023e-06,
-      "loss": 0.9474,
+      "epoch": 0.81,
+      "grad_norm": 15.561197280883789,
+      "learning_rate": 1.4590637158515669e-05,
+      "loss": 1.4414,
       "step": 6466
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 18.08493423461914,
-      "learning_rate": 7.038187832013632e-06,
-      "loss": 1.711,
+      "epoch": 0.81,
+      "grad_norm": 14.853275299072266,
+      "learning_rate": 1.458980044345898e-05,
+      "loss": 3.0884,
       "step": 6467
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 54.114418029785156,
-      "learning_rate": 7.036183221409242e-06,
-      "loss": 0.6773,
+      "epoch": 0.81,
+      "grad_norm": 9.874292373657227,
+      "learning_rate": 1.4588963728402294e-05,
+      "loss": 0.7533,
       "step": 6468
     },
     {
-      "epoch": 1.94,
-      "grad_norm": 13.001557350158691,
-      "learning_rate": 7.034178610804851e-06,
-      "loss": 1.1428,
+      "epoch": 0.81,
+      "grad_norm": 4.604920864105225,
+      "learning_rate": 1.4588127013345606e-05,
+      "loss": 0.4188,
       "step": 6469
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 8.948216438293457,
-      "learning_rate": 7.032174000200461e-06,
-      "loss": 1.2129,
+      "epoch": 0.81,
+      "grad_norm": 19.53866195678711,
+      "learning_rate": 1.4587290298288918e-05,
+      "loss": 3.1209,
       "step": 6470
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 96.37139892578125,
-      "learning_rate": 7.030169389596072e-06,
-      "loss": 1.3139,
+      "epoch": 0.81,
+      "grad_norm": 39.01898956298828,
+      "learning_rate": 1.4586453583232232e-05,
+      "loss": 1.2419,
       "step": 6471
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 14.782844543457031,
-      "learning_rate": 7.0281647789916814e-06,
-      "loss": 1.4799,
+      "epoch": 0.81,
+      "grad_norm": 17.05350685119629,
+      "learning_rate": 1.4585616868175545e-05,
+      "loss": 1.5823,
       "step": 6472
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 14.976436614990234,
-      "learning_rate": 7.0261601683872915e-06,
-      "loss": 1.4801,
+      "epoch": 0.81,
+      "grad_norm": 15.502795219421387,
+      "learning_rate": 1.4584780153118856e-05,
+      "loss": 0.9528,
       "step": 6473
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 26.140769958496094,
-      "learning_rate": 7.024155557782902e-06,
-      "loss": 1.9828,
+      "epoch": 0.81,
+      "grad_norm": 36.6273307800293,
+      "learning_rate": 1.458394343806217e-05,
+      "loss": 3.6041,
       "step": 6474
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 14.462868690490723,
-      "learning_rate": 7.022150947178511e-06,
-      "loss": 1.3346,
+      "epoch": 0.81,
+      "grad_norm": 7.87349271774292,
+      "learning_rate": 1.4583106723005483e-05,
+      "loss": 1.4797,
       "step": 6475
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 12.547101974487305,
-      "learning_rate": 7.020146336574121e-06,
-      "loss": 1.8225,
+      "epoch": 0.81,
+      "grad_norm": 32.351585388183594,
+      "learning_rate": 1.4582270007948793e-05,
+      "loss": 4.1317,
       "step": 6476
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 13.108290672302246,
-      "learning_rate": 7.01814172596973e-06,
-      "loss": 1.7796,
+      "epoch": 0.81,
+      "grad_norm": 23.48708724975586,
+      "learning_rate": 1.4581433292892107e-05,
+      "loss": 1.7281,
       "step": 6477
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 20.673439025878906,
-      "learning_rate": 7.016137115365341e-06,
-      "loss": 1.7685,
+      "epoch": 0.81,
+      "grad_norm": 22.382984161376953,
+      "learning_rate": 1.458059657783542e-05,
+      "loss": 0.6815,
       "step": 6478
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 9.68248462677002,
-      "learning_rate": 7.014132504760951e-06,
-      "loss": 1.1545,
+      "epoch": 0.81,
+      "grad_norm": 5.5734381675720215,
+      "learning_rate": 1.4579759862778732e-05,
+      "loss": 1.6547,
       "step": 6479
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 11.824162483215332,
-      "learning_rate": 7.01212789415656e-06,
-      "loss": 1.1257,
-      "step": 6480
-    },
-    {
-      "epoch": 1.95,
-      "eval_loss": 0.17668664455413818,
-      "eval_runtime": 43.6307,
-      "eval_samples_per_second": 33.898,
-      "eval_steps_per_second": 33.898,
+      "epoch": 0.81,
+      "grad_norm": 17.19704246520996,
+      "learning_rate": 1.4578923147722044e-05,
+      "loss": 1.7212,
       "step": 6480
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 19.132976531982422,
-      "learning_rate": 7.0101232835521705e-06,
-      "loss": 1.4466,
+      "epoch": 0.81,
+      "grad_norm": 9.862728118896484,
+      "learning_rate": 1.4578086432665356e-05,
+      "loss": 0.9488,
       "step": 6481
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 13.578184127807617,
-      "learning_rate": 7.00811867294778e-06,
-      "loss": 1.8283,
+      "epoch": 0.81,
+      "grad_norm": 21.093936920166016,
+      "learning_rate": 1.457724971760867e-05,
+      "loss": 2.7171,
       "step": 6482
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 14.636981010437012,
-      "learning_rate": 7.00611406234339e-06,
-      "loss": 1.5467,
+      "epoch": 0.81,
+      "grad_norm": 20.390226364135742,
+      "learning_rate": 1.4576413002551982e-05,
+      "loss": 1.7474,
       "step": 6483
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 14.252835273742676,
-      "learning_rate": 7.004109451739001e-06,
-      "loss": 1.0799,
+      "epoch": 0.81,
+      "grad_norm": 8.369909286499023,
+      "learning_rate": 1.4575576287495294e-05,
+      "loss": 2.2826,
       "step": 6484
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 7.7530837059021,
-      "learning_rate": 7.002104841134611e-06,
-      "loss": 1.7159,
+      "epoch": 0.81,
+      "grad_norm": 12.084911346435547,
+      "learning_rate": 1.4574739572438608e-05,
+      "loss": 1.048,
       "step": 6485
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 19.219745635986328,
-      "learning_rate": 7.00010023053022e-06,
-      "loss": 1.3202,
+      "epoch": 0.81,
+      "grad_norm": 7.501756191253662,
+      "learning_rate": 1.4573902857381921e-05,
+      "loss": 1.6434,
       "step": 6486
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 10.907150268554688,
-      "learning_rate": 6.99809561992583e-06,
-      "loss": 1.3228,
+      "epoch": 0.81,
+      "grad_norm": 16.92917251586914,
+      "learning_rate": 1.4573066142325232e-05,
+      "loss": 2.7004,
       "step": 6487
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 7.514638423919678,
-      "learning_rate": 6.996091009321439e-06,
-      "loss": 1.1397,
+      "epoch": 0.81,
+      "grad_norm": 7.9938836097717285,
+      "learning_rate": 1.4572229427268545e-05,
+      "loss": 1.5723,
       "step": 6488
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 21.773658752441406,
-      "learning_rate": 6.9940863987170495e-06,
-      "loss": 0.9278,
+      "epoch": 0.81,
+      "grad_norm": 38.44096755981445,
+      "learning_rate": 1.4571392712211859e-05,
+      "loss": 0.8792,
       "step": 6489
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 10.506806373596191,
-      "learning_rate": 6.99208178811266e-06,
-      "loss": 1.709,
+      "epoch": 0.81,
+      "grad_norm": 19.077499389648438,
+      "learning_rate": 1.4570555997155169e-05,
+      "loss": 3.0232,
       "step": 6490
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 15.891654014587402,
-      "learning_rate": 6.99007717750827e-06,
-      "loss": 1.4618,
+      "epoch": 0.81,
+      "grad_norm": 13.681564331054688,
+      "learning_rate": 1.4569719282098483e-05,
+      "loss": 1.5459,
       "step": 6491
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 31.406612396240234,
-      "learning_rate": 6.98807256690388e-06,
-      "loss": 1.977,
+      "epoch": 0.81,
+      "grad_norm": 20.16901397705078,
+      "learning_rate": 1.4568882567041795e-05,
+      "loss": 1.6177,
       "step": 6492
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 20.738195419311523,
-      "learning_rate": 6.986067956299489e-06,
-      "loss": 1.4744,
+      "epoch": 0.81,
+      "grad_norm": 10.907618522644043,
+      "learning_rate": 1.4568045851985108e-05,
+      "loss": 1.2013,
       "step": 6493
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 17.87335968017578,
-      "learning_rate": 6.984063345695099e-06,
-      "loss": 1.8068,
+      "epoch": 0.81,
+      "grad_norm": 24.92099952697754,
+      "learning_rate": 1.456720913692842e-05,
+      "loss": 1.4485,
       "step": 6494
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 13.646010398864746,
-      "learning_rate": 6.982058735090709e-06,
-      "loss": 0.9828,
+      "epoch": 0.82,
+      "grad_norm": 17.624778747558594,
+      "learning_rate": 1.4566372421871732e-05,
+      "loss": 1.7824,
       "step": 6495
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 12.331052780151367,
-      "learning_rate": 6.980054124486319e-06,
-      "loss": 1.6284,
+      "epoch": 0.82,
+      "grad_norm": 10.988775253295898,
+      "learning_rate": 1.4565535706815046e-05,
+      "loss": 1.9106,
       "step": 6496
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 52.41801452636719,
-      "learning_rate": 6.978049513881929e-06,
-      "loss": 0.973,
+      "epoch": 0.82,
+      "grad_norm": 6.560670852661133,
+      "learning_rate": 1.4564698991758358e-05,
+      "loss": 0.2385,
       "step": 6497
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 27.965364456176758,
-      "learning_rate": 6.976044903277539e-06,
-      "loss": 1.0861,
+      "epoch": 0.82,
+      "grad_norm": 18.28946304321289,
+      "learning_rate": 1.456386227670167e-05,
+      "loss": 1.6398,
       "step": 6498
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 25.636484146118164,
-      "learning_rate": 6.974040292673149e-06,
-      "loss": 1.8763,
+      "epoch": 0.82,
+      "grad_norm": 9.239535331726074,
+      "learning_rate": 1.4563025561644983e-05,
+      "loss": 0.9596,
       "step": 6499
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 15.626718521118164,
-      "learning_rate": 6.972035682068759e-06,
-      "loss": 1.2224,
+      "epoch": 0.82,
+      "grad_norm": 42.755859375,
+      "learning_rate": 1.4562188846588297e-05,
+      "loss": 2.6544,
       "step": 6500
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 12.968523979187012,
-      "learning_rate": 6.970031071464368e-06,
-      "loss": 1.4936,
+      "epoch": 0.82,
+      "grad_norm": 23.385692596435547,
+      "learning_rate": 1.4561352131531607e-05,
+      "loss": 2.5512,
       "step": 6501
     },
     {
-      "epoch": 1.95,
-      "grad_norm": 38.640384674072266,
-      "learning_rate": 6.968026460859979e-06,
-      "loss": 1.847,
+      "epoch": 0.82,
+      "grad_norm": 16.858610153198242,
+      "learning_rate": 1.4560515416474921e-05,
+      "loss": 1.8767,
       "step": 6502
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 35.005760192871094,
-      "learning_rate": 6.966021850255589e-06,
-      "loss": 1.8344,
+      "epoch": 0.82,
+      "grad_norm": 13.347295761108398,
+      "learning_rate": 1.4559678701418235e-05,
+      "loss": 1.7668,
       "step": 6503
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 59.90297317504883,
-      "learning_rate": 6.964017239651198e-06,
-      "loss": 1.5804,
+      "epoch": 0.82,
+      "grad_norm": 4.6510329246521,
+      "learning_rate": 1.4558841986361545e-05,
+      "loss": 0.4936,
       "step": 6504
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 8.664277076721191,
-      "learning_rate": 6.962012629046808e-06,
-      "loss": 1.1554,
+      "epoch": 0.82,
+      "grad_norm": 10.455964088439941,
+      "learning_rate": 1.4558005271304859e-05,
+      "loss": 1.5058,
       "step": 6505
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 15.182535171508789,
-      "learning_rate": 6.9600080184424175e-06,
-      "loss": 1.107,
+      "epoch": 0.82,
+      "grad_norm": 18.887109756469727,
+      "learning_rate": 1.455716855624817e-05,
+      "loss": 0.6563,
       "step": 6506
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 25.827924728393555,
-      "learning_rate": 6.958003407838028e-06,
-      "loss": 1.1151,
+      "epoch": 0.82,
+      "grad_norm": 12.829666137695312,
+      "learning_rate": 1.4556331841191484e-05,
+      "loss": 2.7943,
       "step": 6507
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 9.57219409942627,
-      "learning_rate": 6.9559987972336385e-06,
-      "loss": 1.3565,
+      "epoch": 0.82,
+      "grad_norm": 6.674988269805908,
+      "learning_rate": 1.4555495126134796e-05,
+      "loss": 0.8994,
       "step": 6508
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 113.87895202636719,
-      "learning_rate": 6.953994186629248e-06,
-      "loss": 1.4153,
+      "epoch": 0.82,
+      "grad_norm": 25.241641998291016,
+      "learning_rate": 1.4554658411078108e-05,
+      "loss": 2.5382,
       "step": 6509
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 17.828977584838867,
-      "learning_rate": 6.951989576024858e-06,
-      "loss": 1.8861,
+      "epoch": 0.82,
+      "grad_norm": 2.330019474029541,
+      "learning_rate": 1.4553821696021422e-05,
+      "loss": 0.0793,
       "step": 6510
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 27.323060989379883,
-      "learning_rate": 6.949984965420468e-06,
-      "loss": 1.4276,
+      "epoch": 0.82,
+      "grad_norm": 19.723289489746094,
+      "learning_rate": 1.4552984980964732e-05,
+      "loss": 1.4636,
       "step": 6511
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 11.81863021850586,
-      "learning_rate": 6.947980354816077e-06,
-      "loss": 0.785,
+      "epoch": 0.82,
+      "grad_norm": 17.301788330078125,
+      "learning_rate": 1.4552148265908046e-05,
+      "loss": 1.09,
       "step": 6512
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 10.938015937805176,
-      "learning_rate": 6.945975744211687e-06,
-      "loss": 1.096,
+      "epoch": 0.82,
+      "grad_norm": 30.26137351989746,
+      "learning_rate": 1.455131155085136e-05,
+      "loss": 2.0605,
       "step": 6513
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 11.981101036071777,
-      "learning_rate": 6.9439711336072965e-06,
-      "loss": 1.3268,
+      "epoch": 0.82,
+      "grad_norm": 19.499509811401367,
+      "learning_rate": 1.4550474835794673e-05,
+      "loss": 1.6498,
       "step": 6514
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 19.58521842956543,
-      "learning_rate": 6.9419665230029074e-06,
-      "loss": 1.8049,
+      "epoch": 0.82,
+      "grad_norm": 12.36483383178711,
+      "learning_rate": 1.4549638120737983e-05,
+      "loss": 2.8728,
       "step": 6515
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 24.513118743896484,
-      "learning_rate": 6.9399619123985175e-06,
-      "loss": 1.3997,
+      "epoch": 0.82,
+      "grad_norm": 20.424285888671875,
+      "learning_rate": 1.4548801405681297e-05,
+      "loss": 0.8148,
       "step": 6516
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 13.00726318359375,
-      "learning_rate": 6.937957301794127e-06,
-      "loss": 2.1062,
+      "epoch": 0.82,
+      "grad_norm": 8.933712005615234,
+      "learning_rate": 1.454796469062461e-05,
+      "loss": 2.214,
       "step": 6517
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 16.61869239807129,
-      "learning_rate": 6.935952691189737e-06,
-      "loss": 1.4637,
+      "epoch": 0.82,
+      "grad_norm": 28.269676208496094,
+      "learning_rate": 1.454712797556792e-05,
+      "loss": 1.7172,
       "step": 6518
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 18.2260684967041,
-      "learning_rate": 6.933948080585347e-06,
-      "loss": 1.3907,
+      "epoch": 0.82,
+      "grad_norm": 9.066912651062012,
+      "learning_rate": 1.4546291260511234e-05,
+      "loss": 1.6103,
       "step": 6519
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 27.4097957611084,
-      "learning_rate": 6.931943469980956e-06,
-      "loss": 2.0212,
+      "epoch": 0.82,
+      "grad_norm": 17.551225662231445,
+      "learning_rate": 1.4545454545454546e-05,
+      "loss": 1.2698,
       "step": 6520
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 13.984885215759277,
-      "learning_rate": 6.929938859376567e-06,
-      "loss": 0.8537,
+      "epoch": 0.82,
+      "grad_norm": 15.091469764709473,
+      "learning_rate": 1.454461783039786e-05,
+      "loss": 2.0922,
       "step": 6521
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 29.465072631835938,
-      "learning_rate": 6.927934248772177e-06,
-      "loss": 2.082,
+      "epoch": 0.82,
+      "grad_norm": 11.744821548461914,
+      "learning_rate": 1.4543781115341172e-05,
+      "loss": 0.7955,
       "step": 6522
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 13.194611549377441,
-      "learning_rate": 6.925929638167786e-06,
-      "loss": 1.1381,
+      "epoch": 0.82,
+      "grad_norm": 8.435943603515625,
+      "learning_rate": 1.4542944400284484e-05,
+      "loss": 0.6975,
       "step": 6523
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 7.926638603210449,
-      "learning_rate": 6.9239250275633965e-06,
-      "loss": 0.9932,
+      "epoch": 0.82,
+      "grad_norm": 13.73911190032959,
+      "learning_rate": 1.4542107685227798e-05,
+      "loss": 0.9118,
       "step": 6524
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 16.14949607849121,
-      "learning_rate": 6.921920416959006e-06,
-      "loss": 1.4698,
+      "epoch": 0.82,
+      "grad_norm": 8.145610809326172,
+      "learning_rate": 1.4541270970171108e-05,
+      "loss": 0.4499,
       "step": 6525
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 33.72904968261719,
-      "learning_rate": 6.919915806354616e-06,
-      "loss": 3.2966,
+      "epoch": 0.82,
+      "grad_norm": 11.110485076904297,
+      "learning_rate": 1.4540434255114421e-05,
+      "loss": 2.4025,
       "step": 6526
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 8.642800331115723,
-      "learning_rate": 6.917911195750227e-06,
-      "loss": 1.1102,
+      "epoch": 0.82,
+      "grad_norm": 91.34126281738281,
+      "learning_rate": 1.4539597540057735e-05,
+      "loss": 2.0706,
       "step": 6527
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 12.295087814331055,
-      "learning_rate": 6.915906585145836e-06,
-      "loss": 0.9972,
+      "epoch": 0.82,
+      "grad_norm": 16.182498931884766,
+      "learning_rate": 1.4538760825001049e-05,
+      "loss": 1.4164,
       "step": 6528
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 24.457801818847656,
-      "learning_rate": 6.913901974541446e-06,
-      "loss": 1.4332,
+      "epoch": 0.82,
+      "grad_norm": 9.682636260986328,
+      "learning_rate": 1.4537924109944359e-05,
+      "loss": 1.55,
       "step": 6529
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 14.026710510253906,
-      "learning_rate": 6.911897363937055e-06,
-      "loss": 1.8641,
+      "epoch": 0.82,
+      "grad_norm": 16.95926856994629,
+      "learning_rate": 1.4537087394887673e-05,
+      "loss": 1.9937,
       "step": 6530
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 12.17983341217041,
-      "learning_rate": 6.909892753332665e-06,
-      "loss": 1.4611,
+      "epoch": 0.82,
+      "grad_norm": 7.322328567504883,
+      "learning_rate": 1.4536250679830986e-05,
+      "loss": 1.2618,
       "step": 6531
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 12.283072471618652,
-      "learning_rate": 6.9078881427282755e-06,
-      "loss": 1.3947,
+      "epoch": 0.82,
+      "grad_norm": 30.958463668823242,
+      "learning_rate": 1.4535413964774297e-05,
+      "loss": 0.4021,
       "step": 6532
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 17.96526527404785,
-      "learning_rate": 6.9058835321238856e-06,
-      "loss": 1.7153,
+      "epoch": 0.82,
+      "grad_norm": 13.310428619384766,
+      "learning_rate": 1.453457724971761e-05,
+      "loss": 1.0993,
       "step": 6533
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 24.099628448486328,
-      "learning_rate": 6.903878921519496e-06,
-      "loss": 1.2495,
+      "epoch": 0.82,
+      "grad_norm": 15.63068675994873,
+      "learning_rate": 1.4533740534660922e-05,
+      "loss": 1.9267,
       "step": 6534
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 16.107059478759766,
-      "learning_rate": 6.901874310915106e-06,
-      "loss": 1.4728,
+      "epoch": 0.82,
+      "grad_norm": 21.709959030151367,
+      "learning_rate": 1.4532903819604236e-05,
+      "loss": 3.3146,
       "step": 6535
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 15.214371681213379,
-      "learning_rate": 6.899869700310715e-06,
-      "loss": 1.1217,
+      "epoch": 0.82,
+      "grad_norm": 13.965682029724121,
+      "learning_rate": 1.4532067104547548e-05,
+      "loss": 3.9313,
       "step": 6536
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 27.67184829711914,
-      "learning_rate": 6.897865089706325e-06,
-      "loss": 1.4113,
+      "epoch": 0.82,
+      "grad_norm": 14.858061790466309,
+      "learning_rate": 1.453123038949086e-05,
+      "loss": 1.5124,
       "step": 6537
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 20.205726623535156,
-      "learning_rate": 6.895860479101934e-06,
-      "loss": 0.8228,
+      "epoch": 0.82,
+      "grad_norm": 50.09086608886719,
+      "learning_rate": 1.4530393674434173e-05,
+      "loss": 2.489,
       "step": 6538
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 12.976736068725586,
-      "learning_rate": 6.893855868497545e-06,
-      "loss": 1.0787,
+      "epoch": 0.82,
+      "grad_norm": 26.609758377075195,
+      "learning_rate": 1.4529556959377484e-05,
+      "loss": 1.7796,
       "step": 6539
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 141.27496337890625,
-      "learning_rate": 6.891851257893155e-06,
-      "loss": 2.659,
+      "epoch": 0.82,
+      "grad_norm": 10.551162719726562,
+      "learning_rate": 1.4528720244320797e-05,
+      "loss": 1.2116,
       "step": 6540
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 25.662107467651367,
-      "learning_rate": 6.8898466472887645e-06,
-      "loss": 2.3812,
+      "epoch": 0.82,
+      "grad_norm": 13.381997108459473,
+      "learning_rate": 1.4527883529264111e-05,
+      "loss": 0.7883,
       "step": 6541
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 17.610265731811523,
-      "learning_rate": 6.887842036684375e-06,
-      "loss": 1.1345,
+      "epoch": 0.82,
+      "grad_norm": 33.09524917602539,
+      "learning_rate": 1.4527046814207425e-05,
+      "loss": 2.4682,
       "step": 6542
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 13.565105438232422,
-      "learning_rate": 6.885837426079985e-06,
-      "loss": 1.7477,
+      "epoch": 0.82,
+      "grad_norm": 13.487656593322754,
+      "learning_rate": 1.4526210099150735e-05,
+      "loss": 0.9119,
       "step": 6543
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 72.72836303710938,
-      "learning_rate": 6.883832815475594e-06,
-      "loss": 1.0895,
+      "epoch": 0.82,
+      "grad_norm": 13.092456817626953,
+      "learning_rate": 1.4525373384094048e-05,
+      "loss": 3.033,
       "step": 6544
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 11.967425346374512,
-      "learning_rate": 6.881828204871205e-06,
-      "loss": 1.2826,
+      "epoch": 0.82,
+      "grad_norm": 59.662330627441406,
+      "learning_rate": 1.452453666903736e-05,
+      "loss": 2.0694,
       "step": 6545
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 27.99849510192871,
-      "learning_rate": 6.879823594266815e-06,
-      "loss": 1.4163,
+      "epoch": 0.82,
+      "grad_norm": 4.204308986663818,
+      "learning_rate": 1.4523699953980672e-05,
+      "loss": 0.1224,
       "step": 6546
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 17.99080467224121,
-      "learning_rate": 6.877818983662424e-06,
-      "loss": 0.7908,
+      "epoch": 0.82,
+      "grad_norm": 6.319829940795898,
+      "learning_rate": 1.4522863238923986e-05,
+      "loss": 0.9945,
       "step": 6547
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 8.94594955444336,
-      "learning_rate": 6.875814373058034e-06,
-      "loss": 1.5791,
+      "epoch": 0.82,
+      "grad_norm": 14.876279830932617,
+      "learning_rate": 1.4522026523867298e-05,
+      "loss": 1.4276,
       "step": 6548
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 18.120492935180664,
-      "learning_rate": 6.8738097624536435e-06,
-      "loss": 1.5981,
-      "step": 6549
+      "epoch": 0.82,
+      "grad_norm": 11.182560920715332,
+      "learning_rate": 1.4521189808810612e-05,
+      "loss": 1.499,
+      "step": 6549
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 11.887974739074707,
-      "learning_rate": 6.871805151849254e-06,
-      "loss": 2.1877,
+      "epoch": 0.82,
+      "grad_norm": 23.092039108276367,
+      "learning_rate": 1.4520353093753924e-05,
+      "loss": 3.6536,
       "step": 6550
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 14.82646369934082,
-      "learning_rate": 6.8698005412448645e-06,
-      "loss": 1.2886,
+      "epoch": 0.82,
+      "grad_norm": 16.563962936401367,
+      "learning_rate": 1.4519516378697236e-05,
+      "loss": 2.0491,
       "step": 6551
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 15.733491897583008,
-      "learning_rate": 6.867795930640474e-06,
-      "loss": 0.836,
+      "epoch": 0.82,
+      "grad_norm": 11.452547073364258,
+      "learning_rate": 1.451867966364055e-05,
+      "loss": 1.4192,
       "step": 6552
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 20.574148178100586,
-      "learning_rate": 6.865791320036084e-06,
-      "loss": 0.9531,
+      "epoch": 0.82,
+      "grad_norm": 19.012704849243164,
+      "learning_rate": 1.451784294858386e-05,
+      "loss": 2.8846,
       "step": 6553
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 28.415647506713867,
-      "learning_rate": 6.863786709431693e-06,
-      "loss": 2.1882,
+      "epoch": 0.82,
+      "grad_norm": 27.29177474975586,
+      "learning_rate": 1.4517006233527173e-05,
+      "loss": 1.7701,
       "step": 6554
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 17.647674560546875,
-      "learning_rate": 6.861782098827303e-06,
-      "loss": 2.3014,
+      "epoch": 0.82,
+      "grad_norm": 6.170631408691406,
+      "learning_rate": 1.4516169518470487e-05,
+      "loss": 0.7061,
       "step": 6555
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 49.27579879760742,
-      "learning_rate": 6.859777488222913e-06,
-      "loss": 1.5323,
+      "epoch": 0.82,
+      "grad_norm": 8.111652374267578,
+      "learning_rate": 1.45153328034138e-05,
+      "loss": 0.8509,
       "step": 6556
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 19.76232147216797,
-      "learning_rate": 6.8577728776185225e-06,
-      "loss": 1.6397,
+      "epoch": 0.82,
+      "grad_norm": 31.29617691040039,
+      "learning_rate": 1.451449608835711e-05,
+      "loss": 1.1399,
       "step": 6557
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 11.850191116333008,
-      "learning_rate": 6.8557682670141334e-06,
-      "loss": 1.027,
+      "epoch": 0.82,
+      "grad_norm": 9.659380912780762,
+      "learning_rate": 1.4513659373300424e-05,
+      "loss": 1.1278,
       "step": 6558
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 26.760108947753906,
-      "learning_rate": 6.8537636564097435e-06,
-      "loss": 1.7687,
+      "epoch": 0.82,
+      "grad_norm": 9.498473167419434,
+      "learning_rate": 1.4512822658243736e-05,
+      "loss": 1.8005,
       "step": 6559
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 11.1150484085083,
-      "learning_rate": 6.851759045805353e-06,
-      "loss": 1.418,
+      "epoch": 0.82,
+      "grad_norm": 11.763265609741211,
+      "learning_rate": 1.4511985943187048e-05,
+      "loss": 1.965,
       "step": 6560
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 23.722135543823242,
-      "learning_rate": 6.849754435200963e-06,
-      "loss": 1.0756,
+      "epoch": 0.82,
+      "grad_norm": 15.806600570678711,
+      "learning_rate": 1.4511149228130362e-05,
+      "loss": 1.2199,
       "step": 6561
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 17.482393264770508,
-      "learning_rate": 6.847749824596572e-06,
-      "loss": 1.2066,
+      "epoch": 0.82,
+      "grad_norm": 13.062655448913574,
+      "learning_rate": 1.4510312513073674e-05,
+      "loss": 0.9208,
       "step": 6562
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 17.04688835144043,
-      "learning_rate": 6.845745213992182e-06,
-      "loss": 1.4078,
+      "epoch": 0.82,
+      "grad_norm": 6.248857498168945,
+      "learning_rate": 1.4509475798016987e-05,
+      "loss": 0.5726,
       "step": 6563
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 15.257699012756348,
-      "learning_rate": 6.843740603387793e-06,
-      "loss": 1.8723,
+      "epoch": 0.82,
+      "grad_norm": 19.46588706970215,
+      "learning_rate": 1.4508639082960298e-05,
+      "loss": 1.0639,
       "step": 6564
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 13.324009895324707,
-      "learning_rate": 6.841735992783402e-06,
-      "loss": 1.3293,
+      "epoch": 0.82,
+      "grad_norm": 17.309402465820312,
+      "learning_rate": 1.4507802367903611e-05,
+      "loss": 1.03,
       "step": 6565
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 18.284709930419922,
-      "learning_rate": 6.839731382179012e-06,
-      "loss": 1.8599,
+      "epoch": 0.82,
+      "grad_norm": 2.397465229034424,
+      "learning_rate": 1.4506965652846925e-05,
+      "loss": 0.0929,
       "step": 6566
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 13.653274536132812,
-      "learning_rate": 6.837726771574622e-06,
-      "loss": 1.7576,
+      "epoch": 0.82,
+      "grad_norm": 13.140107154846191,
+      "learning_rate": 1.4506128937790235e-05,
+      "loss": 0.843,
       "step": 6567
     },
     {
-      "epoch": 1.97,
-      "grad_norm": 38.0052490234375,
-      "learning_rate": 6.835722160970232e-06,
-      "loss": 1.656,
+      "epoch": 0.82,
+      "grad_norm": 9.642655372619629,
+      "learning_rate": 1.4505292222733549e-05,
+      "loss": 0.7449,
       "step": 6568
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 15.570902824401855,
-      "learning_rate": 6.833717550365842e-06,
-      "loss": 1.3572,
+      "epoch": 0.82,
+      "grad_norm": 18.0646915435791,
+      "learning_rate": 1.4504455507676863e-05,
+      "loss": 2.3933,
       "step": 6569
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 10.287487030029297,
-      "learning_rate": 6.831712939761453e-06,
-      "loss": 1.0965,
+      "epoch": 0.82,
+      "grad_norm": 6.070803165435791,
+      "learning_rate": 1.4503618792620176e-05,
+      "loss": 0.9627,
       "step": 6570
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 21.403841018676758,
-      "learning_rate": 6.829708329157062e-06,
-      "loss": 1.657,
+      "epoch": 0.82,
+      "grad_norm": 10.14867877960205,
+      "learning_rate": 1.4502782077563487e-05,
+      "loss": 1.2322,
       "step": 6571
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 11.690693855285645,
-      "learning_rate": 6.827703718552672e-06,
-      "loss": 1.9188,
+      "epoch": 0.82,
+      "grad_norm": 24.438552856445312,
+      "learning_rate": 1.45019453625068e-05,
+      "loss": 0.8593,
       "step": 6572
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 15.413796424865723,
-      "learning_rate": 6.825699107948281e-06,
-      "loss": 1.1102,
+      "epoch": 0.82,
+      "grad_norm": 10.15174388885498,
+      "learning_rate": 1.4501108647450112e-05,
+      "loss": 0.7729,
       "step": 6573
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 14.91450023651123,
-      "learning_rate": 6.823694497343891e-06,
-      "loss": 1.0896,
+      "epoch": 0.83,
+      "grad_norm": 35.63993835449219,
+      "learning_rate": 1.4500271932393424e-05,
+      "loss": 1.5794,
       "step": 6574
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 12.081074714660645,
-      "learning_rate": 6.821689886739501e-06,
-      "loss": 1.7378,
+      "epoch": 0.83,
+      "grad_norm": 13.550148963928223,
+      "learning_rate": 1.4499435217336738e-05,
+      "loss": 1.2075,
       "step": 6575
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 16.04603385925293,
-      "learning_rate": 6.8196852761351116e-06,
-      "loss": 1.4417,
+      "epoch": 0.83,
+      "grad_norm": 24.06031036376953,
+      "learning_rate": 1.449859850228005e-05,
+      "loss": 3.1595,
       "step": 6576
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 22.271512985229492,
-      "learning_rate": 6.817680665530722e-06,
-      "loss": 1.7555,
+      "epoch": 0.83,
+      "grad_norm": 31.2019100189209,
+      "learning_rate": 1.4497761787223363e-05,
+      "loss": 1.8097,
       "step": 6577
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 13.81209945678711,
-      "learning_rate": 6.815676054926331e-06,
-      "loss": 1.1311,
+      "epoch": 0.83,
+      "grad_norm": 8.17690372467041,
+      "learning_rate": 1.4496925072166674e-05,
+      "loss": 0.6208,
       "step": 6578
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 16.152433395385742,
-      "learning_rate": 6.813671444321941e-06,
-      "loss": 1.1026,
+      "epoch": 0.83,
+      "grad_norm": 19.666261672973633,
+      "learning_rate": 1.4496088357109987e-05,
+      "loss": 1.6847,
       "step": 6579
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 12.994856834411621,
-      "learning_rate": 6.811666833717551e-06,
-      "loss": 1.1081,
+      "epoch": 0.83,
+      "grad_norm": 8.355585098266602,
+      "learning_rate": 1.4495251642053301e-05,
+      "loss": 1.4162,
       "step": 6580
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 12.3396577835083,
-      "learning_rate": 6.80966222311316e-06,
-      "loss": 1.3073,
+      "epoch": 0.83,
+      "grad_norm": 17.03511619567871,
+      "learning_rate": 1.4494414926996611e-05,
+      "loss": 1.1782,
       "step": 6581
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 8.903776168823242,
-      "learning_rate": 6.807657612508771e-06,
-      "loss": 0.9238,
+      "epoch": 0.83,
+      "grad_norm": 11.48168659210205,
+      "learning_rate": 1.4493578211939925e-05,
+      "loss": 1.9596,
       "step": 6582
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 15.200712203979492,
-      "learning_rate": 6.805653001904381e-06,
-      "loss": 1.3466,
+      "epoch": 0.83,
+      "grad_norm": 9.669479370117188,
+      "learning_rate": 1.4492741496883238e-05,
+      "loss": 0.6663,
       "step": 6583
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 19.97858428955078,
-      "learning_rate": 6.8036483912999905e-06,
-      "loss": 1.7572,
+      "epoch": 0.83,
+      "grad_norm": 12.308369636535645,
+      "learning_rate": 1.449190478182655e-05,
+      "loss": 1.0629,
       "step": 6584
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 24.35478401184082,
-      "learning_rate": 6.801643780695601e-06,
-      "loss": 1.929,
+      "epoch": 0.83,
+      "grad_norm": 15.225822448730469,
+      "learning_rate": 1.4491068066769862e-05,
+      "loss": 1.1809,
       "step": 6585
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 30.171939849853516,
-      "learning_rate": 6.79963917009121e-06,
-      "loss": 1.5492,
+      "epoch": 0.83,
+      "grad_norm": 15.169183731079102,
+      "learning_rate": 1.4490231351713176e-05,
+      "loss": 0.9538,
       "step": 6586
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 12.768831253051758,
-      "learning_rate": 6.79763455948682e-06,
-      "loss": 1.0463,
+      "epoch": 0.83,
+      "grad_norm": 5.91043758392334,
+      "learning_rate": 1.4489394636656488e-05,
+      "loss": 0.4872,
       "step": 6587
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 21.945470809936523,
-      "learning_rate": 6.795629948882431e-06,
-      "loss": 2.2842,
+      "epoch": 0.83,
+      "grad_norm": 19.87512969970703,
+      "learning_rate": 1.44885579215998e-05,
+      "loss": 1.5201,
       "step": 6588
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 22.516357421875,
-      "learning_rate": 6.79362533827804e-06,
-      "loss": 2.0211,
+      "epoch": 0.83,
+      "grad_norm": 8.933917045593262,
+      "learning_rate": 1.4487721206543114e-05,
+      "loss": 0.555,
       "step": 6589
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 8.880302429199219,
-      "learning_rate": 6.79162072767365e-06,
-      "loss": 1.2581,
+      "epoch": 0.83,
+      "grad_norm": 7.488944053649902,
+      "learning_rate": 1.4486884491486426e-05,
+      "loss": 0.2726,
       "step": 6590
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 9.477561950683594,
-      "learning_rate": 6.7896161170692594e-06,
-      "loss": 0.7168,
+      "epoch": 0.83,
+      "grad_norm": 12.712186813354492,
+      "learning_rate": 1.448604777642974e-05,
+      "loss": 1.8749,
       "step": 6591
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 21.51420783996582,
-      "learning_rate": 6.7876115064648695e-06,
-      "loss": 1.3912,
+      "epoch": 0.83,
+      "grad_norm": 14.857038497924805,
+      "learning_rate": 1.448521106137305e-05,
+      "loss": 1.1538,
       "step": 6592
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 50.1191291809082,
-      "learning_rate": 6.78560689586048e-06,
-      "loss": 3.0844,
+      "epoch": 0.83,
+      "grad_norm": 12.747661590576172,
+      "learning_rate": 1.4484374346316363e-05,
+      "loss": 0.7235,
       "step": 6593
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 19.533275604248047,
-      "learning_rate": 6.78360228525609e-06,
-      "loss": 2.1365,
+      "epoch": 0.83,
+      "grad_norm": 39.1489372253418,
+      "learning_rate": 1.4483537631259677e-05,
+      "loss": 1.893,
       "step": 6594
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 11.014450073242188,
-      "learning_rate": 6.7815976746517e-06,
-      "loss": 0.7238,
+      "epoch": 0.83,
+      "grad_norm": 9.506407737731934,
+      "learning_rate": 1.4482700916202987e-05,
+      "loss": 2.2091,
       "step": 6595
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 35.398197174072266,
-      "learning_rate": 6.77959306404731e-06,
-      "loss": 2.9231,
+      "epoch": 0.83,
+      "grad_norm": 21.36495590209961,
+      "learning_rate": 1.44818642011463e-05,
+      "loss": 2.8877,
       "step": 6596
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 59.179141998291016,
-      "learning_rate": 6.777588453442919e-06,
-      "loss": 1.4903,
+      "epoch": 0.83,
+      "grad_norm": 12.15341854095459,
+      "learning_rate": 1.4481027486089614e-05,
+      "loss": 2.097,
       "step": 6597
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 26.32767677307129,
-      "learning_rate": 6.775583842838529e-06,
-      "loss": 1.0562,
+      "epoch": 0.83,
+      "grad_norm": 29.05472755432129,
+      "learning_rate": 1.4480190771032926e-05,
+      "loss": 1.5282,
       "step": 6598
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 14.91978931427002,
-      "learning_rate": 6.773579232234138e-06,
-      "loss": 1.8945,
+      "epoch": 0.83,
+      "grad_norm": 12.839364051818848,
+      "learning_rate": 1.4479354055976238e-05,
+      "loss": 1.5512,
       "step": 6599
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 43.29436111450195,
-      "learning_rate": 6.7715746216297485e-06,
-      "loss": 2.4119,
-      "step": 6600
-    },
-    {
-      "epoch": 1.98,
-      "eval_loss": 0.17498844861984253,
-      "eval_runtime": 43.9202,
-      "eval_samples_per_second": 33.675,
-      "eval_steps_per_second": 33.675,
+      "epoch": 0.83,
+      "grad_norm": 7.66991662979126,
+      "learning_rate": 1.4478517340919552e-05,
+      "loss": 1.5824,
       "step": 6600
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 20.57763671875,
-      "learning_rate": 6.7695700110253594e-06,
-      "loss": 1.2593,
+      "epoch": 0.83,
+      "grad_norm": 15.398689270019531,
+      "learning_rate": 1.4477680625862864e-05,
+      "loss": 1.4943,
       "step": 6601
     },
     {
-      "epoch": 1.98,
-      "grad_norm": 11.902490615844727,
-      "learning_rate": 6.767565400420969e-06,
-      "loss": 0.8341,
+      "epoch": 0.83,
+      "grad_norm": 24.575775146484375,
+      "learning_rate": 1.4476843910806176e-05,
+      "loss": 1.7269,
       "step": 6602
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 11.146263122558594,
-      "learning_rate": 6.765560789816579e-06,
-      "loss": 1.3651,
+      "epoch": 0.83,
+      "grad_norm": 11.358213424682617,
+      "learning_rate": 1.4476007195749488e-05,
+      "loss": 1.1742,
       "step": 6603
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 11.864297866821289,
-      "learning_rate": 6.763556179212189e-06,
-      "loss": 0.9559,
+      "epoch": 0.83,
+      "grad_norm": 20.26782989501953,
+      "learning_rate": 1.4475170480692801e-05,
+      "loss": 2.2997,
       "step": 6604
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 12.255757331848145,
-      "learning_rate": 6.761551568607798e-06,
-      "loss": 1.1177,
+      "epoch": 0.83,
+      "grad_norm": 31.358434677124023,
+      "learning_rate": 1.4474333765636113e-05,
+      "loss": 0.9536,
       "step": 6605
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 6.7448506355285645,
-      "learning_rate": 6.759546958003408e-06,
-      "loss": 0.6869,
+      "epoch": 0.83,
+      "grad_norm": 13.94992733001709,
+      "learning_rate": 1.4473497050579425e-05,
+      "loss": 1.7772,
       "step": 6606
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 15.666464805603027,
-      "learning_rate": 6.757542347399019e-06,
-      "loss": 1.7579,
+      "epoch": 0.83,
+      "grad_norm": 11.810144424438477,
+      "learning_rate": 1.4472660335522739e-05,
+      "loss": 0.7789,
       "step": 6607
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 16.473482131958008,
-      "learning_rate": 6.755537736794628e-06,
-      "loss": 0.9031,
+      "epoch": 0.83,
+      "grad_norm": 38.90315246582031,
+      "learning_rate": 1.4471823620466053e-05,
+      "loss": 3.4662,
       "step": 6608
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 12.140952110290527,
-      "learning_rate": 6.753533126190238e-06,
-      "loss": 0.7175,
+      "epoch": 0.83,
+      "grad_norm": 18.91117286682129,
+      "learning_rate": 1.4470986905409363e-05,
+      "loss": 2.3911,
       "step": 6609
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 9.83596134185791,
-      "learning_rate": 6.751528515585848e-06,
-      "loss": 0.8216,
+      "epoch": 0.83,
+      "grad_norm": 13.002034187316895,
+      "learning_rate": 1.4470150190352676e-05,
+      "loss": 1.4097,
       "step": 6610
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 13.17597770690918,
-      "learning_rate": 6.749523904981458e-06,
-      "loss": 1.694,
+      "epoch": 0.83,
+      "grad_norm": 12.609968185424805,
+      "learning_rate": 1.446931347529599e-05,
+      "loss": 2.7201,
       "step": 6611
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 17.648849487304688,
-      "learning_rate": 6.747519294377067e-06,
-      "loss": 2.1423,
+      "epoch": 0.83,
+      "grad_norm": 8.692763328552246,
+      "learning_rate": 1.44684767602393e-05,
+      "loss": 0.9019,
       "step": 6612
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 13.655058860778809,
-      "learning_rate": 6.745514683772678e-06,
-      "loss": 0.9178,
+      "epoch": 0.83,
+      "grad_norm": 20.65419578552246,
+      "learning_rate": 1.4467640045182614e-05,
+      "loss": 1.9726,
       "step": 6613
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 18.347841262817383,
-      "learning_rate": 6.743510073168288e-06,
-      "loss": 1.0027,
+      "epoch": 0.83,
+      "grad_norm": 16.71739959716797,
+      "learning_rate": 1.4466803330125928e-05,
+      "loss": 2.2029,
       "step": 6614
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 26.16404914855957,
-      "learning_rate": 6.741505462563897e-06,
-      "loss": 2.206,
+      "epoch": 0.83,
+      "grad_norm": 10.462072372436523,
+      "learning_rate": 1.446596661506924e-05,
+      "loss": 2.1755,
       "step": 6615
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 7.279219150543213,
-      "learning_rate": 6.739500851959507e-06,
-      "loss": 0.7718,
+      "epoch": 0.83,
+      "grad_norm": 12.887832641601562,
+      "learning_rate": 1.4465129900012552e-05,
+      "loss": 0.7345,
       "step": 6616
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 12.55292797088623,
-      "learning_rate": 6.737496241355117e-06,
-      "loss": 1.8409,
+      "epoch": 0.83,
+      "grad_norm": 9.375746726989746,
+      "learning_rate": 1.4464293184955864e-05,
+      "loss": 1.3041,
       "step": 6617
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 42.25106430053711,
-      "learning_rate": 6.735491630750727e-06,
-      "loss": 2.3152,
+      "epoch": 0.83,
+      "grad_norm": 14.980518341064453,
+      "learning_rate": 1.4463456469899177e-05,
+      "loss": 2.3885,
       "step": 6618
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 14.544368743896484,
-      "learning_rate": 6.7334870201463376e-06,
-      "loss": 1.2898,
+      "epoch": 0.83,
+      "grad_norm": 33.584938049316406,
+      "learning_rate": 1.4462619754842489e-05,
+      "loss": 1.862,
       "step": 6619
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 21.00879669189453,
-      "learning_rate": 6.731482409541948e-06,
-      "loss": 1.4371,
+      "epoch": 0.83,
+      "grad_norm": 1.6788455247879028,
+      "learning_rate": 1.4461783039785801e-05,
+      "loss": 0.0572,
       "step": 6620
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 15.112432479858398,
-      "learning_rate": 6.729477798937557e-06,
-      "loss": 2.2847,
+      "epoch": 0.83,
+      "grad_norm": 9.320537567138672,
+      "learning_rate": 1.4460946324729115e-05,
+      "loss": 0.4605,
       "step": 6621
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 16.390920639038086,
-      "learning_rate": 6.727473188333167e-06,
-      "loss": 1.1366,
+      "epoch": 0.83,
+      "grad_norm": 7.229696750640869,
+      "learning_rate": 1.4460109609672428e-05,
+      "loss": 0.9963,
       "step": 6622
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 17.105491638183594,
-      "learning_rate": 6.725468577728776e-06,
-      "loss": 1.7725,
+      "epoch": 0.83,
+      "grad_norm": 18.225360870361328,
+      "learning_rate": 1.4459272894615739e-05,
+      "loss": 1.2763,
       "step": 6623
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 17.488632202148438,
-      "learning_rate": 6.723463967124386e-06,
-      "loss": 0.5385,
+      "epoch": 0.83,
+      "grad_norm": 3.779670000076294,
+      "learning_rate": 1.4458436179559052e-05,
+      "loss": 0.1906,
       "step": 6624
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 29.78978157043457,
-      "learning_rate": 6.721459356519997e-06,
-      "loss": 2.3833,
+      "epoch": 0.83,
+      "grad_norm": 39.11140441894531,
+      "learning_rate": 1.4457599464502366e-05,
+      "loss": 1.4888,
       "step": 6625
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 15.036394119262695,
-      "learning_rate": 6.7194547459156065e-06,
-      "loss": 1.0212,
+      "epoch": 0.83,
+      "grad_norm": 10.29348373413086,
+      "learning_rate": 1.4456762749445676e-05,
+      "loss": 0.6129,
       "step": 6626
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 17.2657527923584,
-      "learning_rate": 6.7174501353112165e-06,
-      "loss": 1.4752,
+      "epoch": 0.83,
+      "grad_norm": 24.620576858520508,
+      "learning_rate": 1.445592603438899e-05,
+      "loss": 1.8498,
       "step": 6627
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 27.82944679260254,
-      "learning_rate": 6.715445524706826e-06,
-      "loss": 1.741,
+      "epoch": 0.83,
+      "grad_norm": 20.528329849243164,
+      "learning_rate": 1.4455089319332304e-05,
+      "loss": 0.7714,
       "step": 6628
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 14.460271835327148,
-      "learning_rate": 6.713440914102436e-06,
-      "loss": 1.6737,
+      "epoch": 0.83,
+      "grad_norm": 12.630378723144531,
+      "learning_rate": 1.4454252604275615e-05,
+      "loss": 1.9049,
       "step": 6629
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 30.571147918701172,
-      "learning_rate": 6.711436303498046e-06,
-      "loss": 2.2318,
+      "epoch": 0.83,
+      "grad_norm": 21.8435001373291,
+      "learning_rate": 1.4453415889218927e-05,
+      "loss": 3.5011,
       "step": 6630
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 18.02724838256836,
-      "learning_rate": 6.709431692893657e-06,
-      "loss": 1.0354,
+      "epoch": 0.83,
+      "grad_norm": 59.98054122924805,
+      "learning_rate": 1.445257917416224e-05,
+      "loss": 2.5349,
       "step": 6631
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 12.847180366516113,
-      "learning_rate": 6.707427082289266e-06,
-      "loss": 1.5948,
+      "epoch": 0.83,
+      "grad_norm": 7.572198867797852,
+      "learning_rate": 1.4451742459105553e-05,
+      "loss": 0.6822,
       "step": 6632
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 13.379613876342773,
-      "learning_rate": 6.705422471684876e-06,
-      "loss": 1.2021,
+      "epoch": 0.83,
+      "grad_norm": 14.560245513916016,
+      "learning_rate": 1.4450905744048865e-05,
+      "loss": 0.4531,
       "step": 6633
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 20.53904914855957,
-      "learning_rate": 6.7034178610804854e-06,
-      "loss": 1.9106,
+      "epoch": 0.83,
+      "grad_norm": 12.604669570922852,
+      "learning_rate": 1.4450069028992177e-05,
+      "loss": 1.1481,
       "step": 6634
     },
     {
-      "epoch": 1.99,
-      "grad_norm": 23.03459930419922,
-      "learning_rate": 6.7014132504760955e-06,
-      "loss": 1.785,
+      "epoch": 0.83,
+      "grad_norm": 35.92240905761719,
+      "learning_rate": 1.444923231393549e-05,
+      "loss": 1.5445,
       "step": 6635
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 19.824228286743164,
-      "learning_rate": 6.699408639871705e-06,
-      "loss": 0.8569,
+      "epoch": 0.83,
+      "grad_norm": 15.33647346496582,
+      "learning_rate": 1.4448395598878804e-05,
+      "loss": 0.9631,
       "step": 6636
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 59.54680633544922,
-      "learning_rate": 6.697404029267316e-06,
-      "loss": 2.1475,
+      "epoch": 0.83,
+      "grad_norm": 7.366020202636719,
+      "learning_rate": 1.4447558883822115e-05,
+      "loss": 0.3153,
       "step": 6637
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 10.229207992553711,
-      "learning_rate": 6.695399418662926e-06,
-      "loss": 1.1441,
+      "epoch": 0.83,
+      "grad_norm": 9.031330108642578,
+      "learning_rate": 1.4446722168765428e-05,
+      "loss": 0.7969,
       "step": 6638
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 28.467979431152344,
-      "learning_rate": 6.693394808058535e-06,
-      "loss": 1.7708,
+      "epoch": 0.83,
+      "grad_norm": 12.082524299621582,
+      "learning_rate": 1.4445885453708742e-05,
+      "loss": 0.9905,
       "step": 6639
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 10.531692504882812,
-      "learning_rate": 6.691390197454145e-06,
-      "loss": 0.5444,
+      "epoch": 0.83,
+      "grad_norm": 5.6854753494262695,
+      "learning_rate": 1.4445048738652052e-05,
+      "loss": 0.3756,
       "step": 6640
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 13.020025253295898,
-      "learning_rate": 6.689385586849755e-06,
-      "loss": 0.8387,
+      "epoch": 0.83,
+      "grad_norm": 57.63182067871094,
+      "learning_rate": 1.4444212023595366e-05,
+      "loss": 2.161,
       "step": 6641
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 12.514997482299805,
-      "learning_rate": 6.687380976245364e-06,
-      "loss": 1.3939,
+      "epoch": 0.83,
+      "grad_norm": 1.8692203760147095,
+      "learning_rate": 1.444337530853868e-05,
+      "loss": 0.0638,
       "step": 6642
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 27.652118682861328,
-      "learning_rate": 6.6853763656409745e-06,
-      "loss": 3.0772,
+      "epoch": 0.83,
+      "grad_norm": 22.31800651550293,
+      "learning_rate": 1.4442538593481991e-05,
+      "loss": 0.5192,
       "step": 6643
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 24.15280532836914,
-      "learning_rate": 6.6833717550365854e-06,
-      "loss": 1.5129,
+      "epoch": 0.83,
+      "grad_norm": 9.42281436920166,
+      "learning_rate": 1.4441701878425303e-05,
+      "loss": 0.7522,
       "step": 6644
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 13.905855178833008,
-      "learning_rate": 6.681367144432195e-06,
-      "loss": 0.9144,
+      "epoch": 0.83,
+      "grad_norm": 23.117460250854492,
+      "learning_rate": 1.4440865163368615e-05,
+      "loss": 0.7487,
       "step": 6645
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 16.91253662109375,
-      "learning_rate": 6.679362533827805e-06,
-      "loss": 1.5355,
+      "epoch": 0.83,
+      "grad_norm": 26.78668212890625,
+      "learning_rate": 1.4440028448311929e-05,
+      "loss": 3.0379,
       "step": 6646
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 11.888856887817383,
-      "learning_rate": 6.677357923223414e-06,
-      "loss": 1.1719,
+      "epoch": 0.83,
+      "grad_norm": 25.992734909057617,
+      "learning_rate": 1.443919173325524e-05,
+      "loss": 2.8284,
       "step": 6647
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 14.441067695617676,
-      "learning_rate": 6.675353312619024e-06,
-      "loss": 1.3663,
+      "epoch": 0.83,
+      "grad_norm": 11.912419319152832,
+      "learning_rate": 1.4438355018198553e-05,
+      "loss": 2.9013,
       "step": 6648
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 13.287384033203125,
-      "learning_rate": 6.673348702014633e-06,
-      "loss": 0.5206,
+      "epoch": 0.83,
+      "grad_norm": 11.855568885803223,
+      "learning_rate": 1.4437518303141866e-05,
+      "loss": 2.0173,
       "step": 6649
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 12.121244430541992,
-      "learning_rate": 6.671344091410244e-06,
-      "loss": 1.0182,
+      "epoch": 0.83,
+      "grad_norm": 24.56494903564453,
+      "learning_rate": 1.443668158808518e-05,
+      "loss": 3.4956,
       "step": 6650
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 44.00876998901367,
-      "learning_rate": 6.669339480805854e-06,
-      "loss": 0.9455,
+      "epoch": 0.83,
+      "grad_norm": 38.72340774536133,
+      "learning_rate": 1.443584487302849e-05,
+      "loss": 3.564,
       "step": 6651
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 15.983967781066895,
-      "learning_rate": 6.6673348702014636e-06,
-      "loss": 0.8615,
+      "epoch": 0.83,
+      "grad_norm": 21.563867568969727,
+      "learning_rate": 1.4435008157971804e-05,
+      "loss": 2.1634,
       "step": 6652
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 13.022303581237793,
-      "learning_rate": 6.665330259597074e-06,
-      "loss": 1.3803,
+      "epoch": 0.83,
+      "grad_norm": 40.92628860473633,
+      "learning_rate": 1.4434171442915118e-05,
+      "loss": 1.0617,
       "step": 6653
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 57.43462371826172,
-      "learning_rate": 6.663325648992684e-06,
-      "loss": 1.9661,
+      "epoch": 0.84,
+      "grad_norm": 18.901601791381836,
+      "learning_rate": 1.4433334727858428e-05,
+      "loss": 1.821,
       "step": 6654
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 13.801424026489258,
-      "learning_rate": 6.661321038388293e-06,
-      "loss": 1.6653,
+      "epoch": 0.84,
+      "grad_norm": 25.665586471557617,
+      "learning_rate": 1.4432498012801742e-05,
+      "loss": 2.5711,
       "step": 6655
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 14.12394905090332,
-      "learning_rate": 6.659316427783904e-06,
-      "loss": 1.0337,
+      "epoch": 0.84,
+      "grad_norm": 5.861196994781494,
+      "learning_rate": 1.4431661297745053e-05,
+      "loss": 0.3677,
       "step": 6656
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 16.065610885620117,
-      "learning_rate": 6.657311817179514e-06,
-      "loss": 1.6578,
+      "epoch": 0.84,
+      "grad_norm": 7.068995952606201,
+      "learning_rate": 1.4430824582688367e-05,
+      "loss": 0.4365,
       "step": 6657
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 23.195810317993164,
-      "learning_rate": 6.655307206575123e-06,
-      "loss": 2.0227,
+      "epoch": 0.84,
+      "grad_norm": 29.462949752807617,
+      "learning_rate": 1.4429987867631679e-05,
+      "loss": 2.3702,
       "step": 6658
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 39.60324478149414,
-      "learning_rate": 6.653302595970733e-06,
-      "loss": 2.9616,
+      "epoch": 0.84,
+      "grad_norm": 23.18093490600586,
+      "learning_rate": 1.4429151152574991e-05,
+      "loss": 1.9585,
       "step": 6659
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 12.803409576416016,
-      "learning_rate": 6.6512979853663425e-06,
-      "loss": 1.3529,
+      "epoch": 0.84,
+      "grad_norm": 9.60701847076416,
+      "learning_rate": 1.4428314437518305e-05,
+      "loss": 1.2616,
       "step": 6660
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 27.853031158447266,
-      "learning_rate": 6.649293374761953e-06,
-      "loss": 1.2641,
+      "epoch": 0.84,
+      "grad_norm": 23.689254760742188,
+      "learning_rate": 1.4427477722461617e-05,
+      "loss": 2.2917,
       "step": 6661
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 12.550612449645996,
-      "learning_rate": 6.6472887641575636e-06,
-      "loss": 1.032,
+      "epoch": 0.84,
+      "grad_norm": 11.679755210876465,
+      "learning_rate": 1.4426641007404929e-05,
+      "loss": 0.5695,
       "step": 6662
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 14.057964324951172,
-      "learning_rate": 6.645284153553173e-06,
-      "loss": 1.4831,
+      "epoch": 0.84,
+      "grad_norm": 12.400811195373535,
+      "learning_rate": 1.4425804292348242e-05,
+      "loss": 1.4716,
       "step": 6663
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 13.735244750976562,
-      "learning_rate": 6.643279542948783e-06,
-      "loss": 1.842,
+      "epoch": 0.84,
+      "grad_norm": 9.619452476501465,
+      "learning_rate": 1.4424967577291556e-05,
+      "loss": 1.1475,
       "step": 6664
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 10.447699546813965,
-      "learning_rate": 6.641274932344393e-06,
-      "loss": 1.1719,
+      "epoch": 0.84,
+      "grad_norm": 18.736326217651367,
+      "learning_rate": 1.4424130862234866e-05,
+      "loss": 2.2559,
       "step": 6665
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 24.760276794433594,
-      "learning_rate": 6.639270321740002e-06,
-      "loss": 1.0642,
+      "epoch": 0.84,
+      "grad_norm": 21.631404876708984,
+      "learning_rate": 1.442329414717818e-05,
+      "loss": 1.1375,
       "step": 6666
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 17.945220947265625,
-      "learning_rate": 6.637265711135612e-06,
-      "loss": 2.1282,
+      "epoch": 0.84,
+      "grad_norm": 45.52559280395508,
+      "learning_rate": 1.4422457432121493e-05,
+      "loss": 2.2882,
       "step": 6667
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 40.95531463623047,
-      "learning_rate": 6.635261100531223e-06,
-      "loss": 1.8811,
+      "epoch": 0.84,
+      "grad_norm": 25.720808029174805,
+      "learning_rate": 1.4421620717064804e-05,
+      "loss": 2.2039,
       "step": 6668
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 13.15553092956543,
-      "learning_rate": 6.6332564899268325e-06,
-      "loss": 1.134,
+      "epoch": 0.84,
+      "grad_norm": 10.99753475189209,
+      "learning_rate": 1.4420784002008117e-05,
+      "loss": 0.6437,
       "step": 6669
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 20.673992156982422,
-      "learning_rate": 6.6312518793224425e-06,
-      "loss": 1.2125,
+      "epoch": 0.84,
+      "grad_norm": 14.956695556640625,
+      "learning_rate": 1.441994728695143e-05,
+      "loss": 1.6667,
       "step": 6670
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 17.39179801940918,
-      "learning_rate": 6.629247268718052e-06,
-      "loss": 1.7538,
+      "epoch": 0.84,
+      "grad_norm": 17.53224754333496,
+      "learning_rate": 1.4419110571894743e-05,
+      "loss": 2.2026,
       "step": 6671
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 8.030019760131836,
-      "learning_rate": 6.627242658113662e-06,
-      "loss": 1.5265,
+      "epoch": 0.84,
+      "grad_norm": 40.021331787109375,
+      "learning_rate": 1.4418273856838055e-05,
+      "loss": 1.2136,
       "step": 6672
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 23.085996627807617,
-      "learning_rate": 6.625238047509271e-06,
-      "loss": 1.5642,
+      "epoch": 0.84,
+      "grad_norm": 26.441740036010742,
+      "learning_rate": 1.4417437141781367e-05,
+      "loss": 1.3784,
       "step": 6673
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 36.12226104736328,
-      "learning_rate": 6.623233436904882e-06,
-      "loss": 2.836,
+      "epoch": 0.84,
+      "grad_norm": 5.3640336990356445,
+      "learning_rate": 1.441660042672468e-05,
+      "loss": 1.3586,
       "step": 6674
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 16.673858642578125,
-      "learning_rate": 6.621228826300492e-06,
-      "loss": 1.0128,
+      "epoch": 0.84,
+      "grad_norm": 30.53775978088379,
+      "learning_rate": 1.441576371166799e-05,
+      "loss": 3.0845,
       "step": 6675
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 18.04882049560547,
-      "learning_rate": 6.619224215696101e-06,
-      "loss": 1.2667,
+      "epoch": 0.84,
+      "grad_norm": 35.42987823486328,
+      "learning_rate": 1.4414926996611304e-05,
+      "loss": 3.7773,
       "step": 6676
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 14.236947059631348,
-      "learning_rate": 6.6172196050917114e-06,
-      "loss": 1.0438,
+      "epoch": 0.84,
+      "grad_norm": 10.641247749328613,
+      "learning_rate": 1.4414090281554618e-05,
+      "loss": 1.1733,
       "step": 6677
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 64.36849212646484,
-      "learning_rate": 6.6152149944873215e-06,
-      "loss": 2.7021,
+      "epoch": 0.84,
+      "grad_norm": 17.126296997070312,
+      "learning_rate": 1.4413253566497932e-05,
+      "loss": 2.457,
       "step": 6678
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 23.676084518432617,
-      "learning_rate": 6.613210383882931e-06,
-      "loss": 1.0589,
+      "epoch": 0.84,
+      "grad_norm": 16.539026260375977,
+      "learning_rate": 1.4412416851441242e-05,
+      "loss": 1.8305,
       "step": 6679
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 19.20448875427246,
-      "learning_rate": 6.611205773278542e-06,
-      "loss": 1.8695,
+      "epoch": 0.84,
+      "grad_norm": 21.915935516357422,
+      "learning_rate": 1.4411580136384556e-05,
+      "loss": 2.6571,
       "step": 6680
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 10.511244773864746,
-      "learning_rate": 6.609201162674152e-06,
-      "loss": 1.1752,
+      "epoch": 0.84,
+      "grad_norm": 17.56529426574707,
+      "learning_rate": 1.441074342132787e-05,
+      "loss": 1.7428,
       "step": 6681
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 40.56865692138672,
-      "learning_rate": 6.607196552069761e-06,
-      "loss": 1.6004,
+      "epoch": 0.84,
+      "grad_norm": 35.25141525268555,
+      "learning_rate": 1.440990670627118e-05,
+      "loss": 0.7839,
       "step": 6682
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 17.472505569458008,
-      "learning_rate": 6.605191941465371e-06,
-      "loss": 1.5864,
+      "epoch": 0.84,
+      "grad_norm": 8.508606910705566,
+      "learning_rate": 1.4409069991214493e-05,
+      "loss": 0.6937,
       "step": 6683
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 10.834254264831543,
-      "learning_rate": 6.60318733086098e-06,
-      "loss": 1.2419,
+      "epoch": 0.84,
+      "grad_norm": 6.896846294403076,
+      "learning_rate": 1.4408233276157805e-05,
+      "loss": 0.688,
       "step": 6684
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 16.969730377197266,
-      "learning_rate": 6.60118272025659e-06,
-      "loss": 1.2284,
+      "epoch": 0.84,
+      "grad_norm": 10.6935396194458,
+      "learning_rate": 1.4407396561101119e-05,
+      "loss": 1.5449,
       "step": 6685
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 7.831116199493408,
-      "learning_rate": 6.5991781096522e-06,
-      "loss": 1.4888,
+      "epoch": 0.84,
+      "grad_norm": 12.529866218566895,
+      "learning_rate": 1.440655984604443e-05,
+      "loss": 1.0339,
       "step": 6686
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 24.744110107421875,
-      "learning_rate": 6.597173499047811e-06,
-      "loss": 1.6571,
+      "epoch": 0.84,
+      "grad_norm": 11.404264450073242,
+      "learning_rate": 1.4405723130987743e-05,
+      "loss": 1.3663,
       "step": 6687
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 21.574138641357422,
-      "learning_rate": 6.595168888443421e-06,
-      "loss": 2.0631,
+      "epoch": 0.84,
+      "grad_norm": 29.173799514770508,
+      "learning_rate": 1.4404886415931056e-05,
+      "loss": 3.1138,
       "step": 6688
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 11.040058135986328,
-      "learning_rate": 6.593164277839031e-06,
-      "loss": 1.1278,
+      "epoch": 0.84,
+      "grad_norm": 12.170273780822754,
+      "learning_rate": 1.4404049700874367e-05,
+      "loss": 3.3146,
       "step": 6689
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 46.7620735168457,
-      "learning_rate": 6.59115966723464e-06,
-      "loss": 1.4132,
+      "epoch": 0.84,
+      "grad_norm": 15.326367378234863,
+      "learning_rate": 1.440321298581768e-05,
+      "loss": 1.28,
       "step": 6690
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 10.559418678283691,
-      "learning_rate": 6.58915505663025e-06,
-      "loss": 0.9851,
+      "epoch": 0.84,
+      "grad_norm": 23.169282913208008,
+      "learning_rate": 1.4402376270760994e-05,
+      "loss": 2.1622,
       "step": 6691
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 8.289081573486328,
-      "learning_rate": 6.587150446025859e-06,
-      "loss": 1.1853,
+      "epoch": 0.84,
+      "grad_norm": 7.151150703430176,
+      "learning_rate": 1.4401539555704308e-05,
+      "loss": 1.7352,
       "step": 6692
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 8.112937927246094,
-      "learning_rate": 6.58514583542147e-06,
-      "loss": 1.0523,
+      "epoch": 0.84,
+      "grad_norm": 12.488361358642578,
+      "learning_rate": 1.4400702840647618e-05,
+      "loss": 1.8024,
       "step": 6693
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 19.774045944213867,
-      "learning_rate": 6.58314122481708e-06,
-      "loss": 1.6196,
+      "epoch": 0.84,
+      "grad_norm": 16.26409149169922,
+      "learning_rate": 1.4399866125590931e-05,
+      "loss": 1.0197,
       "step": 6694
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 52.585914611816406,
-      "learning_rate": 6.5811366142126896e-06,
-      "loss": 1.0791,
+      "epoch": 0.84,
+      "grad_norm": 10.992413520812988,
+      "learning_rate": 1.4399029410534245e-05,
+      "loss": 1.0422,
       "step": 6695
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 10.37691879272461,
-      "learning_rate": 6.5791320036083e-06,
-      "loss": 0.525,
+      "epoch": 0.84,
+      "grad_norm": 24.888111114501953,
+      "learning_rate": 1.4398192695477555e-05,
+      "loss": 1.2818,
       "step": 6696
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 13.222021102905273,
-      "learning_rate": 6.577127393003909e-06,
-      "loss": 1.0604,
+      "epoch": 0.84,
+      "grad_norm": 63.796077728271484,
+      "learning_rate": 1.4397355980420869e-05,
+      "loss": 2.3386,
       "step": 6697
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 12.805461883544922,
-      "learning_rate": 6.575122782399519e-06,
-      "loss": 1.4706,
+      "epoch": 0.84,
+      "grad_norm": 10.447388648986816,
+      "learning_rate": 1.4396519265364181e-05,
+      "loss": 1.4342,
       "step": 6698
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 43.80728530883789,
-      "learning_rate": 6.57311817179513e-06,
-      "loss": 2.084,
+      "epoch": 0.84,
+      "grad_norm": 19.698040008544922,
+      "learning_rate": 1.4395682550307495e-05,
+      "loss": 1.805,
       "step": 6699
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 15.121885299682617,
-      "learning_rate": 6.571113561190739e-06,
-      "loss": 0.8853,
+      "epoch": 0.84,
+      "grad_norm": 9.697070121765137,
+      "learning_rate": 1.4394845835250807e-05,
+      "loss": 1.255,
       "step": 6700
     },
     {
-      "epoch": 2.01,
-      "grad_norm": 32.659427642822266,
-      "learning_rate": 6.569108950586349e-06,
-      "loss": 1.8597,
+      "epoch": 0.84,
+      "grad_norm": 19.996601104736328,
+      "learning_rate": 1.4394009120194119e-05,
+      "loss": 1.9659,
       "step": 6701
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 26.797901153564453,
-      "learning_rate": 6.567104339981959e-06,
-      "loss": 2.3284,
+      "epoch": 0.84,
+      "grad_norm": 9.914801597595215,
+      "learning_rate": 1.4393172405137432e-05,
+      "loss": 1.7741,
       "step": 6702
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 3.8142130374908447,
-      "learning_rate": 6.5650997293775685e-06,
-      "loss": 0.2054,
+      "epoch": 0.84,
+      "grad_norm": 5.036116123199463,
+      "learning_rate": 1.4392335690080742e-05,
+      "loss": 1.1576,
       "step": 6703
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 15.68226432800293,
-      "learning_rate": 6.563095118773179e-06,
-      "loss": 2.487,
+      "epoch": 0.84,
+      "grad_norm": 118.02140045166016,
+      "learning_rate": 1.4391498975024056e-05,
+      "loss": 0.8999,
       "step": 6704
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 42.038726806640625,
-      "learning_rate": 6.5610905081687896e-06,
-      "loss": 2.1623,
+      "epoch": 0.84,
+      "grad_norm": 9.128668785095215,
+      "learning_rate": 1.439066225996737e-05,
+      "loss": 1.5079,
       "step": 6705
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 19.80816650390625,
-      "learning_rate": 6.559085897564399e-06,
-      "loss": 1.0708,
+      "epoch": 0.84,
+      "grad_norm": 9.483262062072754,
+      "learning_rate": 1.4389825544910683e-05,
+      "loss": 0.8413,
       "step": 6706
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 9.321516990661621,
-      "learning_rate": 6.557081286960009e-06,
-      "loss": 1.605,
+      "epoch": 0.84,
+      "grad_norm": 32.772186279296875,
+      "learning_rate": 1.4388988829853994e-05,
+      "loss": 2.5021,
       "step": 6707
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 159.25819396972656,
-      "learning_rate": 6.555076676355618e-06,
-      "loss": 2.8398,
+      "epoch": 0.84,
+      "grad_norm": 9.163297653198242,
+      "learning_rate": 1.4388152114797307e-05,
+      "loss": 0.6177,
       "step": 6708
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 17.203962326049805,
-      "learning_rate": 6.553072065751228e-06,
-      "loss": 1.4595,
+      "epoch": 0.84,
+      "grad_norm": 19.01010513305664,
+      "learning_rate": 1.438731539974062e-05,
+      "loss": 1.5838,
       "step": 6709
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 17.149423599243164,
-      "learning_rate": 6.5510674551468374e-06,
-      "loss": 1.7006,
+      "epoch": 0.84,
+      "grad_norm": 219.34219360351562,
+      "learning_rate": 1.4386478684683931e-05,
+      "loss": 1.209,
       "step": 6710
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 12.44357967376709,
-      "learning_rate": 6.549062844542448e-06,
-      "loss": 0.8718,
+      "epoch": 0.84,
+      "grad_norm": 8.837409019470215,
+      "learning_rate": 1.4385641969627245e-05,
+      "loss": 0.5343,
       "step": 6711
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 9.552597999572754,
-      "learning_rate": 6.5470582339380585e-06,
-      "loss": 1.4398,
+      "epoch": 0.84,
+      "grad_norm": 10.831384658813477,
+      "learning_rate": 1.4384805254570557e-05,
+      "loss": 1.4185,
       "step": 6712
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 12.471324920654297,
-      "learning_rate": 6.545053623333668e-06,
-      "loss": 2.0504,
+      "epoch": 0.84,
+      "grad_norm": 6.488797664642334,
+      "learning_rate": 1.438396853951387e-05,
+      "loss": 0.1876,
       "step": 6713
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 13.623446464538574,
-      "learning_rate": 6.543049012729278e-06,
-      "loss": 1.7889,
+      "epoch": 0.84,
+      "grad_norm": 18.163673400878906,
+      "learning_rate": 1.438313182445718e-05,
+      "loss": 2.2669,
       "step": 6714
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 8.370439529418945,
-      "learning_rate": 6.541044402124888e-06,
-      "loss": 1.1749,
+      "epoch": 0.84,
+      "grad_norm": 10.061616897583008,
+      "learning_rate": 1.4382295109400494e-05,
+      "loss": 1.0209,
       "step": 6715
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 16.073829650878906,
-      "learning_rate": 6.539039791520497e-06,
-      "loss": 1.3742,
+      "epoch": 0.84,
+      "grad_norm": 10.187827110290527,
+      "learning_rate": 1.4381458394343808e-05,
+      "loss": 0.3215,
       "step": 6716
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 9.64084529876709,
-      "learning_rate": 6.537035180916108e-06,
-      "loss": 1.142,
+      "epoch": 0.84,
+      "grad_norm": 30.014314651489258,
+      "learning_rate": 1.4380621679287118e-05,
+      "loss": 1.3229,
       "step": 6717
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 25.97657585144043,
-      "learning_rate": 6.535030570311718e-06,
-      "loss": 2.3333,
+      "epoch": 0.84,
+      "grad_norm": 11.296913146972656,
+      "learning_rate": 1.4379784964230432e-05,
+      "loss": 1.4596,
       "step": 6718
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 118.77561950683594,
-      "learning_rate": 6.533025959707327e-06,
-      "loss": 1.5912,
+      "epoch": 0.84,
+      "grad_norm": 7.084781646728516,
+      "learning_rate": 1.4378948249173746e-05,
+      "loss": 0.3353,
       "step": 6719
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 14.903034210205078,
-      "learning_rate": 6.5310213491029374e-06,
-      "loss": 1.897,
-      "step": 6720
-    },
-    {
-      "epoch": 2.02,
-      "eval_loss": 0.16902218759059906,
-      "eval_runtime": 43.4322,
-      "eval_samples_per_second": 34.053,
-      "eval_steps_per_second": 34.053,
+      "epoch": 0.84,
+      "grad_norm": 29.677143096923828,
+      "learning_rate": 1.437811153411706e-05,
+      "loss": 0.8754,
       "step": 6720
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 39.25941848754883,
-      "learning_rate": 6.529016738498547e-06,
-      "loss": 1.8256,
+      "epoch": 0.84,
+      "grad_norm": 44.01093673706055,
+      "learning_rate": 1.437727481906037e-05,
+      "loss": 1.5755,
       "step": 6721
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 9.439811706542969,
-      "learning_rate": 6.527012127894157e-06,
-      "loss": 0.9625,
+      "epoch": 0.84,
+      "grad_norm": 7.352614402770996,
+      "learning_rate": 1.4376438104003683e-05,
+      "loss": 0.72,
       "step": 6722
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 12.62756061553955,
-      "learning_rate": 6.525007517289767e-06,
-      "loss": 0.8624,
+      "epoch": 0.84,
+      "grad_norm": 11.98291301727295,
+      "learning_rate": 1.4375601388946995e-05,
+      "loss": 0.597,
       "step": 6723
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 13.67402458190918,
-      "learning_rate": 6.523002906685377e-06,
-      "loss": 0.785,
+      "epoch": 0.84,
+      "grad_norm": 13.343174934387207,
+      "learning_rate": 1.4374764673890307e-05,
+      "loss": 1.7283,
       "step": 6724
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 29.54031753540039,
-      "learning_rate": 6.520998296080987e-06,
-      "loss": 2.0936,
+      "epoch": 0.84,
+      "grad_norm": 12.499581336975098,
+      "learning_rate": 1.437392795883362e-05,
+      "loss": 0.9371,
       "step": 6725
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 13.225194931030273,
-      "learning_rate": 6.518993685476597e-06,
-      "loss": 1.2107,
+      "epoch": 0.84,
+      "grad_norm": 19.109291076660156,
+      "learning_rate": 1.4373091243776933e-05,
+      "loss": 2.2072,
       "step": 6726
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 8.916555404663086,
-      "learning_rate": 6.516989074872206e-06,
-      "loss": 1.0452,
+      "epoch": 0.84,
+      "grad_norm": 8.464346885681152,
+      "learning_rate": 1.4372254528720246e-05,
+      "loss": 1.1365,
       "step": 6727
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 24.967422485351562,
-      "learning_rate": 6.514984464267816e-06,
-      "loss": 2.1105,
+      "epoch": 0.84,
+      "grad_norm": 52.96107482910156,
+      "learning_rate": 1.4371417813663557e-05,
+      "loss": 3.4795,
       "step": 6728
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 14.154083251953125,
-      "learning_rate": 6.512979853663426e-06,
-      "loss": 0.7428,
+      "epoch": 0.84,
+      "grad_norm": 32.23268508911133,
+      "learning_rate": 1.437058109860687e-05,
+      "loss": 2.4121,
       "step": 6729
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 17.682830810546875,
-      "learning_rate": 6.510975243059037e-06,
-      "loss": 1.3675,
+      "epoch": 0.84,
+      "grad_norm": 11.593282699584961,
+      "learning_rate": 1.4369744383550184e-05,
+      "loss": 0.9362,
       "step": 6730
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 12.129364013671875,
-      "learning_rate": 6.508970632454647e-06,
-      "loss": 1.7799,
+      "epoch": 0.84,
+      "grad_norm": 10.491436958312988,
+      "learning_rate": 1.4368907668493494e-05,
+      "loss": 0.9679,
       "step": 6731
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 37.760650634765625,
-      "learning_rate": 6.506966021850256e-06,
-      "loss": 1.1513,
+      "epoch": 0.84,
+      "grad_norm": 14.975268363952637,
+      "learning_rate": 1.4368070953436808e-05,
+      "loss": 1.1917,
       "step": 6732
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 87.66400909423828,
-      "learning_rate": 6.504961411245866e-06,
-      "loss": 1.9592,
+      "epoch": 0.84,
+      "grad_norm": 9.948664665222168,
+      "learning_rate": 1.4367234238380121e-05,
+      "loss": 1.3456,
       "step": 6733
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 12.75828742980957,
-      "learning_rate": 6.502956800641475e-06,
-      "loss": 1.1079,
+      "epoch": 0.85,
+      "grad_norm": 12.157992362976074,
+      "learning_rate": 1.4366397523323435e-05,
+      "loss": 1.9081,
       "step": 6734
     },
     {
-      "epoch": 2.02,
-      "grad_norm": 12.436376571655273,
-      "learning_rate": 6.500952190037085e-06,
-      "loss": 0.8629,
+      "epoch": 0.85,
+      "grad_norm": 20.210771560668945,
+      "learning_rate": 1.4365560808266745e-05,
+      "loss": 2.5774,
       "step": 6735
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 8.984572410583496,
-      "learning_rate": 6.498947579432696e-06,
-      "loss": 0.9553,
+      "epoch": 0.85,
+      "grad_norm": 5.872859477996826,
+      "learning_rate": 1.4364724093210059e-05,
+      "loss": 1.6095,
       "step": 6736
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 22.55635643005371,
-      "learning_rate": 6.4969429688283055e-06,
-      "loss": 1.6062,
+      "epoch": 0.85,
+      "grad_norm": 10.525941848754883,
+      "learning_rate": 1.4363887378153371e-05,
+      "loss": 1.6281,
       "step": 6737
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 13.196118354797363,
-      "learning_rate": 6.4949383582239156e-06,
-      "loss": 0.5885,
+      "epoch": 0.85,
+      "grad_norm": 36.14300537109375,
+      "learning_rate": 1.4363050663096683e-05,
+      "loss": 3.0778,
       "step": 6738
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 46.98003387451172,
-      "learning_rate": 6.492933747619526e-06,
-      "loss": 1.3093,
+      "epoch": 0.85,
+      "grad_norm": 26.05812644958496,
+      "learning_rate": 1.4362213948039997e-05,
+      "loss": 2.2379,
       "step": 6739
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 23.900150299072266,
-      "learning_rate": 6.490929137015135e-06,
-      "loss": 1.4142,
+      "epoch": 0.85,
+      "grad_norm": 18.90000343322754,
+      "learning_rate": 1.4361377232983309e-05,
+      "loss": 1.5087,
       "step": 6740
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 17.741270065307617,
-      "learning_rate": 6.488924526410745e-06,
-      "loss": 1.4268,
+      "epoch": 0.85,
+      "grad_norm": 30.030092239379883,
+      "learning_rate": 1.4360540517926622e-05,
+      "loss": 2.7298,
       "step": 6741
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 26.587739944458008,
-      "learning_rate": 6.486919915806356e-06,
-      "loss": 2.1686,
+      "epoch": 0.85,
+      "grad_norm": 19.3814697265625,
+      "learning_rate": 1.4359703802869932e-05,
+      "loss": 1.3061,
       "step": 6742
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 35.67372131347656,
-      "learning_rate": 6.484915305201965e-06,
-      "loss": 1.6213,
+      "epoch": 0.85,
+      "grad_norm": 82.06686401367188,
+      "learning_rate": 1.4358867087813246e-05,
+      "loss": 0.9033,
       "step": 6743
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 26.185956954956055,
-      "learning_rate": 6.482910694597575e-06,
-      "loss": 1.1309,
+      "epoch": 0.85,
+      "grad_norm": 18.357946395874023,
+      "learning_rate": 1.435803037275656e-05,
+      "loss": 2.0855,
       "step": 6744
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 17.040918350219727,
-      "learning_rate": 6.4809060839931845e-06,
-      "loss": 1.0778,
+      "epoch": 0.85,
+      "grad_norm": 13.921502113342285,
+      "learning_rate": 1.435719365769987e-05,
+      "loss": 1.5051,
       "step": 6745
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 23.627857208251953,
-      "learning_rate": 6.4789014733887946e-06,
-      "loss": 1.1968,
+      "epoch": 0.85,
+      "grad_norm": 6.822644233703613,
+      "learning_rate": 1.4356356942643184e-05,
+      "loss": 1.0913,
       "step": 6746
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 29.324899673461914,
-      "learning_rate": 6.476896862784405e-06,
-      "loss": 1.9221,
+      "epoch": 0.85,
+      "grad_norm": 13.126599311828613,
+      "learning_rate": 1.4355520227586497e-05,
+      "loss": 2.861,
       "step": 6747
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 24.806861877441406,
-      "learning_rate": 6.474892252180015e-06,
-      "loss": 1.0605,
+      "epoch": 0.85,
+      "grad_norm": 11.187861442565918,
+      "learning_rate": 1.435468351252981e-05,
+      "loss": 1.2133,
       "step": 6748
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 32.63029098510742,
-      "learning_rate": 6.472887641575625e-06,
-      "loss": 1.4815,
+      "epoch": 0.85,
+      "grad_norm": 12.776350021362305,
+      "learning_rate": 1.4353846797473121e-05,
+      "loss": 2.4266,
       "step": 6749
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 17.42801284790039,
-      "learning_rate": 6.470883030971235e-06,
-      "loss": 1.6286,
+      "epoch": 0.85,
+      "grad_norm": 135.60122680664062,
+      "learning_rate": 1.4353010082416435e-05,
+      "loss": 4.2787,
       "step": 6750
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 40.37263488769531,
-      "learning_rate": 6.468878420366844e-06,
-      "loss": 1.6338,
+      "epoch": 0.85,
+      "grad_norm": 6.1601433753967285,
+      "learning_rate": 1.4352173367359747e-05,
+      "loss": 1.3039,
       "step": 6751
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 17.74640464782715,
-      "learning_rate": 6.466873809762454e-06,
-      "loss": 1.2096,
+      "epoch": 0.85,
+      "grad_norm": 10.253626823425293,
+      "learning_rate": 1.4351336652303059e-05,
+      "loss": 0.742,
       "step": 6752
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 26.730159759521484,
-      "learning_rate": 6.4648691991580634e-06,
-      "loss": 2.1374,
+      "epoch": 0.85,
+      "grad_norm": 16.65516471862793,
+      "learning_rate": 1.4350499937246372e-05,
+      "loss": 1.8368,
       "step": 6753
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 13.699724197387695,
-      "learning_rate": 6.462864588553674e-06,
-      "loss": 1.7276,
+      "epoch": 0.85,
+      "grad_norm": 7.756155490875244,
+      "learning_rate": 1.4349663222189684e-05,
+      "loss": 0.527,
       "step": 6754
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 26.75263023376465,
-      "learning_rate": 6.4608599779492845e-06,
-      "loss": 1.0525,
+      "epoch": 0.85,
+      "grad_norm": 13.48439884185791,
+      "learning_rate": 1.4348826507132998e-05,
+      "loss": 1.2254,
       "step": 6755
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 24.192272186279297,
-      "learning_rate": 6.458855367344894e-06,
-      "loss": 1.4798,
+      "epoch": 0.85,
+      "grad_norm": 30.818511962890625,
+      "learning_rate": 1.4347989792076308e-05,
+      "loss": 2.1113,
       "step": 6756
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 62.38142013549805,
-      "learning_rate": 6.456850756740504e-06,
-      "loss": 1.8068,
+      "epoch": 0.85,
+      "grad_norm": 7.61986780166626,
+      "learning_rate": 1.4347153077019622e-05,
+      "loss": 1.1408,
       "step": 6757
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 29.921567916870117,
-      "learning_rate": 6.454846146136113e-06,
-      "loss": 1.619,
+      "epoch": 0.85,
+      "grad_norm": 6.420275688171387,
+      "learning_rate": 1.4346316361962936e-05,
+      "loss": 1.0442,
       "step": 6758
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 6.582921981811523,
-      "learning_rate": 6.452841535531723e-06,
-      "loss": 0.648,
+      "epoch": 0.85,
+      "grad_norm": 15.52956485748291,
+      "learning_rate": 1.4345479646906246e-05,
+      "loss": 0.8404,
       "step": 6759
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 13.459547996520996,
-      "learning_rate": 6.450836924927334e-06,
-      "loss": 1.4701,
+      "epoch": 0.85,
+      "grad_norm": 17.56768798828125,
+      "learning_rate": 1.434464293184956e-05,
+      "loss": 1.8449,
       "step": 6760
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 19.61090660095215,
-      "learning_rate": 6.448832314322943e-06,
-      "loss": 1.2505,
+      "epoch": 0.85,
+      "grad_norm": 13.578958511352539,
+      "learning_rate": 1.4343806216792873e-05,
+      "loss": 1.668,
       "step": 6761
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 32.77788543701172,
-      "learning_rate": 6.446827703718553e-06,
-      "loss": 1.9251,
+      "epoch": 0.85,
+      "grad_norm": 21.581754684448242,
+      "learning_rate": 1.4342969501736185e-05,
+      "loss": 1.9299,
       "step": 6762
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 7.951191425323486,
-      "learning_rate": 6.4448230931141634e-06,
-      "loss": 0.7957,
+      "epoch": 0.85,
+      "grad_norm": 14.992815971374512,
+      "learning_rate": 1.4342132786679497e-05,
+      "loss": 2.406,
       "step": 6763
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 13.6834135055542,
-      "learning_rate": 6.442818482509773e-06,
-      "loss": 1.488,
+      "epoch": 0.85,
+      "grad_norm": 24.969329833984375,
+      "learning_rate": 1.434129607162281e-05,
+      "loss": 0.9601,
       "step": 6764
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 26.919858932495117,
-      "learning_rate": 6.440813871905383e-06,
-      "loss": 1.6017,
+      "epoch": 0.85,
+      "grad_norm": 28.413381576538086,
+      "learning_rate": 1.4340459356566123e-05,
+      "loss": 0.5697,
       "step": 6765
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 18.014230728149414,
-      "learning_rate": 6.438809261300992e-06,
-      "loss": 0.8233,
+      "epoch": 0.85,
+      "grad_norm": 10.309455871582031,
+      "learning_rate": 1.4339622641509435e-05,
+      "loss": 0.6994,
       "step": 6766
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 18.699932098388672,
-      "learning_rate": 6.436804650696603e-06,
-      "loss": 1.1873,
+      "epoch": 0.85,
+      "grad_norm": 6.308617115020752,
+      "learning_rate": 1.4338785926452747e-05,
+      "loss": 2.0733,
       "step": 6767
     },
     {
-      "epoch": 2.03,
-      "grad_norm": 24.57878303527832,
-      "learning_rate": 6.434800040092213e-06,
-      "loss": 1.7365,
+      "epoch": 0.85,
+      "grad_norm": 36.197715759277344,
+      "learning_rate": 1.433794921139606e-05,
+      "loss": 0.8404,
       "step": 6768
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 30.225631713867188,
-      "learning_rate": 6.432795429487822e-06,
-      "loss": 2.7789,
+      "epoch": 0.85,
+      "grad_norm": 14.402820587158203,
+      "learning_rate": 1.4337112496339374e-05,
+      "loss": 1.3064,
       "step": 6769
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 22.14498519897461,
-      "learning_rate": 6.430790818883432e-06,
-      "loss": 1.6425,
+      "epoch": 0.85,
+      "grad_norm": 3.842750310897827,
+      "learning_rate": 1.4336275781282684e-05,
+      "loss": 0.3571,
       "step": 6770
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 12.216958045959473,
-      "learning_rate": 6.428786208279042e-06,
-      "loss": 1.7486,
+      "epoch": 0.85,
+      "grad_norm": 12.000794410705566,
+      "learning_rate": 1.4335439066225998e-05,
+      "loss": 1.7991,
       "step": 6771
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 26.501882553100586,
-      "learning_rate": 6.426781597674652e-06,
-      "loss": 1.3613,
+      "epoch": 0.85,
+      "grad_norm": 15.14625358581543,
+      "learning_rate": 1.4334602351169311e-05,
+      "loss": 0.6939,
       "step": 6772
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 30.44036293029785,
-      "learning_rate": 6.424776987070263e-06,
-      "loss": 1.2315,
+      "epoch": 0.85,
+      "grad_norm": 9.255779266357422,
+      "learning_rate": 1.4333765636112622e-05,
+      "loss": 0.6928,
       "step": 6773
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 19.72119140625,
-      "learning_rate": 6.422772376465873e-06,
-      "loss": 1.8577,
+      "epoch": 0.85,
+      "grad_norm": 6.127439022064209,
+      "learning_rate": 1.4332928921055935e-05,
+      "loss": 0.4097,
       "step": 6774
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 11.062511444091797,
-      "learning_rate": 6.420767765861482e-06,
-      "loss": 1.4272,
+      "epoch": 0.85,
+      "grad_norm": 17.041950225830078,
+      "learning_rate": 1.4332092205999249e-05,
+      "loss": 1.0672,
       "step": 6775
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 16.308460235595703,
-      "learning_rate": 6.418763155257092e-06,
-      "loss": 1.3599,
+      "epoch": 0.85,
+      "grad_norm": 8.400264739990234,
+      "learning_rate": 1.4331255490942561e-05,
+      "loss": 1.4374,
       "step": 6776
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 35.35619354248047,
-      "learning_rate": 6.416758544652701e-06,
-      "loss": 1.477,
+      "epoch": 0.85,
+      "grad_norm": 50.31614303588867,
+      "learning_rate": 1.4330418775885873e-05,
+      "loss": 3.6345,
       "step": 6777
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 13.431120872497559,
-      "learning_rate": 6.414753934048311e-06,
-      "loss": 0.9098,
+      "epoch": 0.85,
+      "grad_norm": 43.917266845703125,
+      "learning_rate": 1.4329582060829187e-05,
+      "loss": 2.2039,
       "step": 6778
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 22.69354820251465,
-      "learning_rate": 6.412749323443922e-06,
-      "loss": 1.3123,
+      "epoch": 0.85,
+      "grad_norm": 17.45854949951172,
+      "learning_rate": 1.4328745345772498e-05,
+      "loss": 1.9558,
       "step": 6779
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 41.717098236083984,
-      "learning_rate": 6.4107447128395315e-06,
-      "loss": 2.688,
+      "epoch": 0.85,
+      "grad_norm": 18.78993034362793,
+      "learning_rate": 1.432790863071581e-05,
+      "loss": 1.9115,
       "step": 6780
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 9.498014450073242,
-      "learning_rate": 6.4087401022351416e-06,
-      "loss": 0.8678,
+      "epoch": 0.85,
+      "grad_norm": 9.703263282775879,
+      "learning_rate": 1.4327071915659122e-05,
+      "loss": 0.6795,
       "step": 6781
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 10.49351692199707,
-      "learning_rate": 6.406735491630751e-06,
-      "loss": 1.8695,
+      "epoch": 0.85,
+      "grad_norm": 8.62363052368164,
+      "learning_rate": 1.4326235200602436e-05,
+      "loss": 1.3623,
       "step": 6782
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 31.190580368041992,
-      "learning_rate": 6.404730881026361e-06,
-      "loss": 1.78,
+      "epoch": 0.85,
+      "grad_norm": 103.20475769042969,
+      "learning_rate": 1.432539848554575e-05,
+      "loss": 1.1805,
       "step": 6783
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 15.785634994506836,
-      "learning_rate": 6.402726270421971e-06,
-      "loss": 1.0891,
+      "epoch": 0.85,
+      "grad_norm": 23.955724716186523,
+      "learning_rate": 1.432456177048906e-05,
+      "loss": 1.0631,
       "step": 6784
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 14.310815811157227,
-      "learning_rate": 6.400721659817581e-06,
-      "loss": 0.9196,
+      "epoch": 0.85,
+      "grad_norm": 37.74839782714844,
+      "learning_rate": 1.4323725055432374e-05,
+      "loss": 1.071,
       "step": 6785
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 38.41286087036133,
-      "learning_rate": 6.398717049213191e-06,
-      "loss": 2.4826,
+      "epoch": 0.85,
+      "grad_norm": 46.22028732299805,
+      "learning_rate": 1.4322888340375687e-05,
+      "loss": 2.6372,
       "step": 6786
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 10.610367774963379,
-      "learning_rate": 6.396712438608801e-06,
-      "loss": 1.0001,
+      "epoch": 0.85,
+      "grad_norm": 45.86208724975586,
+      "learning_rate": 1.4322051625318998e-05,
+      "loss": 1.0468,
       "step": 6787
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 9.231528282165527,
-      "learning_rate": 6.3947078280044105e-06,
-      "loss": 1.0759,
+      "epoch": 0.85,
+      "grad_norm": 18.5753116607666,
+      "learning_rate": 1.4321214910262311e-05,
+      "loss": 2.4317,
       "step": 6788
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 7.4241862297058105,
-      "learning_rate": 6.3927032174000206e-06,
-      "loss": 1.9806,
+      "epoch": 0.85,
+      "grad_norm": 6.768474578857422,
+      "learning_rate": 1.4320378195205625e-05,
+      "loss": 0.7415,
       "step": 6789
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 14.655599594116211,
-      "learning_rate": 6.39069860679563e-06,
-      "loss": 1.3411,
+      "epoch": 0.85,
+      "grad_norm": 14.827455520629883,
+      "learning_rate": 1.4319541480148937e-05,
+      "loss": 1.5113,
       "step": 6790
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 10.586833000183105,
-      "learning_rate": 6.388693996191241e-06,
-      "loss": 0.9932,
+      "epoch": 0.85,
+      "grad_norm": 7.561862945556641,
+      "learning_rate": 1.4318704765092249e-05,
+      "loss": 2.9066,
       "step": 6791
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 19.10291290283203,
-      "learning_rate": 6.386689385586851e-06,
-      "loss": 1.5976,
+      "epoch": 0.85,
+      "grad_norm": 46.11296844482422,
+      "learning_rate": 1.4317868050035562e-05,
+      "loss": 2.2379,
       "step": 6792
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 15.874801635742188,
-      "learning_rate": 6.38468477498246e-06,
-      "loss": 0.897,
+      "epoch": 0.85,
+      "grad_norm": 9.779031753540039,
+      "learning_rate": 1.4317031334978874e-05,
+      "loss": 1.3426,
       "step": 6793
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 13.638111114501953,
-      "learning_rate": 6.38268016437807e-06,
-      "loss": 0.7186,
+      "epoch": 0.85,
+      "grad_norm": 28.658103942871094,
+      "learning_rate": 1.4316194619922186e-05,
+      "loss": 2.5177,
       "step": 6794
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 9.389366149902344,
-      "learning_rate": 6.380675553773679e-06,
-      "loss": 0.9544,
+      "epoch": 0.85,
+      "grad_norm": 66.79032897949219,
+      "learning_rate": 1.4315357904865498e-05,
+      "loss": 1.5142,
       "step": 6795
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 26.145200729370117,
-      "learning_rate": 6.3786709431692894e-06,
-      "loss": 1.3817,
+      "epoch": 0.85,
+      "grad_norm": 8.954390525817871,
+      "learning_rate": 1.4314521189808812e-05,
+      "loss": 0.8566,
       "step": 6796
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 12.559453010559082,
-      "learning_rate": 6.3766663325649e-06,
-      "loss": 1.5977,
+      "epoch": 0.85,
+      "grad_norm": 10.869063377380371,
+      "learning_rate": 1.4313684474752126e-05,
+      "loss": 0.6127,
       "step": 6797
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 9.843866348266602,
-      "learning_rate": 6.37466172196051e-06,
-      "loss": 0.9535,
+      "epoch": 0.85,
+      "grad_norm": 25.037322998046875,
+      "learning_rate": 1.4312847759695436e-05,
+      "loss": 2.1371,
       "step": 6798
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 15.275118827819824,
-      "learning_rate": 6.37265711135612e-06,
-      "loss": 1.3151,
+      "epoch": 0.85,
+      "grad_norm": 4.422179698944092,
+      "learning_rate": 1.431201104463875e-05,
+      "loss": 0.2207,
       "step": 6799
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 18.97870635986328,
-      "learning_rate": 6.37065250075173e-06,
-      "loss": 1.6633,
+      "epoch": 0.85,
+      "grad_norm": 16.942899703979492,
+      "learning_rate": 1.4311174329582063e-05,
+      "loss": 2.3062,
       "step": 6800
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 11.981043815612793,
-      "learning_rate": 6.368647890147339e-06,
-      "loss": 1.0551,
+      "epoch": 0.85,
+      "eval_loss": 0.09935930371284485,
+      "eval_runtime": 111.9226,
+      "eval_samples_per_second": 31.647,
+      "eval_steps_per_second": 31.647,
+      "step": 6800
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 17.58791732788086,
+      "learning_rate": 1.4310337614525373e-05,
+      "loss": 2.8723,
       "step": 6801
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 25.228471755981445,
-      "learning_rate": 6.366643279542949e-06,
-      "loss": 1.4122,
+      "epoch": 0.85,
+      "grad_norm": 10.952619552612305,
+      "learning_rate": 1.4309500899468687e-05,
+      "loss": 0.6372,
       "step": 6802
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 46.05463790893555,
-      "learning_rate": 6.36463866893856e-06,
-      "loss": 2.192,
+      "epoch": 0.85,
+      "grad_norm": 30.083515167236328,
+      "learning_rate": 1.4308664184412e-05,
+      "loss": 2.126,
       "step": 6803
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 28.93854522705078,
-      "learning_rate": 6.362634058334169e-06,
-      "loss": 1.235,
+      "epoch": 0.85,
+      "grad_norm": 29.942514419555664,
+      "learning_rate": 1.4307827469355313e-05,
+      "loss": 0.8289,
       "step": 6804
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 22.576128005981445,
-      "learning_rate": 6.360629447729779e-06,
-      "loss": 1.61,
+      "epoch": 0.85,
+      "grad_norm": 26.02439308166504,
+      "learning_rate": 1.4306990754298625e-05,
+      "loss": 1.4562,
       "step": 6805
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 8.201756477355957,
-      "learning_rate": 6.358624837125389e-06,
-      "loss": 1.3716,
+      "epoch": 0.85,
+      "grad_norm": 13.239129066467285,
+      "learning_rate": 1.4306154039241938e-05,
+      "loss": 2.6482,
       "step": 6806
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 9.097603797912598,
-      "learning_rate": 6.356620226520999e-06,
-      "loss": 0.69,
+      "epoch": 0.85,
+      "grad_norm": 14.090408325195312,
+      "learning_rate": 1.430531732418525e-05,
+      "loss": 0.621,
       "step": 6807
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 26.398914337158203,
-      "learning_rate": 6.354615615916609e-06,
-      "loss": 2.6075,
+      "epoch": 0.85,
+      "grad_norm": 4.832688331604004,
+      "learning_rate": 1.4304480609128562e-05,
+      "loss": 0.4012,
       "step": 6808
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 15.047880172729492,
-      "learning_rate": 6.352611005312218e-06,
-      "loss": 1.2919,
+      "epoch": 0.85,
+      "grad_norm": 21.516300201416016,
+      "learning_rate": 1.4303643894071874e-05,
+      "loss": 4.5126,
       "step": 6809
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 42.7448616027832,
-      "learning_rate": 6.350606394707829e-06,
-      "loss": 1.1902,
+      "epoch": 0.85,
+      "grad_norm": 6.897380828857422,
+      "learning_rate": 1.4302807179015188e-05,
+      "loss": 0.1666,
       "step": 6810
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 29.7187557220459,
-      "learning_rate": 6.348601784103439e-06,
-      "loss": 1.9144,
+      "epoch": 0.85,
+      "grad_norm": 14.608335494995117,
+      "learning_rate": 1.4301970463958501e-05,
+      "loss": 1.9012,
       "step": 6811
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 13.130889892578125,
-      "learning_rate": 6.346597173499048e-06,
-      "loss": 0.9165,
+      "epoch": 0.85,
+      "grad_norm": 9.011201858520508,
+      "learning_rate": 1.4301133748901812e-05,
+      "loss": 1.2681,
       "step": 6812
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 14.863086700439453,
-      "learning_rate": 6.344592562894658e-06,
-      "loss": 2.3026,
+      "epoch": 0.86,
+      "grad_norm": 10.01158618927002,
+      "learning_rate": 1.4300297033845125e-05,
+      "loss": 0.1179,
       "step": 6813
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 31.419349670410156,
-      "learning_rate": 6.342587952290268e-06,
-      "loss": 1.929,
+      "epoch": 0.86,
+      "grad_norm": 11.291976928710938,
+      "learning_rate": 1.4299460318788439e-05,
+      "loss": 1.6185,
       "step": 6814
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 16.03536605834961,
-      "learning_rate": 6.340583341685878e-06,
-      "loss": 1.0579,
+      "epoch": 0.86,
+      "grad_norm": 19.130647659301758,
+      "learning_rate": 1.429862360373175e-05,
+      "loss": 1.3798,
       "step": 6815
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 9.93125057220459,
-      "learning_rate": 6.338578731081489e-06,
-      "loss": 1.0248,
+      "epoch": 0.86,
+      "grad_norm": 8.56436538696289,
+      "learning_rate": 1.4297786888675063e-05,
+      "loss": 0.6916,
       "step": 6816
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 20.598365783691406,
-      "learning_rate": 6.336574120477098e-06,
-      "loss": 1.0575,
+      "epoch": 0.86,
+      "grad_norm": 14.949737548828125,
+      "learning_rate": 1.4296950173618376e-05,
+      "loss": 2.0076,
       "step": 6817
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 24.11957359313965,
-      "learning_rate": 6.334569509872708e-06,
-      "loss": 2.0264,
+      "epoch": 0.86,
+      "grad_norm": 37.16676330566406,
+      "learning_rate": 1.4296113458561688e-05,
+      "loss": 2.9857,
       "step": 6818
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 12.065218925476074,
-      "learning_rate": 6.332564899268317e-06,
-      "loss": 0.6793,
+      "epoch": 0.86,
+      "grad_norm": 68.93708038330078,
+      "learning_rate": 1.4295276743505e-05,
+      "loss": 1.1956,
       "step": 6819
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 43.54728698730469,
-      "learning_rate": 6.330560288663927e-06,
-      "loss": 2.746,
+      "epoch": 0.86,
+      "grad_norm": 67.95276641845703,
+      "learning_rate": 1.4294440028448312e-05,
+      "loss": 2.1929,
       "step": 6820
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 12.573387145996094,
-      "learning_rate": 6.328555678059537e-06,
-      "loss": 1.0912,
+      "epoch": 0.86,
+      "grad_norm": 47.011775970458984,
+      "learning_rate": 1.4293603313391626e-05,
+      "loss": 1.3765,
       "step": 6821
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 53.12962341308594,
-      "learning_rate": 6.326551067455147e-06,
-      "loss": 2.5651,
+      "epoch": 0.86,
+      "grad_norm": 6.818073749542236,
+      "learning_rate": 1.4292766598334938e-05,
+      "loss": 0.2283,
       "step": 6822
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 27.892974853515625,
-      "learning_rate": 6.3245464568507575e-06,
-      "loss": 2.1102,
+      "epoch": 0.86,
+      "grad_norm": 4.801296710968018,
+      "learning_rate": 1.429192988327825e-05,
+      "loss": 0.4125,
       "step": 6823
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 11.939574241638184,
-      "learning_rate": 6.3225418462463676e-06,
-      "loss": 1.0358,
+      "epoch": 0.86,
+      "grad_norm": 17.18333625793457,
+      "learning_rate": 1.4291093168221564e-05,
+      "loss": 2.2435,
       "step": 6824
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 15.528552055358887,
-      "learning_rate": 6.320537235641977e-06,
-      "loss": 1.43,
+      "epoch": 0.86,
+      "grad_norm": 11.65798568725586,
+      "learning_rate": 1.4290256453164877e-05,
+      "loss": 0.8295,
       "step": 6825
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 19.028608322143555,
-      "learning_rate": 6.318532625037587e-06,
-      "loss": 1.3059,
+      "epoch": 0.86,
+      "grad_norm": 31.394832611083984,
+      "learning_rate": 1.4289419738108187e-05,
+      "loss": 2.3899,
       "step": 6826
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 29.945940017700195,
-      "learning_rate": 6.316528014433196e-06,
-      "loss": 1.4528,
+      "epoch": 0.86,
+      "grad_norm": 6.690638065338135,
+      "learning_rate": 1.4288583023051501e-05,
+      "loss": 0.5558,
       "step": 6827
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 8.454259872436523,
-      "learning_rate": 6.314523403828807e-06,
-      "loss": 0.9366,
+      "epoch": 0.86,
+      "grad_norm": 27.214235305786133,
+      "learning_rate": 1.4287746307994815e-05,
+      "loss": 3.1397,
       "step": 6828
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 19.378828048706055,
-      "learning_rate": 6.312518793224417e-06,
-      "loss": 0.8679,
+      "epoch": 0.86,
+      "grad_norm": 9.921982765197754,
+      "learning_rate": 1.4286909592938125e-05,
+      "loss": 0.5252,
       "step": 6829
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 20.16804313659668,
-      "learning_rate": 6.310514182620026e-06,
-      "loss": 1.6893,
+      "epoch": 0.86,
+      "grad_norm": 30.677799224853516,
+      "learning_rate": 1.4286072877881439e-05,
+      "loss": 1.8776,
       "step": 6830
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 9.753872871398926,
-      "learning_rate": 6.3085095720156365e-06,
-      "loss": 1.1253,
+      "epoch": 0.86,
+      "grad_norm": 10.252274513244629,
+      "learning_rate": 1.4285236162824752e-05,
+      "loss": 0.8843,
       "step": 6831
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 14.549400329589844,
-      "learning_rate": 6.306504961411246e-06,
-      "loss": 1.3079,
+      "epoch": 0.86,
+      "grad_norm": 8.96366024017334,
+      "learning_rate": 1.4284399447768064e-05,
+      "loss": 1.1557,
       "step": 6832
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 15.794536590576172,
-      "learning_rate": 6.304500350806856e-06,
-      "loss": 1.394,
+      "epoch": 0.86,
+      "grad_norm": 14.817914962768555,
+      "learning_rate": 1.4283562732711376e-05,
+      "loss": 1.5293,
       "step": 6833
     },
     {
-      "epoch": 2.05,
-      "grad_norm": 32.083927154541016,
-      "learning_rate": 6.302495740202467e-06,
-      "loss": 1.716,
+      "epoch": 0.86,
+      "grad_norm": 112.65931701660156,
+      "learning_rate": 1.4282726017654688e-05,
+      "loss": 1.856,
       "step": 6834
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 12.402176856994629,
-      "learning_rate": 6.300491129598077e-06,
-      "loss": 1.0654,
+      "epoch": 0.86,
+      "grad_norm": 3.1464476585388184,
+      "learning_rate": 1.4281889302598002e-05,
+      "loss": 0.2797,
       "step": 6835
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 16.35983657836914,
-      "learning_rate": 6.298486518993686e-06,
-      "loss": 1.1802,
+      "epoch": 0.86,
+      "grad_norm": 5.00979471206665,
+      "learning_rate": 1.4281052587541314e-05,
+      "loss": 0.4282,
       "step": 6836
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 27.28036117553711,
-      "learning_rate": 6.296481908389296e-06,
-      "loss": 1.8658,
+      "epoch": 0.86,
+      "grad_norm": 12.247176170349121,
+      "learning_rate": 1.4280215872484626e-05,
+      "loss": 0.8076,
       "step": 6837
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 10.746267318725586,
-      "learning_rate": 6.294477297784905e-06,
-      "loss": 1.2586,
+      "epoch": 0.86,
+      "grad_norm": 10.779017448425293,
+      "learning_rate": 1.427937915742794e-05,
+      "loss": 1.221,
       "step": 6838
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 27.72111701965332,
-      "learning_rate": 6.2924726871805155e-06,
-      "loss": 2.0948,
+      "epoch": 0.86,
+      "grad_norm": 9.00314998626709,
+      "learning_rate": 1.4278542442371253e-05,
+      "loss": 0.4914,
       "step": 6839
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 13.532320976257324,
-      "learning_rate": 6.290468076576126e-06,
-      "loss": 1.2246,
-      "step": 6840
-    },
-    {
-      "epoch": 2.06,
-      "eval_loss": 0.18061862885951996,
-      "eval_runtime": 43.8871,
-      "eval_samples_per_second": 33.7,
-      "eval_steps_per_second": 33.7,
+      "epoch": 0.86,
+      "grad_norm": 4.5170817375183105,
+      "learning_rate": 1.4277705727314563e-05,
+      "loss": 0.5171,
       "step": 6840
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 23.459665298461914,
-      "learning_rate": 6.288463465971736e-06,
-      "loss": 1.4494,
+      "epoch": 0.86,
+      "grad_norm": 21.79302215576172,
+      "learning_rate": 1.4276869012257877e-05,
+      "loss": 2.1364,
       "step": 6841
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 13.611770629882812,
-      "learning_rate": 6.286458855367346e-06,
-      "loss": 1.1819,
+      "epoch": 0.86,
+      "grad_norm": 7.483713150024414,
+      "learning_rate": 1.427603229720119e-05,
+      "loss": 1.3484,
       "step": 6842
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 25.924118041992188,
-      "learning_rate": 6.284454244762955e-06,
-      "loss": 1.0744,
+      "epoch": 0.86,
+      "grad_norm": 23.12000846862793,
+      "learning_rate": 1.4275195582144501e-05,
+      "loss": 1.1458,
       "step": 6843
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 19.138996124267578,
-      "learning_rate": 6.282449634158565e-06,
-      "loss": 1.2561,
+      "epoch": 0.86,
+      "grad_norm": 11.738572120666504,
+      "learning_rate": 1.4274358867087814e-05,
+      "loss": 2.4691,
       "step": 6844
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 46.55351638793945,
-      "learning_rate": 6.280445023554175e-06,
-      "loss": 1.4002,
+      "epoch": 0.86,
+      "grad_norm": 10.765612602233887,
+      "learning_rate": 1.4273522152031128e-05,
+      "loss": 1.4455,
       "step": 6845
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 16.169979095458984,
-      "learning_rate": 6.278440412949785e-06,
-      "loss": 1.2913,
+      "epoch": 0.86,
+      "grad_norm": 12.555647850036621,
+      "learning_rate": 1.427268543697444e-05,
+      "loss": 1.2412,
       "step": 6846
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 37.923133850097656,
-      "learning_rate": 6.276435802345395e-06,
-      "loss": 1.5886,
+      "epoch": 0.86,
+      "grad_norm": 13.866482734680176,
+      "learning_rate": 1.4271848721917752e-05,
+      "loss": 2.5328,
       "step": 6847
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 57.48485565185547,
-      "learning_rate": 6.274431191741005e-06,
-      "loss": 2.0952,
+      "epoch": 0.86,
+      "grad_norm": 26.960792541503906,
+      "learning_rate": 1.4271012006861064e-05,
+      "loss": 2.052,
       "step": 6848
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 13.789143562316895,
-      "learning_rate": 6.272426581136615e-06,
-      "loss": 1.712,
+      "epoch": 0.86,
+      "grad_norm": 23.56977081298828,
+      "learning_rate": 1.4270175291804378e-05,
+      "loss": 1.7812,
       "step": 6849
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 21.859037399291992,
-      "learning_rate": 6.270421970532225e-06,
-      "loss": 2.0063,
+      "epoch": 0.86,
+      "grad_norm": 18.891651153564453,
+      "learning_rate": 1.426933857674769e-05,
+      "loss": 2.1884,
       "step": 6850
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 13.258837699890137,
-      "learning_rate": 6.268417359927834e-06,
-      "loss": 1.2955,
+      "epoch": 0.86,
+      "grad_norm": 5.035281181335449,
+      "learning_rate": 1.4268501861691002e-05,
+      "loss": 1.1694,
       "step": 6851
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 8.517374038696289,
-      "learning_rate": 6.266412749323444e-06,
-      "loss": 1.0369,
+      "epoch": 0.86,
+      "grad_norm": 15.985159873962402,
+      "learning_rate": 1.4267665146634315e-05,
+      "loss": 1.0514,
       "step": 6852
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 10.359166145324707,
-      "learning_rate": 6.264408138719055e-06,
-      "loss": 1.0215,
+      "epoch": 0.86,
+      "grad_norm": 26.73896598815918,
+      "learning_rate": 1.4266828431577629e-05,
+      "loss": 1.9858,
       "step": 6853
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 28.84781837463379,
-      "learning_rate": 6.262403528114664e-06,
-      "loss": 1.9747,
+      "epoch": 0.86,
+      "grad_norm": 42.217872619628906,
+      "learning_rate": 1.4265991716520939e-05,
+      "loss": 3.1336,
       "step": 6854
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 11.506865501403809,
-      "learning_rate": 6.260398917510274e-06,
-      "loss": 1.1076,
+      "epoch": 0.86,
+      "grad_norm": 6.821484565734863,
+      "learning_rate": 1.4265155001464253e-05,
+      "loss": 0.6554,
       "step": 6855
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 15.466108322143555,
-      "learning_rate": 6.2583943069058835e-06,
-      "loss": 1.2779,
+      "epoch": 0.86,
+      "grad_norm": 45.5909538269043,
+      "learning_rate": 1.4264318286407566e-05,
+      "loss": 1.9567,
       "step": 6856
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 31.154294967651367,
-      "learning_rate": 6.256389696301494e-06,
-      "loss": 1.9285,
+      "epoch": 0.86,
+      "grad_norm": 21.700092315673828,
+      "learning_rate": 1.4263481571350877e-05,
+      "loss": 1.7945,
       "step": 6857
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 15.335009574890137,
-      "learning_rate": 6.254385085697104e-06,
-      "loss": 1.6327,
+      "epoch": 0.86,
+      "grad_norm": 7.3325676918029785,
+      "learning_rate": 1.426264485629419e-05,
+      "loss": 0.6593,
       "step": 6858
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 14.675117492675781,
-      "learning_rate": 6.252380475092715e-06,
-      "loss": 1.644,
+      "epoch": 0.86,
+      "grad_norm": 11.090742111206055,
+      "learning_rate": 1.4261808141237502e-05,
+      "loss": 1.1509,
       "step": 6859
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 133.6639862060547,
-      "learning_rate": 6.250375864488324e-06,
-      "loss": 1.1432,
+      "epoch": 0.86,
+      "grad_norm": 35.66590118408203,
+      "learning_rate": 1.4260971426180816e-05,
+      "loss": 2.1461,
       "step": 6860
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 39.901222229003906,
-      "learning_rate": 6.248371253883934e-06,
-      "loss": 2.097,
+      "epoch": 0.86,
+      "grad_norm": 18.339378356933594,
+      "learning_rate": 1.4260134711124128e-05,
+      "loss": 0.7108,
       "step": 6861
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 13.691509246826172,
-      "learning_rate": 6.246366643279543e-06,
-      "loss": 1.6336,
+      "epoch": 0.86,
+      "grad_norm": 16.616304397583008,
+      "learning_rate": 1.425929799606744e-05,
+      "loss": 0.9501,
       "step": 6862
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 14.485090255737305,
-      "learning_rate": 6.244362032675153e-06,
-      "loss": 1.3259,
+      "epoch": 0.86,
+      "grad_norm": 18.746173858642578,
+      "learning_rate": 1.4258461281010753e-05,
+      "loss": 2.2851,
       "step": 6863
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 72.5770492553711,
-      "learning_rate": 6.2423574220707625e-06,
-      "loss": 1.6907,
+      "epoch": 0.86,
+      "grad_norm": 12.186095237731934,
+      "learning_rate": 1.4257624565954065e-05,
+      "loss": 2.7731,
       "step": 6864
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 10.634366989135742,
-      "learning_rate": 6.240352811466373e-06,
-      "loss": 0.5027,
+      "epoch": 0.86,
+      "grad_norm": 9.447904586791992,
+      "learning_rate": 1.4256787850897377e-05,
+      "loss": 1.1061,
       "step": 6865
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 33.018367767333984,
-      "learning_rate": 6.2383482008619835e-06,
-      "loss": 1.7409,
+      "epoch": 0.86,
+      "grad_norm": 26.08285903930664,
+      "learning_rate": 1.4255951135840691e-05,
+      "loss": 1.4773,
       "step": 6866
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 9.623235702514648,
-      "learning_rate": 6.236343590257593e-06,
-      "loss": 1.715,
+      "epoch": 0.86,
+      "grad_norm": 8.051258087158203,
+      "learning_rate": 1.4255114420784005e-05,
+      "loss": 0.8141,
       "step": 6867
     },
     {
-      "epoch": 2.06,
-      "grad_norm": 47.73927307128906,
-      "learning_rate": 6.234338979653203e-06,
-      "loss": 1.4383,
+      "epoch": 0.86,
+      "grad_norm": 16.438636779785156,
+      "learning_rate": 1.4254277705727315e-05,
+      "loss": 1.0134,
       "step": 6868
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 9.162986755371094,
-      "learning_rate": 6.232334369048813e-06,
-      "loss": 1.1415,
+      "epoch": 0.86,
+      "grad_norm": 4.74124002456665,
+      "learning_rate": 1.4253440990670629e-05,
+      "loss": 0.3712,
       "step": 6869
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 41.81623077392578,
-      "learning_rate": 6.230329758444422e-06,
-      "loss": 1.6832,
+      "epoch": 0.86,
+      "grad_norm": 18.44993782043457,
+      "learning_rate": 1.4252604275613942e-05,
+      "loss": 0.8501,
       "step": 6870
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 21.047195434570312,
-      "learning_rate": 6.228325147840033e-06,
-      "loss": 1.202,
+      "epoch": 0.86,
+      "grad_norm": 12.221519470214844,
+      "learning_rate": 1.4251767560557253e-05,
+      "loss": 0.8011,
       "step": 6871
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 19.954036712646484,
-      "learning_rate": 6.226320537235643e-06,
-      "loss": 1.4934,
+      "epoch": 0.86,
+      "grad_norm": 23.90431785583496,
+      "learning_rate": 1.4250930845500566e-05,
+      "loss": 1.0791,
       "step": 6872
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 24.60007095336914,
-      "learning_rate": 6.224315926631252e-06,
-      "loss": 0.9536,
+      "epoch": 0.86,
+      "grad_norm": 13.108061790466309,
+      "learning_rate": 1.4250094130443878e-05,
+      "loss": 3.2827,
       "step": 6873
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 23.833621978759766,
-      "learning_rate": 6.2223113160268625e-06,
-      "loss": 2.561,
+      "epoch": 0.86,
+      "grad_norm": 31.300273895263672,
+      "learning_rate": 1.4249257415387192e-05,
+      "loss": 2.3229,
       "step": 6874
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 51.6407585144043,
-      "learning_rate": 6.220306705422472e-06,
-      "loss": 1.8729,
+      "epoch": 0.86,
+      "grad_norm": 107.48026275634766,
+      "learning_rate": 1.4248420700330504e-05,
+      "loss": 2.0077,
       "step": 6875
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 9.72765827178955,
-      "learning_rate": 6.218302094818082e-06,
-      "loss": 1.6889,
+      "epoch": 0.86,
+      "grad_norm": 13.970173835754395,
+      "learning_rate": 1.4247583985273816e-05,
+      "loss": 1.0402,
       "step": 6876
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 86.259765625,
-      "learning_rate": 6.216297484213693e-06,
-      "loss": 1.6229,
+      "epoch": 0.86,
+      "grad_norm": 3.0541088581085205,
+      "learning_rate": 1.424674727021713e-05,
+      "loss": 0.4837,
       "step": 6877
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 12.195944786071777,
-      "learning_rate": 6.214292873609302e-06,
-      "loss": 1.7568,
+      "epoch": 0.86,
+      "grad_norm": 7.691309928894043,
+      "learning_rate": 1.424591055516044e-05,
+      "loss": 0.3931,
       "step": 6878
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 41.19993591308594,
-      "learning_rate": 6.212288263004912e-06,
-      "loss": 1.2259,
+      "epoch": 0.86,
+      "grad_norm": 15.071650505065918,
+      "learning_rate": 1.4245073840103753e-05,
+      "loss": 3.4964,
       "step": 6879
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 10.510581970214844,
-      "learning_rate": 6.210283652400521e-06,
-      "loss": 1.1598,
+      "epoch": 0.86,
+      "grad_norm": 7.250248908996582,
+      "learning_rate": 1.4244237125047067e-05,
+      "loss": 0.6171,
       "step": 6880
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 14.523313522338867,
-      "learning_rate": 6.208279041796131e-06,
-      "loss": 1.3958,
+      "epoch": 0.86,
+      "grad_norm": 60.21806716918945,
+      "learning_rate": 1.424340040999038e-05,
+      "loss": 0.9972,
       "step": 6881
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 11.1996431350708,
-      "learning_rate": 6.2062744311917415e-06,
-      "loss": 1.6634,
+      "epoch": 0.86,
+      "grad_norm": 9.885921478271484,
+      "learning_rate": 1.424256369493369e-05,
+      "loss": 1.5922,
       "step": 6882
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 17.171239852905273,
-      "learning_rate": 6.2042698205873515e-06,
-      "loss": 1.8742,
+      "epoch": 0.86,
+      "grad_norm": 23.861791610717773,
+      "learning_rate": 1.4241726979877004e-05,
+      "loss": 1.9084,
       "step": 6883
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 83.33916473388672,
-      "learning_rate": 6.202265209982962e-06,
-      "loss": 2.0191,
+      "epoch": 0.86,
+      "grad_norm": 35.16427993774414,
+      "learning_rate": 1.4240890264820318e-05,
+      "loss": 1.6943,
       "step": 6884
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 13.033934593200684,
-      "learning_rate": 6.200260599378572e-06,
-      "loss": 0.9672,
+      "epoch": 0.86,
+      "grad_norm": 23.21710777282715,
+      "learning_rate": 1.4240053549763628e-05,
+      "loss": 3.1607,
       "step": 6885
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 25.13782501220703,
-      "learning_rate": 6.198255988774181e-06,
-      "loss": 1.5467,
+      "epoch": 0.86,
+      "grad_norm": 17.348012924194336,
+      "learning_rate": 1.4239216834706942e-05,
+      "loss": 1.9241,
       "step": 6886
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 129.21408081054688,
-      "learning_rate": 6.196251378169791e-06,
-      "loss": 1.8225,
+      "epoch": 0.86,
+      "grad_norm": 25.596208572387695,
+      "learning_rate": 1.4238380119650254e-05,
+      "loss": 1.6988,
       "step": 6887
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 15.809123992919922,
-      "learning_rate": 6.1942467675654e-06,
-      "loss": 1.1815,
+      "epoch": 0.86,
+      "grad_norm": 15.840264320373535,
+      "learning_rate": 1.4237543404593568e-05,
+      "loss": 1.1699,
       "step": 6888
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 10.88241195678711,
-      "learning_rate": 6.19224215696101e-06,
-      "loss": 0.5932,
+      "epoch": 0.86,
+      "grad_norm": 37.30739974975586,
+      "learning_rate": 1.423670668953688e-05,
+      "loss": 1.8944,
       "step": 6889
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 17.281024932861328,
-      "learning_rate": 6.190237546356621e-06,
-      "loss": 0.6785,
+      "epoch": 0.86,
+      "grad_norm": 17.949352264404297,
+      "learning_rate": 1.4235869974480192e-05,
+      "loss": 1.7828,
       "step": 6890
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 78.56420135498047,
-      "learning_rate": 6.1882329357522305e-06,
-      "loss": 2.5001,
+      "epoch": 0.86,
+      "grad_norm": 28.773000717163086,
+      "learning_rate": 1.4235033259423505e-05,
+      "loss": 2.6329,
       "step": 6891
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 7.025608539581299,
-      "learning_rate": 6.186228325147841e-06,
-      "loss": 0.9645,
+      "epoch": 0.86,
+      "grad_norm": 14.005388259887695,
+      "learning_rate": 1.4234196544366815e-05,
+      "loss": 2.0188,
       "step": 6892
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 17.67293930053711,
-      "learning_rate": 6.184223714543451e-06,
-      "loss": 1.5476,
+      "epoch": 0.87,
+      "grad_norm": 38.49642562866211,
+      "learning_rate": 1.4233359829310129e-05,
+      "loss": 1.6906,
       "step": 6893
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 54.84214782714844,
-      "learning_rate": 6.18221910393906e-06,
-      "loss": 2.5763,
+      "epoch": 0.87,
+      "grad_norm": 11.821557998657227,
+      "learning_rate": 1.4232523114253443e-05,
+      "loss": 3.0551,
       "step": 6894
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 16.025243759155273,
-      "learning_rate": 6.18021449333467e-06,
-      "loss": 1.0716,
+      "epoch": 0.87,
+      "grad_norm": 8.1752290725708,
+      "learning_rate": 1.4231686399196756e-05,
+      "loss": 1.6771,
       "step": 6895
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 19.868698120117188,
-      "learning_rate": 6.178209882730281e-06,
-      "loss": 1.6816,
+      "epoch": 0.87,
+      "grad_norm": 8.746302604675293,
+      "learning_rate": 1.4230849684140067e-05,
+      "loss": 1.7391,
       "step": 6896
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 23.872350692749023,
-      "learning_rate": 6.17620527212589e-06,
-      "loss": 1.1166,
+      "epoch": 0.87,
+      "grad_norm": 17.55376625061035,
+      "learning_rate": 1.423001296908338e-05,
+      "loss": 0.5621,
       "step": 6897
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 17.220922470092773,
-      "learning_rate": 6.1742006615215e-06,
-      "loss": 1.601,
+      "epoch": 0.87,
+      "grad_norm": 17.566560745239258,
+      "learning_rate": 1.4229176254026694e-05,
+      "loss": 2.2356,
       "step": 6898
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 37.89704895019531,
-      "learning_rate": 6.1721960509171095e-06,
-      "loss": 1.2833,
+      "epoch": 0.87,
+      "grad_norm": 7.742341995239258,
+      "learning_rate": 1.4228339538970004e-05,
+      "loss": 0.8694,
       "step": 6899
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 10.233612060546875,
-      "learning_rate": 6.17019144031272e-06,
-      "loss": 1.137,
+      "epoch": 0.87,
+      "grad_norm": 7.490888595581055,
+      "learning_rate": 1.4227502823913318e-05,
+      "loss": 0.4476,
       "step": 6900
     },
     {
-      "epoch": 2.07,
-      "grad_norm": 14.788456916809082,
-      "learning_rate": 6.168186829708329e-06,
-      "loss": 0.7968,
+      "epoch": 0.87,
+      "grad_norm": 26.35576057434082,
+      "learning_rate": 1.422666610885663e-05,
+      "loss": 2.3157,
       "step": 6901
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 47.135398864746094,
-      "learning_rate": 6.16618221910394e-06,
-      "loss": 1.8272,
+      "epoch": 0.87,
+      "grad_norm": 21.355043411254883,
+      "learning_rate": 1.4225829393799943e-05,
+      "loss": 1.2947,
       "step": 6902
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 48.95859146118164,
-      "learning_rate": 6.16417760849955e-06,
-      "loss": 1.8223,
+      "epoch": 0.87,
+      "grad_norm": 21.20218849182129,
+      "learning_rate": 1.4224992678743255e-05,
+      "loss": 0.7256,
       "step": 6903
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 9.25454044342041,
-      "learning_rate": 6.162172997895159e-06,
-      "loss": 0.9862,
+      "epoch": 0.87,
+      "grad_norm": 9.560347557067871,
+      "learning_rate": 1.4224155963686567e-05,
+      "loss": 1.951,
       "step": 6904
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 34.636192321777344,
-      "learning_rate": 6.160168387290769e-06,
-      "loss": 1.5629,
+      "epoch": 0.87,
+      "grad_norm": 13.457576751708984,
+      "learning_rate": 1.4223319248629881e-05,
+      "loss": 2.0152,
       "step": 6905
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 51.81803512573242,
-      "learning_rate": 6.158163776686379e-06,
-      "loss": 1.5237,
+      "epoch": 0.87,
+      "grad_norm": 9.960187911987305,
+      "learning_rate": 1.4222482533573191e-05,
+      "loss": 0.695,
       "step": 6906
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 39.07495880126953,
-      "learning_rate": 6.1561591660819885e-06,
-      "loss": 1.1899,
+      "epoch": 0.87,
+      "grad_norm": 14.246359825134277,
+      "learning_rate": 1.4221645818516505e-05,
+      "loss": 2.6998,
       "step": 6907
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 22.247758865356445,
-      "learning_rate": 6.154154555477599e-06,
-      "loss": 0.9568,
+      "epoch": 0.87,
+      "grad_norm": 140.53585815429688,
+      "learning_rate": 1.4220809103459819e-05,
+      "loss": 4.2545,
       "step": 6908
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 15.606700897216797,
-      "learning_rate": 6.1521499448732095e-06,
-      "loss": 1.5251,
+      "epoch": 0.87,
+      "grad_norm": 16.900049209594727,
+      "learning_rate": 1.4219972388403132e-05,
+      "loss": 0.7686,
       "step": 6909
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 17.546934127807617,
-      "learning_rate": 6.150145334268819e-06,
-      "loss": 1.1845,
+      "epoch": 0.87,
+      "grad_norm": 6.372493267059326,
+      "learning_rate": 1.4219135673346442e-05,
+      "loss": 0.5945,
       "step": 6910
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 7.191950798034668,
-      "learning_rate": 6.148140723664429e-06,
-      "loss": 0.7725,
+      "epoch": 0.87,
+      "grad_norm": 11.708452224731445,
+      "learning_rate": 1.4218298958289756e-05,
+      "loss": 1.1837,
       "step": 6911
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 10.01009464263916,
-      "learning_rate": 6.146136113060038e-06,
-      "loss": 1.0729,
+      "epoch": 0.87,
+      "grad_norm": 17.008995056152344,
+      "learning_rate": 1.4217462243233068e-05,
+      "loss": 2.3779,
       "step": 6912
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 9.743487358093262,
-      "learning_rate": 6.144131502455648e-06,
-      "loss": 1.2371,
+      "epoch": 0.87,
+      "grad_norm": 5.525969982147217,
+      "learning_rate": 1.421662552817638e-05,
+      "loss": 1.4673,
       "step": 6913
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 21.309419631958008,
-      "learning_rate": 6.142126891851259e-06,
-      "loss": 2.0328,
+      "epoch": 0.87,
+      "grad_norm": 23.5950984954834,
+      "learning_rate": 1.4215788813119694e-05,
+      "loss": 1.7786,
       "step": 6914
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 14.136860847473145,
-      "learning_rate": 6.140122281246868e-06,
-      "loss": 1.4086,
+      "epoch": 0.87,
+      "grad_norm": 71.16661834716797,
+      "learning_rate": 1.4214952098063006e-05,
+      "loss": 2.6548,
       "step": 6915
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 12.133606910705566,
-      "learning_rate": 6.138117670642478e-06,
-      "loss": 1.3393,
+      "epoch": 0.87,
+      "grad_norm": 6.5101823806762695,
+      "learning_rate": 1.421411538300632e-05,
+      "loss": 0.6794,
       "step": 6916
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 8.155012130737305,
-      "learning_rate": 6.136113060038088e-06,
-      "loss": 0.8114,
+      "epoch": 0.87,
+      "grad_norm": 22.508495330810547,
+      "learning_rate": 1.4213278667949631e-05,
+      "loss": 2.1745,
       "step": 6917
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 14.2897367477417,
-      "learning_rate": 6.134108449433698e-06,
-      "loss": 1.0815,
+      "epoch": 0.87,
+      "grad_norm": 9.747279167175293,
+      "learning_rate": 1.4212441952892943e-05,
+      "loss": 1.7265,
       "step": 6918
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 37.01668930053711,
-      "learning_rate": 6.132103838829308e-06,
-      "loss": 2.4772,
+      "epoch": 0.87,
+      "grad_norm": 8.583632469177246,
+      "learning_rate": 1.4211605237836257e-05,
+      "loss": 0.9745,
       "step": 6919
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 9.246944427490234,
-      "learning_rate": 6.130099228224919e-06,
-      "loss": 1.0533,
+      "epoch": 0.87,
+      "grad_norm": 10.896392822265625,
+      "learning_rate": 1.4210768522779567e-05,
+      "loss": 1.7416,
       "step": 6920
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 14.934267044067383,
-      "learning_rate": 6.128094617620528e-06,
-      "loss": 0.8619,
+      "epoch": 0.87,
+      "grad_norm": 11.444214820861816,
+      "learning_rate": 1.420993180772288e-05,
+      "loss": 1.5348,
       "step": 6921
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 25.984840393066406,
-      "learning_rate": 6.126090007016138e-06,
-      "loss": 1.3991,
+      "epoch": 0.87,
+      "grad_norm": 14.957980155944824,
+      "learning_rate": 1.4209095092666194e-05,
+      "loss": 2.9994,
       "step": 6922
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 23.823701858520508,
-      "learning_rate": 6.124085396411747e-06,
-      "loss": 2.2563,
+      "epoch": 0.87,
+      "grad_norm": 13.002206802368164,
+      "learning_rate": 1.4208258377609508e-05,
+      "loss": 1.1721,
       "step": 6923
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 25.616357803344727,
-      "learning_rate": 6.122080785807357e-06,
-      "loss": 1.559,
+      "epoch": 0.87,
+      "grad_norm": 29.35884666442871,
+      "learning_rate": 1.4207421662552818e-05,
+      "loss": 2.1527,
       "step": 6924
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 25.483346939086914,
-      "learning_rate": 6.120076175202967e-06,
-      "loss": 1.8079,
+      "epoch": 0.87,
+      "grad_norm": 38.53581237792969,
+      "learning_rate": 1.4206584947496132e-05,
+      "loss": 3.7084,
       "step": 6925
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 26.09974479675293,
-      "learning_rate": 6.1180715645985775e-06,
-      "loss": 1.8173,
+      "epoch": 0.87,
+      "grad_norm": 9.266407012939453,
+      "learning_rate": 1.4205748232439444e-05,
+      "loss": 1.0747,
       "step": 6926
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 33.826454162597656,
-      "learning_rate": 6.116066953994188e-06,
-      "loss": 1.2576,
+      "epoch": 0.87,
+      "grad_norm": 12.204514503479004,
+      "learning_rate": 1.4204911517382756e-05,
+      "loss": 1.8189,
       "step": 6927
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 22.255189895629883,
-      "learning_rate": 6.114062343389797e-06,
-      "loss": 1.5849,
+      "epoch": 0.87,
+      "grad_norm": 21.560924530029297,
+      "learning_rate": 1.420407480232607e-05,
+      "loss": 1.6065,
       "step": 6928
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 23.203243255615234,
-      "learning_rate": 6.112057732785407e-06,
-      "loss": 2.0366,
+      "epoch": 0.87,
+      "grad_norm": 10.391530990600586,
+      "learning_rate": 1.4203238087269381e-05,
+      "loss": 0.4697,
       "step": 6929
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 48.67564010620117,
-      "learning_rate": 6.110053122181017e-06,
-      "loss": 1.4752,
+      "epoch": 0.87,
+      "grad_norm": 8.712636947631836,
+      "learning_rate": 1.4202401372212695e-05,
+      "loss": 0.5286,
       "step": 6930
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 13.649922370910645,
-      "learning_rate": 6.108048511576626e-06,
-      "loss": 1.0061,
+      "epoch": 0.87,
+      "grad_norm": 5.031757354736328,
+      "learning_rate": 1.4201564657156005e-05,
+      "loss": 0.0881,
       "step": 6931
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 15.346248626708984,
-      "learning_rate": 6.106043900972236e-06,
-      "loss": 0.975,
+      "epoch": 0.87,
+      "grad_norm": 12.193156242370605,
+      "learning_rate": 1.4200727942099319e-05,
+      "loss": 1.2649,
       "step": 6932
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 43.33583450317383,
-      "learning_rate": 6.104039290367847e-06,
-      "loss": 1.8835,
+      "epoch": 0.87,
+      "grad_norm": 16.03141975402832,
+      "learning_rate": 1.4199891227042633e-05,
+      "loss": 1.6772,
       "step": 6933
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 11.801863670349121,
-      "learning_rate": 6.1020346797634565e-06,
-      "loss": 0.9421,
+      "epoch": 0.87,
+      "grad_norm": 56.04813003540039,
+      "learning_rate": 1.4199054511985943e-05,
+      "loss": 3.0297,
       "step": 6934
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 13.316075325012207,
-      "learning_rate": 6.100030069159067e-06,
-      "loss": 0.9621,
+      "epoch": 0.87,
+      "grad_norm": 20.556060791015625,
+      "learning_rate": 1.4198217796929257e-05,
+      "loss": 1.6633,
       "step": 6935
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 7.643853187561035,
-      "learning_rate": 6.098025458554676e-06,
-      "loss": 0.7109,
+      "epoch": 0.87,
+      "grad_norm": 55.4222412109375,
+      "learning_rate": 1.419738108187257e-05,
+      "loss": 2.2368,
       "step": 6936
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 6.656645774841309,
-      "learning_rate": 6.096020847950286e-06,
-      "loss": 0.9677,
+      "epoch": 0.87,
+      "grad_norm": 25.066200256347656,
+      "learning_rate": 1.4196544366815884e-05,
+      "loss": 0.8717,
       "step": 6937
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 8.41319751739502,
-      "learning_rate": 6.094016237345895e-06,
-      "loss": 1.4541,
+      "epoch": 0.87,
+      "grad_norm": 11.717459678649902,
+      "learning_rate": 1.4195707651759194e-05,
+      "loss": 3.939,
       "step": 6938
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 17.172225952148438,
-      "learning_rate": 6.092011626741506e-06,
-      "loss": 1.2853,
+      "epoch": 0.87,
+      "grad_norm": 18.704364776611328,
+      "learning_rate": 1.4194870936702508e-05,
+      "loss": 2.1806,
       "step": 6939
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 13.03164005279541,
-      "learning_rate": 6.090007016137116e-06,
-      "loss": 1.2991,
+      "epoch": 0.87,
+      "grad_norm": 18.1362361907959,
+      "learning_rate": 1.419403422164582e-05,
+      "loss": 1.377,
       "step": 6940
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 26.83146095275879,
-      "learning_rate": 6.088002405532725e-06,
-      "loss": 1.2144,
+      "epoch": 0.87,
+      "grad_norm": 18.96443748474121,
+      "learning_rate": 1.4193197506589132e-05,
+      "loss": 1.608,
       "step": 6941
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 15.566588401794434,
-      "learning_rate": 6.0859977949283355e-06,
-      "loss": 1.1742,
+      "epoch": 0.87,
+      "grad_norm": 10.305618286132812,
+      "learning_rate": 1.4192360791532445e-05,
+      "loss": 0.4654,
       "step": 6942
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 11.731130599975586,
-      "learning_rate": 6.083993184323946e-06,
-      "loss": 1.5685,
+      "epoch": 0.87,
+      "grad_norm": 12.450068473815918,
+      "learning_rate": 1.4191524076475757e-05,
+      "loss": 1.6787,
       "step": 6943
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 23.45868492126465,
-      "learning_rate": 6.081988573719555e-06,
-      "loss": 1.1652,
+      "epoch": 0.87,
+      "grad_norm": 6.444420337677002,
+      "learning_rate": 1.4190687361419071e-05,
+      "loss": 0.5719,
       "step": 6944
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 30.693378448486328,
-      "learning_rate": 6.079983963115166e-06,
-      "loss": 1.1192,
+      "epoch": 0.87,
+      "grad_norm": 10.12495231628418,
+      "learning_rate": 1.4189850646362381e-05,
+      "loss": 0.7943,
       "step": 6945
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 76.05176544189453,
-      "learning_rate": 6.077979352510776e-06,
-      "loss": 1.03,
+      "epoch": 0.87,
+      "grad_norm": 12.811422348022461,
+      "learning_rate": 1.4189013931305695e-05,
+      "loss": 1.4826,
       "step": 6946
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 42.937931060791016,
-      "learning_rate": 6.075974741906385e-06,
-      "loss": 2.9801,
+      "epoch": 0.87,
+      "grad_norm": 27.980676651000977,
+      "learning_rate": 1.4188177216249009e-05,
+      "loss": 1.9857,
       "step": 6947
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 21.51824188232422,
-      "learning_rate": 6.073970131301995e-06,
-      "loss": 1.3744,
+      "epoch": 0.87,
+      "grad_norm": 23.33842658996582,
+      "learning_rate": 1.4187340501192319e-05,
+      "loss": 1.7549,
       "step": 6948
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 13.035652160644531,
-      "learning_rate": 6.071965520697604e-06,
-      "loss": 1.2832,
+      "epoch": 0.87,
+      "grad_norm": 8.448101043701172,
+      "learning_rate": 1.4186503786135632e-05,
+      "loss": 1.0385,
       "step": 6949
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 7.819190502166748,
-      "learning_rate": 6.0699609100932145e-06,
-      "loss": 1.3225,
+      "epoch": 0.87,
+      "grad_norm": 18.32587242126465,
+      "learning_rate": 1.4185667071078946e-05,
+      "loss": 2.4694,
       "step": 6950
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 10.307256698608398,
-      "learning_rate": 6.067956299488825e-06,
-      "loss": 1.1998,
+      "epoch": 0.87,
+      "grad_norm": 14.72725772857666,
+      "learning_rate": 1.418483035602226e-05,
+      "loss": 1.1245,
       "step": 6951
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 43.879268646240234,
-      "learning_rate": 6.065951688884435e-06,
-      "loss": 1.2899,
+      "epoch": 0.87,
+      "grad_norm": 10.395641326904297,
+      "learning_rate": 1.418399364096557e-05,
+      "loss": 2.1181,
       "step": 6952
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 14.263348579406738,
-      "learning_rate": 6.063947078280045e-06,
-      "loss": 0.8165,
+      "epoch": 0.87,
+      "grad_norm": 17.333349227905273,
+      "learning_rate": 1.4183156925908884e-05,
+      "loss": 2.2361,
       "step": 6953
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 19.4408016204834,
-      "learning_rate": 6.061942467675655e-06,
-      "loss": 1.2222,
+      "epoch": 0.87,
+      "grad_norm": 18.26038932800293,
+      "learning_rate": 1.4182320210852196e-05,
+      "loss": 1.1122,
       "step": 6954
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 21.61421012878418,
-      "learning_rate": 6.059937857071264e-06,
-      "loss": 1.6328,
+      "epoch": 0.87,
+      "grad_norm": 7.8660478591918945,
+      "learning_rate": 1.4181483495795508e-05,
+      "loss": 0.234,
       "step": 6955
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 59.171634674072266,
-      "learning_rate": 6.057933246466874e-06,
-      "loss": 2.2851,
+      "epoch": 0.87,
+      "grad_norm": 12.084300994873047,
+      "learning_rate": 1.4180646780738821e-05,
+      "loss": 0.5299,
       "step": 6956
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 17.417770385742188,
-      "learning_rate": 6.055928635862485e-06,
-      "loss": 1.5558,
+      "epoch": 0.87,
+      "grad_norm": 16.44608497619629,
+      "learning_rate": 1.4179810065682133e-05,
+      "loss": 1.3773,
       "step": 6957
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 21.351945877075195,
-      "learning_rate": 6.053924025258094e-06,
-      "loss": 1.9198,
+      "epoch": 0.87,
+      "grad_norm": 10.259683609008789,
+      "learning_rate": 1.4178973350625445e-05,
+      "loss": 0.7847,
       "step": 6958
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 21.162813186645508,
-      "learning_rate": 6.051919414653704e-06,
-      "loss": 0.8733,
+      "epoch": 0.87,
+      "grad_norm": 6.972939968109131,
+      "learning_rate": 1.4178136635568757e-05,
+      "loss": 0.7622,
       "step": 6959
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 28.082653045654297,
-      "learning_rate": 6.049914804049314e-06,
-      "loss": 1.3162,
-      "step": 6960
-    },
-    {
-      "epoch": 2.09,
-      "eval_loss": 0.1762227714061737,
-      "eval_runtime": 43.6018,
-      "eval_samples_per_second": 33.921,
-      "eval_steps_per_second": 33.921,
+      "epoch": 0.87,
+      "grad_norm": 10.767699241638184,
+      "learning_rate": 1.417729992051207e-05,
+      "loss": 1.5511,
       "step": 6960
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 17.02800941467285,
-      "learning_rate": 6.047910193444924e-06,
-      "loss": 1.4449,
+      "epoch": 0.87,
+      "grad_norm": 15.990443229675293,
+      "learning_rate": 1.4176463205455384e-05,
+      "loss": 2.1373,
       "step": 6961
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 48.51543045043945,
-      "learning_rate": 6.045905582840533e-06,
-      "loss": 1.0376,
+      "epoch": 0.87,
+      "grad_norm": 9.765302658081055,
+      "learning_rate": 1.4175626490398695e-05,
+      "loss": 1.1673,
       "step": 6962
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 30.203258514404297,
-      "learning_rate": 6.043900972236144e-06,
-      "loss": 1.8892,
+      "epoch": 0.87,
+      "grad_norm": 16.047128677368164,
+      "learning_rate": 1.4174789775342008e-05,
+      "loss": 3.8891,
       "step": 6963
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 22.062679290771484,
-      "learning_rate": 6.041896361631754e-06,
-      "loss": 2.034,
+      "epoch": 0.87,
+      "grad_norm": 75.53546142578125,
+      "learning_rate": 1.4173953060285322e-05,
+      "loss": 2.9341,
       "step": 6964
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 29.935195922851562,
-      "learning_rate": 6.039891751027363e-06,
-      "loss": 2.1552,
+      "epoch": 0.87,
+      "grad_norm": 39.25362014770508,
+      "learning_rate": 1.4173116345228632e-05,
+      "loss": 0.6573,
       "step": 6965
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 21.121536254882812,
-      "learning_rate": 6.037887140422973e-06,
-      "loss": 1.848,
+      "epoch": 0.87,
+      "grad_norm": 28.128896713256836,
+      "learning_rate": 1.4172279630171946e-05,
+      "loss": 1.7481,
       "step": 6966
     },
     {
-      "epoch": 2.09,
-      "grad_norm": 13.469535827636719,
-      "learning_rate": 6.035882529818583e-06,
-      "loss": 0.8718,
+      "epoch": 0.87,
+      "grad_norm": 33.34860610961914,
+      "learning_rate": 1.417144291511526e-05,
+      "loss": 2.1332,
       "step": 6967
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 9.78769302368164,
-      "learning_rate": 6.033877919214193e-06,
-      "loss": 0.5627,
+      "epoch": 0.87,
+      "grad_norm": 14.426985740661621,
+      "learning_rate": 1.4170606200058571e-05,
+      "loss": 1.993,
       "step": 6968
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 11.262333869934082,
-      "learning_rate": 6.0318733086098035e-06,
-      "loss": 1.4755,
+      "epoch": 0.87,
+      "grad_norm": 12.120182037353516,
+      "learning_rate": 1.4169769485001883e-05,
+      "loss": 2.3124,
       "step": 6969
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 21.476552963256836,
-      "learning_rate": 6.029868698005414e-06,
-      "loss": 0.9482,
+      "epoch": 0.87,
+      "grad_norm": 9.713573455810547,
+      "learning_rate": 1.4168932769945195e-05,
+      "loss": 0.6764,
       "step": 6970
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 8.83089542388916,
-      "learning_rate": 6.027864087401023e-06,
-      "loss": 0.9499,
+      "epoch": 0.87,
+      "grad_norm": 10.640819549560547,
+      "learning_rate": 1.4168096054888509e-05,
+      "loss": 1.2914,
       "step": 6971
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 16.572052001953125,
-      "learning_rate": 6.025859476796633e-06,
-      "loss": 1.4945,
+      "epoch": 0.87,
+      "grad_norm": 9.241393089294434,
+      "learning_rate": 1.4167259339831821e-05,
+      "loss": 1.2839,
       "step": 6972
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 94.96258544921875,
-      "learning_rate": 6.023854866192242e-06,
-      "loss": 1.655,
+      "epoch": 0.88,
+      "grad_norm": 15.63196086883545,
+      "learning_rate": 1.4166422624775133e-05,
+      "loss": 0.7556,
       "step": 6973
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 11.988658905029297,
-      "learning_rate": 6.021850255587852e-06,
-      "loss": 1.4834,
+      "epoch": 0.88,
+      "grad_norm": 15.385232925415039,
+      "learning_rate": 1.4165585909718447e-05,
+      "loss": 1.7352,
       "step": 6974
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 15.83250904083252,
-      "learning_rate": 6.0198456449834615e-06,
-      "loss": 1.7121,
+      "epoch": 0.88,
+      "grad_norm": 54.85871505737305,
+      "learning_rate": 1.416474919466176e-05,
+      "loss": 2.9882,
       "step": 6975
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 14.637907981872559,
-      "learning_rate": 6.0178410343790724e-06,
-      "loss": 1.849,
+      "epoch": 0.88,
+      "grad_norm": 18.45826530456543,
+      "learning_rate": 1.416391247960507e-05,
+      "loss": 1.0526,
       "step": 6976
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 42.469970703125,
-      "learning_rate": 6.0158364237746825e-06,
-      "loss": 1.3548,
+      "epoch": 0.88,
+      "grad_norm": 33.669708251953125,
+      "learning_rate": 1.4163075764548384e-05,
+      "loss": 1.7136,
       "step": 6977
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 14.263885498046875,
-      "learning_rate": 6.013831813170293e-06,
-      "loss": 1.5441,
+      "epoch": 0.88,
+      "grad_norm": 16.6716251373291,
+      "learning_rate": 1.4162239049491698e-05,
+      "loss": 1.7772,
       "step": 6978
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 91.35665130615234,
-      "learning_rate": 6.011827202565902e-06,
-      "loss": 2.4573,
+      "epoch": 0.88,
+      "grad_norm": 10.420305252075195,
+      "learning_rate": 1.4161402334435008e-05,
+      "loss": 3.1408,
       "step": 6979
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 26.59601402282715,
-      "learning_rate": 6.009822591961512e-06,
-      "loss": 1.0937,
+      "epoch": 0.88,
+      "grad_norm": 30.392881393432617,
+      "learning_rate": 1.4160565619378322e-05,
+      "loss": 1.545,
       "step": 6980
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 21.7692928314209,
-      "learning_rate": 6.007817981357121e-06,
-      "loss": 1.4581,
+      "epoch": 0.88,
+      "grad_norm": 7.688721656799316,
+      "learning_rate": 1.4159728904321635e-05,
+      "loss": 1.6216,
       "step": 6981
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 55.85102081298828,
-      "learning_rate": 6.005813370752732e-06,
-      "loss": 1.5007,
+      "epoch": 0.88,
+      "grad_norm": 8.282111167907715,
+      "learning_rate": 1.4158892189264947e-05,
+      "loss": 1.3615,
       "step": 6982
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 17.059696197509766,
-      "learning_rate": 6.003808760148342e-06,
-      "loss": 1.2327,
+      "epoch": 0.88,
+      "grad_norm": 20.921794891357422,
+      "learning_rate": 1.415805547420826e-05,
+      "loss": 2.6633,
       "step": 6983
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 20.334264755249023,
-      "learning_rate": 6.001804149543951e-06,
-      "loss": 1.1726,
+      "epoch": 0.88,
+      "grad_norm": 16.57208251953125,
+      "learning_rate": 1.4157218759151571e-05,
+      "loss": 2.6769,
       "step": 6984
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 14.4188871383667,
-      "learning_rate": 5.9997995389395615e-06,
-      "loss": 1.4264,
+      "epoch": 0.88,
+      "grad_norm": 13.83458137512207,
+      "learning_rate": 1.4156382044094885e-05,
+      "loss": 2.5668,
       "step": 6985
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 10.726988792419434,
-      "learning_rate": 5.997794928335171e-06,
-      "loss": 0.912,
+      "epoch": 0.88,
+      "grad_norm": 9.996246337890625,
+      "learning_rate": 1.4155545329038197e-05,
+      "loss": 1.924,
       "step": 6986
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 22.09978675842285,
-      "learning_rate": 5.995790317730781e-06,
-      "loss": 1.9006,
+      "epoch": 0.88,
+      "grad_norm": 14.711870193481445,
+      "learning_rate": 1.4154708613981509e-05,
+      "loss": 1.1225,
       "step": 6987
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 25.827951431274414,
-      "learning_rate": 5.993785707126392e-06,
-      "loss": 2.7776,
+      "epoch": 0.88,
+      "grad_norm": 13.016510009765625,
+      "learning_rate": 1.4153871898924822e-05,
+      "loss": 0.9249,
       "step": 6988
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 43.5711784362793,
-      "learning_rate": 5.991781096522001e-06,
-      "loss": 1.5516,
+      "epoch": 0.88,
+      "grad_norm": 11.098211288452148,
+      "learning_rate": 1.4153035183868136e-05,
+      "loss": 1.3117,
       "step": 6989
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 20.165645599365234,
-      "learning_rate": 5.989776485917611e-06,
-      "loss": 1.1189,
+      "epoch": 0.88,
+      "grad_norm": 22.908828735351562,
+      "learning_rate": 1.4152198468811446e-05,
+      "loss": 1.8384,
       "step": 6990
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 23.63228988647461,
-      "learning_rate": 5.987771875313221e-06,
-      "loss": 1.4044,
+      "epoch": 0.88,
+      "grad_norm": 18.00226593017578,
+      "learning_rate": 1.415136175375476e-05,
+      "loss": 2.6107,
       "step": 6991
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 18.94355583190918,
-      "learning_rate": 5.98576726470883e-06,
-      "loss": 1.3975,
+      "epoch": 0.88,
+      "grad_norm": 19.675580978393555,
+      "learning_rate": 1.4150525038698074e-05,
+      "loss": 0.7767,
       "step": 6992
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 43.25507736206055,
-      "learning_rate": 5.9837626541044405e-06,
-      "loss": 2.5693,
+      "epoch": 0.88,
+      "grad_norm": 5.079214096069336,
+      "learning_rate": 1.4149688323641384e-05,
+      "loss": 0.1675,
       "step": 6993
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 24.190704345703125,
-      "learning_rate": 5.981758043500051e-06,
-      "loss": 1.0735,
+      "epoch": 0.88,
+      "grad_norm": 18.402210235595703,
+      "learning_rate": 1.4148851608584697e-05,
+      "loss": 1.213,
       "step": 6994
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 16.01865577697754,
-      "learning_rate": 5.979753432895661e-06,
-      "loss": 1.2435,
+      "epoch": 0.88,
+      "grad_norm": 24.991798400878906,
+      "learning_rate": 1.4148014893528011e-05,
+      "loss": 1.6542,
       "step": 6995
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 16.649044036865234,
-      "learning_rate": 5.977748822291271e-06,
-      "loss": 1.3125,
+      "epoch": 0.88,
+      "grad_norm": 9.81252384185791,
+      "learning_rate": 1.4147178178471323e-05,
+      "loss": 1.3761,
       "step": 6996
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 17.755905151367188,
-      "learning_rate": 5.97574421168688e-06,
-      "loss": 1.6912,
+      "epoch": 0.88,
+      "grad_norm": 4.908636569976807,
+      "learning_rate": 1.4146341463414635e-05,
+      "loss": 0.2464,
       "step": 6997
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 18.950183868408203,
-      "learning_rate": 5.97373960108249e-06,
-      "loss": 1.9294,
+      "epoch": 0.88,
+      "grad_norm": 7.581484317779541,
+      "learning_rate": 1.4145504748357947e-05,
+      "loss": 1.0555,
       "step": 6998
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 9.899605751037598,
-      "learning_rate": 5.971734990478099e-06,
-      "loss": 1.6834,
+      "epoch": 0.88,
+      "grad_norm": 11.803753852844238,
+      "learning_rate": 1.414466803330126e-05,
+      "loss": 1.0861,
       "step": 6999
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 16.555978775024414,
-      "learning_rate": 5.96973037987371e-06,
-      "loss": 1.773,
+      "epoch": 0.88,
+      "grad_norm": 15.46438980102539,
+      "learning_rate": 1.4143831318244573e-05,
+      "loss": 1.4876,
       "step": 7000
     },
     {
-      "epoch": 2.1,
-      "grad_norm": 9.918785095214844,
-      "learning_rate": 5.96772576926932e-06,
-      "loss": 0.8201,
+      "epoch": 0.88,
+      "grad_norm": 18.067228317260742,
+      "learning_rate": 1.4142994603187885e-05,
+      "loss": 2.3218,
       "step": 7001
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 7.417164325714111,
-      "learning_rate": 5.9657211586649295e-06,
-      "loss": 0.8608,
+      "epoch": 0.88,
+      "grad_norm": 34.01871109008789,
+      "learning_rate": 1.4142157888131198e-05,
+      "loss": 1.8282,
       "step": 7002
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 31.227968215942383,
-      "learning_rate": 5.96371654806054e-06,
-      "loss": 1.3737,
+      "epoch": 0.88,
+      "grad_norm": 12.05657958984375,
+      "learning_rate": 1.4141321173074512e-05,
+      "loss": 1.8803,
       "step": 7003
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 16.146133422851562,
-      "learning_rate": 5.96171193745615e-06,
-      "loss": 1.6191,
+      "epoch": 0.88,
+      "grad_norm": 12.193592071533203,
+      "learning_rate": 1.4140484458017822e-05,
+      "loss": 2.3793,
       "step": 7004
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 11.509698867797852,
-      "learning_rate": 5.959707326851759e-06,
-      "loss": 1.118,
+      "epoch": 0.88,
+      "grad_norm": 29.061677932739258,
+      "learning_rate": 1.4139647742961136e-05,
+      "loss": 1.6299,
       "step": 7005
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 11.928789138793945,
-      "learning_rate": 5.95770271624737e-06,
-      "loss": 1.3745,
+      "epoch": 0.88,
+      "grad_norm": 7.4007039070129395,
+      "learning_rate": 1.413881102790445e-05,
+      "loss": 0.6984,
       "step": 7006
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 12.056342124938965,
-      "learning_rate": 5.95569810564298e-06,
-      "loss": 2.4071,
+      "epoch": 0.88,
+      "grad_norm": 9.471341133117676,
+      "learning_rate": 1.413797431284776e-05,
+      "loss": 0.9251,
       "step": 7007
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 15.146832466125488,
-      "learning_rate": 5.953693495038589e-06,
-      "loss": 0.6288,
+      "epoch": 0.88,
+      "grad_norm": 7.935895919799805,
+      "learning_rate": 1.4137137597791073e-05,
+      "loss": 0.7665,
       "step": 7008
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 15.896178245544434,
-      "learning_rate": 5.951688884434199e-06,
-      "loss": 1.3848,
+      "epoch": 0.88,
+      "grad_norm": 145.23446655273438,
+      "learning_rate": 1.4136300882734387e-05,
+      "loss": 2.0616,
       "step": 7009
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 20.91379737854004,
-      "learning_rate": 5.9496842738298085e-06,
-      "loss": 1.9552,
+      "epoch": 0.88,
+      "grad_norm": 11.415495872497559,
+      "learning_rate": 1.4135464167677699e-05,
+      "loss": 1.2726,
       "step": 7010
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 15.576189041137695,
-      "learning_rate": 5.947679663225419e-06,
-      "loss": 1.7155,
+      "epoch": 0.88,
+      "grad_norm": 31.87625503540039,
+      "learning_rate": 1.4134627452621011e-05,
+      "loss": 1.1281,
       "step": 7011
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 12.269247055053711,
-      "learning_rate": 5.9456750526210295e-06,
-      "loss": 1.1991,
+      "epoch": 0.88,
+      "grad_norm": 10.819061279296875,
+      "learning_rate": 1.4133790737564323e-05,
+      "loss": 1.6783,
       "step": 7012
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 11.45450210571289,
-      "learning_rate": 5.943670442016639e-06,
-      "loss": 0.6404,
+      "epoch": 0.88,
+      "grad_norm": 10.764213562011719,
+      "learning_rate": 1.4132954022507636e-05,
+      "loss": 1.2949,
       "step": 7013
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 14.839483261108398,
-      "learning_rate": 5.941665831412249e-06,
-      "loss": 1.1984,
+      "epoch": 0.88,
+      "grad_norm": 16.699296951293945,
+      "learning_rate": 1.4132117307450948e-05,
+      "loss": 0.6814,
       "step": 7014
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 11.097734451293945,
-      "learning_rate": 5.939661220807859e-06,
-      "loss": 1.0033,
+      "epoch": 0.88,
+      "grad_norm": 10.548154830932617,
+      "learning_rate": 1.413128059239426e-05,
+      "loss": 1.881,
       "step": 7015
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 23.61093521118164,
-      "learning_rate": 5.937656610203468e-06,
-      "loss": 2.0833,
+      "epoch": 0.88,
+      "grad_norm": 22.957895278930664,
+      "learning_rate": 1.4130443877337574e-05,
+      "loss": 1.9368,
       "step": 7016
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 16.99240493774414,
-      "learning_rate": 5.935651999599078e-06,
-      "loss": 1.4891,
+      "epoch": 0.88,
+      "grad_norm": 10.620356559753418,
+      "learning_rate": 1.4129607162280888e-05,
+      "loss": 0.5651,
       "step": 7017
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 11.620327949523926,
-      "learning_rate": 5.9336473889946875e-06,
-      "loss": 0.9176,
+      "epoch": 0.88,
+      "grad_norm": 12.291309356689453,
+      "learning_rate": 1.4128770447224198e-05,
+      "loss": 1.1731,
       "step": 7018
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 17.769229888916016,
-      "learning_rate": 5.9316427783902984e-06,
-      "loss": 2.0986,
+      "epoch": 0.88,
+      "grad_norm": 30.018396377563477,
+      "learning_rate": 1.4127933732167512e-05,
+      "loss": 1.7424,
       "step": 7019
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 9.187108039855957,
-      "learning_rate": 5.9296381677859085e-06,
-      "loss": 1.0068,
+      "epoch": 0.88,
+      "grad_norm": 12.34261703491211,
+      "learning_rate": 1.4127097017110825e-05,
+      "loss": 1.4332,
       "step": 7020
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 47.74100112915039,
-      "learning_rate": 5.927633557181518e-06,
-      "loss": 1.3823,
+      "epoch": 0.88,
+      "grad_norm": 22.039567947387695,
+      "learning_rate": 1.4126260302054136e-05,
+      "loss": 1.5204,
       "step": 7021
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 20.690021514892578,
-      "learning_rate": 5.925628946577128e-06,
-      "loss": 1.5737,
+      "epoch": 0.88,
+      "grad_norm": 9.983137130737305,
+      "learning_rate": 1.412542358699745e-05,
+      "loss": 1.5573,
       "step": 7022
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 19.90296745300293,
-      "learning_rate": 5.923624335972737e-06,
-      "loss": 1.6142,
+      "epoch": 0.88,
+      "grad_norm": 7.264036178588867,
+      "learning_rate": 1.4124586871940761e-05,
+      "loss": 0.2537,
       "step": 7023
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 19.526081085205078,
-      "learning_rate": 5.921619725368347e-06,
-      "loss": 0.9043,
+      "epoch": 0.88,
+      "grad_norm": 10.880606651306152,
+      "learning_rate": 1.4123750156884075e-05,
+      "loss": 2.5847,
       "step": 7024
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 47.775001525878906,
-      "learning_rate": 5.919615114763958e-06,
-      "loss": 3.4549,
+      "epoch": 0.88,
+      "grad_norm": 23.634489059448242,
+      "learning_rate": 1.4122913441827387e-05,
+      "loss": 2.3404,
       "step": 7025
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 12.722251892089844,
-      "learning_rate": 5.917610504159567e-06,
-      "loss": 1.7378,
+      "epoch": 0.88,
+      "grad_norm": 13.010921478271484,
+      "learning_rate": 1.4122076726770699e-05,
+      "loss": 1.2087,
       "step": 7026
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 10.309690475463867,
-      "learning_rate": 5.915605893555177e-06,
-      "loss": 0.9647,
+      "epoch": 0.88,
+      "grad_norm": 11.797823905944824,
+      "learning_rate": 1.4121240011714012e-05,
+      "loss": 2.3684,
       "step": 7027
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 10.273581504821777,
-      "learning_rate": 5.9136012829507875e-06,
-      "loss": 1.6892,
+      "epoch": 0.88,
+      "grad_norm": 6.033276557922363,
+      "learning_rate": 1.4120403296657324e-05,
+      "loss": 0.6981,
       "step": 7028
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 16.655803680419922,
-      "learning_rate": 5.911596672346397e-06,
-      "loss": 1.6853,
+      "epoch": 0.88,
+      "grad_norm": 12.963159561157227,
+      "learning_rate": 1.4119566581600636e-05,
+      "loss": 2.9143,
       "step": 7029
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 23.953218460083008,
-      "learning_rate": 5.909592061742007e-06,
-      "loss": 0.5064,
+      "epoch": 0.88,
+      "grad_norm": 25.27248764038086,
+      "learning_rate": 1.411872986654395e-05,
+      "loss": 4.1087,
       "step": 7030
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 21.488235473632812,
-      "learning_rate": 5.907587451137618e-06,
-      "loss": 1.7174,
+      "epoch": 0.88,
+      "grad_norm": 11.717084884643555,
+      "learning_rate": 1.4117893151487264e-05,
+      "loss": 1.4052,
       "step": 7031
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 54.67894744873047,
-      "learning_rate": 5.905582840533227e-06,
-      "loss": 1.7804,
+      "epoch": 0.88,
+      "grad_norm": 9.638819694519043,
+      "learning_rate": 1.4117056436430574e-05,
+      "loss": 2.5262,
       "step": 7032
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 10.905550956726074,
-      "learning_rate": 5.903578229928837e-06,
-      "loss": 0.7826,
+      "epoch": 0.88,
+      "grad_norm": 4.7968220710754395,
+      "learning_rate": 1.4116219721373887e-05,
+      "loss": 0.3063,
       "step": 7033
     },
     {
-      "epoch": 2.11,
-      "grad_norm": 6.627830982208252,
-      "learning_rate": 5.901573619324446e-06,
-      "loss": 0.9134,
+      "epoch": 0.88,
+      "grad_norm": 15.35245132446289,
+      "learning_rate": 1.4115383006317201e-05,
+      "loss": 1.6842,
       "step": 7034
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 12.987330436706543,
-      "learning_rate": 5.899569008720056e-06,
-      "loss": 1.4485,
+      "epoch": 0.88,
+      "grad_norm": 3.629316806793213,
+      "learning_rate": 1.4114546291260511e-05,
+      "loss": 0.0851,
       "step": 7035
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 14.574166297912598,
-      "learning_rate": 5.897564398115666e-06,
-      "loss": 1.1571,
+      "epoch": 0.88,
+      "grad_norm": 15.7343168258667,
+      "learning_rate": 1.4113709576203825e-05,
+      "loss": 2.3484,
       "step": 7036
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 12.813142776489258,
-      "learning_rate": 5.8955597875112766e-06,
-      "loss": 2.2213,
+      "epoch": 0.88,
+      "grad_norm": 9.407702445983887,
+      "learning_rate": 1.4112872861147137e-05,
+      "loss": 2.2536,
       "step": 7037
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 96.96940612792969,
-      "learning_rate": 5.893555176906887e-06,
-      "loss": 2.6402,
+      "epoch": 0.88,
+      "grad_norm": 15.683849334716797,
+      "learning_rate": 1.411203614609045e-05,
+      "loss": 1.682,
       "step": 7038
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 22.448095321655273,
-      "learning_rate": 5.891550566302497e-06,
-      "loss": 1.2935,
+      "epoch": 0.88,
+      "grad_norm": 29.45321273803711,
+      "learning_rate": 1.4111199431033763e-05,
+      "loss": 1.3847,
       "step": 7039
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 8.147926330566406,
-      "learning_rate": 5.889545955698106e-06,
-      "loss": 0.8031,
+      "epoch": 0.88,
+      "grad_norm": 12.613272666931152,
+      "learning_rate": 1.4110362715977075e-05,
+      "loss": 1.7653,
       "step": 7040
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 16.05350685119629,
-      "learning_rate": 5.887541345093716e-06,
-      "loss": 1.6077,
+      "epoch": 0.88,
+      "grad_norm": 13.921170234680176,
+      "learning_rate": 1.4109526000920388e-05,
+      "loss": 2.4739,
       "step": 7041
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 6.752038955688477,
-      "learning_rate": 5.885536734489325e-06,
-      "loss": 0.6485,
+      "epoch": 0.88,
+      "grad_norm": 9.824431419372559,
+      "learning_rate": 1.4108689285863698e-05,
+      "loss": 1.2665,
       "step": 7042
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 13.922531127929688,
-      "learning_rate": 5.883532123884936e-06,
-      "loss": 1.3103,
+      "epoch": 0.88,
+      "grad_norm": 31.29667854309082,
+      "learning_rate": 1.4107852570807012e-05,
+      "loss": 1.5647,
       "step": 7043
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 26.871532440185547,
-      "learning_rate": 5.881527513280546e-06,
-      "loss": 1.4686,
+      "epoch": 0.88,
+      "grad_norm": 57.366641998291016,
+      "learning_rate": 1.4107015855750326e-05,
+      "loss": 0.898,
       "step": 7044
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 10.8401460647583,
-      "learning_rate": 5.8795229026761555e-06,
-      "loss": 1.5021,
+      "epoch": 0.88,
+      "grad_norm": 8.376490592956543,
+      "learning_rate": 1.410617914069364e-05,
+      "loss": 0.7108,
       "step": 7045
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 22.18703842163086,
-      "learning_rate": 5.877518292071766e-06,
-      "loss": 1.2032,
+      "epoch": 0.88,
+      "grad_norm": 52.11063766479492,
+      "learning_rate": 1.410534242563695e-05,
+      "loss": 2.0481,
       "step": 7046
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 14.264338493347168,
-      "learning_rate": 5.875513681467375e-06,
-      "loss": 1.3754,
+      "epoch": 0.88,
+      "grad_norm": 55.179683685302734,
+      "learning_rate": 1.4104505710580263e-05,
+      "loss": 1.5647,
       "step": 7047
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 14.007697105407715,
-      "learning_rate": 5.873509070862985e-06,
-      "loss": 1.549,
+      "epoch": 0.88,
+      "grad_norm": 9.608619689941406,
+      "learning_rate": 1.4103668995523577e-05,
+      "loss": 0.7384,
       "step": 7048
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 45.7745361328125,
-      "learning_rate": 5.871504460258596e-06,
-      "loss": 1.5497,
+      "epoch": 0.88,
+      "grad_norm": 9.80874252319336,
+      "learning_rate": 1.4102832280466887e-05,
+      "loss": 1.4948,
       "step": 7049
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 13.82027530670166,
-      "learning_rate": 5.869499849654205e-06,
-      "loss": 0.9325,
+      "epoch": 0.88,
+      "grad_norm": 7.465558052062988,
+      "learning_rate": 1.41019955654102e-05,
+      "loss": 0.4291,
       "step": 7050
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 28.13387107849121,
-      "learning_rate": 5.867495239049815e-06,
-      "loss": 1.5999,
+      "epoch": 0.88,
+      "grad_norm": 5.3900628089904785,
+      "learning_rate": 1.4101158850353513e-05,
+      "loss": 0.4701,
       "step": 7051
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 6.187619209289551,
-      "learning_rate": 5.865490628445425e-06,
-      "loss": 0.843,
+      "epoch": 0.89,
+      "grad_norm": 14.090367317199707,
+      "learning_rate": 1.4100322135296826e-05,
+      "loss": 0.8991,
       "step": 7052
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 16.584333419799805,
-      "learning_rate": 5.8634860178410345e-06,
-      "loss": 1.3149,
+      "epoch": 0.89,
+      "grad_norm": 32.191429138183594,
+      "learning_rate": 1.4099485420240138e-05,
+      "loss": 1.2037,
       "step": 7053
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 36.140869140625,
-      "learning_rate": 5.861481407236645e-06,
-      "loss": 1.3137,
+      "epoch": 0.89,
+      "grad_norm": 5.541016578674316,
+      "learning_rate": 1.409864870518345e-05,
+      "loss": 0.098,
       "step": 7054
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 25.52848243713379,
-      "learning_rate": 5.8594767966322555e-06,
-      "loss": 2.0762,
+      "epoch": 0.89,
+      "grad_norm": 10.426650047302246,
+      "learning_rate": 1.4097811990126764e-05,
+      "loss": 0.9164,
       "step": 7055
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 36.224937438964844,
-      "learning_rate": 5.857472186027865e-06,
-      "loss": 1.4368,
+      "epoch": 0.89,
+      "grad_norm": 30.81728172302246,
+      "learning_rate": 1.4096975275070074e-05,
+      "loss": 1.9322,
       "step": 7056
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 37.73203659057617,
-      "learning_rate": 5.855467575423475e-06,
-      "loss": 2.0182,
+      "epoch": 0.89,
+      "grad_norm": 29.294023513793945,
+      "learning_rate": 1.4096138560013388e-05,
+      "loss": 1.4222,
       "step": 7057
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 10.022035598754883,
-      "learning_rate": 5.853462964819084e-06,
-      "loss": 0.9173,
+      "epoch": 0.89,
+      "grad_norm": 21.741195678710938,
+      "learning_rate": 1.4095301844956702e-05,
+      "loss": 1.363,
       "step": 7058
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 31.495861053466797,
-      "learning_rate": 5.851458354214694e-06,
-      "loss": 1.8068,
+      "epoch": 0.89,
+      "grad_norm": 13.830698013305664,
+      "learning_rate": 1.4094465129900015e-05,
+      "loss": 1.0887,
       "step": 7059
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 13.925558090209961,
-      "learning_rate": 5.849453743610303e-06,
-      "loss": 0.8992,
+      "epoch": 0.89,
+      "grad_norm": 2.4803571701049805,
+      "learning_rate": 1.4093628414843325e-05,
+      "loss": 0.0894,
       "step": 7060
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 45.48153305053711,
-      "learning_rate": 5.8474491330059135e-06,
-      "loss": 2.8715,
+      "epoch": 0.89,
+      "grad_norm": 18.780452728271484,
+      "learning_rate": 1.4092791699786639e-05,
+      "loss": 1.6977,
       "step": 7061
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 19.373300552368164,
-      "learning_rate": 5.8454445224015244e-06,
-      "loss": 1.1757,
+      "epoch": 0.89,
+      "grad_norm": 17.498626708984375,
+      "learning_rate": 1.4091954984729953e-05,
+      "loss": 2.0894,
       "step": 7062
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 20.65883445739746,
-      "learning_rate": 5.8434399117971345e-06,
-      "loss": 2.2168,
+      "epoch": 0.89,
+      "grad_norm": 20.703948974609375,
+      "learning_rate": 1.4091118269673263e-05,
+      "loss": 2.0404,
       "step": 7063
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 18.880720138549805,
-      "learning_rate": 5.841435301192744e-06,
-      "loss": 1.2185,
+      "epoch": 0.89,
+      "grad_norm": 6.733626842498779,
+      "learning_rate": 1.4090281554616577e-05,
+      "loss": 1.989,
       "step": 7064
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 14.757368087768555,
-      "learning_rate": 5.839430690588354e-06,
-      "loss": 1.3524,
+      "epoch": 0.89,
+      "grad_norm": 12.881988525390625,
+      "learning_rate": 1.4089444839559889e-05,
+      "loss": 0.8604,
       "step": 7065
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 14.642279624938965,
-      "learning_rate": 5.837426079983963e-06,
-      "loss": 2.0154,
+      "epoch": 0.89,
+      "grad_norm": 5.169025897979736,
+      "learning_rate": 1.4088608124503202e-05,
+      "loss": 0.5154,
       "step": 7066
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 7.192394733428955,
-      "learning_rate": 5.835421469379573e-06,
-      "loss": 0.3944,
+      "epoch": 0.89,
+      "grad_norm": 15.80472469329834,
+      "learning_rate": 1.4087771409446514e-05,
+      "loss": 1.5132,
       "step": 7067
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 33.72096252441406,
-      "learning_rate": 5.833416858775184e-06,
-      "loss": 1.6128,
+      "epoch": 0.89,
+      "grad_norm": 15.307950973510742,
+      "learning_rate": 1.4086934694389826e-05,
+      "loss": 1.4989,
       "step": 7068
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 18.03380584716797,
-      "learning_rate": 5.831412248170793e-06,
-      "loss": 1.4521,
+      "epoch": 0.89,
+      "grad_norm": 17.048763275146484,
+      "learning_rate": 1.408609797933314e-05,
+      "loss": 1.9225,
       "step": 7069
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 10.494623184204102,
-      "learning_rate": 5.829407637566403e-06,
-      "loss": 1.3512,
+      "epoch": 0.89,
+      "grad_norm": 14.885255813598633,
+      "learning_rate": 1.408526126427645e-05,
+      "loss": 1.6576,
       "step": 7070
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 12.263495445251465,
-      "learning_rate": 5.827403026962013e-06,
-      "loss": 1.0091,
+      "epoch": 0.89,
+      "grad_norm": 9.93848991394043,
+      "learning_rate": 1.4084424549219764e-05,
+      "loss": 1.1178,
       "step": 7071
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 10.31989860534668,
-      "learning_rate": 5.825398416357623e-06,
-      "loss": 1.4169,
+      "epoch": 0.89,
+      "grad_norm": 10.413420677185059,
+      "learning_rate": 1.4083587834163077e-05,
+      "loss": 1.8969,
       "step": 7072
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 24.252532958984375,
-      "learning_rate": 5.823393805753233e-06,
-      "loss": 1.5791,
+      "epoch": 0.89,
+      "grad_norm": 12.487821578979492,
+      "learning_rate": 1.4082751119106391e-05,
+      "loss": 0.8172,
       "step": 7073
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 9.647172927856445,
-      "learning_rate": 5.821389195148843e-06,
-      "loss": 0.7097,
+      "epoch": 0.89,
+      "grad_norm": 5.916619300842285,
+      "learning_rate": 1.4081914404049701e-05,
+      "loss": 0.4773,
       "step": 7074
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 39.26737976074219,
-      "learning_rate": 5.819384584544453e-06,
-      "loss": 1.2926,
+      "epoch": 0.89,
+      "grad_norm": 13.073233604431152,
+      "learning_rate": 1.4081077688993015e-05,
+      "loss": 0.8163,
       "step": 7075
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 18.434120178222656,
-      "learning_rate": 5.817379973940063e-06,
-      "loss": 1.684,
+      "epoch": 0.89,
+      "grad_norm": 6.288042068481445,
+      "learning_rate": 1.4080240973936327e-05,
+      "loss": 0.9977,
       "step": 7076
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 23.794525146484375,
-      "learning_rate": 5.815375363335672e-06,
-      "loss": 1.4986,
+      "epoch": 0.89,
+      "grad_norm": 38.35702133178711,
+      "learning_rate": 1.4079404258879639e-05,
+      "loss": 2.5409,
       "step": 7077
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 23.152494430541992,
-      "learning_rate": 5.813370752731282e-06,
-      "loss": 1.7962,
+      "epoch": 0.89,
+      "grad_norm": 23.470169067382812,
+      "learning_rate": 1.4078567543822953e-05,
+      "loss": 1.3125,
       "step": 7078
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 10.018393516540527,
-      "learning_rate": 5.811366142126892e-06,
-      "loss": 0.6564,
+      "epoch": 0.89,
+      "grad_norm": 16.05512046813965,
+      "learning_rate": 1.4077730828766264e-05,
+      "loss": 2.153,
       "step": 7079
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 14.856311798095703,
-      "learning_rate": 5.8093615315225026e-06,
-      "loss": 1.3591,
-      "step": 7080
-    },
-    {
-      "epoch": 2.13,
-      "eval_loss": 0.1809202879667282,
-      "eval_runtime": 43.5269,
-      "eval_samples_per_second": 33.979,
-      "eval_steps_per_second": 33.979,
+      "epoch": 0.89,
+      "grad_norm": 12.977189064025879,
+      "learning_rate": 1.4076894113709578e-05,
+      "loss": 2.1821,
       "step": 7080
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 14.052577018737793,
-      "learning_rate": 5.807356920918113e-06,
-      "loss": 0.7851,
+      "epoch": 0.89,
+      "grad_norm": 14.016845703125,
+      "learning_rate": 1.4076057398652888e-05,
+      "loss": 3.3767,
       "step": 7081
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 49.35251998901367,
-      "learning_rate": 5.805352310313722e-06,
-      "loss": 2.7815,
+      "epoch": 0.89,
+      "grad_norm": 21.714128494262695,
+      "learning_rate": 1.4075220683596202e-05,
+      "loss": 1.9269,
       "step": 7082
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 48.668617248535156,
-      "learning_rate": 5.803347699709332e-06,
-      "loss": 1.4805,
+      "epoch": 0.89,
+      "grad_norm": 10.526546478271484,
+      "learning_rate": 1.4074383968539516e-05,
+      "loss": 2.438,
       "step": 7083
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 9.275986671447754,
-      "learning_rate": 5.801343089104941e-06,
-      "loss": 0.7976,
+      "epoch": 0.89,
+      "grad_norm": 12.505184173583984,
+      "learning_rate": 1.4073547253482826e-05,
+      "loss": 1.74,
       "step": 7084
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 50.608848571777344,
-      "learning_rate": 5.799338478500551e-06,
-      "loss": 1.9386,
+      "epoch": 0.89,
+      "grad_norm": 22.044870376586914,
+      "learning_rate": 1.407271053842614e-05,
+      "loss": 1.5427,
       "step": 7085
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 18.53827476501465,
-      "learning_rate": 5.797333867896162e-06,
-      "loss": 1.2232,
+      "epoch": 0.89,
+      "grad_norm": 16.35655403137207,
+      "learning_rate": 1.4071873823369453e-05,
+      "loss": 3.4506,
       "step": 7086
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 20.711294174194336,
-      "learning_rate": 5.7953292572917715e-06,
-      "loss": 2.057,
+      "epoch": 0.89,
+      "grad_norm": 26.269641876220703,
+      "learning_rate": 1.4071037108312767e-05,
+      "loss": 1.656,
       "step": 7087
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 12.552864074707031,
-      "learning_rate": 5.7933246466873815e-06,
-      "loss": 0.9687,
+      "epoch": 0.89,
+      "grad_norm": 14.767579078674316,
+      "learning_rate": 1.4070200393256077e-05,
+      "loss": 2.7777,
       "step": 7088
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 20.43316078186035,
-      "learning_rate": 5.791320036082992e-06,
-      "loss": 1.6212,
+      "epoch": 0.89,
+      "grad_norm": 15.188955307006836,
+      "learning_rate": 1.406936367819939e-05,
+      "loss": 1.2754,
       "step": 7089
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 8.86861801147461,
-      "learning_rate": 5.789315425478601e-06,
-      "loss": 1.2358,
+      "epoch": 0.89,
+      "grad_norm": 13.957823753356934,
+      "learning_rate": 1.4068526963142703e-05,
+      "loss": 1.8468,
       "step": 7090
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 17.865428924560547,
-      "learning_rate": 5.787310814874211e-06,
-      "loss": 1.8776,
+      "epoch": 0.89,
+      "grad_norm": 6.557285308837891,
+      "learning_rate": 1.4067690248086015e-05,
+      "loss": 0.657,
       "step": 7091
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 10.841514587402344,
-      "learning_rate": 5.785306204269822e-06,
-      "loss": 1.0038,
+      "epoch": 0.89,
+      "grad_norm": 12.158381462097168,
+      "learning_rate": 1.4066853533029328e-05,
+      "loss": 2.1722,
       "step": 7092
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 10.26215934753418,
-      "learning_rate": 5.783301593665431e-06,
-      "loss": 0.7371,
+      "epoch": 0.89,
+      "grad_norm": 7.080685615539551,
+      "learning_rate": 1.406601681797264e-05,
+      "loss": 0.5531,
       "step": 7093
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 19.391332626342773,
-      "learning_rate": 5.781296983061041e-06,
-      "loss": 1.335,
+      "epoch": 0.89,
+      "grad_norm": 16.063068389892578,
+      "learning_rate": 1.4065180102915954e-05,
+      "loss": 1.5258,
       "step": 7094
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 84.89678955078125,
-      "learning_rate": 5.7792923724566504e-06,
-      "loss": 3.5264,
+      "epoch": 0.89,
+      "grad_norm": 24.38238525390625,
+      "learning_rate": 1.4064343387859264e-05,
+      "loss": 0.9849,
       "step": 7095
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 18.809005737304688,
-      "learning_rate": 5.7772877618522605e-06,
-      "loss": 1.2989,
+      "epoch": 0.89,
+      "grad_norm": 11.94936466217041,
+      "learning_rate": 1.4063506672802578e-05,
+      "loss": 0.9163,
       "step": 7096
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 17.723434448242188,
-      "learning_rate": 5.775283151247871e-06,
-      "loss": 0.9432,
+      "epoch": 0.89,
+      "grad_norm": 13.647477149963379,
+      "learning_rate": 1.4062669957745892e-05,
+      "loss": 1.0582,
       "step": 7097
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 11.459822654724121,
-      "learning_rate": 5.77327854064348e-06,
-      "loss": 0.796,
+      "epoch": 0.89,
+      "grad_norm": 17.50379180908203,
+      "learning_rate": 1.4061833242689202e-05,
+      "loss": 1.1992,
       "step": 7098
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 10.62411880493164,
-      "learning_rate": 5.771273930039091e-06,
-      "loss": 2.0095,
+      "epoch": 0.89,
+      "grad_norm": 23.42878532409668,
+      "learning_rate": 1.4060996527632515e-05,
+      "loss": 1.5985,
       "step": 7099
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 17.644634246826172,
-      "learning_rate": 5.769269319434701e-06,
-      "loss": 1.3346,
+      "epoch": 0.89,
+      "grad_norm": 14.225120544433594,
+      "learning_rate": 1.4060159812575829e-05,
+      "loss": 1.7414,
       "step": 7100
     },
     {
-      "epoch": 2.13,
-      "grad_norm": 16.67945671081543,
-      "learning_rate": 5.76726470883031e-06,
-      "loss": 1.6617,
+      "epoch": 0.89,
+      "grad_norm": 16.80451011657715,
+      "learning_rate": 1.4059323097519143e-05,
+      "loss": 1.5058,
       "step": 7101
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 12.336973190307617,
-      "learning_rate": 5.76526009822592e-06,
-      "loss": 1.2088,
+      "epoch": 0.89,
+      "grad_norm": 13.066004753112793,
+      "learning_rate": 1.4058486382462453e-05,
+      "loss": 2.9879,
       "step": 7102
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 43.285831451416016,
-      "learning_rate": 5.763255487621529e-06,
-      "loss": 1.8144,
+      "epoch": 0.89,
+      "grad_norm": 11.759817123413086,
+      "learning_rate": 1.4057649667405767e-05,
+      "loss": 2.7269,
       "step": 7103
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 95.72914123535156,
-      "learning_rate": 5.7612508770171395e-06,
-      "loss": 1.4038,
+      "epoch": 0.89,
+      "grad_norm": 8.385038375854492,
+      "learning_rate": 1.4056812952349079e-05,
+      "loss": 1.5329,
       "step": 7104
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 37.060298919677734,
-      "learning_rate": 5.7592462664127504e-06,
-      "loss": 1.6265,
+      "epoch": 0.89,
+      "grad_norm": 5.424639701843262,
+      "learning_rate": 1.405597623729239e-05,
+      "loss": 0.2796,
       "step": 7105
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 45.99290466308594,
-      "learning_rate": 5.75724165580836e-06,
-      "loss": 1.7691,
+      "epoch": 0.89,
+      "grad_norm": 7.860196113586426,
+      "learning_rate": 1.4055139522235704e-05,
+      "loss": 0.5247,
       "step": 7106
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 15.260518074035645,
-      "learning_rate": 5.75523704520397e-06,
-      "loss": 1.7976,
+      "epoch": 0.89,
+      "grad_norm": 8.838438987731934,
+      "learning_rate": 1.4054302807179016e-05,
+      "loss": 1.2771,
       "step": 7107
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 18.973817825317383,
-      "learning_rate": 5.753232434599579e-06,
-      "loss": 1.355,
+      "epoch": 0.89,
+      "grad_norm": 17.758642196655273,
+      "learning_rate": 1.405346609212233e-05,
+      "loss": 2.9798,
       "step": 7108
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 16.729829788208008,
-      "learning_rate": 5.751227823995189e-06,
-      "loss": 1.2276,
+      "epoch": 0.89,
+      "grad_norm": 8.714234352111816,
+      "learning_rate": 1.405262937706564e-05,
+      "loss": 0.5166,
       "step": 7109
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 15.583281517028809,
-      "learning_rate": 5.749223213390799e-06,
-      "loss": 1.3231,
+      "epoch": 0.89,
+      "grad_norm": 13.17325210571289,
+      "learning_rate": 1.4051792662008954e-05,
+      "loss": 1.4872,
       "step": 7110
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 9.61965274810791,
-      "learning_rate": 5.747218602786409e-06,
-      "loss": 0.6578,
+      "epoch": 0.89,
+      "grad_norm": 15.605276107788086,
+      "learning_rate": 1.4050955946952267e-05,
+      "loss": 2.4161,
       "step": 7111
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 14.255253791809082,
-      "learning_rate": 5.745213992182019e-06,
-      "loss": 1.0521,
+      "epoch": 0.89,
+      "grad_norm": 16.567859649658203,
+      "learning_rate": 1.4050119231895578e-05,
+      "loss": 0.7589,
       "step": 7112
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 28.25699234008789,
-      "learning_rate": 5.743209381577629e-06,
-      "loss": 1.7292,
+      "epoch": 0.89,
+      "grad_norm": 6.031264781951904,
+      "learning_rate": 1.4049282516838891e-05,
+      "loss": 0.5424,
       "step": 7113
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 24.005878448486328,
-      "learning_rate": 5.741204770973239e-06,
-      "loss": 1.3648,
+      "epoch": 0.89,
+      "grad_norm": 12.151751518249512,
+      "learning_rate": 1.4048445801782205e-05,
+      "loss": 1.4286,
       "step": 7114
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 55.37303924560547,
-      "learning_rate": 5.739200160368849e-06,
-      "loss": 1.104,
+      "epoch": 0.89,
+      "grad_norm": 14.095078468322754,
+      "learning_rate": 1.4047609086725517e-05,
+      "loss": 1.2149,
       "step": 7115
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 18.499359130859375,
-      "learning_rate": 5.737195549764458e-06,
-      "loss": 1.4972,
+      "epoch": 0.89,
+      "grad_norm": 15.963891983032227,
+      "learning_rate": 1.4046772371668829e-05,
+      "loss": 2.6274,
       "step": 7116
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 14.160520553588867,
-      "learning_rate": 5.735190939160069e-06,
-      "loss": 1.2106,
+      "epoch": 0.89,
+      "grad_norm": 7.442634582519531,
+      "learning_rate": 1.4045935656612142e-05,
+      "loss": 0.7399,
       "step": 7117
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 12.180475234985352,
-      "learning_rate": 5.733186328555679e-06,
-      "loss": 1.4281,
+      "epoch": 0.89,
+      "grad_norm": 25.207595825195312,
+      "learning_rate": 1.4045098941555454e-05,
+      "loss": 2.1868,
       "step": 7118
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 10.075895309448242,
-      "learning_rate": 5.731181717951288e-06,
-      "loss": 0.811,
+      "epoch": 0.89,
+      "grad_norm": 26.329158782958984,
+      "learning_rate": 1.4044262226498766e-05,
+      "loss": 1.7922,
       "step": 7119
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 62.43848419189453,
-      "learning_rate": 5.729177107346898e-06,
-      "loss": 2.8725,
+      "epoch": 0.89,
+      "grad_norm": 8.771462440490723,
+      "learning_rate": 1.404342551144208e-05,
+      "loss": 1.5612,
       "step": 7120
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 23.31486701965332,
-      "learning_rate": 5.7271724967425076e-06,
-      "loss": 2.8676,
+      "epoch": 0.89,
+      "grad_norm": 2.6278157234191895,
+      "learning_rate": 1.4042588796385392e-05,
+      "loss": 0.0416,
       "step": 7121
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 13.410110473632812,
-      "learning_rate": 5.725167886138118e-06,
-      "loss": 1.0656,
+      "epoch": 0.89,
+      "grad_norm": 6.591731071472168,
+      "learning_rate": 1.4041752081328706e-05,
+      "loss": 0.9422,
       "step": 7122
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 25.776451110839844,
-      "learning_rate": 5.7231632755337286e-06,
-      "loss": 1.9623,
+      "epoch": 0.89,
+      "grad_norm": 7.92766809463501,
+      "learning_rate": 1.4040915366272016e-05,
+      "loss": 1.6853,
       "step": 7123
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 19.853778839111328,
-      "learning_rate": 5.721158664929339e-06,
-      "loss": 1.2182,
+      "epoch": 0.89,
+      "grad_norm": 6.853774547576904,
+      "learning_rate": 1.404007865121533e-05,
+      "loss": 0.7023,
       "step": 7124
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 30.225332260131836,
-      "learning_rate": 5.719154054324948e-06,
-      "loss": 1.4418,
+      "epoch": 0.89,
+      "grad_norm": 8.716609954833984,
+      "learning_rate": 1.4039241936158643e-05,
+      "loss": 0.6673,
       "step": 7125
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 9.551996231079102,
-      "learning_rate": 5.717149443720558e-06,
-      "loss": 1.1343,
+      "epoch": 0.89,
+      "grad_norm": 21.26909637451172,
+      "learning_rate": 1.4038405221101953e-05,
+      "loss": 1.9829,
       "step": 7126
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 33.873294830322266,
-      "learning_rate": 5.715144833116167e-06,
-      "loss": 1.9743,
+      "epoch": 0.89,
+      "grad_norm": 8.954400062561035,
+      "learning_rate": 1.4037568506045267e-05,
+      "loss": 1.3099,
       "step": 7127
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 11.439074516296387,
-      "learning_rate": 5.713140222511777e-06,
-      "loss": 1.1761,
+      "epoch": 0.89,
+      "grad_norm": 17.29623794555664,
+      "learning_rate": 1.403673179098858e-05,
+      "loss": 2.3307,
       "step": 7128
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 11.673768043518066,
-      "learning_rate": 5.711135611907388e-06,
-      "loss": 1.4744,
+      "epoch": 0.89,
+      "grad_norm": 17.295703887939453,
+      "learning_rate": 1.4035895075931893e-05,
+      "loss": 2.1599,
       "step": 7129
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 22.201738357543945,
-      "learning_rate": 5.7091310013029975e-06,
-      "loss": 2.8592,
+      "epoch": 0.89,
+      "grad_norm": 24.35731315612793,
+      "learning_rate": 1.4035058360875205e-05,
+      "loss": 2.1767,
       "step": 7130
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 11.81099796295166,
-      "learning_rate": 5.7071263906986075e-06,
-      "loss": 1.2218,
+      "epoch": 0.89,
+      "grad_norm": 20.482467651367188,
+      "learning_rate": 1.4034221645818518e-05,
+      "loss": 1.7421,
       "step": 7131
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 26.6683406829834,
-      "learning_rate": 5.705121780094217e-06,
-      "loss": 1.2941,
+      "epoch": 0.9,
+      "grad_norm": 27.635377883911133,
+      "learning_rate": 1.403338493076183e-05,
+      "loss": 1.744,
       "step": 7132
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 8.380386352539062,
-      "learning_rate": 5.703117169489827e-06,
-      "loss": 0.6369,
+      "epoch": 0.9,
+      "grad_norm": 10.110763549804688,
+      "learning_rate": 1.4032548215705142e-05,
+      "loss": 0.9675,
       "step": 7133
     },
     {
-      "epoch": 2.14,
-      "grad_norm": 11.767404556274414,
-      "learning_rate": 5.701112558885437e-06,
-      "loss": 1.0829,
+      "epoch": 0.9,
+      "grad_norm": 13.34918212890625,
+      "learning_rate": 1.4031711500648454e-05,
+      "loss": 2.1284,
       "step": 7134
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 25.861289978027344,
-      "learning_rate": 5.699107948281047e-06,
-      "loss": 1.9822,
+      "epoch": 0.9,
+      "grad_norm": 5.477113246917725,
+      "learning_rate": 1.4030874785591768e-05,
+      "loss": 0.4138,
       "step": 7135
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 27.93609619140625,
-      "learning_rate": 5.697103337676657e-06,
-      "loss": 1.3616,
+      "epoch": 0.9,
+      "grad_norm": 6.456885814666748,
+      "learning_rate": 1.4030038070535081e-05,
+      "loss": 0.3358,
       "step": 7136
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 87.64994812011719,
-      "learning_rate": 5.695098727072267e-06,
-      "loss": 2.3206,
+      "epoch": 0.9,
+      "grad_norm": 10.445297241210938,
+      "learning_rate": 1.4029201355478392e-05,
+      "loss": 1.6127,
       "step": 7137
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 11.817543029785156,
-      "learning_rate": 5.6930941164678764e-06,
-      "loss": 1.2847,
+      "epoch": 0.9,
+      "grad_norm": 13.392292976379395,
+      "learning_rate": 1.4028364640421705e-05,
+      "loss": 1.4171,
       "step": 7138
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 21.869726181030273,
-      "learning_rate": 5.6910895058634865e-06,
-      "loss": 1.2728,
+      "epoch": 0.9,
+      "grad_norm": 13.771772384643555,
+      "learning_rate": 1.4027527925365019e-05,
+      "loss": 1.5265,
       "step": 7139
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 12.761932373046875,
-      "learning_rate": 5.689084895259096e-06,
-      "loss": 1.5109,
+      "epoch": 0.9,
+      "grad_norm": 20.047332763671875,
+      "learning_rate": 1.402669121030833e-05,
+      "loss": 2.2711,
       "step": 7140
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 10.641894340515137,
-      "learning_rate": 5.687080284654706e-06,
-      "loss": 0.7452,
+      "epoch": 0.9,
+      "grad_norm": 21.45895004272461,
+      "learning_rate": 1.4025854495251643e-05,
+      "loss": 3.7054,
       "step": 7141
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 17.009889602661133,
-      "learning_rate": 5.685075674050317e-06,
-      "loss": 1.8117,
+      "epoch": 0.9,
+      "grad_norm": 11.379473686218262,
+      "learning_rate": 1.4025017780194957e-05,
+      "loss": 2.3151,
       "step": 7142
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 24.577238082885742,
-      "learning_rate": 5.683071063445926e-06,
-      "loss": 1.1146,
+      "epoch": 0.9,
+      "grad_norm": 75.81961822509766,
+      "learning_rate": 1.4024181065138269e-05,
+      "loss": 1.9507,
       "step": 7143
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 35.608741760253906,
-      "learning_rate": 5.681066452841536e-06,
-      "loss": 2.1795,
+      "epoch": 0.9,
+      "grad_norm": 17.682466506958008,
+      "learning_rate": 1.402334435008158e-05,
+      "loss": 0.8294,
       "step": 7144
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 11.501810073852539,
-      "learning_rate": 5.679061842237145e-06,
-      "loss": 0.9263,
+      "epoch": 0.9,
+      "grad_norm": 8.107851028442383,
+      "learning_rate": 1.4022507635024894e-05,
+      "loss": 0.8469,
       "step": 7145
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 17.368040084838867,
-      "learning_rate": 5.6770572316327554e-06,
-      "loss": 1.4102,
+      "epoch": 0.9,
+      "grad_norm": 28.90208625793457,
+      "learning_rate": 1.4021670919968206e-05,
+      "loss": 1.9015,
       "step": 7146
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 16.95510482788086,
-      "learning_rate": 5.6750526210283655e-06,
-      "loss": 1.601,
+      "epoch": 0.9,
+      "grad_norm": 14.991819381713867,
+      "learning_rate": 1.4020834204911518e-05,
+      "loss": 2.5976,
       "step": 7147
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 8.293194770812988,
-      "learning_rate": 5.673048010423976e-06,
-      "loss": 0.8742,
+      "epoch": 0.9,
+      "grad_norm": 13.504834175109863,
+      "learning_rate": 1.401999748985483e-05,
+      "loss": 1.9669,
       "step": 7148
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 6.959547519683838,
-      "learning_rate": 5.671043399819586e-06,
-      "loss": 0.4787,
+      "epoch": 0.9,
+      "grad_norm": 14.985677719116211,
+      "learning_rate": 1.4019160774798144e-05,
+      "loss": 1.2278,
       "step": 7149
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 15.539986610412598,
-      "learning_rate": 5.669038789215196e-06,
-      "loss": 0.8455,
+      "epoch": 0.9,
+      "grad_norm": 31.02518081665039,
+      "learning_rate": 1.4018324059741457e-05,
+      "loss": 1.9601,
       "step": 7150
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 11.356565475463867,
-      "learning_rate": 5.667034178610805e-06,
-      "loss": 1.1054,
+      "epoch": 0.9,
+      "grad_norm": 13.735119819641113,
+      "learning_rate": 1.4017487344684768e-05,
+      "loss": 1.2967,
       "step": 7151
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 25.823801040649414,
-      "learning_rate": 5.665029568006415e-06,
-      "loss": 1.7274,
+      "epoch": 0.9,
+      "grad_norm": 7.824489116668701,
+      "learning_rate": 1.4016650629628081e-05,
+      "loss": 0.6478,
       "step": 7152
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 43.120635986328125,
-      "learning_rate": 5.663024957402024e-06,
-      "loss": 1.6869,
+      "epoch": 0.9,
+      "grad_norm": 11.406153678894043,
+      "learning_rate": 1.4015813914571395e-05,
+      "loss": 1.9932,
       "step": 7153
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 8.684465408325195,
-      "learning_rate": 5.661020346797635e-06,
-      "loss": 1.2637,
+      "epoch": 0.9,
+      "grad_norm": 10.797396659851074,
+      "learning_rate": 1.4014977199514705e-05,
+      "loss": 1.707,
       "step": 7154
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 23.894262313842773,
-      "learning_rate": 5.659015736193245e-06,
-      "loss": 1.4771,
+      "epoch": 0.9,
+      "grad_norm": 34.46938705444336,
+      "learning_rate": 1.4014140484458019e-05,
+      "loss": 2.7764,
       "step": 7155
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 26.122549057006836,
-      "learning_rate": 5.6570111255888546e-06,
-      "loss": 1.3112,
+      "epoch": 0.9,
+      "grad_norm": 59.19340896606445,
+      "learning_rate": 1.4013303769401332e-05,
+      "loss": 1.8641,
       "step": 7156
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 20.096174240112305,
-      "learning_rate": 5.655006514984465e-06,
-      "loss": 1.0569,
+      "epoch": 0.9,
+      "grad_norm": 12.268207550048828,
+      "learning_rate": 1.4012467054344644e-05,
+      "loss": 0.8285,
       "step": 7157
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 36.696006774902344,
-      "learning_rate": 5.653001904380075e-06,
-      "loss": 1.7493,
+      "epoch": 0.9,
+      "grad_norm": 4.307581424713135,
+      "learning_rate": 1.4011630339287956e-05,
+      "loss": 0.4835,
       "step": 7158
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 28.245203018188477,
-      "learning_rate": 5.650997293775684e-06,
-      "loss": 1.5218,
+      "epoch": 0.9,
+      "grad_norm": 18.47773551940918,
+      "learning_rate": 1.401079362423127e-05,
+      "loss": 1.4313,
       "step": 7159
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 27.659273147583008,
-      "learning_rate": 5.648992683171295e-06,
-      "loss": 2.6867,
+      "epoch": 0.9,
+      "grad_norm": 4.660049915313721,
+      "learning_rate": 1.4009956909174582e-05,
+      "loss": 1.4403,
       "step": 7160
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 47.04728317260742,
-      "learning_rate": 5.646988072566905e-06,
-      "loss": 1.1954,
+      "epoch": 0.9,
+      "grad_norm": 13.854349136352539,
+      "learning_rate": 1.4009120194117894e-05,
+      "loss": 1.8415,
       "step": 7161
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 17.212331771850586,
-      "learning_rate": 5.644983461962514e-06,
-      "loss": 2.1451,
+      "epoch": 0.9,
+      "grad_norm": 13.846010208129883,
+      "learning_rate": 1.4008283479061206e-05,
+      "loss": 1.4329,
       "step": 7162
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 12.564343452453613,
-      "learning_rate": 5.642978851358124e-06,
-      "loss": 1.1914,
+      "epoch": 0.9,
+      "grad_norm": 35.755775451660156,
+      "learning_rate": 1.400744676400452e-05,
+      "loss": 1.4303,
       "step": 7163
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 38.6114501953125,
-      "learning_rate": 5.6409742407537336e-06,
-      "loss": 0.9675,
+      "epoch": 0.9,
+      "grad_norm": 77.38518524169922,
+      "learning_rate": 1.4006610048947833e-05,
+      "loss": 1.8845,
       "step": 7164
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 49.109031677246094,
-      "learning_rate": 5.638969630149344e-06,
-      "loss": 1.4241,
+      "epoch": 0.9,
+      "grad_norm": 8.316797256469727,
+      "learning_rate": 1.4005773333891143e-05,
+      "loss": 0.9716,
       "step": 7165
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 17.366100311279297,
-      "learning_rate": 5.6369650195449546e-06,
-      "loss": 1.9756,
+      "epoch": 0.9,
+      "grad_norm": 17.104446411132812,
+      "learning_rate": 1.4004936618834457e-05,
+      "loss": 0.42,
       "step": 7166
     },
     {
-      "epoch": 2.15,
-      "grad_norm": 79.63975524902344,
-      "learning_rate": 5.634960408940564e-06,
-      "loss": 1.9999,
+      "epoch": 0.9,
+      "grad_norm": 24.48944664001465,
+      "learning_rate": 1.400409990377777e-05,
+      "loss": 1.9745,
       "step": 7167
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 28.03964614868164,
-      "learning_rate": 5.632955798336174e-06,
-      "loss": 1.4169,
+      "epoch": 0.9,
+      "grad_norm": 10.361266136169434,
+      "learning_rate": 1.4003263188721081e-05,
+      "loss": 2.0676,
       "step": 7168
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 23.203266143798828,
-      "learning_rate": 5.630951187731783e-06,
-      "loss": 1.3344,
+      "epoch": 0.9,
+      "grad_norm": 13.4504976272583,
+      "learning_rate": 1.4002426473664395e-05,
+      "loss": 1.0771,
       "step": 7169
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 12.224784851074219,
-      "learning_rate": 5.628946577127393e-06,
-      "loss": 1.4286,
+      "epoch": 0.9,
+      "grad_norm": 18.236623764038086,
+      "learning_rate": 1.4001589758607708e-05,
+      "loss": 0.8684,
       "step": 7170
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 9.047876358032227,
-      "learning_rate": 5.626941966523003e-06,
-      "loss": 1.1898,
+      "epoch": 0.9,
+      "grad_norm": 32.698753356933594,
+      "learning_rate": 1.400075304355102e-05,
+      "loss": 1.0423,
       "step": 7171
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 22.1667423248291,
-      "learning_rate": 5.624937355918613e-06,
-      "loss": 1.945,
+      "epoch": 0.9,
+      "grad_norm": 7.705559253692627,
+      "learning_rate": 1.3999916328494332e-05,
+      "loss": 0.4881,
       "step": 7172
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 13.911121368408203,
-      "learning_rate": 5.6229327453142235e-06,
-      "loss": 1.3742,
+      "epoch": 0.9,
+      "grad_norm": 8.77570629119873,
+      "learning_rate": 1.3999079613437646e-05,
+      "loss": 0.7506,
       "step": 7173
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 9.241662979125977,
-      "learning_rate": 5.6209281347098335e-06,
-      "loss": 1.0261,
+      "epoch": 0.9,
+      "grad_norm": 11.598416328430176,
+      "learning_rate": 1.3998242898380958e-05,
+      "loss": 0.2903,
       "step": 7174
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 13.640268325805664,
-      "learning_rate": 5.618923524105443e-06,
-      "loss": 1.1987,
+      "epoch": 0.9,
+      "grad_norm": 18.047786712646484,
+      "learning_rate": 1.399740618332427e-05,
+      "loss": 1.9002,
       "step": 7175
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 21.423112869262695,
-      "learning_rate": 5.616918913501053e-06,
-      "loss": 1.722,
+      "epoch": 0.9,
+      "grad_norm": 15.605576515197754,
+      "learning_rate": 1.3996569468267582e-05,
+      "loss": 2.7821,
       "step": 7176
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 33.87396240234375,
-      "learning_rate": 5.614914302896662e-06,
-      "loss": 0.9787,
+      "epoch": 0.9,
+      "grad_norm": 8.118613243103027,
+      "learning_rate": 1.3995732753210895e-05,
+      "loss": 1.4705,
       "step": 7177
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 29.940101623535156,
-      "learning_rate": 5.612909692292273e-06,
-      "loss": 1.792,
+      "epoch": 0.9,
+      "grad_norm": 32.597618103027344,
+      "learning_rate": 1.3994896038154209e-05,
+      "loss": 2.3296,
       "step": 7178
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 22.763090133666992,
-      "learning_rate": 5.610905081687883e-06,
-      "loss": 1.3948,
+      "epoch": 0.9,
+      "grad_norm": 11.029437065124512,
+      "learning_rate": 1.399405932309752e-05,
+      "loss": 2.8688,
       "step": 7179
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 33.94498062133789,
-      "learning_rate": 5.608900471083492e-06,
-      "loss": 2.2237,
+      "epoch": 0.9,
+      "grad_norm": 4.349131107330322,
+      "learning_rate": 1.3993222608040833e-05,
+      "loss": 0.579,
       "step": 7180
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 15.50265884399414,
-      "learning_rate": 5.6068958604791024e-06,
-      "loss": 1.5329,
+      "epoch": 0.9,
+      "grad_norm": 18.68450164794922,
+      "learning_rate": 1.3992385892984147e-05,
+      "loss": 2.8652,
       "step": 7181
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 20.567745208740234,
-      "learning_rate": 5.6048912498747125e-06,
-      "loss": 1.7991,
+      "epoch": 0.9,
+      "grad_norm": 14.697657585144043,
+      "learning_rate": 1.3991549177927457e-05,
+      "loss": 0.5772,
       "step": 7182
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 38.924346923828125,
-      "learning_rate": 5.602886639270322e-06,
-      "loss": 1.8058,
+      "epoch": 0.9,
+      "grad_norm": 3.4043476581573486,
+      "learning_rate": 1.399071246287077e-05,
+      "loss": 0.1265,
       "step": 7183
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 23.95490264892578,
-      "learning_rate": 5.600882028665932e-06,
-      "loss": 0.9036,
+      "epoch": 0.9,
+      "grad_norm": 18.97923469543457,
+      "learning_rate": 1.3989875747814084e-05,
+      "loss": 0.9864,
       "step": 7184
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 8.675360679626465,
-      "learning_rate": 5.598877418061543e-06,
-      "loss": 1.1445,
+      "epoch": 0.9,
+      "grad_norm": 13.734219551086426,
+      "learning_rate": 1.3989039032757396e-05,
+      "loss": 1.0469,
       "step": 7185
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 52.8515510559082,
-      "learning_rate": 5.596872807457152e-06,
-      "loss": 0.9069,
+      "epoch": 0.9,
+      "grad_norm": 9.58630084991455,
+      "learning_rate": 1.3988202317700708e-05,
+      "loss": 0.1549,
       "step": 7186
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 14.209603309631348,
-      "learning_rate": 5.594868196852762e-06,
-      "loss": 1.1668,
+      "epoch": 0.9,
+      "grad_norm": 13.013339042663574,
+      "learning_rate": 1.398736560264402e-05,
+      "loss": 1.3083,
       "step": 7187
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 19.132238388061523,
-      "learning_rate": 5.592863586248371e-06,
-      "loss": 1.6934,
+      "epoch": 0.9,
+      "grad_norm": 20.65224266052246,
+      "learning_rate": 1.3986528887587334e-05,
+      "loss": 1.9961,
       "step": 7188
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 16.72909927368164,
-      "learning_rate": 5.5908589756439814e-06,
-      "loss": 1.127,
+      "epoch": 0.9,
+      "grad_norm": 58.49673080444336,
+      "learning_rate": 1.3985692172530646e-05,
+      "loss": 2.1291,
       "step": 7189
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 11.741925239562988,
-      "learning_rate": 5.588854365039591e-06,
-      "loss": 1.9914,
+      "epoch": 0.9,
+      "grad_norm": 21.04233741760254,
+      "learning_rate": 1.3984855457473958e-05,
+      "loss": 3.0153,
       "step": 7190
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 34.14512634277344,
-      "learning_rate": 5.586849754435202e-06,
-      "loss": 2.0267,
+      "epoch": 0.9,
+      "grad_norm": 56.784305572509766,
+      "learning_rate": 1.3984018742417271e-05,
+      "loss": 1.2907,
       "step": 7191
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 15.906356811523438,
-      "learning_rate": 5.584845143830812e-06,
-      "loss": 1.2969,
+      "epoch": 0.9,
+      "grad_norm": 18.404029846191406,
+      "learning_rate": 1.3983182027360585e-05,
+      "loss": 1.9775,
       "step": 7192
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 14.884824752807617,
-      "learning_rate": 5.582840533226421e-06,
-      "loss": 1.6604,
+      "epoch": 0.9,
+      "grad_norm": 16.816679000854492,
+      "learning_rate": 1.3982345312303895e-05,
+      "loss": 1.1258,
       "step": 7193
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 14.613624572753906,
-      "learning_rate": 5.580835922622031e-06,
-      "loss": 1.3751,
+      "epoch": 0.9,
+      "grad_norm": 4.080275535583496,
+      "learning_rate": 1.3981508597247209e-05,
+      "loss": 0.1642,
       "step": 7194
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 14.830942153930664,
-      "learning_rate": 5.578831312017641e-06,
-      "loss": 1.5889,
+      "epoch": 0.9,
+      "grad_norm": 10.518272399902344,
+      "learning_rate": 1.3980671882190522e-05,
+      "loss": 2.8347,
       "step": 7195
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 17.539030075073242,
-      "learning_rate": 5.57682670141325e-06,
-      "loss": 1.0911,
+      "epoch": 0.9,
+      "grad_norm": 14.956340789794922,
+      "learning_rate": 1.3979835167133833e-05,
+      "loss": 1.6647,
       "step": 7196
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 28.30144500732422,
-      "learning_rate": 5.574822090808861e-06,
-      "loss": 1.2038,
+      "epoch": 0.9,
+      "grad_norm": 9.525736808776855,
+      "learning_rate": 1.3978998452077146e-05,
+      "loss": 1.8221,
       "step": 7197
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 11.865046501159668,
-      "learning_rate": 5.572817480204471e-06,
-      "loss": 1.2387,
+      "epoch": 0.9,
+      "grad_norm": 47.34003829956055,
+      "learning_rate": 1.397816173702046e-05,
+      "loss": 2.235,
       "step": 7198
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 29.797569274902344,
-      "learning_rate": 5.5708128696000806e-06,
-      "loss": 2.5909,
+      "epoch": 0.9,
+      "grad_norm": 12.209076881408691,
+      "learning_rate": 1.3977325021963772e-05,
+      "loss": 1.2154,
       "step": 7199
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 14.200777053833008,
-      "learning_rate": 5.568808258995691e-06,
-      "loss": 1.5443,
+      "epoch": 0.9,
+      "grad_norm": 15.106338500976562,
+      "learning_rate": 1.3976488306907084e-05,
+      "loss": 2.5594,
       "step": 7200
     },
     {
-      "epoch": 2.16,
-      "eval_loss": 0.1803688108921051,
-      "eval_runtime": 43.5403,
-      "eval_samples_per_second": 33.969,
-      "eval_steps_per_second": 33.969,
+      "epoch": 0.9,
+      "eval_loss": 0.09446447342634201,
+      "eval_runtime": 156.8041,
+      "eval_samples_per_second": 22.589,
+      "eval_steps_per_second": 22.589,
       "step": 7200
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 8.243300437927246,
-      "learning_rate": 5.5668036483913e-06,
-      "loss": 1.2393,
+      "epoch": 0.9,
+      "grad_norm": 25.125381469726562,
+      "learning_rate": 1.3975651591850396e-05,
+      "loss": 3.3564,
       "step": 7201
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 29.610637664794922,
-      "learning_rate": 5.56479903778691e-06,
-      "loss": 0.9743,
+      "epoch": 0.9,
+      "grad_norm": 9.072014808654785,
+      "learning_rate": 1.397481487679371e-05,
+      "loss": 1.421,
       "step": 7202
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 30.562345504760742,
-      "learning_rate": 5.562794427182521e-06,
-      "loss": 1.5775,
+      "epoch": 0.9,
+      "grad_norm": 4.650312900543213,
+      "learning_rate": 1.3973978161737021e-05,
+      "loss": 0.3649,
       "step": 7203
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 13.141053199768066,
-      "learning_rate": 5.56078981657813e-06,
-      "loss": 1.026,
+      "epoch": 0.9,
+      "grad_norm": 9.013452529907227,
+      "learning_rate": 1.3973141446680333e-05,
+      "loss": 0.7415,
       "step": 7204
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 15.844094276428223,
-      "learning_rate": 5.55878520597374e-06,
-      "loss": 1.8327,
+      "epoch": 0.9,
+      "grad_norm": 57.821529388427734,
+      "learning_rate": 1.3972304731623647e-05,
+      "loss": 2.823,
       "step": 7205
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 7.878782272338867,
-      "learning_rate": 5.5567805953693495e-06,
-      "loss": 1.0848,
+      "epoch": 0.9,
+      "grad_norm": 9.371427536010742,
+      "learning_rate": 1.397146801656696e-05,
+      "loss": 1.2171,
       "step": 7206
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 12.858576774597168,
-      "learning_rate": 5.5547759847649596e-06,
-      "loss": 1.5242,
+      "epoch": 0.9,
+      "grad_norm": 18.87268829345703,
+      "learning_rate": 1.3970631301510271e-05,
+      "loss": 1.9732,
       "step": 7207
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 11.919968605041504,
-      "learning_rate": 5.55277137416057e-06,
-      "loss": 1.6723,
+      "epoch": 0.9,
+      "grad_norm": 27.41904067993164,
+      "learning_rate": 1.3969794586453585e-05,
+      "loss": 2.1656,
       "step": 7208
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 11.811442375183105,
-      "learning_rate": 5.5507667635561806e-06,
-      "loss": 1.1354,
+      "epoch": 0.9,
+      "grad_norm": 17.43402862548828,
+      "learning_rate": 1.3968957871396898e-05,
+      "loss": 1.4941,
       "step": 7209
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 32.790809631347656,
-      "learning_rate": 5.54876215295179e-06,
-      "loss": 1.136,
+      "epoch": 0.9,
+      "grad_norm": 11.491718292236328,
+      "learning_rate": 1.3968121156340208e-05,
+      "loss": 1.061,
       "step": 7210
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 29.143482208251953,
-      "learning_rate": 5.5467575423474e-06,
-      "loss": 1.3966,
+      "epoch": 0.9,
+      "grad_norm": 23.266481399536133,
+      "learning_rate": 1.3967284441283522e-05,
+      "loss": 1.5969,
       "step": 7211
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 24.455890655517578,
-      "learning_rate": 5.544752931743009e-06,
-      "loss": 2.0205,
+      "epoch": 0.91,
+      "grad_norm": 14.703265190124512,
+      "learning_rate": 1.3966447726226836e-05,
+      "loss": 2.4698,
       "step": 7212
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 23.328998565673828,
-      "learning_rate": 5.542748321138619e-06,
-      "loss": 1.6471,
+      "epoch": 0.91,
+      "grad_norm": 133.14865112304688,
+      "learning_rate": 1.3965611011170148e-05,
+      "loss": 2.1416,
       "step": 7213
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 26.35028076171875,
-      "learning_rate": 5.5407437105342285e-06,
-      "loss": 1.5245,
+      "epoch": 0.91,
+      "grad_norm": 4.905088424682617,
+      "learning_rate": 1.396477429611346e-05,
+      "loss": 0.1839,
       "step": 7214
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 23.13437843322754,
-      "learning_rate": 5.538739099929839e-06,
-      "loss": 1.7795,
+      "epoch": 0.91,
+      "grad_norm": 8.868611335754395,
+      "learning_rate": 1.3963937581056772e-05,
+      "loss": 3.0066,
       "step": 7215
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 16.176660537719727,
-      "learning_rate": 5.5367344893254495e-06,
-      "loss": 1.5189,
+      "epoch": 0.91,
+      "grad_norm": 36.57992935180664,
+      "learning_rate": 1.3963100866000085e-05,
+      "loss": 2.9071,
       "step": 7216
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 16.43119239807129,
-      "learning_rate": 5.534729878721059e-06,
-      "loss": 1.0649,
+      "epoch": 0.91,
+      "grad_norm": 27.707067489624023,
+      "learning_rate": 1.3962264150943397e-05,
+      "loss": 1.2687,
       "step": 7217
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 16.297130584716797,
-      "learning_rate": 5.532725268116669e-06,
-      "loss": 1.0049,
+      "epoch": 0.91,
+      "grad_norm": 25.992782592773438,
+      "learning_rate": 1.396142743588671e-05,
+      "loss": 2.2674,
       "step": 7218
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 11.09915542602539,
-      "learning_rate": 5.530720657512279e-06,
-      "loss": 1.3301,
+      "epoch": 0.91,
+      "grad_norm": 34.82918167114258,
+      "learning_rate": 1.3960590720830023e-05,
+      "loss": 2.9348,
       "step": 7219
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 10.667206764221191,
-      "learning_rate": 5.528716046907888e-06,
-      "loss": 1.0879,
+      "epoch": 0.91,
+      "grad_norm": 26.297290802001953,
+      "learning_rate": 1.3959754005773336e-05,
+      "loss": 1.9831,
       "step": 7220
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 22.059289932250977,
-      "learning_rate": 5.526711436303499e-06,
-      "loss": 1.7633,
+      "epoch": 0.91,
+      "grad_norm": 11.515657424926758,
+      "learning_rate": 1.3958917290716647e-05,
+      "loss": 1.78,
       "step": 7221
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 45.51456069946289,
-      "learning_rate": 5.524706825699109e-06,
-      "loss": 1.451,
+      "epoch": 0.91,
+      "grad_norm": 11.170564651489258,
+      "learning_rate": 1.395808057565996e-05,
+      "loss": 2.2037,
       "step": 7222
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 13.462637901306152,
-      "learning_rate": 5.522702215094718e-06,
-      "loss": 1.4225,
+      "epoch": 0.91,
+      "grad_norm": 84.04158020019531,
+      "learning_rate": 1.3957243860603274e-05,
+      "loss": 3.4347,
       "step": 7223
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 11.555397033691406,
-      "learning_rate": 5.5206976044903284e-06,
-      "loss": 1.2796,
+      "epoch": 0.91,
+      "grad_norm": 38.162410736083984,
+      "learning_rate": 1.3956407145546584e-05,
+      "loss": 2.0202,
       "step": 7224
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 25.498672485351562,
-      "learning_rate": 5.518692993885938e-06,
-      "loss": 1.1889,
+      "epoch": 0.91,
+      "grad_norm": 20.224584579467773,
+      "learning_rate": 1.3955570430489898e-05,
+      "loss": 1.7124,
       "step": 7225
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 18.293188095092773,
-      "learning_rate": 5.516688383281548e-06,
-      "loss": 2.2397,
+      "epoch": 0.91,
+      "grad_norm": 17.18352508544922,
+      "learning_rate": 1.395473371543321e-05,
+      "loss": 3.1194,
       "step": 7226
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 17.86231231689453,
-      "learning_rate": 5.514683772677157e-06,
-      "loss": 1.5173,
+      "epoch": 0.91,
+      "grad_norm": 10.034878730773926,
+      "learning_rate": 1.3953897000376524e-05,
+      "loss": 0.6442,
       "step": 7227
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 18.436603546142578,
-      "learning_rate": 5.512679162072768e-06,
-      "loss": 1.2884,
+      "epoch": 0.91,
+      "grad_norm": 24.862689971923828,
+      "learning_rate": 1.3953060285319836e-05,
+      "loss": 1.5621,
       "step": 7228
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 10.068883895874023,
-      "learning_rate": 5.510674551468378e-06,
-      "loss": 1.314,
+      "epoch": 0.91,
+      "grad_norm": 26.202434539794922,
+      "learning_rate": 1.3952223570263147e-05,
+      "loss": 0.9027,
       "step": 7229
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 9.923480987548828,
-      "learning_rate": 5.508669940863987e-06,
-      "loss": 0.8318,
+      "epoch": 0.91,
+      "grad_norm": 19.527454376220703,
+      "learning_rate": 1.3951386855206461e-05,
+      "loss": 2.2706,
       "step": 7230
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 30.2728214263916,
-      "learning_rate": 5.506665330259597e-06,
-      "loss": 1.7006,
+      "epoch": 0.91,
+      "grad_norm": 29.45587158203125,
+      "learning_rate": 1.3950550140149773e-05,
+      "loss": 1.4363,
       "step": 7231
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 24.514514923095703,
-      "learning_rate": 5.5046607196552074e-06,
-      "loss": 1.2372,
+      "epoch": 0.91,
+      "grad_norm": 17.59807777404785,
+      "learning_rate": 1.3949713425093085e-05,
+      "loss": 2.2094,
       "step": 7232
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 8.947182655334473,
-      "learning_rate": 5.502656109050817e-06,
-      "loss": 0.6531,
+      "epoch": 0.91,
+      "grad_norm": 6.8975934982299805,
+      "learning_rate": 1.3948876710036399e-05,
+      "loss": 0.938,
       "step": 7233
     },
     {
-      "epoch": 2.17,
-      "grad_norm": 32.466487884521484,
-      "learning_rate": 5.500651498446428e-06,
-      "loss": 0.9527,
+      "epoch": 0.91,
+      "grad_norm": 75.04706573486328,
+      "learning_rate": 1.3948039994979712e-05,
+      "loss": 2.0924,
       "step": 7234
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 19.44408416748047,
-      "learning_rate": 5.498646887842038e-06,
-      "loss": 1.8595,
+      "epoch": 0.91,
+      "grad_norm": 8.471443176269531,
+      "learning_rate": 1.3947203279923023e-05,
+      "loss": 1.807,
       "step": 7235
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 15.075602531433105,
-      "learning_rate": 5.496642277237647e-06,
-      "loss": 1.2715,
+      "epoch": 0.91,
+      "grad_norm": 23.299184799194336,
+      "learning_rate": 1.3946366564866336e-05,
+      "loss": 3.5616,
       "step": 7236
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 40.67919921875,
-      "learning_rate": 5.494637666633257e-06,
-      "loss": 1.1642,
+      "epoch": 0.91,
+      "grad_norm": 10.661001205444336,
+      "learning_rate": 1.394552984980965e-05,
+      "loss": 0.9546,
       "step": 7237
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 11.373068809509277,
-      "learning_rate": 5.492633056028866e-06,
-      "loss": 0.8346,
+      "epoch": 0.91,
+      "grad_norm": 28.18711280822754,
+      "learning_rate": 1.394469313475296e-05,
+      "loss": 1.6205,
       "step": 7238
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 19.530393600463867,
-      "learning_rate": 5.490628445424476e-06,
-      "loss": 1.1305,
+      "epoch": 0.91,
+      "grad_norm": 177.2775115966797,
+      "learning_rate": 1.3943856419696274e-05,
+      "loss": 3.1622,
       "step": 7239
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 93.64016723632812,
-      "learning_rate": 5.488623834820087e-06,
-      "loss": 1.6901,
+      "epoch": 0.91,
+      "grad_norm": 10.07783031463623,
+      "learning_rate": 1.3943019704639586e-05,
+      "loss": 1.1788,
       "step": 7240
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 26.258949279785156,
-      "learning_rate": 5.4866192242156965e-06,
-      "loss": 2.7283,
+      "epoch": 0.91,
+      "grad_norm": 24.80470848083496,
+      "learning_rate": 1.39421829895829e-05,
+      "loss": 2.455,
       "step": 7241
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 23.18985366821289,
-      "learning_rate": 5.4846146136113066e-06,
-      "loss": 1.3425,
+      "epoch": 0.91,
+      "grad_norm": 12.35865306854248,
+      "learning_rate": 1.3941346274526211e-05,
+      "loss": 2.0769,
       "step": 7242
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 18.908994674682617,
-      "learning_rate": 5.482610003006917e-06,
-      "loss": 1.075,
+      "epoch": 0.91,
+      "grad_norm": 22.02276039123535,
+      "learning_rate": 1.3940509559469523e-05,
+      "loss": 1.3337,
       "step": 7243
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 16.2775936126709,
-      "learning_rate": 5.480605392402526e-06,
-      "loss": 1.6765,
+      "epoch": 0.91,
+      "grad_norm": 16.321331024169922,
+      "learning_rate": 1.3939672844412837e-05,
+      "loss": 1.0,
       "step": 7244
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 22.71320915222168,
-      "learning_rate": 5.478600781798136e-06,
-      "loss": 0.7902,
+      "epoch": 0.91,
+      "grad_norm": 34.85578918457031,
+      "learning_rate": 1.3938836129356147e-05,
+      "loss": 1.4231,
       "step": 7245
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 27.827980041503906,
-      "learning_rate": 5.476596171193747e-06,
-      "loss": 1.4328,
+      "epoch": 0.91,
+      "grad_norm": 19.227888107299805,
+      "learning_rate": 1.3937999414299461e-05,
+      "loss": 1.4348,
       "step": 7246
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 11.23440170288086,
-      "learning_rate": 5.474591560589356e-06,
-      "loss": 1.3235,
+      "epoch": 0.91,
+      "grad_norm": 22.612197875976562,
+      "learning_rate": 1.3937162699242775e-05,
+      "loss": 1.4956,
       "step": 7247
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 32.618648529052734,
-      "learning_rate": 5.472586949984966e-06,
-      "loss": 1.3066,
+      "epoch": 0.91,
+      "grad_norm": 27.75803565979004,
+      "learning_rate": 1.3936325984186088e-05,
+      "loss": 2.9314,
       "step": 7248
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 41.39394760131836,
-      "learning_rate": 5.4705823393805755e-06,
-      "loss": 1.2173,
+      "epoch": 0.91,
+      "grad_norm": 13.859046936035156,
+      "learning_rate": 1.3935489269129398e-05,
+      "loss": 3.4463,
       "step": 7249
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 17.75341796875,
-      "learning_rate": 5.4685777287761856e-06,
-      "loss": 0.9759,
+      "epoch": 0.91,
+      "grad_norm": 31.465036392211914,
+      "learning_rate": 1.3934652554072712e-05,
+      "loss": 3.0091,
       "step": 7250
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 24.715322494506836,
-      "learning_rate": 5.466573118171795e-06,
-      "loss": 2.2409,
+      "epoch": 0.91,
+      "grad_norm": 12.641586303710938,
+      "learning_rate": 1.3933815839016026e-05,
+      "loss": 1.5988,
       "step": 7251
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 16.27878761291504,
-      "learning_rate": 5.464568507567406e-06,
-      "loss": 1.0532,
+      "epoch": 0.91,
+      "grad_norm": 13.063556671142578,
+      "learning_rate": 1.3932979123959336e-05,
+      "loss": 1.3257,
       "step": 7252
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 31.52488899230957,
-      "learning_rate": 5.462563896963016e-06,
-      "loss": 2.1667,
+      "epoch": 0.91,
+      "grad_norm": 8.62407112121582,
+      "learning_rate": 1.393214240890265e-05,
+      "loss": 0.7926,
       "step": 7253
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 24.314443588256836,
-      "learning_rate": 5.460559286358625e-06,
-      "loss": 1.6379,
+      "epoch": 0.91,
+      "grad_norm": 28.4606990814209,
+      "learning_rate": 1.3931305693845962e-05,
+      "loss": 2.8975,
       "step": 7254
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 9.153544425964355,
-      "learning_rate": 5.458554675754235e-06,
-      "loss": 0.9792,
+      "epoch": 0.91,
+      "grad_norm": 6.050575256347656,
+      "learning_rate": 1.3930468978789275e-05,
+      "loss": 0.6473,
       "step": 7255
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 15.48164176940918,
-      "learning_rate": 5.456550065149845e-06,
-      "loss": 1.3478,
+      "epoch": 0.91,
+      "grad_norm": 15.265398979187012,
+      "learning_rate": 1.3929632263732587e-05,
+      "loss": 2.0768,
       "step": 7256
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 13.745574951171875,
-      "learning_rate": 5.4545454545454545e-06,
-      "loss": 1.0092,
+      "epoch": 0.91,
+      "grad_norm": 19.988567352294922,
+      "learning_rate": 1.3928795548675899e-05,
+      "loss": 2.2212,
       "step": 7257
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 42.99429702758789,
-      "learning_rate": 5.452540843941065e-06,
-      "loss": 2.0053,
+      "epoch": 0.91,
+      "grad_norm": 9.324673652648926,
+      "learning_rate": 1.3927958833619213e-05,
+      "loss": 0.4111,
       "step": 7258
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 15.831052780151367,
-      "learning_rate": 5.4505362333366755e-06,
-      "loss": 1.2616,
+      "epoch": 0.91,
+      "grad_norm": 8.651508331298828,
+      "learning_rate": 1.3927122118562523e-05,
+      "loss": 1.4244,
       "step": 7259
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 18.406320571899414,
-      "learning_rate": 5.448531622732285e-06,
-      "loss": 1.7931,
+      "epoch": 0.91,
+      "grad_norm": 14.281431198120117,
+      "learning_rate": 1.3926285403505837e-05,
+      "loss": 1.139,
       "step": 7260
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 23.25520133972168,
-      "learning_rate": 5.446527012127895e-06,
-      "loss": 1.7849,
+      "epoch": 0.91,
+      "grad_norm": 13.728886604309082,
+      "learning_rate": 1.392544868844915e-05,
+      "loss": 1.0722,
       "step": 7261
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 11.698332786560059,
-      "learning_rate": 5.444522401523504e-06,
-      "loss": 1.0173,
+      "epoch": 0.91,
+      "grad_norm": 10.251960754394531,
+      "learning_rate": 1.3924611973392464e-05,
+      "loss": 0.8638,
       "step": 7262
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 18.50562286376953,
-      "learning_rate": 5.442517790919114e-06,
-      "loss": 1.0716,
-      "step": 7263
+      "epoch": 0.91,
+      "grad_norm": 20.055835723876953,
+      "learning_rate": 1.3923775258335774e-05,
+      "loss": 1.4993,
+      "step": 7263
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 18.679840087890625,
-      "learning_rate": 5.440513180314725e-06,
-      "loss": 1.331,
+      "epoch": 0.91,
+      "grad_norm": 11.874016761779785,
+      "learning_rate": 1.3922938543279088e-05,
+      "loss": 2.348,
       "step": 7264
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 11.41555118560791,
-      "learning_rate": 5.438508569710334e-06,
-      "loss": 1.2174,
+      "epoch": 0.91,
+      "grad_norm": 14.539461135864258,
+      "learning_rate": 1.3922101828222402e-05,
+      "loss": 1.5973,
       "step": 7265
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 12.062613487243652,
-      "learning_rate": 5.436503959105944e-06,
-      "loss": 1.7347,
+      "epoch": 0.91,
+      "grad_norm": 7.348511219024658,
+      "learning_rate": 1.3921265113165712e-05,
+      "loss": 0.9858,
       "step": 7266
     },
     {
-      "epoch": 2.18,
-      "grad_norm": 12.057971000671387,
-      "learning_rate": 5.4344993485015544e-06,
-      "loss": 1.2392,
+      "epoch": 0.91,
+      "grad_norm": 9.626068115234375,
+      "learning_rate": 1.3920428398109025e-05,
+      "loss": 1.9385,
       "step": 7267
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 14.793939590454102,
-      "learning_rate": 5.432494737897164e-06,
-      "loss": 1.7844,
+      "epoch": 0.91,
+      "grad_norm": 14.306622505187988,
+      "learning_rate": 1.3919591683052337e-05,
+      "loss": 0.9388,
       "step": 7268
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 20.210912704467773,
-      "learning_rate": 5.430490127292774e-06,
-      "loss": 2.0136,
+      "epoch": 0.91,
+      "grad_norm": 35.246219635009766,
+      "learning_rate": 1.3918754967995651e-05,
+      "loss": 1.5638,
       "step": 7269
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 6.897406578063965,
-      "learning_rate": 5.428485516688383e-06,
-      "loss": 1.0915,
+      "epoch": 0.91,
+      "grad_norm": 13.283602714538574,
+      "learning_rate": 1.3917918252938963e-05,
+      "loss": 1.0215,
       "step": 7270
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 12.593406677246094,
-      "learning_rate": 5.426480906083994e-06,
-      "loss": 0.7125,
+      "epoch": 0.91,
+      "grad_norm": 16.750701904296875,
+      "learning_rate": 1.3917081537882275e-05,
+      "loss": 3.2107,
       "step": 7271
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 12.755428314208984,
-      "learning_rate": 5.424476295479604e-06,
-      "loss": 1.9477,
+      "epoch": 0.91,
+      "grad_norm": 8.930619239807129,
+      "learning_rate": 1.3916244822825589e-05,
+      "loss": 1.3279,
       "step": 7272
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 7.096426486968994,
-      "learning_rate": 5.422471684875213e-06,
-      "loss": 0.718,
+      "epoch": 0.91,
+      "grad_norm": 9.9955472946167,
+      "learning_rate": 1.3915408107768899e-05,
+      "loss": 0.43,
       "step": 7273
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 8.153369903564453,
-      "learning_rate": 5.420467074270823e-06,
-      "loss": 0.7129,
+      "epoch": 0.91,
+      "grad_norm": 21.53797149658203,
+      "learning_rate": 1.3914571392712213e-05,
+      "loss": 1.2248,
       "step": 7274
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 14.67286491394043,
-      "learning_rate": 5.418462463666433e-06,
-      "loss": 1.6797,
+      "epoch": 0.91,
+      "grad_norm": 14.853739738464355,
+      "learning_rate": 1.3913734677655526e-05,
+      "loss": 1.7979,
       "step": 7275
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 18.533056259155273,
-      "learning_rate": 5.416457853062043e-06,
-      "loss": 1.1951,
+      "epoch": 0.91,
+      "grad_norm": 27.4403018951416,
+      "learning_rate": 1.391289796259884e-05,
+      "loss": 4.5861,
       "step": 7276
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 8.047659873962402,
-      "learning_rate": 5.414453242457654e-06,
-      "loss": 0.6167,
+      "epoch": 0.91,
+      "grad_norm": 15.481956481933594,
+      "learning_rate": 1.391206124754215e-05,
+      "loss": 1.9291,
       "step": 7277
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 11.328049659729004,
-      "learning_rate": 5.412448631853263e-06,
-      "loss": 0.9758,
+      "epoch": 0.91,
+      "grad_norm": 15.197737693786621,
+      "learning_rate": 1.3911224532485464e-05,
+      "loss": 2.6118,
       "step": 7278
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 23.584749221801758,
-      "learning_rate": 5.410444021248873e-06,
-      "loss": 1.7252,
+      "epoch": 0.91,
+      "grad_norm": 16.19619369506836,
+      "learning_rate": 1.3910387817428776e-05,
+      "loss": 0.5395,
       "step": 7279
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 22.34720230102539,
-      "learning_rate": 5.408439410644483e-06,
-      "loss": 2.0207,
+      "epoch": 0.91,
+      "grad_norm": 13.692214012145996,
+      "learning_rate": 1.3909551102372088e-05,
+      "loss": 0.9755,
       "step": 7280
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 19.402429580688477,
-      "learning_rate": 5.406434800040092e-06,
-      "loss": 1.2057,
+      "epoch": 0.91,
+      "grad_norm": 9.371413230895996,
+      "learning_rate": 1.3908714387315401e-05,
+      "loss": 1.5356,
       "step": 7281
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 16.989200592041016,
-      "learning_rate": 5.404430189435702e-06,
-      "loss": 1.9234,
+      "epoch": 0.91,
+      "grad_norm": 24.087139129638672,
+      "learning_rate": 1.3907877672258713e-05,
+      "loss": 1.9896,
       "step": 7282
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 26.748077392578125,
-      "learning_rate": 5.402425578831313e-06,
-      "loss": 1.8995,
+      "epoch": 0.91,
+      "grad_norm": 22.373689651489258,
+      "learning_rate": 1.3907040957202027e-05,
+      "loss": 2.4051,
       "step": 7283
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 97.96756744384766,
-      "learning_rate": 5.4004209682269225e-06,
-      "loss": 2.0797,
+      "epoch": 0.91,
+      "grad_norm": 15.469829559326172,
+      "learning_rate": 1.3906204242145339e-05,
+      "loss": 1.7109,
       "step": 7284
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 20.710670471191406,
-      "learning_rate": 5.398416357622533e-06,
-      "loss": 1.0238,
+      "epoch": 0.91,
+      "grad_norm": 28.993919372558594,
+      "learning_rate": 1.390536752708865e-05,
+      "loss": 1.6556,
       "step": 7285
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 13.581433296203613,
-      "learning_rate": 5.396411747018142e-06,
-      "loss": 1.6866,
+      "epoch": 0.91,
+      "grad_norm": 5.3975348472595215,
+      "learning_rate": 1.3904530812031964e-05,
+      "loss": 1.0122,
       "step": 7286
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 12.058156967163086,
-      "learning_rate": 5.394407136413752e-06,
-      "loss": 1.0288,
+      "epoch": 0.91,
+      "grad_norm": 15.712648391723633,
+      "learning_rate": 1.3903694096975275e-05,
+      "loss": 3.0246,
       "step": 7287
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 21.06386375427246,
-      "learning_rate": 5.392402525809361e-06,
-      "loss": 1.271,
+      "epoch": 0.91,
+      "grad_norm": 9.079512596130371,
+      "learning_rate": 1.3902857381918588e-05,
+      "loss": 0.5973,
       "step": 7288
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 51.95856857299805,
-      "learning_rate": 5.390397915204972e-06,
-      "loss": 1.5959,
+      "epoch": 0.91,
+      "grad_norm": 15.587929725646973,
+      "learning_rate": 1.3902020666861902e-05,
+      "loss": 1.6151,
       "step": 7289
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 39.71308898925781,
-      "learning_rate": 5.388393304600582e-06,
-      "loss": 1.7519,
+      "epoch": 0.91,
+      "grad_norm": 10.63734245300293,
+      "learning_rate": 1.3901183951805216e-05,
+      "loss": 1.9477,
       "step": 7290
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 22.637428283691406,
-      "learning_rate": 5.386388693996191e-06,
-      "loss": 2.0241,
+      "epoch": 0.92,
+      "grad_norm": 5.688929557800293,
+      "learning_rate": 1.3900347236748526e-05,
+      "loss": 0.3181,
       "step": 7291
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 13.722464561462402,
-      "learning_rate": 5.3843840833918015e-06,
-      "loss": 1.0318,
+      "epoch": 0.92,
+      "grad_norm": 10.356765747070312,
+      "learning_rate": 1.389951052169184e-05,
+      "loss": 0.6252,
       "step": 7292
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 35.6817512512207,
-      "learning_rate": 5.3823794727874116e-06,
-      "loss": 1.7191,
+      "epoch": 0.92,
+      "grad_norm": 7.640063762664795,
+      "learning_rate": 1.3898673806635152e-05,
+      "loss": 1.0457,
       "step": 7293
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 13.845057487487793,
-      "learning_rate": 5.380374862183021e-06,
-      "loss": 1.3774,
+      "epoch": 0.92,
+      "grad_norm": 3.4225265979766846,
+      "learning_rate": 1.3897837091578463e-05,
+      "loss": 0.0955,
       "step": 7294
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 12.214203834533691,
-      "learning_rate": 5.378370251578632e-06,
-      "loss": 1.5962,
+      "epoch": 0.92,
+      "grad_norm": 8.284340858459473,
+      "learning_rate": 1.3897000376521777e-05,
+      "loss": 2.0904,
       "step": 7295
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 66.19721984863281,
-      "learning_rate": 5.376365640974242e-06,
-      "loss": 1.2639,
+      "epoch": 0.92,
+      "grad_norm": 2.8512842655181885,
+      "learning_rate": 1.3896163661465089e-05,
+      "loss": 0.1162,
       "step": 7296
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 14.872382164001465,
-      "learning_rate": 5.374361030369851e-06,
-      "loss": 1.6655,
+      "epoch": 0.92,
+      "grad_norm": 10.442544937133789,
+      "learning_rate": 1.3895326946408403e-05,
+      "loss": 0.6149,
       "step": 7297
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 19.328413009643555,
-      "learning_rate": 5.372356419765461e-06,
-      "loss": 1.3506,
+      "epoch": 0.92,
+      "grad_norm": 9.65011978149414,
+      "learning_rate": 1.3894490231351713e-05,
+      "loss": 1.4331,
       "step": 7298
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 12.651726722717285,
-      "learning_rate": 5.37035180916107e-06,
-      "loss": 1.2916,
+      "epoch": 0.92,
+      "grad_norm": 14.805377006530762,
+      "learning_rate": 1.3893653516295027e-05,
+      "loss": 1.3727,
       "step": 7299
     },
     {
-      "epoch": 2.19,
-      "grad_norm": 13.568867683410645,
-      "learning_rate": 5.3683471985566805e-06,
-      "loss": 1.116,
+      "epoch": 0.92,
+      "grad_norm": 32.2829704284668,
+      "learning_rate": 1.389281680123834e-05,
+      "loss": 2.195,
       "step": 7300
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 17.879594802856445,
-      "learning_rate": 5.366342587952291e-06,
-      "loss": 1.9263,
+      "epoch": 0.92,
+      "grad_norm": 34.021888732910156,
+      "learning_rate": 1.389198008618165e-05,
+      "loss": 1.14,
       "step": 7301
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 16.369518280029297,
-      "learning_rate": 5.364337977347901e-06,
-      "loss": 0.7172,
+      "epoch": 0.92,
+      "grad_norm": 7.26022481918335,
+      "learning_rate": 1.3891143371124964e-05,
+      "loss": 0.6746,
       "step": 7302
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 37.3539924621582,
-      "learning_rate": 5.362333366743511e-06,
-      "loss": 2.3236,
+      "epoch": 0.92,
+      "grad_norm": 4.94225549697876,
+      "learning_rate": 1.3890306656068278e-05,
+      "loss": 1.5535,
       "step": 7303
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 28.66567611694336,
-      "learning_rate": 5.360328756139121e-06,
-      "loss": 1.4089,
+      "epoch": 0.92,
+      "grad_norm": 38.85111618041992,
+      "learning_rate": 1.3889469941011592e-05,
+      "loss": 5.2981,
       "step": 7304
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 15.510285377502441,
-      "learning_rate": 5.35832414553473e-06,
-      "loss": 1.4704,
+      "epoch": 0.92,
+      "grad_norm": 18.075780868530273,
+      "learning_rate": 1.3888633225954902e-05,
+      "loss": 2.4893,
       "step": 7305
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 18.096101760864258,
-      "learning_rate": 5.35631953493034e-06,
-      "loss": 1.029,
+      "epoch": 0.92,
+      "grad_norm": 8.066834449768066,
+      "learning_rate": 1.3887796510898215e-05,
+      "loss": 1.8651,
       "step": 7306
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 34.01200866699219,
-      "learning_rate": 5.354314924325949e-06,
-      "loss": 1.732,
+      "epoch": 0.92,
+      "grad_norm": 15.736470222473145,
+      "learning_rate": 1.3886959795841527e-05,
+      "loss": 1.8323,
       "step": 7307
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 12.730477333068848,
-      "learning_rate": 5.35231031372156e-06,
-      "loss": 2.1386,
+      "epoch": 0.92,
+      "grad_norm": 20.7080078125,
+      "learning_rate": 1.388612308078484e-05,
+      "loss": 2.635,
       "step": 7308
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 17.955995559692383,
-      "learning_rate": 5.35030570311717e-06,
-      "loss": 0.6812,
+      "epoch": 0.92,
+      "grad_norm": 8.999063491821289,
+      "learning_rate": 1.3885286365728153e-05,
+      "loss": 1.009,
       "step": 7309
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 9.278726577758789,
-      "learning_rate": 5.34830109251278e-06,
-      "loss": 0.5906,
+      "epoch": 0.92,
+      "grad_norm": 96.46437072753906,
+      "learning_rate": 1.3884449650671465e-05,
+      "loss": 1.8903,
       "step": 7310
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 7.9487199783325195,
-      "learning_rate": 5.34629648190839e-06,
-      "loss": 0.6598,
+      "epoch": 0.92,
+      "grad_norm": 91.24198150634766,
+      "learning_rate": 1.3883612935614779e-05,
+      "loss": 1.8324,
       "step": 7311
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 14.72791576385498,
-      "learning_rate": 5.344291871303999e-06,
-      "loss": 0.7821,
+      "epoch": 0.92,
+      "grad_norm": 18.01479721069336,
+      "learning_rate": 1.3882776220558089e-05,
+      "loss": 0.8534,
       "step": 7312
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 8.070356369018555,
-      "learning_rate": 5.342287260699609e-06,
-      "loss": 0.7893,
+      "epoch": 0.92,
+      "grad_norm": 12.081615447998047,
+      "learning_rate": 1.3881939505501402e-05,
+      "loss": 0.997,
       "step": 7313
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 14.87585163116455,
-      "learning_rate": 5.34028265009522e-06,
-      "loss": 1.5793,
+      "epoch": 0.92,
+      "grad_norm": 16.82686424255371,
+      "learning_rate": 1.3881102790444716e-05,
+      "loss": 2.0204,
       "step": 7314
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 119.86707305908203,
-      "learning_rate": 5.338278039490829e-06,
-      "loss": 2.2575,
+      "epoch": 0.92,
+      "grad_norm": 4.8888115882873535,
+      "learning_rate": 1.3880266075388026e-05,
+      "loss": 0.2367,
       "step": 7315
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 22.031450271606445,
-      "learning_rate": 5.336273428886439e-06,
-      "loss": 0.9648,
+      "epoch": 0.92,
+      "grad_norm": 11.908449172973633,
+      "learning_rate": 1.387942936033134e-05,
+      "loss": 1.0558,
       "step": 7316
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 20.54271697998047,
-      "learning_rate": 5.334268818282049e-06,
-      "loss": 1.7682,
+      "epoch": 0.92,
+      "grad_norm": 10.127660751342773,
+      "learning_rate": 1.3878592645274654e-05,
+      "loss": 1.0307,
       "step": 7317
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 18.402219772338867,
-      "learning_rate": 5.332264207677659e-06,
-      "loss": 1.3206,
+      "epoch": 0.92,
+      "grad_norm": 15.890748023986816,
+      "learning_rate": 1.3877755930217964e-05,
+      "loss": 0.8469,
       "step": 7318
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 21.85274887084961,
-      "learning_rate": 5.330259597073269e-06,
-      "loss": 2.4046,
+      "epoch": 0.92,
+      "grad_norm": 11.59317398071289,
+      "learning_rate": 1.3876919215161278e-05,
+      "loss": 2.238,
       "step": 7319
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 12.441575050354004,
-      "learning_rate": 5.32825498646888e-06,
-      "loss": 1.8915,
-      "step": 7320
-    },
-    {
-      "epoch": 2.2,
-      "eval_loss": 0.18911099433898926,
-      "eval_runtime": 43.726,
-      "eval_samples_per_second": 33.824,
-      "eval_steps_per_second": 33.824,
+      "epoch": 0.92,
+      "grad_norm": 16.340787887573242,
+      "learning_rate": 1.3876082500104591e-05,
+      "loss": 1.6672,
       "step": 7320
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 18.565292358398438,
-      "learning_rate": 5.326250375864489e-06,
-      "loss": 1.4864,
+      "epoch": 0.92,
+      "grad_norm": 900.1279907226562,
+      "learning_rate": 1.3875245785047903e-05,
+      "loss": 1.7338,
       "step": 7321
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 21.362342834472656,
-      "learning_rate": 5.324245765260099e-06,
-      "loss": 1.0767,
+      "epoch": 0.92,
+      "grad_norm": 23.428512573242188,
+      "learning_rate": 1.3874409069991215e-05,
+      "loss": 2.7145,
       "step": 7322
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 15.389693260192871,
-      "learning_rate": 5.322241154655708e-06,
-      "loss": 0.9276,
+      "epoch": 0.92,
+      "grad_norm": 38.12934494018555,
+      "learning_rate": 1.3873572354934529e-05,
+      "loss": 3.1097,
       "step": 7323
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 83.4190673828125,
-      "learning_rate": 5.320236544051318e-06,
-      "loss": 2.4627,
+      "epoch": 0.92,
+      "grad_norm": 5.672229290008545,
+      "learning_rate": 1.387273563987784e-05,
+      "loss": 0.4034,
       "step": 7324
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 17.144039154052734,
-      "learning_rate": 5.3182319334469275e-06,
-      "loss": 1.6204,
+      "epoch": 0.92,
+      "grad_norm": 9.384552001953125,
+      "learning_rate": 1.3871898924821153e-05,
+      "loss": 0.4092,
       "step": 7325
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 12.209066390991211,
-      "learning_rate": 5.316227322842538e-06,
-      "loss": 0.9967,
+      "epoch": 0.92,
+      "grad_norm": 4.673786640167236,
+      "learning_rate": 1.3871062209764465e-05,
+      "loss": 0.23,
       "step": 7326
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 20.075319290161133,
-      "learning_rate": 5.3142227122381485e-06,
-      "loss": 2.2126,
+      "epoch": 0.92,
+      "grad_norm": 18.703664779663086,
+      "learning_rate": 1.3870225494707778e-05,
+      "loss": 1.3684,
       "step": 7327
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 22.937063217163086,
-      "learning_rate": 5.312218101633759e-06,
-      "loss": 1.583,
+      "epoch": 0.92,
+      "grad_norm": 9.275733947753906,
+      "learning_rate": 1.3869388779651092e-05,
+      "loss": 1.0947,
       "step": 7328
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 9.0242338180542,
-      "learning_rate": 5.310213491029368e-06,
-      "loss": 0.6662,
+      "epoch": 0.92,
+      "grad_norm": 24.407047271728516,
+      "learning_rate": 1.3868552064594402e-05,
+      "loss": 1.3104,
       "step": 7329
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 8.918848037719727,
-      "learning_rate": 5.308208880424978e-06,
-      "loss": 0.8654,
+      "epoch": 0.92,
+      "grad_norm": 7.646800518035889,
+      "learning_rate": 1.3867715349537716e-05,
+      "loss": 1.5059,
       "step": 7330
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 32.961360931396484,
-      "learning_rate": 5.306204269820587e-06,
-      "loss": 1.8518,
+      "epoch": 0.92,
+      "grad_norm": 28.678197860717773,
+      "learning_rate": 1.386687863448103e-05,
+      "loss": 2.3387,
       "step": 7331
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 16.80774688720703,
-      "learning_rate": 5.304199659216198e-06,
-      "loss": 1.4526,
+      "epoch": 0.92,
+      "grad_norm": 9.313803672790527,
+      "learning_rate": 1.386604191942434e-05,
+      "loss": 0.3099,
       "step": 7332
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 86.86822509765625,
-      "learning_rate": 5.302195048611808e-06,
-      "loss": 1.354,
+      "epoch": 0.92,
+      "grad_norm": 16.865198135375977,
+      "learning_rate": 1.3865205204367653e-05,
+      "loss": 2.1941,
       "step": 7333
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 18.777795791625977,
-      "learning_rate": 5.300190438007417e-06,
-      "loss": 1.2628,
+      "epoch": 0.92,
+      "grad_norm": 37.41999435424805,
+      "learning_rate": 1.3864368489310967e-05,
+      "loss": 1.9056,
       "step": 7334
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 47.588802337646484,
-      "learning_rate": 5.2981858274030275e-06,
-      "loss": 3.0136,
+      "epoch": 0.92,
+      "grad_norm": 10.510849952697754,
+      "learning_rate": 1.3863531774254279e-05,
+      "loss": 0.8305,
       "step": 7335
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 41.61444091796875,
-      "learning_rate": 5.296181216798637e-06,
-      "loss": 2.3059,
+      "epoch": 0.92,
+      "grad_norm": 53.111297607421875,
+      "learning_rate": 1.3862695059197591e-05,
+      "loss": 2.8198,
       "step": 7336
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 16.625478744506836,
-      "learning_rate": 5.294176606194247e-06,
-      "loss": 1.0123,
+      "epoch": 0.92,
+      "grad_norm": 11.218560218811035,
+      "learning_rate": 1.3861858344140905e-05,
+      "loss": 1.5149,
       "step": 7337
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 50.9505729675293,
-      "learning_rate": 5.292171995589858e-06,
-      "loss": 1.1573,
+      "epoch": 0.92,
+      "grad_norm": 29.03372573852539,
+      "learning_rate": 1.3861021629084217e-05,
+      "loss": 1.3805,
       "step": 7338
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 3.5535824298858643,
-      "learning_rate": 5.290167384985467e-06,
-      "loss": 0.5366,
+      "epoch": 0.92,
+      "grad_norm": 43.508018493652344,
+      "learning_rate": 1.3860184914027529e-05,
+      "loss": 2.5458,
       "step": 7339
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 13.707948684692383,
-      "learning_rate": 5.288162774381077e-06,
-      "loss": 1.2388,
+      "epoch": 0.92,
+      "grad_norm": 9.611120223999023,
+      "learning_rate": 1.385934819897084e-05,
+      "loss": 1.8637,
       "step": 7340
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 117.4101333618164,
-      "learning_rate": 5.286158163776687e-06,
-      "loss": 2.3683,
+      "epoch": 0.92,
+      "grad_norm": 12.637951850891113,
+      "learning_rate": 1.3858511483914154e-05,
+      "loss": 3.2302,
       "step": 7341
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 13.597126960754395,
-      "learning_rate": 5.284153553172296e-06,
-      "loss": 1.0199,
+      "epoch": 0.92,
+      "grad_norm": 16.153526306152344,
+      "learning_rate": 1.3857674768857468e-05,
+      "loss": 1.8928,
       "step": 7342
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 8.304295539855957,
-      "learning_rate": 5.2821489425679065e-06,
-      "loss": 0.7169,
+      "epoch": 0.92,
+      "grad_norm": 21.75115394592285,
+      "learning_rate": 1.3856838053800778e-05,
+      "loss": 3.125,
       "step": 7343
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 14.256132125854492,
-      "learning_rate": 5.280144331963517e-06,
-      "loss": 1.2964,
+      "epoch": 0.92,
+      "grad_norm": 14.656881332397461,
+      "learning_rate": 1.3856001338744092e-05,
+      "loss": 1.3364,
       "step": 7344
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 86.91548919677734,
-      "learning_rate": 5.278139721359127e-06,
-      "loss": 3.997,
+      "epoch": 0.92,
+      "grad_norm": 52.3955078125,
+      "learning_rate": 1.3855164623687405e-05,
+      "loss": 2.6731,
       "step": 7345
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 16.045759201049805,
-      "learning_rate": 5.276135110754737e-06,
-      "loss": 1.1771,
+      "epoch": 0.92,
+      "grad_norm": 9.987492561340332,
+      "learning_rate": 1.3854327908630716e-05,
+      "loss": 0.8823,
       "step": 7346
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 39.05559158325195,
-      "learning_rate": 5.274130500150346e-06,
-      "loss": 2.0468,
+      "epoch": 0.92,
+      "grad_norm": 6.1971330642700195,
+      "learning_rate": 1.385349119357403e-05,
+      "loss": 0.6373,
       "step": 7347
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 13.659476280212402,
-      "learning_rate": 5.272125889545956e-06,
-      "loss": 0.9715,
+      "epoch": 0.92,
+      "grad_norm": 24.01615333557129,
+      "learning_rate": 1.3852654478517343e-05,
+      "loss": 1.7604,
       "step": 7348
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 12.945648193359375,
-      "learning_rate": 5.270121278941565e-06,
-      "loss": 1.0217,
+      "epoch": 0.92,
+      "grad_norm": 5.612430572509766,
+      "learning_rate": 1.3851817763460655e-05,
+      "loss": 0.3099,
       "step": 7349
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 11.518617630004883,
-      "learning_rate": 5.268116668337175e-06,
-      "loss": 1.2031,
+      "epoch": 0.92,
+      "grad_norm": 13.868998527526855,
+      "learning_rate": 1.3850981048403967e-05,
+      "loss": 1.3927,
       "step": 7350
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 43.542816162109375,
-      "learning_rate": 5.266112057732786e-06,
-      "loss": 1.8437,
+      "epoch": 0.92,
+      "grad_norm": 9.475567817687988,
+      "learning_rate": 1.3850144333347279e-05,
+      "loss": 0.7834,
       "step": 7351
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 9.149428367614746,
-      "learning_rate": 5.2641074471283955e-06,
-      "loss": 1.1645,
+      "epoch": 0.92,
+      "grad_norm": 10.77379322052002,
+      "learning_rate": 1.3849307618290592e-05,
+      "loss": 1.2543,
       "step": 7352
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 17.6467227935791,
-      "learning_rate": 5.262102836524006e-06,
-      "loss": 1.3694,
+      "epoch": 0.92,
+      "grad_norm": 9.363557815551758,
+      "learning_rate": 1.3848470903233904e-05,
+      "loss": 2.3459,
       "step": 7353
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 30.845457077026367,
-      "learning_rate": 5.260098225919616e-06,
-      "loss": 1.3288,
+      "epoch": 0.92,
+      "grad_norm": 17.62056541442871,
+      "learning_rate": 1.3847634188177216e-05,
+      "loss": 3.9656,
       "step": 7354
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 17.407968521118164,
-      "learning_rate": 5.258093615315225e-06,
-      "loss": 1.3238,
+      "epoch": 0.92,
+      "grad_norm": 7.625229358673096,
+      "learning_rate": 1.384679747312053e-05,
+      "loss": 1.9232,
       "step": 7355
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 22.961395263671875,
-      "learning_rate": 5.256089004710835e-06,
-      "loss": 1.1421,
+      "epoch": 0.92,
+      "grad_norm": 10.988778114318848,
+      "learning_rate": 1.3845960758063844e-05,
+      "loss": 1.0116,
       "step": 7356
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 21.641342163085938,
-      "learning_rate": 5.254084394106446e-06,
-      "loss": 1.2616,
+      "epoch": 0.92,
+      "grad_norm": 15.370214462280273,
+      "learning_rate": 1.3845124043007154e-05,
+      "loss": 3.1325,
       "step": 7357
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 16.428573608398438,
-      "learning_rate": 5.252079783502055e-06,
-      "loss": 0.9203,
+      "epoch": 0.92,
+      "grad_norm": 11.344792366027832,
+      "learning_rate": 1.3844287327950468e-05,
+      "loss": 0.529,
       "step": 7358
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 17.370075225830078,
-      "learning_rate": 5.250075172897665e-06,
-      "loss": 1.0207,
+      "epoch": 0.92,
+      "grad_norm": 11.936365127563477,
+      "learning_rate": 1.3843450612893781e-05,
+      "loss": 1.2153,
       "step": 7359
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 17.009870529174805,
-      "learning_rate": 5.2480705622932745e-06,
-      "loss": 1.3327,
+      "epoch": 0.92,
+      "grad_norm": 12.046215057373047,
+      "learning_rate": 1.3842613897837091e-05,
+      "loss": 0.9478,
       "step": 7360
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 16.25709342956543,
-      "learning_rate": 5.246065951688885e-06,
-      "loss": 1.274,
+      "epoch": 0.92,
+      "grad_norm": 11.978967666625977,
+      "learning_rate": 1.3841777182780405e-05,
+      "loss": 1.6368,
       "step": 7361
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 11.702857971191406,
-      "learning_rate": 5.244061341084495e-06,
-      "loss": 1.3384,
+      "epoch": 0.92,
+      "grad_norm": 23.105283737182617,
+      "learning_rate": 1.3840940467723719e-05,
+      "loss": 3.2074,
       "step": 7362
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 39.857627868652344,
-      "learning_rate": 5.242056730480105e-06,
-      "loss": 1.6856,
+      "epoch": 0.92,
+      "grad_norm": 10.976641654968262,
+      "learning_rate": 1.384010375266703e-05,
+      "loss": 1.0288,
       "step": 7363
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 34.70151138305664,
-      "learning_rate": 5.240052119875715e-06,
-      "loss": 1.7488,
+      "epoch": 0.92,
+      "grad_norm": 12.47696590423584,
+      "learning_rate": 1.3839267037610343e-05,
+      "loss": 1.0648,
       "step": 7364
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 8.398064613342285,
-      "learning_rate": 5.238047509271325e-06,
-      "loss": 1.1744,
+      "epoch": 0.92,
+      "grad_norm": 3.1748385429382324,
+      "learning_rate": 1.3838430322553655e-05,
+      "loss": 0.1766,
       "step": 7365
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 16.08351707458496,
-      "learning_rate": 5.236042898666934e-06,
-      "loss": 1.2976,
+      "epoch": 0.92,
+      "grad_norm": 134.07736206054688,
+      "learning_rate": 1.3837593607496968e-05,
+      "loss": 2.144,
       "step": 7366
     },
     {
-      "epoch": 2.21,
-      "grad_norm": 41.76036834716797,
-      "learning_rate": 5.234038288062544e-06,
-      "loss": 2.1369,
+      "epoch": 0.92,
+      "grad_norm": 8.017218589782715,
+      "learning_rate": 1.383675689244028e-05,
+      "loss": 1.0889,
       "step": 7367
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 14.112876892089844,
-      "learning_rate": 5.2320336774581535e-06,
-      "loss": 0.5372,
+      "epoch": 0.92,
+      "grad_norm": 12.834436416625977,
+      "learning_rate": 1.3835920177383592e-05,
+      "loss": 0.5733,
       "step": 7368
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 25.881088256835938,
-      "learning_rate": 5.230029066853764e-06,
-      "loss": 1.3301,
+      "epoch": 0.92,
+      "grad_norm": 11.939716339111328,
+      "learning_rate": 1.3835083462326906e-05,
+      "loss": 1.137,
       "step": 7369
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 28.427711486816406,
-      "learning_rate": 5.2280244562493745e-06,
-      "loss": 1.6783,
+      "epoch": 0.92,
+      "grad_norm": 15.541251182556152,
+      "learning_rate": 1.383424674727022e-05,
+      "loss": 1.0705,
       "step": 7370
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 10.222917556762695,
-      "learning_rate": 5.226019845644984e-06,
-      "loss": 0.9323,
+      "epoch": 0.93,
+      "grad_norm": 10.923202514648438,
+      "learning_rate": 1.383341003221353e-05,
+      "loss": 0.8229,
       "step": 7371
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 8.820756912231445,
-      "learning_rate": 5.224015235040594e-06,
-      "loss": 0.9893,
+      "epoch": 0.93,
+      "grad_norm": 7.435032367706299,
+      "learning_rate": 1.3832573317156843e-05,
+      "loss": 0.5059,
       "step": 7372
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 17.932085037231445,
-      "learning_rate": 5.222010624436203e-06,
-      "loss": 1.7065,
+      "epoch": 0.93,
+      "grad_norm": 19.814422607421875,
+      "learning_rate": 1.3831736602100157e-05,
+      "loss": 1.4652,
       "step": 7373
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 11.802325248718262,
-      "learning_rate": 5.220006013831813e-06,
-      "loss": 1.3393,
+      "epoch": 0.93,
+      "grad_norm": 12.966021537780762,
+      "learning_rate": 1.3830899887043467e-05,
+      "loss": 1.8792,
       "step": 7374
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 23.23512077331543,
-      "learning_rate": 5.218001403227424e-06,
-      "loss": 1.1122,
+      "epoch": 0.93,
+      "grad_norm": 27.098787307739258,
+      "learning_rate": 1.3830063171986781e-05,
+      "loss": 2.7767,
       "step": 7375
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 15.658203125,
-      "learning_rate": 5.215996792623033e-06,
-      "loss": 1.3305,
+      "epoch": 0.93,
+      "grad_norm": 14.493932723999023,
+      "learning_rate": 1.3829226456930095e-05,
+      "loss": 1.2419,
       "step": 7376
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 39.932865142822266,
-      "learning_rate": 5.213992182018643e-06,
-      "loss": 1.7015,
+      "epoch": 0.93,
+      "grad_norm": 15.625213623046875,
+      "learning_rate": 1.3828389741873407e-05,
+      "loss": 1.3691,
       "step": 7377
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 33.233642578125,
-      "learning_rate": 5.2119875714142535e-06,
-      "loss": 1.5578,
+      "epoch": 0.93,
+      "grad_norm": 18.71935272216797,
+      "learning_rate": 1.3827553026816719e-05,
+      "loss": 1.6166,
       "step": 7378
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 12.168890953063965,
-      "learning_rate": 5.209982960809863e-06,
-      "loss": 0.9835,
+      "epoch": 0.93,
+      "grad_norm": 10.90803050994873,
+      "learning_rate": 1.382671631176003e-05,
+      "loss": 1.3807,
       "step": 7379
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 9.322324752807617,
-      "learning_rate": 5.207978350205473e-06,
-      "loss": 0.8729,
+      "epoch": 0.93,
+      "grad_norm": 8.382681846618652,
+      "learning_rate": 1.3825879596703344e-05,
+      "loss": 1.7445,
       "step": 7380
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 32.88364791870117,
-      "learning_rate": 5.205973739601084e-06,
-      "loss": 1.5136,
+      "epoch": 0.93,
+      "grad_norm": 6.501852035522461,
+      "learning_rate": 1.3825042881646656e-05,
+      "loss": 1.4699,
       "step": 7381
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 24.65951156616211,
-      "learning_rate": 5.203969128996693e-06,
-      "loss": 1.2119,
+      "epoch": 0.93,
+      "grad_norm": 9.208329200744629,
+      "learning_rate": 1.3824206166589968e-05,
+      "loss": 1.6787,
       "step": 7382
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 9.384584426879883,
-      "learning_rate": 5.201964518392303e-06,
-      "loss": 0.8526,
+      "epoch": 0.93,
+      "grad_norm": 25.36642837524414,
+      "learning_rate": 1.3823369451533282e-05,
+      "loss": 3.3833,
       "step": 7383
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 16.68515396118164,
-      "learning_rate": 5.199959907787912e-06,
-      "loss": 1.6815,
+      "epoch": 0.93,
+      "grad_norm": 11.165738105773926,
+      "learning_rate": 1.3822532736476595e-05,
+      "loss": 0.2827,
       "step": 7384
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 12.431203842163086,
-      "learning_rate": 5.197955297183522e-06,
-      "loss": 0.8398,
+      "epoch": 0.93,
+      "grad_norm": 12.452773094177246,
+      "learning_rate": 1.3821696021419906e-05,
+      "loss": 1.4942,
       "step": 7385
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 10.749804496765137,
-      "learning_rate": 5.1959506865791325e-06,
-      "loss": 1.6542,
+      "epoch": 0.93,
+      "grad_norm": 18.42108154296875,
+      "learning_rate": 1.382085930636322e-05,
+      "loss": 2.2891,
       "step": 7386
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 19.671411514282227,
-      "learning_rate": 5.1939460759747425e-06,
-      "loss": 1.1745,
+      "epoch": 0.93,
+      "grad_norm": 16.08868980407715,
+      "learning_rate": 1.3820022591306533e-05,
+      "loss": 1.7406,
       "step": 7387
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 15.82059383392334,
-      "learning_rate": 5.191941465370353e-06,
-      "loss": 1.5952,
+      "epoch": 0.93,
+      "grad_norm": 21.774072647094727,
+      "learning_rate": 1.3819185876249843e-05,
+      "loss": 1.0848,
       "step": 7388
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 11.592869758605957,
-      "learning_rate": 5.189936854765963e-06,
-      "loss": 1.375,
+      "epoch": 0.93,
+      "grad_norm": 11.0586519241333,
+      "learning_rate": 1.3818349161193157e-05,
+      "loss": 1.4377,
       "step": 7389
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 12.64307975769043,
-      "learning_rate": 5.187932244161572e-06,
-      "loss": 1.5403,
+      "epoch": 0.93,
+      "grad_norm": 8.787914276123047,
+      "learning_rate": 1.3817512446136469e-05,
+      "loss": 1.1182,
       "step": 7390
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 21.732891082763672,
-      "learning_rate": 5.185927633557182e-06,
-      "loss": 0.9123,
+      "epoch": 0.93,
+      "grad_norm": 28.798927307128906,
+      "learning_rate": 1.3816675731079782e-05,
+      "loss": 1.9551,
       "step": 7391
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 11.803874969482422,
-      "learning_rate": 5.183923022952791e-06,
-      "loss": 1.1493,
+      "epoch": 0.93,
+      "grad_norm": 20.054466247558594,
+      "learning_rate": 1.3815839016023094e-05,
+      "loss": 1.446,
       "step": 7392
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 13.172371864318848,
-      "learning_rate": 5.181918412348401e-06,
-      "loss": 1.5287,
+      "epoch": 0.93,
+      "grad_norm": 14.390433311462402,
+      "learning_rate": 1.3815002300966406e-05,
+      "loss": 1.6343,
       "step": 7393
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 11.935776710510254,
-      "learning_rate": 5.179913801744012e-06,
-      "loss": 1.0215,
+      "epoch": 0.93,
+      "grad_norm": 27.4454345703125,
+      "learning_rate": 1.381416558590972e-05,
+      "loss": 1.9062,
       "step": 7394
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 41.79684829711914,
-      "learning_rate": 5.1779091911396215e-06,
-      "loss": 2.0082,
+      "epoch": 0.93,
+      "grad_norm": 22.983104705810547,
+      "learning_rate": 1.3813328870853032e-05,
+      "loss": 1.524,
       "step": 7395
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 20.29804229736328,
-      "learning_rate": 5.175904580535232e-06,
-      "loss": 1.9343,
+      "epoch": 0.93,
+      "grad_norm": 18.85066032409668,
+      "learning_rate": 1.3812492155796344e-05,
+      "loss": 1.1134,
       "step": 7396
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 13.398682594299316,
-      "learning_rate": 5.173899969930841e-06,
-      "loss": 1.1503,
+      "epoch": 0.93,
+      "grad_norm": 31.238264083862305,
+      "learning_rate": 1.3811655440739658e-05,
+      "loss": 2.4208,
       "step": 7397
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 24.813034057617188,
-      "learning_rate": 5.171895359326451e-06,
-      "loss": 1.0149,
+      "epoch": 0.93,
+      "grad_norm": 8.082986831665039,
+      "learning_rate": 1.3810818725682971e-05,
+      "loss": 0.1792,
       "step": 7398
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 23.43954849243164,
-      "learning_rate": 5.169890748722061e-06,
-      "loss": 1.2993,
+      "epoch": 0.93,
+      "grad_norm": 6.717072486877441,
+      "learning_rate": 1.3809982010626281e-05,
+      "loss": 1.0453,
       "step": 7399
     },
     {
-      "epoch": 2.22,
-      "grad_norm": 14.955592155456543,
-      "learning_rate": 5.167886138117671e-06,
-      "loss": 1.3491,
+      "epoch": 0.93,
+      "grad_norm": 8.783857345581055,
+      "learning_rate": 1.3809145295569595e-05,
+      "loss": 0.2954,
       "step": 7400
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 15.617820739746094,
-      "learning_rate": 5.165881527513281e-06,
-      "loss": 1.1608,
+      "epoch": 0.93,
+      "grad_norm": 13.640789031982422,
+      "learning_rate": 1.3808308580512909e-05,
+      "loss": 0.9781,
       "step": 7401
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 18.086448669433594,
-      "learning_rate": 5.163876916908891e-06,
-      "loss": 1.4179,
+      "epoch": 0.93,
+      "grad_norm": 10.328807830810547,
+      "learning_rate": 1.3807471865456219e-05,
+      "loss": 0.3612,
       "step": 7402
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 7.104720115661621,
-      "learning_rate": 5.1618723063045005e-06,
-      "loss": 0.8867,
+      "epoch": 0.93,
+      "grad_norm": 17.736278533935547,
+      "learning_rate": 1.3806635150399533e-05,
+      "loss": 0.8795,
       "step": 7403
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 15.77996826171875,
-      "learning_rate": 5.159867695700111e-06,
-      "loss": 1.276,
+      "epoch": 0.93,
+      "grad_norm": 20.107175827026367,
+      "learning_rate": 1.3805798435342845e-05,
+      "loss": 2.9603,
       "step": 7404
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 28.867000579833984,
-      "learning_rate": 5.15786308509572e-06,
-      "loss": 1.6279,
+      "epoch": 0.93,
+      "grad_norm": 33.18934631347656,
+      "learning_rate": 1.3804961720286158e-05,
+      "loss": 1.4749,
       "step": 7405
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 26.46651268005371,
-      "learning_rate": 5.155858474491331e-06,
-      "loss": 1.2349,
+      "epoch": 0.93,
+      "grad_norm": 11.525893211364746,
+      "learning_rate": 1.380412500522947e-05,
+      "loss": 0.6962,
       "step": 7406
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 15.890294075012207,
-      "learning_rate": 5.153853863886941e-06,
-      "loss": 2.3934,
+      "epoch": 0.93,
+      "grad_norm": 16.433094024658203,
+      "learning_rate": 1.3803288290172782e-05,
+      "loss": 1.3322,
       "step": 7407
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 37.84567642211914,
-      "learning_rate": 5.15184925328255e-06,
-      "loss": 1.5725,
+      "epoch": 0.93,
+      "grad_norm": 22.894237518310547,
+      "learning_rate": 1.3802451575116096e-05,
+      "loss": 4.3966,
       "step": 7408
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 13.57693862915039,
-      "learning_rate": 5.14984464267816e-06,
-      "loss": 1.4944,
+      "epoch": 0.93,
+      "grad_norm": 11.382543563842773,
+      "learning_rate": 1.3801614860059406e-05,
+      "loss": 3.1286,
       "step": 7409
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 16.91181755065918,
-      "learning_rate": 5.147840032073769e-06,
-      "loss": 1.7839,
+      "epoch": 0.93,
+      "grad_norm": 38.796470642089844,
+      "learning_rate": 1.380077814500272e-05,
+      "loss": 1.9953,
       "step": 7410
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 36.64265441894531,
-      "learning_rate": 5.1458354214693795e-06,
-      "loss": 2.3358,
+      "epoch": 0.93,
+      "grad_norm": 13.195964813232422,
+      "learning_rate": 1.3799941429946033e-05,
+      "loss": 2.0658,
       "step": 7411
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 5.110405921936035,
-      "learning_rate": 5.14383081086499e-06,
-      "loss": 0.6339,
+      "epoch": 0.93,
+      "grad_norm": 11.612751960754395,
+      "learning_rate": 1.3799104714889347e-05,
+      "loss": 2.575,
       "step": 7412
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 13.546368598937988,
-      "learning_rate": 5.1418262002606005e-06,
-      "loss": 2.2864,
+      "epoch": 0.93,
+      "grad_norm": 22.106855392456055,
+      "learning_rate": 1.3798267999832657e-05,
+      "loss": 0.7899,
       "step": 7413
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 13.818781852722168,
-      "learning_rate": 5.13982158965621e-06,
-      "loss": 1.8962,
+      "epoch": 0.93,
+      "grad_norm": 7.899169921875,
+      "learning_rate": 1.3797431284775971e-05,
+      "loss": 0.7671,
       "step": 7414
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 8.545878410339355,
-      "learning_rate": 5.13781697905182e-06,
-      "loss": 1.9361,
+      "epoch": 0.93,
+      "grad_norm": 33.73710632324219,
+      "learning_rate": 1.3796594569719285e-05,
+      "loss": 1.829,
       "step": 7415
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 22.833742141723633,
-      "learning_rate": 5.135812368447429e-06,
-      "loss": 1.869,
+      "epoch": 0.93,
+      "grad_norm": 27.88739776611328,
+      "learning_rate": 1.3795757854662595e-05,
+      "loss": 1.6476,
       "step": 7416
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 48.628639221191406,
-      "learning_rate": 5.133807757843039e-06,
-      "loss": 1.6338,
+      "epoch": 0.93,
+      "grad_norm": 16.797008514404297,
+      "learning_rate": 1.3794921139605908e-05,
+      "loss": 1.0908,
       "step": 7417
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 18.07792091369629,
-      "learning_rate": 5.13180314723865e-06,
-      "loss": 1.0774,
+      "epoch": 0.93,
+      "grad_norm": 12.012316703796387,
+      "learning_rate": 1.379408442454922e-05,
+      "loss": 1.7921,
       "step": 7418
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 12.404952049255371,
-      "learning_rate": 5.129798536634259e-06,
-      "loss": 1.8888,
+      "epoch": 0.93,
+      "grad_norm": 17.799327850341797,
+      "learning_rate": 1.3793247709492534e-05,
+      "loss": 2.1023,
       "step": 7419
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 16.634307861328125,
-      "learning_rate": 5.127793926029869e-06,
-      "loss": 0.9383,
+      "epoch": 0.93,
+      "grad_norm": 10.95718765258789,
+      "learning_rate": 1.3792410994435846e-05,
+      "loss": 0.8405,
       "step": 7420
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 13.299745559692383,
-      "learning_rate": 5.125789315425479e-06,
-      "loss": 1.4448,
+      "epoch": 0.93,
+      "grad_norm": 18.754972457885742,
+      "learning_rate": 1.3791574279379158e-05,
+      "loss": 2.3038,
       "step": 7421
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 20.849573135375977,
-      "learning_rate": 5.123784704821089e-06,
-      "loss": 1.8726,
+      "epoch": 0.93,
+      "grad_norm": 12.339506149291992,
+      "learning_rate": 1.3790737564322472e-05,
+      "loss": 0.9753,
       "step": 7422
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 69.56287384033203,
-      "learning_rate": 5.121780094216699e-06,
-      "loss": 2.2076,
+      "epoch": 0.93,
+      "grad_norm": 12.315550804138184,
+      "learning_rate": 1.3789900849265782e-05,
+      "loss": 1.1979,
       "step": 7423
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 13.498839378356934,
-      "learning_rate": 5.119775483612309e-06,
-      "loss": 0.9413,
+      "epoch": 0.93,
+      "grad_norm": 8.54345989227295,
+      "learning_rate": 1.3789064134209096e-05,
+      "loss": 0.7829,
       "step": 7424
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 18.645475387573242,
-      "learning_rate": 5.117770873007919e-06,
-      "loss": 1.2546,
+      "epoch": 0.93,
+      "grad_norm": 14.872376441955566,
+      "learning_rate": 1.378822741915241e-05,
+      "loss": 1.2126,
       "step": 7425
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 24.639263153076172,
-      "learning_rate": 5.115766262403529e-06,
-      "loss": 1.5101,
+      "epoch": 0.93,
+      "grad_norm": 25.59175682067871,
+      "learning_rate": 1.3787390704095723e-05,
+      "loss": 1.7746,
       "step": 7426
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 9.19277572631836,
-      "learning_rate": 5.113761651799138e-06,
-      "loss": 0.789,
+      "epoch": 0.93,
+      "grad_norm": 19.434415817260742,
+      "learning_rate": 1.3786553989039033e-05,
+      "loss": 1.5769,
       "step": 7427
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 13.205931663513184,
-      "learning_rate": 5.111757041194748e-06,
-      "loss": 1.2832,
+      "epoch": 0.93,
+      "grad_norm": 7.455471038818359,
+      "learning_rate": 1.3785717273982347e-05,
+      "loss": 0.4915,
       "step": 7428
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 26.456480026245117,
-      "learning_rate": 5.109752430590358e-06,
-      "loss": 2.1971,
+      "epoch": 0.93,
+      "grad_norm": 7.419821262359619,
+      "learning_rate": 1.378488055892566e-05,
+      "loss": 1.1604,
       "step": 7429
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 16.407329559326172,
-      "learning_rate": 5.1077478199859685e-06,
-      "loss": 1.7726,
+      "epoch": 0.93,
+      "grad_norm": 24.802993774414062,
+      "learning_rate": 1.378404384386897e-05,
+      "loss": 1.9481,
       "step": 7430
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 24.08245277404785,
-      "learning_rate": 5.105743209381579e-06,
-      "loss": 1.5051,
+      "epoch": 0.93,
+      "grad_norm": 16.577354431152344,
+      "learning_rate": 1.3783207128812284e-05,
+      "loss": 1.7758,
       "step": 7431
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 42.648616790771484,
-      "learning_rate": 5.103738598777188e-06,
-      "loss": 1.9396,
+      "epoch": 0.93,
+      "grad_norm": 15.920984268188477,
+      "learning_rate": 1.3782370413755596e-05,
+      "loss": 1.6791,
       "step": 7432
     },
     {
-      "epoch": 2.23,
-      "grad_norm": 21.788301467895508,
-      "learning_rate": 5.101733988172798e-06,
-      "loss": 1.0098,
+      "epoch": 0.93,
+      "grad_norm": 23.824922561645508,
+      "learning_rate": 1.378153369869891e-05,
+      "loss": 2.0636,
       "step": 7433
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 19.175148010253906,
-      "learning_rate": 5.099729377568407e-06,
-      "loss": 1.4408,
+      "epoch": 0.93,
+      "grad_norm": 13.970311164855957,
+      "learning_rate": 1.3780696983642222e-05,
+      "loss": 1.3183,
       "step": 7434
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 33.719520568847656,
-      "learning_rate": 5.097724766964017e-06,
-      "loss": 1.6809,
+      "epoch": 0.93,
+      "grad_norm": 9.506698608398438,
+      "learning_rate": 1.3779860268585534e-05,
+      "loss": 0.5083,
       "step": 7435
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 22.389249801635742,
-      "learning_rate": 5.095720156359627e-06,
-      "loss": 1.5647,
+      "epoch": 0.93,
+      "grad_norm": 13.299325942993164,
+      "learning_rate": 1.3779023553528847e-05,
+      "loss": 1.4974,
       "step": 7436
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 21.30531120300293,
-      "learning_rate": 5.0937155457552374e-06,
-      "loss": 1.5027,
+      "epoch": 0.93,
+      "grad_norm": 52.766910552978516,
+      "learning_rate": 1.3778186838472158e-05,
+      "loss": 0.9813,
       "step": 7437
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 26.76338768005371,
-      "learning_rate": 5.0917109351508475e-06,
-      "loss": 1.3216,
+      "epoch": 0.93,
+      "grad_norm": 10.32703971862793,
+      "learning_rate": 1.3777350123415471e-05,
+      "loss": 0.9801,
       "step": 7438
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 13.98495101928711,
-      "learning_rate": 5.089706324546458e-06,
-      "loss": 1.1699,
+      "epoch": 0.93,
+      "grad_norm": 8.354700088500977,
+      "learning_rate": 1.3776513408358785e-05,
+      "loss": 0.8647,
       "step": 7439
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 6.971924304962158,
-      "learning_rate": 5.087701713942067e-06,
-      "loss": 0.8873,
-      "step": 7440
-    },
-    {
-      "epoch": 2.24,
-      "eval_loss": 0.1698659360408783,
-      "eval_runtime": 43.7492,
-      "eval_samples_per_second": 33.806,
-      "eval_steps_per_second": 33.806,
+      "epoch": 0.93,
+      "grad_norm": 29.22992515563965,
+      "learning_rate": 1.3775676693302099e-05,
+      "loss": 1.803,
       "step": 7440
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 15.718365669250488,
-      "learning_rate": 5.085697103337677e-06,
-      "loss": 1.535,
+      "epoch": 0.93,
+      "grad_norm": 13.670393943786621,
+      "learning_rate": 1.3774839978245409e-05,
+      "loss": 1.4286,
       "step": 7441
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 23.17116928100586,
-      "learning_rate": 5.083692492733286e-06,
-      "loss": 1.6606,
+      "epoch": 0.93,
+      "grad_norm": 10.728265762329102,
+      "learning_rate": 1.3774003263188723e-05,
+      "loss": 0.8971,
       "step": 7442
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 17.73175621032715,
-      "learning_rate": 5.081687882128897e-06,
-      "loss": 1.6878,
+      "epoch": 0.93,
+      "grad_norm": 16.651935577392578,
+      "learning_rate": 1.3773166548132035e-05,
+      "loss": 1.7557,
       "step": 7443
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 19.423385620117188,
-      "learning_rate": 5.079683271524507e-06,
-      "loss": 1.3193,
+      "epoch": 0.93,
+      "grad_norm": 71.80274200439453,
+      "learning_rate": 1.3772329833075346e-05,
+      "loss": 4.096,
       "step": 7444
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 11.797534942626953,
-      "learning_rate": 5.077678660920116e-06,
-      "loss": 0.881,
+      "epoch": 0.93,
+      "grad_norm": 25.78373146057129,
+      "learning_rate": 1.377149311801866e-05,
+      "loss": 1.9425,
       "step": 7445
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 15.763280868530273,
-      "learning_rate": 5.0756740503157265e-06,
-      "loss": 1.4587,
+      "epoch": 0.93,
+      "grad_norm": 11.401162147521973,
+      "learning_rate": 1.3770656402961972e-05,
+      "loss": 1.6871,
       "step": 7446
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 8.128504753112793,
-      "learning_rate": 5.073669439711337e-06,
-      "loss": 1.0937,
+      "epoch": 0.93,
+      "grad_norm": 7.95670223236084,
+      "learning_rate": 1.3769819687905286e-05,
+      "loss": 0.9672,
       "step": 7447
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 31.446063995361328,
-      "learning_rate": 5.071664829106946e-06,
-      "loss": 1.4584,
+      "epoch": 0.93,
+      "grad_norm": 8.02076530456543,
+      "learning_rate": 1.3768982972848598e-05,
+      "loss": 0.2893,
       "step": 7448
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 13.80343246459961,
-      "learning_rate": 5.069660218502557e-06,
-      "loss": 0.7605,
+      "epoch": 0.93,
+      "grad_norm": 13.573423385620117,
+      "learning_rate": 1.376814625779191e-05,
+      "loss": 2.2369,
       "step": 7449
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 10.658504486083984,
-      "learning_rate": 5.067655607898167e-06,
-      "loss": 0.8793,
+      "epoch": 0.93,
+      "grad_norm": 10.555697441101074,
+      "learning_rate": 1.3767309542735223e-05,
+      "loss": 0.9838,
       "step": 7450
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 12.806272506713867,
-      "learning_rate": 5.065650997293776e-06,
-      "loss": 0.9486,
+      "epoch": 0.94,
+      "grad_norm": 10.912386894226074,
+      "learning_rate": 1.3766472827678534e-05,
+      "loss": 0.9101,
       "step": 7451
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 9.60067081451416,
-      "learning_rate": 5.063646386689386e-06,
-      "loss": 1.0777,
+      "epoch": 0.94,
+      "grad_norm": 7.684079170227051,
+      "learning_rate": 1.3765636112621847e-05,
+      "loss": 0.5299,
       "step": 7452
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 10.402573585510254,
-      "learning_rate": 5.061641776084995e-06,
-      "loss": 1.0899,
+      "epoch": 0.94,
+      "grad_norm": 86.8624496459961,
+      "learning_rate": 1.3764799397565161e-05,
+      "loss": 2.2287,
       "step": 7453
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 24.5078182220459,
-      "learning_rate": 5.0596371654806055e-06,
-      "loss": 1.6828,
+      "epoch": 0.94,
+      "grad_norm": 41.78707504272461,
+      "learning_rate": 1.3763962682508475e-05,
+      "loss": 3.1811,
       "step": 7454
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 20.464935302734375,
-      "learning_rate": 5.057632554876216e-06,
-      "loss": 1.5368,
+      "epoch": 0.94,
+      "grad_norm": 13.535687446594238,
+      "learning_rate": 1.3763125967451785e-05,
+      "loss": 1.2221,
       "step": 7455
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 13.39626693725586,
-      "learning_rate": 5.055627944271826e-06,
-      "loss": 0.9746,
+      "epoch": 0.94,
+      "grad_norm": 9.792207717895508,
+      "learning_rate": 1.3762289252395098e-05,
+      "loss": 1.3552,
       "step": 7456
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 18.0714168548584,
-      "learning_rate": 5.053623333667436e-06,
-      "loss": 1.0704,
+      "epoch": 0.94,
+      "grad_norm": 68.6406478881836,
+      "learning_rate": 1.376145253733841e-05,
+      "loss": 2.4895,
       "step": 7457
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 33.03631591796875,
-      "learning_rate": 5.051618723063045e-06,
-      "loss": 1.4787,
+      "epoch": 0.94,
+      "grad_norm": 15.665657043457031,
+      "learning_rate": 1.3760615822281722e-05,
+      "loss": 1.4168,
       "step": 7458
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 101.09317779541016,
-      "learning_rate": 5.049614112458655e-06,
-      "loss": 1.9596,
+      "epoch": 0.94,
+      "grad_norm": 7.849665641784668,
+      "learning_rate": 1.3759779107225036e-05,
+      "loss": 0.8555,
       "step": 7459
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 27.06547737121582,
-      "learning_rate": 5.047609501854265e-06,
-      "loss": 0.824,
+      "epoch": 0.94,
+      "grad_norm": 20.467613220214844,
+      "learning_rate": 1.3758942392168348e-05,
+      "loss": 3.5343,
       "step": 7460
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 26.54560089111328,
-      "learning_rate": 5.045604891249875e-06,
-      "loss": 1.0234,
+      "epoch": 0.94,
+      "grad_norm": 12.852072715759277,
+      "learning_rate": 1.3758105677111662e-05,
+      "loss": 0.4474,
       "step": 7461
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 29.58421516418457,
-      "learning_rate": 5.043600280645485e-06,
-      "loss": 1.8553,
+      "epoch": 0.94,
+      "grad_norm": 30.18389320373535,
+      "learning_rate": 1.3757268962054972e-05,
+      "loss": 1.0194,
       "step": 7462
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 11.534109115600586,
-      "learning_rate": 5.041595670041095e-06,
-      "loss": 0.9154,
+      "epoch": 0.94,
+      "grad_norm": 3.9626660346984863,
+      "learning_rate": 1.3756432246998285e-05,
+      "loss": 0.2651,
       "step": 7463
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 23.17076301574707,
-      "learning_rate": 5.039591059436705e-06,
-      "loss": 2.0129,
+      "epoch": 0.94,
+      "grad_norm": 23.92144775390625,
+      "learning_rate": 1.3755595531941599e-05,
+      "loss": 2.8444,
       "step": 7464
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 19.68193244934082,
-      "learning_rate": 5.037586448832315e-06,
-      "loss": 1.0759,
+      "epoch": 0.94,
+      "grad_norm": 13.733891487121582,
+      "learning_rate": 1.375475881688491e-05,
+      "loss": 2.9688,
       "step": 7465
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 21.87065887451172,
-      "learning_rate": 5.035581838227924e-06,
-      "loss": 0.9128,
+      "epoch": 0.94,
+      "grad_norm": 9.05319595336914,
+      "learning_rate": 1.3753922101828223e-05,
+      "loss": 1.8021,
       "step": 7466
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 21.481231689453125,
-      "learning_rate": 5.033577227623535e-06,
-      "loss": 1.7024,
+      "epoch": 0.94,
+      "grad_norm": 9.432863235473633,
+      "learning_rate": 1.3753085386771537e-05,
+      "loss": 0.703,
       "step": 7467
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 11.491585731506348,
-      "learning_rate": 5.031572617019145e-06,
-      "loss": 1.0131,
+      "epoch": 0.94,
+      "grad_norm": 16.356163024902344,
+      "learning_rate": 1.375224867171485e-05,
+      "loss": 1.6557,
       "step": 7468
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 14.809111595153809,
-      "learning_rate": 5.029568006414754e-06,
-      "loss": 1.3544,
+      "epoch": 0.94,
+      "grad_norm": 19.758914947509766,
+      "learning_rate": 1.375141195665816e-05,
+      "loss": 2.7752,
       "step": 7469
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 31.751813888549805,
-      "learning_rate": 5.027563395810364e-06,
-      "loss": 1.477,
+      "epoch": 0.94,
+      "grad_norm": 12.722827911376953,
+      "learning_rate": 1.3750575241601474e-05,
+      "loss": 2.161,
       "step": 7470
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 24.474321365356445,
-      "learning_rate": 5.025558785205974e-06,
-      "loss": 1.5531,
+      "epoch": 0.94,
+      "grad_norm": 12.439787864685059,
+      "learning_rate": 1.3749738526544786e-05,
+      "loss": 0.9437,
       "step": 7471
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 182.55990600585938,
-      "learning_rate": 5.023554174601584e-06,
-      "loss": 2.9409,
+      "epoch": 0.94,
+      "grad_norm": 6.628915309906006,
+      "learning_rate": 1.3748901811488098e-05,
+      "loss": 1.4233,
       "step": 7472
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 17.08757972717285,
-      "learning_rate": 5.0215495639971945e-06,
-      "loss": 1.093,
+      "epoch": 0.94,
+      "grad_norm": 55.125999450683594,
+      "learning_rate": 1.3748065096431412e-05,
+      "loss": 3.3859,
       "step": 7473
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 18.484947204589844,
-      "learning_rate": 5.019544953392805e-06,
-      "loss": 1.7707,
+      "epoch": 0.94,
+      "grad_norm": 15.76113510131836,
+      "learning_rate": 1.3747228381374724e-05,
+      "loss": 1.7849,
       "step": 7474
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 14.49228572845459,
-      "learning_rate": 5.017540342788414e-06,
-      "loss": 1.4408,
+      "epoch": 0.94,
+      "grad_norm": 12.657258033752441,
+      "learning_rate": 1.3746391666318037e-05,
+      "loss": 0.6101,
       "step": 7475
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 10.136235237121582,
-      "learning_rate": 5.015535732184024e-06,
-      "loss": 1.0075,
+      "epoch": 0.94,
+      "grad_norm": 15.167543411254883,
+      "learning_rate": 1.3745554951261348e-05,
+      "loss": 1.2373,
       "step": 7476
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 63.96528244018555,
-      "learning_rate": 5.013531121579633e-06,
-      "loss": 2.0736,
+      "epoch": 0.94,
+      "grad_norm": 17.01559829711914,
+      "learning_rate": 1.3744718236204661e-05,
+      "loss": 2.4644,
       "step": 7477
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 10.56208610534668,
-      "learning_rate": 5.011526510975243e-06,
-      "loss": 1.0639,
+      "epoch": 0.94,
+      "grad_norm": 12.297941207885742,
+      "learning_rate": 1.3743881521147975e-05,
+      "loss": 0.6074,
       "step": 7478
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 11.87157154083252,
-      "learning_rate": 5.0095219003708525e-06,
-      "loss": 1.0202,
+      "epoch": 0.94,
+      "grad_norm": 12.612531661987305,
+      "learning_rate": 1.3743044806091285e-05,
+      "loss": 1.2503,
       "step": 7479
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 11.656047821044922,
-      "learning_rate": 5.0075172897664634e-06,
-      "loss": 1.3296,
+      "epoch": 0.94,
+      "grad_norm": 9.722151756286621,
+      "learning_rate": 1.3742208091034599e-05,
+      "loss": 1.8202,
       "step": 7480
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 20.969064712524414,
-      "learning_rate": 5.0055126791620735e-06,
-      "loss": 2.0833,
+      "epoch": 0.94,
+      "grad_norm": 14.149904251098633,
+      "learning_rate": 1.3741371375977913e-05,
+      "loss": 1.3875,
       "step": 7481
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 12.001461029052734,
-      "learning_rate": 5.003508068557683e-06,
-      "loss": 1.13,
+      "epoch": 0.94,
+      "grad_norm": 7.271524429321289,
+      "learning_rate": 1.3740534660921226e-05,
+      "loss": 1.1477,
       "step": 7482
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 17.097755432128906,
-      "learning_rate": 5.001503457953293e-06,
-      "loss": 1.3623,
+      "epoch": 0.94,
+      "grad_norm": 23.882055282592773,
+      "learning_rate": 1.3739697945864536e-05,
+      "loss": 2.6625,
       "step": 7483
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 10.47364330291748,
-      "learning_rate": 4.999498847348903e-06,
-      "loss": 0.7631,
+      "epoch": 0.94,
+      "grad_norm": 17.38764762878418,
+      "learning_rate": 1.373886123080785e-05,
+      "loss": 1.6953,
       "step": 7484
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 11.359280586242676,
-      "learning_rate": 4.997494236744513e-06,
-      "loss": 0.6724,
+      "epoch": 0.94,
+      "grad_norm": 11.968402862548828,
+      "learning_rate": 1.3738024515751162e-05,
+      "loss": 1.5998,
       "step": 7485
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 10.474048614501953,
-      "learning_rate": 4.995489626140122e-06,
-      "loss": 0.5766,
+      "epoch": 0.94,
+      "grad_norm": 8.602836608886719,
+      "learning_rate": 1.3737187800694474e-05,
+      "loss": 1.1245,
       "step": 7486
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 14.837994575500488,
-      "learning_rate": 4.993485015535733e-06,
-      "loss": 1.1248,
+      "epoch": 0.94,
+      "grad_norm": 6.8582234382629395,
+      "learning_rate": 1.3736351085637788e-05,
+      "loss": 1.5747,
       "step": 7487
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 18.054821014404297,
-      "learning_rate": 4.991480404931342e-06,
-      "loss": 1.0936,
+      "epoch": 0.94,
+      "grad_norm": 21.258777618408203,
+      "learning_rate": 1.37355143705811e-05,
+      "loss": 1.6977,
       "step": 7488
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 9.805904388427734,
-      "learning_rate": 4.9894757943269525e-06,
-      "loss": 0.8442,
+      "epoch": 0.94,
+      "grad_norm": 53.06081771850586,
+      "learning_rate": 1.3734677655524413e-05,
+      "loss": 1.1751,
       "step": 7489
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 44.7296257019043,
-      "learning_rate": 4.987471183722563e-06,
-      "loss": 2.3775,
+      "epoch": 0.94,
+      "grad_norm": 41.308082580566406,
+      "learning_rate": 1.3733840940467724e-05,
+      "loss": 2.9062,
       "step": 7490
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 13.354528427124023,
-      "learning_rate": 4.985466573118173e-06,
-      "loss": 1.3387,
+      "epoch": 0.94,
+      "grad_norm": 14.199646949768066,
+      "learning_rate": 1.3733004225411037e-05,
+      "loss": 2.1871,
       "step": 7491
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 24.598257064819336,
-      "learning_rate": 4.983461962513782e-06,
-      "loss": 2.0694,
+      "epoch": 0.94,
+      "grad_norm": 14.047327041625977,
+      "learning_rate": 1.373216751035435e-05,
+      "loss": 2.9026,
       "step": 7492
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 28.982959747314453,
-      "learning_rate": 4.981457351909392e-06,
-      "loss": 1.3952,
+      "epoch": 0.94,
+      "grad_norm": 23.0893497467041,
+      "learning_rate": 1.3731330795297661e-05,
+      "loss": 1.0065,
       "step": 7493
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 16.613788604736328,
-      "learning_rate": 4.979452741305002e-06,
-      "loss": 1.1931,
+      "epoch": 0.94,
+      "grad_norm": 15.064163208007812,
+      "learning_rate": 1.3730494080240975e-05,
+      "loss": 2.4648,
       "step": 7494
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 11.446743965148926,
-      "learning_rate": 4.977448130700611e-06,
-      "loss": 0.8703,
+      "epoch": 0.94,
+      "grad_norm": 8.875805854797363,
+      "learning_rate": 1.3729657365184288e-05,
+      "loss": 1.6665,
       "step": 7495
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 14.696474075317383,
-      "learning_rate": 4.975443520096221e-06,
-      "loss": 1.3185,
+      "epoch": 0.94,
+      "grad_norm": 14.805878639221191,
+      "learning_rate": 1.37288206501276e-05,
+      "loss": 2.3215,
       "step": 7496
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 33.95481491088867,
-      "learning_rate": 4.9734389094918315e-06,
-      "loss": 1.3152,
+      "epoch": 0.94,
+      "grad_norm": 9.885756492614746,
+      "learning_rate": 1.3727983935070912e-05,
+      "loss": 1.0313,
       "step": 7497
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 9.814773559570312,
-      "learning_rate": 4.9714342988874416e-06,
-      "loss": 1.4784,
+      "epoch": 0.94,
+      "grad_norm": 18.31471824645996,
+      "learning_rate": 1.3727147220014226e-05,
+      "loss": 2.0022,
       "step": 7498
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 31.25359344482422,
-      "learning_rate": 4.969429688283051e-06,
-      "loss": 1.9966,
+      "epoch": 0.94,
+      "grad_norm": 15.420527458190918,
+      "learning_rate": 1.3726310504957538e-05,
+      "loss": 0.7821,
       "step": 7499
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 28.215229034423828,
-      "learning_rate": 4.967425077678662e-06,
-      "loss": 2.6453,
+      "epoch": 0.94,
+      "grad_norm": 7.836379051208496,
+      "learning_rate": 1.372547378990085e-05,
+      "loss": 1.149,
       "step": 7500
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 16.87133026123047,
-      "learning_rate": 4.965420467074271e-06,
-      "loss": 1.0434,
+      "epoch": 0.94,
+      "grad_norm": 9.965003967285156,
+      "learning_rate": 1.3724637074844162e-05,
+      "loss": 0.1831,
       "step": 7501
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 16.18372344970703,
-      "learning_rate": 4.963415856469881e-06,
-      "loss": 1.9418,
+      "epoch": 0.94,
+      "grad_norm": 19.751415252685547,
+      "learning_rate": 1.3723800359787475e-05,
+      "loss": 0.9685,
       "step": 7502
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 10.651266098022461,
-      "learning_rate": 4.961411245865491e-06,
-      "loss": 1.5416,
+      "epoch": 0.94,
+      "grad_norm": 22.532337188720703,
+      "learning_rate": 1.3722963644730789e-05,
+      "loss": 1.5096,
       "step": 7503
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 37.711185455322266,
-      "learning_rate": 4.959406635261101e-06,
-      "loss": 1.3423,
+      "epoch": 0.94,
+      "grad_norm": 37.046485900878906,
+      "learning_rate": 1.37221269296741e-05,
+      "loss": 1.573,
       "step": 7504
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 21.620492935180664,
-      "learning_rate": 4.9574020246567105e-06,
-      "loss": 1.435,
+      "epoch": 0.94,
+      "grad_norm": 27.697509765625,
+      "learning_rate": 1.3721290214617413e-05,
+      "loss": 1.5354,
       "step": 7505
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 52.260536193847656,
-      "learning_rate": 4.9553974140523205e-06,
-      "loss": 1.8419,
+      "epoch": 0.94,
+      "grad_norm": 8.985095024108887,
+      "learning_rate": 1.3720453499560727e-05,
+      "loss": 1.4237,
       "step": 7506
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 21.462879180908203,
-      "learning_rate": 4.953392803447931e-06,
-      "loss": 0.8966,
+      "epoch": 0.94,
+      "grad_norm": 5.892858028411865,
+      "learning_rate": 1.3719616784504037e-05,
+      "loss": 1.0621,
       "step": 7507
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 9.330449104309082,
-      "learning_rate": 4.951388192843541e-06,
-      "loss": 0.8809,
+      "epoch": 0.94,
+      "grad_norm": 14.162527084350586,
+      "learning_rate": 1.371878006944735e-05,
+      "loss": 3.3544,
       "step": 7508
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 31.02956771850586,
-      "learning_rate": 4.949383582239151e-06,
-      "loss": 1.8236,
+      "epoch": 0.94,
+      "grad_norm": 21.386547088623047,
+      "learning_rate": 1.3717943354390664e-05,
+      "loss": 2.5531,
       "step": 7509
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 19.69917869567871,
-      "learning_rate": 4.94737897163476e-06,
-      "loss": 1.4695,
+      "epoch": 0.94,
+      "grad_norm": 6.791244029998779,
+      "learning_rate": 1.3717106639333976e-05,
+      "loss": 1.5847,
       "step": 7510
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 13.709826469421387,
-      "learning_rate": 4.94537436103037e-06,
-      "loss": 0.9547,
+      "epoch": 0.94,
+      "grad_norm": 15.625334739685059,
+      "learning_rate": 1.3716269924277288e-05,
+      "loss": 2.3537,
       "step": 7511
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 14.561116218566895,
-      "learning_rate": 4.94336975042598e-06,
-      "loss": 2.0196,
+      "epoch": 0.94,
+      "grad_norm": 14.458335876464844,
+      "learning_rate": 1.3715433209220602e-05,
+      "loss": 1.6641,
       "step": 7512
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 12.211623191833496,
-      "learning_rate": 4.94136513982159e-06,
-      "loss": 1.0027,
+      "epoch": 0.94,
+      "grad_norm": 7.3591389656066895,
+      "learning_rate": 1.3714596494163914e-05,
+      "loss": 0.5241,
       "step": 7513
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 94.37039947509766,
-      "learning_rate": 4.9393605292171995e-06,
-      "loss": 1.9356,
+      "epoch": 0.94,
+      "grad_norm": 11.480550765991211,
+      "learning_rate": 1.3713759779107226e-05,
+      "loss": 1.361,
       "step": 7514
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 14.454056739807129,
-      "learning_rate": 4.9373559186128105e-06,
-      "loss": 1.0738,
+      "epoch": 0.94,
+      "grad_norm": 19.844825744628906,
+      "learning_rate": 1.3712923064050538e-05,
+      "loss": 1.3203,
       "step": 7515
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 88.53865051269531,
-      "learning_rate": 4.93535130800842e-06,
-      "loss": 1.7283,
+      "epoch": 0.94,
+      "grad_norm": 32.37255096435547,
+      "learning_rate": 1.3712086348993851e-05,
+      "loss": 2.0873,
       "step": 7516
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 225.10995483398438,
-      "learning_rate": 4.93334669740403e-06,
-      "loss": 1.8414,
+      "epoch": 0.94,
+      "grad_norm": 11.962457656860352,
+      "learning_rate": 1.3711249633937165e-05,
+      "loss": 1.2032,
       "step": 7517
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 16.295265197753906,
-      "learning_rate": 4.93134208679964e-06,
-      "loss": 1.7194,
+      "epoch": 0.94,
+      "grad_norm": 11.72451114654541,
+      "learning_rate": 1.3710412918880475e-05,
+      "loss": 1.5265,
       "step": 7518
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 102.90641021728516,
-      "learning_rate": 4.929337476195249e-06,
-      "loss": 1.1804,
+      "epoch": 0.94,
+      "grad_norm": 11.540800094604492,
+      "learning_rate": 1.3709576203823789e-05,
+      "loss": 1.6858,
       "step": 7519
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 22.84174156188965,
-      "learning_rate": 4.927332865590859e-06,
-      "loss": 1.0631,
+      "epoch": 0.94,
+      "grad_norm": 4.787374973297119,
+      "learning_rate": 1.3708739488767102e-05,
+      "loss": 0.5134,
       "step": 7520
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 29.34662628173828,
-      "learning_rate": 4.925328254986469e-06,
-      "loss": 1.2535,
+      "epoch": 0.94,
+      "grad_norm": 20.71868133544922,
+      "learning_rate": 1.3707902773710413e-05,
+      "loss": 1.6976,
       "step": 7521
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 11.276557922363281,
-      "learning_rate": 4.923323644382079e-06,
-      "loss": 1.1659,
+      "epoch": 0.94,
+      "grad_norm": 5.886478424072266,
+      "learning_rate": 1.3707066058653726e-05,
+      "loss": 0.7904,
       "step": 7522
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 14.418801307678223,
-      "learning_rate": 4.921319033777689e-06,
-      "loss": 1.9511,
+      "epoch": 0.94,
+      "grad_norm": 21.81075096130371,
+      "learning_rate": 1.370622934359704e-05,
+      "loss": 3.5599,
       "step": 7523
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 20.165433883666992,
-      "learning_rate": 4.9193144231732995e-06,
-      "loss": 1.258,
+      "epoch": 0.94,
+      "grad_norm": 9.835027694702148,
+      "learning_rate": 1.3705392628540352e-05,
+      "loss": 0.9441,
       "step": 7524
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 15.483535766601562,
-      "learning_rate": 4.917309812568909e-06,
-      "loss": 1.0741,
+      "epoch": 0.94,
+      "grad_norm": 21.744504928588867,
+      "learning_rate": 1.3704555913483664e-05,
+      "loss": 2.4696,
       "step": 7525
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 23.793718338012695,
-      "learning_rate": 4.915305201964519e-06,
-      "loss": 1.4441,
+      "epoch": 0.94,
+      "grad_norm": 52.457340240478516,
+      "learning_rate": 1.3703719198426978e-05,
+      "loss": 1.0649,
       "step": 7526
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 23.1822452545166,
-      "learning_rate": 4.913300591360129e-06,
-      "loss": 1.3305,
+      "epoch": 0.94,
+      "grad_norm": 13.692154884338379,
+      "learning_rate": 1.370288248337029e-05,
+      "loss": 0.7034,
       "step": 7527
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 18.903297424316406,
-      "learning_rate": 4.911295980755739e-06,
-      "loss": 1.1903,
+      "epoch": 0.94,
+      "grad_norm": 17.706499099731445,
+      "learning_rate": 1.3702045768313602e-05,
+      "loss": 2.0586,
       "step": 7528
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 18.78779411315918,
-      "learning_rate": 4.909291370151348e-06,
-      "loss": 1.7469,
+      "epoch": 0.94,
+      "grad_norm": 7.35901403427124,
+      "learning_rate": 1.3701209053256913e-05,
+      "loss": 1.4902,
       "step": 7529
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 15.670270919799805,
-      "learning_rate": 4.907286759546958e-06,
-      "loss": 1.1513,
+      "epoch": 0.95,
+      "grad_norm": 9.223427772521973,
+      "learning_rate": 1.3700372338200227e-05,
+      "loss": 1.7911,
       "step": 7530
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 22.992828369140625,
-      "learning_rate": 4.905282148942568e-06,
-      "loss": 0.9966,
+      "epoch": 0.95,
+      "grad_norm": 21.43293571472168,
+      "learning_rate": 1.369953562314354e-05,
+      "loss": 1.225,
       "step": 7531
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 16.692344665527344,
-      "learning_rate": 4.9032775383381785e-06,
-      "loss": 1.3511,
+      "epoch": 0.95,
+      "grad_norm": 14.161319732666016,
+      "learning_rate": 1.3698698908086851e-05,
+      "loss": 1.5195,
       "step": 7532
     },
     {
-      "epoch": 2.26,
-      "grad_norm": 17.386524200439453,
-      "learning_rate": 4.901272927733789e-06,
-      "loss": 1.4587,
+      "epoch": 0.95,
+      "grad_norm": 16.956146240234375,
+      "learning_rate": 1.3697862193030165e-05,
+      "loss": 1.6587,
       "step": 7533
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 18.931062698364258,
-      "learning_rate": 4.899268317129398e-06,
-      "loss": 2.273,
+      "epoch": 0.95,
+      "grad_norm": 3.7091715335845947,
+      "learning_rate": 1.3697025477973478e-05,
+      "loss": 0.1733,
       "step": 7534
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 86.52537536621094,
-      "learning_rate": 4.897263706525008e-06,
-      "loss": 3.4529,
+      "epoch": 0.95,
+      "grad_norm": 5.567541599273682,
+      "learning_rate": 1.3696188762916789e-05,
+      "loss": 0.3337,
       "step": 7535
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 9.628124237060547,
-      "learning_rate": 4.895259095920617e-06,
-      "loss": 0.8626,
+      "epoch": 0.95,
+      "grad_norm": 16.873165130615234,
+      "learning_rate": 1.3695352047860102e-05,
+      "loss": 1.8054,
       "step": 7536
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 10.610788345336914,
-      "learning_rate": 4.893254485316228e-06,
-      "loss": 1.2612,
+      "epoch": 0.95,
+      "grad_norm": 10.89290714263916,
+      "learning_rate": 1.3694515332803416e-05,
+      "loss": 0.8305,
       "step": 7537
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 15.068879127502441,
-      "learning_rate": 4.891249874711837e-06,
-      "loss": 1.2702,
+      "epoch": 0.95,
+      "grad_norm": 43.21706008911133,
+      "learning_rate": 1.3693678617746728e-05,
+      "loss": 1.1597,
       "step": 7538
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 25.841157913208008,
-      "learning_rate": 4.889245264107447e-06,
-      "loss": 1.5596,
+      "epoch": 0.95,
+      "grad_norm": 5.436285972595215,
+      "learning_rate": 1.369284190269004e-05,
+      "loss": 0.44,
       "step": 7539
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 13.19726276397705,
-      "learning_rate": 4.8872406535030575e-06,
-      "loss": 1.1938,
+      "epoch": 0.95,
+      "grad_norm": 13.519515037536621,
+      "learning_rate": 1.3692005187633353e-05,
+      "loss": 1.7031,
       "step": 7540
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 30.624975204467773,
-      "learning_rate": 4.8852360428986676e-06,
-      "loss": 2.6475,
+      "epoch": 0.95,
+      "grad_norm": 21.97677993774414,
+      "learning_rate": 1.3691168472576665e-05,
+      "loss": 1.8174,
       "step": 7541
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 17.774322509765625,
-      "learning_rate": 4.883231432294277e-06,
-      "loss": 0.8152,
+      "epoch": 0.95,
+      "grad_norm": 35.823272705078125,
+      "learning_rate": 1.3690331757519977e-05,
+      "loss": 2.8414,
       "step": 7542
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 13.231738090515137,
-      "learning_rate": 4.881226821689887e-06,
-      "loss": 1.0322,
+      "epoch": 0.95,
+      "grad_norm": 10.470995903015137,
+      "learning_rate": 1.368949504246329e-05,
+      "loss": 1.1076,
       "step": 7543
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 18.530324935913086,
-      "learning_rate": 4.879222211085497e-06,
-      "loss": 1.6454,
+      "epoch": 0.95,
+      "grad_norm": 19.127220153808594,
+      "learning_rate": 1.3688658327406603e-05,
+      "loss": 1.3538,
       "step": 7544
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 32.340023040771484,
-      "learning_rate": 4.877217600481107e-06,
-      "loss": 1.1846,
+      "epoch": 0.95,
+      "grad_norm": 57.466224670410156,
+      "learning_rate": 1.3687821612349917e-05,
+      "loss": 2.9337,
       "step": 7545
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 13.608917236328125,
-      "learning_rate": 4.875212989876717e-06,
-      "loss": 1.307,
+      "epoch": 0.95,
+      "grad_norm": 35.5211067199707,
+      "learning_rate": 1.3686984897293227e-05,
+      "loss": 2.0792,
       "step": 7546
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 33.83692932128906,
-      "learning_rate": 4.873208379272326e-06,
-      "loss": 1.4187,
+      "epoch": 0.95,
+      "grad_norm": 5.022341728210449,
+      "learning_rate": 1.368614818223654e-05,
+      "loss": 0.4209,
       "step": 7547
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 41.42520523071289,
-      "learning_rate": 4.8712037686679365e-06,
-      "loss": 1.5752,
+      "epoch": 0.95,
+      "grad_norm": 13.89211654663086,
+      "learning_rate": 1.3685311467179854e-05,
+      "loss": 1.5118,
       "step": 7548
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 7.894377708435059,
-      "learning_rate": 4.8691991580635466e-06,
-      "loss": 1.1328,
+      "epoch": 0.95,
+      "grad_norm": 12.629494667053223,
+      "learning_rate": 1.3684474752123164e-05,
+      "loss": 1.1239,
       "step": 7549
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 22.25946807861328,
-      "learning_rate": 4.867194547459157e-06,
-      "loss": 2.0451,
+      "epoch": 0.95,
+      "grad_norm": 16.415287017822266,
+      "learning_rate": 1.3683638037066478e-05,
+      "loss": 2.3969,
       "step": 7550
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 49.385562896728516,
-      "learning_rate": 4.865189936854766e-06,
-      "loss": 2.2722,
+      "epoch": 0.95,
+      "grad_norm": 22.39629554748535,
+      "learning_rate": 1.3682801322009792e-05,
+      "loss": 2.0676,
       "step": 7551
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 13.534929275512695,
-      "learning_rate": 4.863185326250377e-06,
-      "loss": 0.7596,
+      "epoch": 0.95,
+      "grad_norm": 17.810590744018555,
+      "learning_rate": 1.3681964606953104e-05,
+      "loss": 2.9491,
       "step": 7552
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 20.882999420166016,
-      "learning_rate": 4.861180715645986e-06,
-      "loss": 1.9043,
+      "epoch": 0.95,
+      "grad_norm": 12.154153823852539,
+      "learning_rate": 1.3681127891896416e-05,
+      "loss": 0.3778,
       "step": 7553
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 9.814292907714844,
-      "learning_rate": 4.859176105041596e-06,
-      "loss": 1.0012,
+      "epoch": 0.95,
+      "grad_norm": 7.335411071777344,
+      "learning_rate": 1.3680291176839728e-05,
+      "loss": 1.0368,
       "step": 7554
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 23.16642951965332,
-      "learning_rate": 4.857171494437206e-06,
-      "loss": 1.4161,
+      "epoch": 0.95,
+      "grad_norm": 37.25257873535156,
+      "learning_rate": 1.3679454461783041e-05,
+      "loss": 1.8574,
       "step": 7555
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 41.616703033447266,
-      "learning_rate": 4.8551668838328154e-06,
-      "loss": 2.3166,
+      "epoch": 0.95,
+      "grad_norm": 12.6161470413208,
+      "learning_rate": 1.3678617746726353e-05,
+      "loss": 1.5904,
       "step": 7556
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 14.239105224609375,
-      "learning_rate": 4.8531622732284255e-06,
-      "loss": 1.1065,
+      "epoch": 0.95,
+      "grad_norm": 11.606048583984375,
+      "learning_rate": 1.3677781031669665e-05,
+      "loss": 0.53,
       "step": 7557
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 8.06943130493164,
-      "learning_rate": 4.851157662624036e-06,
-      "loss": 0.6586,
+      "epoch": 0.95,
+      "grad_norm": 14.20860767364502,
+      "learning_rate": 1.3676944316612979e-05,
+      "loss": 1.2397,
       "step": 7558
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 16.735288619995117,
-      "learning_rate": 4.849153052019646e-06,
-      "loss": 0.4483,
+      "epoch": 0.95,
+      "grad_norm": 9.279431343078613,
+      "learning_rate": 1.3676107601556292e-05,
+      "loss": 1.2488,
       "step": 7559
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 13.807044982910156,
-      "learning_rate": 4.847148441415255e-06,
-      "loss": 1.3318,
-      "step": 7560
-    },
-    {
-      "epoch": 2.27,
-      "eval_loss": 0.1706262081861496,
-      "eval_runtime": 43.7986,
-      "eval_samples_per_second": 33.768,
-      "eval_steps_per_second": 33.768,
+      "epoch": 0.95,
+      "grad_norm": 11.00720500946045,
+      "learning_rate": 1.3675270886499603e-05,
+      "loss": 0.8454,
       "step": 7560
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 18.019256591796875,
-      "learning_rate": 4.845143830810866e-06,
-      "loss": 1.3919,
+      "epoch": 0.95,
+      "grad_norm": 10.767986297607422,
+      "learning_rate": 1.3674434171442916e-05,
+      "loss": 1.8124,
       "step": 7561
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 16.097368240356445,
-      "learning_rate": 4.843139220206475e-06,
-      "loss": 1.5764,
+      "epoch": 0.95,
+      "grad_norm": 21.558813095092773,
+      "learning_rate": 1.367359745638623e-05,
+      "loss": 0.5925,
       "step": 7562
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 13.278281211853027,
-      "learning_rate": 4.841134609602085e-06,
-      "loss": 1.0933,
+      "epoch": 0.95,
+      "grad_norm": 12.875472068786621,
+      "learning_rate": 1.367276074132954e-05,
+      "loss": 2.0704,
       "step": 7563
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 10.068906784057617,
-      "learning_rate": 4.839129998997695e-06,
-      "loss": 0.8741,
+      "epoch": 0.95,
+      "grad_norm": 60.924991607666016,
+      "learning_rate": 1.3671924026272854e-05,
+      "loss": 1.8117,
       "step": 7564
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 16.215852737426758,
-      "learning_rate": 4.837125388393305e-06,
-      "loss": 2.1497,
+      "epoch": 0.95,
+      "grad_norm": 11.496397972106934,
+      "learning_rate": 1.3671087311216168e-05,
+      "loss": 3.8066,
       "step": 7565
     },
     {
-      "epoch": 2.27,
-      "grad_norm": 11.731410026550293,
-      "learning_rate": 4.835120777788915e-06,
-      "loss": 1.7863,
+      "epoch": 0.95,
+      "grad_norm": 9.729010581970215,
+      "learning_rate": 1.367025059615948e-05,
+      "loss": 1.7768,
       "step": 7566
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 10.969109535217285,
-      "learning_rate": 4.833116167184525e-06,
-      "loss": 0.8401,
+      "epoch": 0.95,
+      "grad_norm": 14.498326301574707,
+      "learning_rate": 1.3669413881102791e-05,
+      "loss": 1.6837,
       "step": 7567
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 10.927252769470215,
-      "learning_rate": 4.831111556580135e-06,
-      "loss": 1.4045,
+      "epoch": 0.95,
+      "grad_norm": 11.079991340637207,
+      "learning_rate": 1.3668577166046103e-05,
+      "loss": 2.1257,
       "step": 7568
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 41.4652214050293,
-      "learning_rate": 4.829106945975745e-06,
-      "loss": 3.292,
+      "epoch": 0.95,
+      "grad_norm": 16.234020233154297,
+      "learning_rate": 1.3667740450989417e-05,
+      "loss": 1.7121,
       "step": 7569
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 8.78602123260498,
-      "learning_rate": 4.827102335371355e-06,
-      "loss": 1.6186,
+      "epoch": 0.95,
+      "grad_norm": 7.767157077789307,
+      "learning_rate": 1.3666903735932729e-05,
+      "loss": 0.791,
       "step": 7570
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 30.659353256225586,
-      "learning_rate": 4.825097724766964e-06,
-      "loss": 1.3395,
+      "epoch": 0.95,
+      "grad_norm": 10.612770080566406,
+      "learning_rate": 1.3666067020876041e-05,
+      "loss": 1.6558,
       "step": 7571
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 9.962449073791504,
-      "learning_rate": 4.823093114162574e-06,
-      "loss": 1.2939,
+      "epoch": 0.95,
+      "grad_norm": 17.550827026367188,
+      "learning_rate": 1.3665230305819355e-05,
+      "loss": 1.4342,
       "step": 7572
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 10.435197830200195,
-      "learning_rate": 4.821088503558184e-06,
-      "loss": 1.0703,
+      "epoch": 0.95,
+      "grad_norm": 19.115427017211914,
+      "learning_rate": 1.3664393590762668e-05,
+      "loss": 2.5511,
       "step": 7573
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 20.504138946533203,
-      "learning_rate": 4.819083892953794e-06,
-      "loss": 1.5566,
+      "epoch": 0.95,
+      "grad_norm": 26.63739585876465,
+      "learning_rate": 1.3663556875705979e-05,
+      "loss": 1.2658,
       "step": 7574
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 15.632285118103027,
-      "learning_rate": 4.817079282349404e-06,
-      "loss": 1.2294,
+      "epoch": 0.95,
+      "grad_norm": 4.140101432800293,
+      "learning_rate": 1.3662720160649292e-05,
+      "loss": 1.0531,
       "step": 7575
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 15.574999809265137,
-      "learning_rate": 4.815074671745015e-06,
-      "loss": 1.4429,
+      "epoch": 0.95,
+      "grad_norm": 9.40150260925293,
+      "learning_rate": 1.3661883445592606e-05,
+      "loss": 1.5241,
       "step": 7576
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 10.924356460571289,
-      "learning_rate": 4.813070061140624e-06,
-      "loss": 1.0845,
+      "epoch": 0.95,
+      "grad_norm": 10.271926879882812,
+      "learning_rate": 1.3661046730535916e-05,
+      "loss": 1.0597,
       "step": 7577
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 24.5418758392334,
-      "learning_rate": 4.811065450536234e-06,
-      "loss": 1.175,
+      "epoch": 0.95,
+      "grad_norm": 13.581698417663574,
+      "learning_rate": 1.366021001547923e-05,
+      "loss": 2.2879,
       "step": 7578
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 29.236740112304688,
-      "learning_rate": 4.809060839931843e-06,
-      "loss": 0.9208,
+      "epoch": 0.95,
+      "grad_norm": 11.900755882263184,
+      "learning_rate": 1.3659373300422543e-05,
+      "loss": 1.3348,
       "step": 7579
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 11.928509712219238,
-      "learning_rate": 4.807056229327453e-06,
-      "loss": 0.7309,
+      "epoch": 0.95,
+      "grad_norm": 20.219989776611328,
+      "learning_rate": 1.3658536585365855e-05,
+      "loss": 2.6691,
       "step": 7580
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 31.78061294555664,
-      "learning_rate": 4.805051618723063e-06,
-      "loss": 1.4187,
+      "epoch": 0.95,
+      "grad_norm": 21.341461181640625,
+      "learning_rate": 1.3657699870309167e-05,
+      "loss": 2.4714,
       "step": 7581
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 18.418073654174805,
-      "learning_rate": 4.803047008118673e-06,
-      "loss": 2.4135,
+      "epoch": 0.95,
+      "grad_norm": 10.618818283081055,
+      "learning_rate": 1.365686315525248e-05,
+      "loss": 0.8738,
       "step": 7582
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 27.633852005004883,
-      "learning_rate": 4.8010423975142835e-06,
-      "loss": 1.7401,
+      "epoch": 0.95,
+      "grad_norm": 12.787776947021484,
+      "learning_rate": 1.3656026440195793e-05,
+      "loss": 1.32,
       "step": 7583
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 48.656333923339844,
-      "learning_rate": 4.799037786909893e-06,
-      "loss": 1.6712,
+      "epoch": 0.95,
+      "grad_norm": 5.652602672576904,
+      "learning_rate": 1.3655189725139105e-05,
+      "loss": 0.6757,
       "step": 7584
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 15.722077369689941,
-      "learning_rate": 4.797033176305503e-06,
-      "loss": 1.2471,
+      "epoch": 0.95,
+      "grad_norm": 12.455671310424805,
+      "learning_rate": 1.3654353010082417e-05,
+      "loss": 0.6094,
       "step": 7585
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 8.848787307739258,
-      "learning_rate": 4.795028565701113e-06,
-      "loss": 0.993,
+      "epoch": 0.95,
+      "grad_norm": 9.646262168884277,
+      "learning_rate": 1.365351629502573e-05,
+      "loss": 1.026,
       "step": 7586
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 15.616233825683594,
-      "learning_rate": 4.793023955096723e-06,
-      "loss": 1.3871,
+      "epoch": 0.95,
+      "grad_norm": 14.054344177246094,
+      "learning_rate": 1.3652679579969044e-05,
+      "loss": 2.245,
       "step": 7587
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 24.031694412231445,
-      "learning_rate": 4.791019344492332e-06,
-      "loss": 1.2137,
+      "epoch": 0.95,
+      "grad_norm": 16.830785751342773,
+      "learning_rate": 1.3651842864912354e-05,
+      "loss": 2.6661,
       "step": 7588
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 11.082610130310059,
-      "learning_rate": 4.789014733887943e-06,
-      "loss": 1.1946,
+      "epoch": 0.95,
+      "grad_norm": 21.882461547851562,
+      "learning_rate": 1.3651006149855668e-05,
+      "loss": 2.4278,
       "step": 7589
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 18.813764572143555,
-      "learning_rate": 4.787010123283552e-06,
-      "loss": 1.1478,
+      "epoch": 0.95,
+      "grad_norm": 11.206304550170898,
+      "learning_rate": 1.3650169434798982e-05,
+      "loss": 1.0506,
       "step": 7590
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 20.931020736694336,
-      "learning_rate": 4.7850055126791625e-06,
-      "loss": 1.3695,
+      "epoch": 0.95,
+      "grad_norm": 16.634540557861328,
+      "learning_rate": 1.3649332719742292e-05,
+      "loss": 2.2103,
       "step": 7591
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 15.224300384521484,
-      "learning_rate": 4.7830009020747726e-06,
-      "loss": 1.0471,
+      "epoch": 0.95,
+      "grad_norm": 6.956151962280273,
+      "learning_rate": 1.3648496004685606e-05,
+      "loss": 1.965,
       "step": 7592
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 13.662450790405273,
-      "learning_rate": 4.780996291470383e-06,
-      "loss": 1.0492,
+      "epoch": 0.95,
+      "grad_norm": 11.869979858398438,
+      "learning_rate": 1.364765928962892e-05,
+      "loss": 0.9426,
       "step": 7593
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 11.05825424194336,
-      "learning_rate": 4.778991680865992e-06,
-      "loss": 1.4997,
+      "epoch": 0.95,
+      "grad_norm": 9.617378234863281,
+      "learning_rate": 1.3646822574572231e-05,
+      "loss": 0.8194,
       "step": 7594
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 14.814659118652344,
-      "learning_rate": 4.776987070261602e-06,
-      "loss": 1.5903,
+      "epoch": 0.95,
+      "grad_norm": 11.88089370727539,
+      "learning_rate": 1.3645985859515543e-05,
+      "loss": 0.4161,
       "step": 7595
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 20.72893524169922,
-      "learning_rate": 4.774982459657212e-06,
-      "loss": 0.9354,
+      "epoch": 0.95,
+      "grad_norm": 32.03239059448242,
+      "learning_rate": 1.3645149144458855e-05,
+      "loss": 1.8739,
       "step": 7596
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 16.968223571777344,
-      "learning_rate": 4.772977849052821e-06,
-      "loss": 1.7869,
+      "epoch": 0.95,
+      "grad_norm": 10.117530822753906,
+      "learning_rate": 1.3644312429402169e-05,
+      "loss": 0.7434,
       "step": 7597
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 12.797143936157227,
-      "learning_rate": 4.770973238448432e-06,
-      "loss": 1.4657,
+      "epoch": 0.95,
+      "grad_norm": 9.655041694641113,
+      "learning_rate": 1.364347571434548e-05,
+      "loss": 0.6933,
       "step": 7598
     },
     {
-      "epoch": 2.28,
-      "grad_norm": 278.7440185546875,
-      "learning_rate": 4.7689686278440414e-06,
-      "loss": 1.1522,
+      "epoch": 0.95,
+      "grad_norm": 15.120823860168457,
+      "learning_rate": 1.3642638999288793e-05,
+      "loss": 0.9372,
       "step": 7599
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 11.309410095214844,
-      "learning_rate": 4.7669640172396515e-06,
-      "loss": 1.6979,
+      "epoch": 0.95,
+      "grad_norm": 29.873746871948242,
+      "learning_rate": 1.3641802284232106e-05,
+      "loss": 2.1381,
       "step": 7600
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 17.630992889404297,
-      "learning_rate": 4.764959406635262e-06,
-      "loss": 1.2774,
+      "epoch": 0.95,
+      "eval_loss": 0.10611134022474289,
+      "eval_runtime": 154.8741,
+      "eval_samples_per_second": 22.87,
+      "eval_steps_per_second": 22.87,
+      "step": 7600
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 17.3528995513916,
+      "learning_rate": 1.364096556917542e-05,
+      "loss": 1.4973,
       "step": 7601
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 64.68533325195312,
-      "learning_rate": 4.762954796030872e-06,
-      "loss": 1.6083,
+      "epoch": 0.95,
+      "grad_norm": 24.112226486206055,
+      "learning_rate": 1.364012885411873e-05,
+      "loss": 1.4021,
       "step": 7602
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 12.072052001953125,
-      "learning_rate": 4.760950185426481e-06,
-      "loss": 0.8717,
+      "epoch": 0.95,
+      "grad_norm": 21.563318252563477,
+      "learning_rate": 1.3639292139062044e-05,
+      "loss": 1.032,
       "step": 7603
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 13.834127426147461,
-      "learning_rate": 4.758945574822091e-06,
-      "loss": 1.385,
+      "epoch": 0.95,
+      "grad_norm": 4.783206939697266,
+      "learning_rate": 1.3638455424005358e-05,
+      "loss": 0.2047,
       "step": 7604
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 18.952781677246094,
-      "learning_rate": 4.756940964217701e-06,
-      "loss": 1.2853,
+      "epoch": 0.95,
+      "grad_norm": 5.902559757232666,
+      "learning_rate": 1.3637618708948668e-05,
+      "loss": 1.2226,
       "step": 7605
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 18.030406951904297,
-      "learning_rate": 4.754936353613311e-06,
-      "loss": 1.6859,
+      "epoch": 0.95,
+      "grad_norm": 12.657173156738281,
+      "learning_rate": 1.3636781993891981e-05,
+      "loss": 1.155,
       "step": 7606
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 35.28142166137695,
-      "learning_rate": 4.752931743008921e-06,
-      "loss": 2.1156,
+      "epoch": 0.95,
+      "grad_norm": 34.17289733886719,
+      "learning_rate": 1.3635945278835293e-05,
+      "loss": 2.0155,
       "step": 7607
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 28.663076400756836,
-      "learning_rate": 4.7509271324045305e-06,
-      "loss": 2.1559,
+      "epoch": 0.95,
+      "grad_norm": 6.413711071014404,
+      "learning_rate": 1.3635108563778607e-05,
+      "loss": 0.7814,
       "step": 7608
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 11.263956069946289,
-      "learning_rate": 4.748922521800141e-06,
-      "loss": 0.9984,
+      "epoch": 0.95,
+      "grad_norm": 9.342193603515625,
+      "learning_rate": 1.3634271848721919e-05,
+      "loss": 0.6112,
       "step": 7609
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 9.154500007629395,
-      "learning_rate": 4.746917911195751e-06,
-      "loss": 1.3324,
+      "epoch": 0.96,
+      "grad_norm": 23.81905174255371,
+      "learning_rate": 1.3633435133665231e-05,
+      "loss": 1.6003,
       "step": 7610
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 18.66733169555664,
-      "learning_rate": 4.744913300591361e-06,
-      "loss": 0.6866,
+      "epoch": 0.96,
+      "grad_norm": 116.54635620117188,
+      "learning_rate": 1.3632598418608545e-05,
+      "loss": 3.7441,
       "step": 7611
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 22.89703369140625,
-      "learning_rate": 4.74290868998697e-06,
-      "loss": 2.2213,
+      "epoch": 0.96,
+      "grad_norm": 22.9312744140625,
+      "learning_rate": 1.3631761703551855e-05,
+      "loss": 1.9241,
       "step": 7612
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 9.362771987915039,
-      "learning_rate": 4.740904079382581e-06,
-      "loss": 1.069,
+      "epoch": 0.96,
+      "grad_norm": 78.7513427734375,
+      "learning_rate": 1.3630924988495168e-05,
+      "loss": 1.4171,
       "step": 7613
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 100.19771575927734,
-      "learning_rate": 4.73889946877819e-06,
-      "loss": 1.1066,
+      "epoch": 0.96,
+      "grad_norm": 5.988093852996826,
+      "learning_rate": 1.3630088273438482e-05,
+      "loss": 0.489,
       "step": 7614
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 12.872753143310547,
-      "learning_rate": 4.7368948581738e-06,
-      "loss": 1.3306,
+      "epoch": 0.96,
+      "grad_norm": 26.50092887878418,
+      "learning_rate": 1.3629251558381796e-05,
+      "loss": 0.8656,
       "step": 7615
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 13.48827075958252,
-      "learning_rate": 4.73489024756941e-06,
-      "loss": 0.8473,
+      "epoch": 0.96,
+      "grad_norm": 11.180814743041992,
+      "learning_rate": 1.3628414843325106e-05,
+      "loss": 1.6053,
       "step": 7616
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 8.565572738647461,
-      "learning_rate": 4.7328856369650204e-06,
-      "loss": 0.7963,
+      "epoch": 0.96,
+      "grad_norm": 39.32979202270508,
+      "learning_rate": 1.362757812826842e-05,
+      "loss": 1.2344,
       "step": 7617
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 16.844417572021484,
-      "learning_rate": 4.73088102636063e-06,
-      "loss": 0.9599,
+      "epoch": 0.96,
+      "grad_norm": 33.10567855834961,
+      "learning_rate": 1.3626741413211733e-05,
+      "loss": 2.0224,
       "step": 7618
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 16.240650177001953,
-      "learning_rate": 4.72887641575624e-06,
-      "loss": 1.0792,
+      "epoch": 0.96,
+      "grad_norm": 9.995561599731445,
+      "learning_rate": 1.3625904698155044e-05,
+      "loss": 1.8112,
       "step": 7619
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 13.393976211547852,
-      "learning_rate": 4.72687180515185e-06,
-      "loss": 1.1918,
+      "epoch": 0.96,
+      "grad_norm": 20.029680252075195,
+      "learning_rate": 1.3625067983098357e-05,
+      "loss": 1.5666,
       "step": 7620
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 27.26254653930664,
-      "learning_rate": 4.724867194547459e-06,
-      "loss": 1.5554,
+      "epoch": 0.96,
+      "grad_norm": 12.913830757141113,
+      "learning_rate": 1.362423126804167e-05,
+      "loss": 3.1843,
       "step": 7621
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 19.264162063598633,
-      "learning_rate": 4.722862583943069e-06,
-      "loss": 1.0877,
+      "epoch": 0.96,
+      "grad_norm": 15.697308540344238,
+      "learning_rate": 1.3623394552984983e-05,
+      "loss": 2.1712,
       "step": 7622
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 14.326976776123047,
-      "learning_rate": 4.720857973338679e-06,
-      "loss": 1.4076,
+      "epoch": 0.96,
+      "grad_norm": 8.030444145202637,
+      "learning_rate": 1.3622557837928295e-05,
+      "loss": 0.3436,
       "step": 7623
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 15.193740844726562,
-      "learning_rate": 4.718853362734289e-06,
-      "loss": 1.6355,
+      "epoch": 0.96,
+      "grad_norm": 5.636011600494385,
+      "learning_rate": 1.3621721122871607e-05,
+      "loss": 0.7249,
       "step": 7624
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 9.33708381652832,
-      "learning_rate": 4.7168487521298986e-06,
-      "loss": 1.4984,
+      "epoch": 0.96,
+      "grad_norm": 11.708727836608887,
+      "learning_rate": 1.362088440781492e-05,
+      "loss": 0.9984,
       "step": 7625
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 84.12486267089844,
-      "learning_rate": 4.7148441415255095e-06,
-      "loss": 1.9932,
+      "epoch": 0.96,
+      "grad_norm": 14.825302124023438,
+      "learning_rate": 1.362004769275823e-05,
+      "loss": 1.047,
       "step": 7626
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 21.830821990966797,
-      "learning_rate": 4.712839530921119e-06,
-      "loss": 1.4148,
+      "epoch": 0.96,
+      "grad_norm": 16.708148956298828,
+      "learning_rate": 1.3619210977701544e-05,
+      "loss": 2.2662,
       "step": 7627
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 11.36988353729248,
-      "learning_rate": 4.710834920316729e-06,
-      "loss": 1.3706,
+      "epoch": 0.96,
+      "grad_norm": 27.896203994750977,
+      "learning_rate": 1.3618374262644858e-05,
+      "loss": 1.142,
       "step": 7628
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 4.701085567474365,
-      "learning_rate": 4.708830309712339e-06,
-      "loss": 0.4887,
+      "epoch": 0.96,
+      "grad_norm": 9.181163787841797,
+      "learning_rate": 1.3617537547588172e-05,
+      "loss": 1.565,
       "step": 7629
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 40.99164962768555,
-      "learning_rate": 4.706825699107949e-06,
-      "loss": 3.0745,
+      "epoch": 0.96,
+      "grad_norm": 40.83127975463867,
+      "learning_rate": 1.3616700832531482e-05,
+      "loss": 1.8924,
       "step": 7630
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 46.76643753051758,
-      "learning_rate": 4.704821088503558e-06,
-      "loss": 1.5348,
+      "epoch": 0.96,
+      "grad_norm": 10.27275276184082,
+      "learning_rate": 1.3615864117474796e-05,
+      "loss": 1.2299,
       "step": 7631
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 24.31828498840332,
-      "learning_rate": 4.702816477899168e-06,
-      "loss": 1.1564,
+      "epoch": 0.96,
+      "grad_norm": 13.874832153320312,
+      "learning_rate": 1.361502740241811e-05,
+      "loss": 1.5729,
       "step": 7632
     },
     {
-      "epoch": 2.29,
-      "grad_norm": 32.74114227294922,
-      "learning_rate": 4.700811867294778e-06,
-      "loss": 0.9132,
+      "epoch": 0.96,
+      "grad_norm": 15.549036026000977,
+      "learning_rate": 1.361419068736142e-05,
+      "loss": 1.6644,
       "step": 7633
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 36.050968170166016,
-      "learning_rate": 4.6988072566903885e-06,
-      "loss": 1.0512,
+      "epoch": 0.96,
+      "grad_norm": 10.457554817199707,
+      "learning_rate": 1.3613353972304733e-05,
+      "loss": 1.2516,
       "step": 7634
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 11.422075271606445,
-      "learning_rate": 4.6968026460859986e-06,
-      "loss": 1.2059,
+      "epoch": 0.96,
+      "grad_norm": 9.416202545166016,
+      "learning_rate": 1.3612517257248045e-05,
+      "loss": 1.1955,
       "step": 7635
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 35.184654235839844,
-      "learning_rate": 4.694798035481608e-06,
-      "loss": 1.4813,
+      "epoch": 0.96,
+      "grad_norm": 15.376522064208984,
+      "learning_rate": 1.3611680542191359e-05,
+      "loss": 1.8816,
       "step": 7636
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 45.07157516479492,
-      "learning_rate": 4.692793424877218e-06,
-      "loss": 1.2941,
+      "epoch": 0.96,
+      "grad_norm": 7.528784275054932,
+      "learning_rate": 1.361084382713467e-05,
+      "loss": 0.8651,
       "step": 7637
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 37.6901969909668,
-      "learning_rate": 4.690788814272828e-06,
-      "loss": 3.1126,
+      "epoch": 0.96,
+      "grad_norm": 20.97710418701172,
+      "learning_rate": 1.3610007112077983e-05,
+      "loss": 1.6182,
       "step": 7638
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 21.26262664794922,
-      "learning_rate": 4.688784203668438e-06,
-      "loss": 1.3052,
+      "epoch": 0.96,
+      "grad_norm": 29.202354431152344,
+      "learning_rate": 1.3609170397021296e-05,
+      "loss": 2.1021,
       "step": 7639
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 26.654037475585938,
-      "learning_rate": 4.686779593064047e-06,
-      "loss": 1.6718,
+      "epoch": 0.96,
+      "grad_norm": 6.581846237182617,
+      "learning_rate": 1.3608333681964607e-05,
+      "loss": 0.3949,
       "step": 7640
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 19.267364501953125,
-      "learning_rate": 4.684774982459657e-06,
-      "loss": 1.1093,
+      "epoch": 0.96,
+      "grad_norm": 20.662389755249023,
+      "learning_rate": 1.360749696690792e-05,
+      "loss": 3.3764,
       "step": 7641
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 11.072561264038086,
-      "learning_rate": 4.6827703718552674e-06,
-      "loss": 1.1142,
+      "epoch": 0.96,
+      "grad_norm": 17.10767364501953,
+      "learning_rate": 1.3606660251851234e-05,
+      "loss": 2.4709,
       "step": 7642
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 17.212007522583008,
-      "learning_rate": 4.6807657612508775e-06,
-      "loss": 1.1688,
+      "epoch": 0.96,
+      "grad_norm": 9.228837013244629,
+      "learning_rate": 1.3605823536794547e-05,
+      "loss": 0.6915,
       "step": 7643
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 14.088850021362305,
-      "learning_rate": 4.678761150646488e-06,
-      "loss": 1.0786,
+      "epoch": 0.96,
+      "grad_norm": 11.638644218444824,
+      "learning_rate": 1.3604986821737858e-05,
+      "loss": 1.1213,
       "step": 7644
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 8.411274909973145,
-      "learning_rate": 4.676756540042097e-06,
-      "loss": 1.6672,
+      "epoch": 0.96,
+      "grad_norm": 18.108949661254883,
+      "learning_rate": 1.3604150106681171e-05,
+      "loss": 1.7587,
       "step": 7645
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 10.209150314331055,
-      "learning_rate": 4.674751929437707e-06,
-      "loss": 0.6725,
+      "epoch": 0.96,
+      "grad_norm": 20.314796447753906,
+      "learning_rate": 1.3603313391624483e-05,
+      "loss": 2.1261,
       "step": 7646
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 8.190345764160156,
-      "learning_rate": 4.672747318833317e-06,
-      "loss": 0.7018,
+      "epoch": 0.96,
+      "grad_norm": 3.3429651260375977,
+      "learning_rate": 1.3602476676567795e-05,
+      "loss": 0.0715,
       "step": 7647
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 17.66701316833496,
-      "learning_rate": 4.670742708228927e-06,
-      "loss": 1.9426,
+      "epoch": 0.96,
+      "grad_norm": 24.380075454711914,
+      "learning_rate": 1.3601639961511109e-05,
+      "loss": 1.8533,
       "step": 7648
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 29.10671043395996,
-      "learning_rate": 4.668738097624536e-06,
-      "loss": 1.8796,
+      "epoch": 0.96,
+      "grad_norm": 10.046948432922363,
+      "learning_rate": 1.3600803246454421e-05,
+      "loss": 0.2984,
       "step": 7649
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 10.14902114868164,
-      "learning_rate": 4.666733487020147e-06,
-      "loss": 0.9493,
+      "epoch": 0.96,
+      "grad_norm": 12.797595977783203,
+      "learning_rate": 1.3599966531397735e-05,
+      "loss": 0.838,
       "step": 7650
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 21.747249603271484,
-      "learning_rate": 4.6647288764157565e-06,
-      "loss": 0.8706,
+      "epoch": 0.96,
+      "grad_norm": 8.366950035095215,
+      "learning_rate": 1.3599129816341046e-05,
+      "loss": 0.4222,
       "step": 7651
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 52.81174087524414,
-      "learning_rate": 4.662724265811367e-06,
-      "loss": 2.5637,
+      "epoch": 0.96,
+      "grad_norm": 17.715896606445312,
+      "learning_rate": 1.3598293101284358e-05,
+      "loss": 1.1354,
       "step": 7652
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 21.7281494140625,
-      "learning_rate": 4.660719655206977e-06,
-      "loss": 1.5396,
+      "epoch": 0.96,
+      "grad_norm": 35.103858947753906,
+      "learning_rate": 1.3597456386227672e-05,
+      "loss": 3.3299,
       "step": 7653
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 9.083329200744629,
-      "learning_rate": 4.658715044602587e-06,
-      "loss": 0.701,
+      "epoch": 0.96,
+      "grad_norm": 8.418503761291504,
+      "learning_rate": 1.3596619671170982e-05,
+      "loss": 0.4461,
       "step": 7654
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 8.001774787902832,
-      "learning_rate": 4.656710433998196e-06,
-      "loss": 0.9427,
+      "epoch": 0.96,
+      "grad_norm": 8.805988311767578,
+      "learning_rate": 1.3595782956114296e-05,
+      "loss": 0.6346,
       "step": 7655
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 8.815278053283691,
-      "learning_rate": 4.654705823393806e-06,
-      "loss": 1.6543,
+      "epoch": 0.96,
+      "grad_norm": 37.53118133544922,
+      "learning_rate": 1.359494624105761e-05,
+      "loss": 1.512,
       "step": 7656
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 37.267616271972656,
-      "learning_rate": 4.652701212789416e-06,
-      "loss": 0.9176,
+      "epoch": 0.96,
+      "grad_norm": 27.297487258911133,
+      "learning_rate": 1.3594109526000923e-05,
+      "loss": 2.6313,
       "step": 7657
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 32.767486572265625,
-      "learning_rate": 4.650696602185025e-06,
-      "loss": 1.6066,
+      "epoch": 0.96,
+      "grad_norm": 4.831480979919434,
+      "learning_rate": 1.3593272810944234e-05,
+      "loss": 0.2401,
       "step": 7658
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 15.666519165039062,
-      "learning_rate": 4.648691991580636e-06,
-      "loss": 1.3438,
+      "epoch": 0.96,
+      "grad_norm": 9.539240837097168,
+      "learning_rate": 1.3592436095887547e-05,
+      "loss": 1.9658,
       "step": 7659
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 31.18556785583496,
-      "learning_rate": 4.646687380976246e-06,
-      "loss": 2.5968,
+      "epoch": 0.96,
+      "grad_norm": 18.462636947631836,
+      "learning_rate": 1.359159938083086e-05,
+      "loss": 2.1001,
       "step": 7660
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 21.61691665649414,
-      "learning_rate": 4.644682770371856e-06,
-      "loss": 1.7081,
+      "epoch": 0.96,
+      "grad_norm": 71.91975402832031,
+      "learning_rate": 1.3590762665774171e-05,
+      "loss": 3.0899,
       "step": 7661
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 6.294021129608154,
-      "learning_rate": 4.642678159767465e-06,
-      "loss": 0.5644,
+      "epoch": 0.96,
+      "grad_norm": 18.084308624267578,
+      "learning_rate": 1.3589925950717485e-05,
+      "loss": 0.8901,
       "step": 7662
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 17.326904296875,
-      "learning_rate": 4.640673549163076e-06,
-      "loss": 1.2007,
+      "epoch": 0.96,
+      "grad_norm": 19.814807891845703,
+      "learning_rate": 1.3589089235660797e-05,
+      "loss": 1.5361,
       "step": 7663
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 7.6418986320495605,
-      "learning_rate": 4.638668938558685e-06,
-      "loss": 1.5581,
+      "epoch": 0.96,
+      "grad_norm": 13.233928680419922,
+      "learning_rate": 1.358825252060411e-05,
+      "loss": 2.3141,
       "step": 7664
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 36.48931121826172,
-      "learning_rate": 4.636664327954295e-06,
-      "loss": 1.8396,
+      "epoch": 0.96,
+      "grad_norm": 11.917679786682129,
+      "learning_rate": 1.358741580554742e-05,
+      "loss": 2.5707,
       "step": 7665
     },
     {
-      "epoch": 2.3,
-      "grad_norm": 12.859515190124512,
-      "learning_rate": 4.634659717349905e-06,
-      "loss": 1.6651,
+      "epoch": 0.96,
+      "grad_norm": 14.535420417785645,
+      "learning_rate": 1.3586579090490734e-05,
+      "loss": 0.7456,
       "step": 7666
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 19.518774032592773,
-      "learning_rate": 4.632655106745515e-06,
-      "loss": 1.0046,
+      "epoch": 0.96,
+      "grad_norm": 23.91950035095215,
+      "learning_rate": 1.3585742375434048e-05,
+      "loss": 2.8434,
       "step": 7667
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 17.942481994628906,
-      "learning_rate": 4.6306504961411246e-06,
-      "loss": 1.7626,
+      "epoch": 0.96,
+      "grad_norm": 9.482237815856934,
+      "learning_rate": 1.3584905660377358e-05,
+      "loss": 0.607,
       "step": 7668
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 17.79322624206543,
-      "learning_rate": 4.628645885536735e-06,
-      "loss": 0.8779,
+      "epoch": 0.96,
+      "grad_norm": 31.98341178894043,
+      "learning_rate": 1.3584068945320672e-05,
+      "loss": 2.7799,
       "step": 7669
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 17.742807388305664,
-      "learning_rate": 4.626641274932345e-06,
-      "loss": 0.9714,
+      "epoch": 0.96,
+      "grad_norm": 15.53034782409668,
+      "learning_rate": 1.3583232230263985e-05,
+      "loss": 3.3731,
       "step": 7670
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 14.361837387084961,
-      "learning_rate": 4.624636664327955e-06,
-      "loss": 1.1075,
+      "epoch": 0.96,
+      "grad_norm": 11.994458198547363,
+      "learning_rate": 1.3582395515207299e-05,
+      "loss": 3.1805,
       "step": 7671
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 19.260719299316406,
-      "learning_rate": 4.622632053723565e-06,
-      "loss": 2.0738,
+      "epoch": 0.96,
+      "grad_norm": 13.355704307556152,
+      "learning_rate": 1.358155880015061e-05,
+      "loss": 1.1367,
       "step": 7672
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 19.078481674194336,
-      "learning_rate": 4.620627443119174e-06,
-      "loss": 1.3898,
+      "epoch": 0.96,
+      "grad_norm": 95.1605453491211,
+      "learning_rate": 1.3580722085093923e-05,
+      "loss": 1.2786,
       "step": 7673
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 11.892518043518066,
-      "learning_rate": 4.618622832514784e-06,
-      "loss": 1.4389,
+      "epoch": 0.96,
+      "grad_norm": 12.803677558898926,
+      "learning_rate": 1.3579885370037235e-05,
+      "loss": 1.5783,
       "step": 7674
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 37.109336853027344,
-      "learning_rate": 4.616618221910394e-06,
-      "loss": 1.5463,
+      "epoch": 0.96,
+      "grad_norm": 10.651666641235352,
+      "learning_rate": 1.3579048654980547e-05,
+      "loss": 0.6602,
       "step": 7675
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 35.61430358886719,
-      "learning_rate": 4.614613611306004e-06,
-      "loss": 1.2612,
+      "epoch": 0.96,
+      "grad_norm": 8.594017028808594,
+      "learning_rate": 1.357821193992386e-05,
+      "loss": 0.7785,
       "step": 7676
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 10.030492782592773,
-      "learning_rate": 4.612609000701614e-06,
-      "loss": 1.0428,
+      "epoch": 0.96,
+      "grad_norm": 8.036545753479004,
+      "learning_rate": 1.3577375224867173e-05,
+      "loss": 0.5393,
       "step": 7677
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 32.46242141723633,
-      "learning_rate": 4.6106043900972246e-06,
-      "loss": 1.2366,
+      "epoch": 0.96,
+      "grad_norm": 12.696398735046387,
+      "learning_rate": 1.3576538509810485e-05,
+      "loss": 1.9092,
       "step": 7678
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 63.76512908935547,
-      "learning_rate": 4.608599779492834e-06,
-      "loss": 1.8638,
+      "epoch": 0.96,
+      "grad_norm": 45.462074279785156,
+      "learning_rate": 1.3575701794753796e-05,
+      "loss": 2.7158,
       "step": 7679
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 21.762617111206055,
-      "learning_rate": 4.606595168888444e-06,
-      "loss": 1.7875,
-      "step": 7680
-    },
-    {
-      "epoch": 2.31,
-      "eval_loss": 0.16951656341552734,
-      "eval_runtime": 43.7895,
-      "eval_samples_per_second": 33.775,
-      "eval_steps_per_second": 33.775,
+      "epoch": 0.96,
+      "grad_norm": 16.446996688842773,
+      "learning_rate": 1.357486507969711e-05,
+      "loss": 1.7265,
       "step": 7680
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 10.09822940826416,
-      "learning_rate": 4.604590558284054e-06,
-      "loss": 1.1142,
+      "epoch": 0.96,
+      "grad_norm": 10.649942398071289,
+      "learning_rate": 1.3574028364640424e-05,
+      "loss": 2.2305,
       "step": 7681
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 10.086823463439941,
-      "learning_rate": 4.602585947679663e-06,
-      "loss": 0.8318,
+      "epoch": 0.96,
+      "grad_norm": 15.220854759216309,
+      "learning_rate": 1.3573191649583734e-05,
+      "loss": 1.5713,
       "step": 7682
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 11.713811874389648,
-      "learning_rate": 4.600581337075273e-06,
-      "loss": 0.9897,
+      "epoch": 0.96,
+      "grad_norm": 11.791414260864258,
+      "learning_rate": 1.3572354934527048e-05,
+      "loss": 1.2042,
       "step": 7683
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 20.750566482543945,
-      "learning_rate": 4.598576726470883e-06,
-      "loss": 1.5894,
+      "epoch": 0.96,
+      "grad_norm": 21.493785858154297,
+      "learning_rate": 1.3571518219470361e-05,
+      "loss": 2.3257,
       "step": 7684
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 17.660737991333008,
-      "learning_rate": 4.5965721158664935e-06,
-      "loss": 1.4457,
+      "epoch": 0.96,
+      "grad_norm": 10.986868858337402,
+      "learning_rate": 1.3570681504413672e-05,
+      "loss": 0.9414,
       "step": 7685
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 11.275466918945312,
-      "learning_rate": 4.594567505262103e-06,
-      "loss": 1.0097,
+      "epoch": 0.96,
+      "grad_norm": 11.292412757873535,
+      "learning_rate": 1.3569844789356985e-05,
+      "loss": 1.0636,
       "step": 7686
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 9.963224411010742,
-      "learning_rate": 4.592562894657714e-06,
-      "loss": 1.7383,
+      "epoch": 0.96,
+      "grad_norm": 26.578144073486328,
+      "learning_rate": 1.3569008074300299e-05,
+      "loss": 1.8814,
       "step": 7687
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 5.169731140136719,
-      "learning_rate": 4.590558284053323e-06,
-      "loss": 0.5528,
+      "epoch": 0.96,
+      "grad_norm": 8.212207794189453,
+      "learning_rate": 1.356817135924361e-05,
+      "loss": 1.4536,
       "step": 7688
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 10.396340370178223,
-      "learning_rate": 4.588553673448933e-06,
-      "loss": 0.967,
+      "epoch": 0.96,
+      "grad_norm": 22.602903366088867,
+      "learning_rate": 1.3567334644186923e-05,
+      "loss": 1.8578,
       "step": 7689
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 15.52597427368164,
-      "learning_rate": 4.586549062844543e-06,
-      "loss": 0.9266,
+      "epoch": 0.97,
+      "grad_norm": 8.634163856506348,
+      "learning_rate": 1.3566497929130236e-05,
+      "loss": 1.0865,
       "step": 7690
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 12.798748016357422,
-      "learning_rate": 4.584544452240153e-06,
-      "loss": 1.288,
+      "epoch": 0.97,
+      "grad_norm": 13.972951889038086,
+      "learning_rate": 1.3565661214073548e-05,
+      "loss": 1.5003,
       "step": 7691
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 21.744783401489258,
-      "learning_rate": 4.582539841635762e-06,
-      "loss": 1.3389,
+      "epoch": 0.97,
+      "grad_norm": 9.140802383422852,
+      "learning_rate": 1.356482449901686e-05,
+      "loss": 0.5627,
       "step": 7692
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 7.416219711303711,
-      "learning_rate": 4.5805352310313724e-06,
-      "loss": 0.9147,
+      "epoch": 0.97,
+      "grad_norm": 11.660378456115723,
+      "learning_rate": 1.3563987783960172e-05,
+      "loss": 1.4375,
       "step": 7693
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 27.035898208618164,
-      "learning_rate": 4.5785306204269825e-06,
-      "loss": 2.5926,
+      "epoch": 0.97,
+      "grad_norm": 17.761133193969727,
+      "learning_rate": 1.3563151068903486e-05,
+      "loss": 2.2885,
       "step": 7694
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 18.482648849487305,
-      "learning_rate": 4.576526009822593e-06,
-      "loss": 2.276,
+      "epoch": 0.97,
+      "grad_norm": 10.578373908996582,
+      "learning_rate": 1.35623143538468e-05,
+      "loss": 2.5974,
       "step": 7695
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 52.933494567871094,
-      "learning_rate": 4.574521399218203e-06,
-      "loss": 2.2703,
+      "epoch": 0.97,
+      "grad_norm": 12.733040809631348,
+      "learning_rate": 1.356147763879011e-05,
+      "loss": 2.3853,
       "step": 7696
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 11.743091583251953,
-      "learning_rate": 4.572516788613812e-06,
-      "loss": 0.9347,
+      "epoch": 0.97,
+      "grad_norm": 10.997481346130371,
+      "learning_rate": 1.3560640923733424e-05,
+      "loss": 0.7155,
       "step": 7697
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 16.696788787841797,
-      "learning_rate": 4.570512178009422e-06,
-      "loss": 1.3391,
+      "epoch": 0.97,
+      "grad_norm": 11.658534049987793,
+      "learning_rate": 1.3559804208676737e-05,
+      "loss": 0.583,
       "step": 7698
     },
     {
-      "epoch": 2.31,
-      "grad_norm": 23.81439971923828,
-      "learning_rate": 4.568507567405032e-06,
-      "loss": 2.2063,
+      "epoch": 0.97,
+      "grad_norm": 71.93724822998047,
+      "learning_rate": 1.3558967493620047e-05,
+      "loss": 2.1163,
       "step": 7699
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 11.924534797668457,
-      "learning_rate": 4.566502956800642e-06,
-      "loss": 1.0567,
+      "epoch": 0.97,
+      "grad_norm": 15.736628532409668,
+      "learning_rate": 1.3558130778563361e-05,
+      "loss": 0.5738,
       "step": 7700
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 15.285096168518066,
-      "learning_rate": 4.564498346196251e-06,
-      "loss": 1.6988,
+      "epoch": 0.97,
+      "grad_norm": 7.046619892120361,
+      "learning_rate": 1.3557294063506675e-05,
+      "loss": 1.2248,
       "step": 7701
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 14.276802062988281,
-      "learning_rate": 4.562493735591862e-06,
-      "loss": 0.7809,
+      "epoch": 0.97,
+      "grad_norm": 15.224453926086426,
+      "learning_rate": 1.3556457348449987e-05,
+      "loss": 1.245,
       "step": 7702
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 19.88568687438965,
-      "learning_rate": 4.560489124987472e-06,
-      "loss": 1.8969,
+      "epoch": 0.97,
+      "grad_norm": 6.766767978668213,
+      "learning_rate": 1.3555620633393299e-05,
+      "loss": 2.0219,
       "step": 7703
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 22.174179077148438,
-      "learning_rate": 4.558484514383082e-06,
-      "loss": 1.4157,
+      "epoch": 0.97,
+      "grad_norm": 10.02777099609375,
+      "learning_rate": 1.3554783918336612e-05,
+      "loss": 1.544,
       "step": 7704
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 26.93759536743164,
-      "learning_rate": 4.556479903778691e-06,
-      "loss": 0.7634,
+      "epoch": 0.97,
+      "grad_norm": 36.131717681884766,
+      "learning_rate": 1.3553947203279924e-05,
+      "loss": 2.2543,
       "step": 7705
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 14.25698184967041,
-      "learning_rate": 4.554475293174301e-06,
-      "loss": 0.9706,
+      "epoch": 0.97,
+      "grad_norm": 17.951614379882812,
+      "learning_rate": 1.3553110488223236e-05,
+      "loss": 1.875,
       "step": 7706
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 24.950761795043945,
-      "learning_rate": 4.552470682569911e-06,
-      "loss": 1.938,
+      "epoch": 0.97,
+      "grad_norm": 32.99787139892578,
+      "learning_rate": 1.3552273773166548e-05,
+      "loss": 1.0474,
       "step": 7707
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 7.129110336303711,
-      "learning_rate": 4.550466071965521e-06,
-      "loss": 0.538,
+      "epoch": 0.97,
+      "grad_norm": 16.231891632080078,
+      "learning_rate": 1.3551437058109862e-05,
+      "loss": 0.8705,
       "step": 7708
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 10.459171295166016,
-      "learning_rate": 4.548461461361131e-06,
-      "loss": 1.3554,
+      "epoch": 0.97,
+      "grad_norm": 11.002676963806152,
+      "learning_rate": 1.3550600343053175e-05,
+      "loss": 2.3549,
       "step": 7709
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 26.577316284179688,
-      "learning_rate": 4.5464568507567405e-06,
-      "loss": 2.1107,
+      "epoch": 0.97,
+      "grad_norm": 15.858479499816895,
+      "learning_rate": 1.3549763627996486e-05,
+      "loss": 1.044,
       "step": 7710
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 87.70563507080078,
-      "learning_rate": 4.5444522401523506e-06,
-      "loss": 1.3952,
+      "epoch": 0.97,
+      "grad_norm": 13.546462059020996,
+      "learning_rate": 1.35489269129398e-05,
+      "loss": 1.4793,
       "step": 7711
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 10.34211254119873,
-      "learning_rate": 4.542447629547961e-06,
-      "loss": 0.7507,
+      "epoch": 0.97,
+      "grad_norm": 5.651560306549072,
+      "learning_rate": 1.3548090197883113e-05,
+      "loss": 0.34,
       "step": 7712
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 33.580265045166016,
-      "learning_rate": 4.540443018943571e-06,
-      "loss": 1.8236,
+      "epoch": 0.97,
+      "grad_norm": 21.489553451538086,
+      "learning_rate": 1.3547253482826423e-05,
+      "loss": 2.1118,
       "step": 7713
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 13.009629249572754,
-      "learning_rate": 4.53843840833918e-06,
-      "loss": 1.2249,
+      "epoch": 0.97,
+      "grad_norm": 41.191070556640625,
+      "learning_rate": 1.3546416767769737e-05,
+      "loss": 3.4881,
       "step": 7714
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 8.66175365447998,
-      "learning_rate": 4.536433797734791e-06,
-      "loss": 0.9531,
+      "epoch": 0.97,
+      "grad_norm": 24.51534080505371,
+      "learning_rate": 1.354558005271305e-05,
+      "loss": 1.7158,
       "step": 7715
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 17.59461784362793,
-      "learning_rate": 4.5344291871304e-06,
-      "loss": 1.6012,
+      "epoch": 0.97,
+      "grad_norm": 8.854673385620117,
+      "learning_rate": 1.3544743337656363e-05,
+      "loss": 0.5715,
       "step": 7716
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 26.786701202392578,
-      "learning_rate": 4.53242457652601e-06,
-      "loss": 1.3363,
+      "epoch": 0.97,
+      "grad_norm": 34.15606689453125,
+      "learning_rate": 1.3543906622599674e-05,
+      "loss": 1.631,
       "step": 7717
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 10.355632781982422,
-      "learning_rate": 4.53041996592162e-06,
-      "loss": 0.8374,
+      "epoch": 0.97,
+      "grad_norm": 8.685201644897461,
+      "learning_rate": 1.3543069907542986e-05,
+      "loss": 0.8118,
       "step": 7718
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 13.772407531738281,
-      "learning_rate": 4.52841535531723e-06,
-      "loss": 1.9609,
+      "epoch": 0.97,
+      "grad_norm": 9.957879066467285,
+      "learning_rate": 1.35422331924863e-05,
+      "loss": 0.7961,
       "step": 7719
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 10.235556602478027,
-      "learning_rate": 4.52641074471284e-06,
-      "loss": 1.4712,
+      "epoch": 0.97,
+      "grad_norm": 27.317323684692383,
+      "learning_rate": 1.3541396477429612e-05,
+      "loss": 1.1692,
       "step": 7720
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 37.70382308959961,
-      "learning_rate": 4.52440613410845e-06,
-      "loss": 1.8212,
+      "epoch": 0.97,
+      "grad_norm": 6.7141618728637695,
+      "learning_rate": 1.3540559762372924e-05,
+      "loss": 1.0255,
       "step": 7721
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 16.85622215270996,
-      "learning_rate": 4.52240152350406e-06,
-      "loss": 1.3294,
+      "epoch": 0.97,
+      "grad_norm": 14.603387832641602,
+      "learning_rate": 1.3539723047316238e-05,
+      "loss": 2.6129,
       "step": 7722
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 13.376880645751953,
-      "learning_rate": 4.520396912899669e-06,
-      "loss": 0.8711,
+      "epoch": 0.97,
+      "grad_norm": 11.722940444946289,
+      "learning_rate": 1.3538886332259551e-05,
+      "loss": 0.8843,
       "step": 7723
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 24.79469871520996,
-      "learning_rate": 4.51839230229528e-06,
-      "loss": 1.125,
+      "epoch": 0.97,
+      "grad_norm": 107.16452026367188,
+      "learning_rate": 1.3538049617202862e-05,
+      "loss": 2.1597,
       "step": 7724
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 21.333232879638672,
-      "learning_rate": 4.516387691690889e-06,
-      "loss": 1.7883,
+      "epoch": 0.97,
+      "grad_norm": 36.03486251831055,
+      "learning_rate": 1.3537212902146175e-05,
+      "loss": 2.2156,
       "step": 7725
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 11.4044771194458,
-      "learning_rate": 4.514383081086499e-06,
-      "loss": 1.1805,
+      "epoch": 0.97,
+      "grad_norm": 21.66221046447754,
+      "learning_rate": 1.3536376187089489e-05,
+      "loss": 1.7477,
       "step": 7726
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 51.108253479003906,
-      "learning_rate": 4.512378470482109e-06,
-      "loss": 1.9263,
+      "epoch": 0.97,
+      "grad_norm": 21.030221939086914,
+      "learning_rate": 1.3535539472032799e-05,
+      "loss": 1.8737,
       "step": 7727
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 21.30112075805664,
-      "learning_rate": 4.5103738598777195e-06,
-      "loss": 1.3295,
+      "epoch": 0.97,
+      "grad_norm": 8.940326690673828,
+      "learning_rate": 1.3534702756976113e-05,
+      "loss": 1.1877,
       "step": 7728
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 26.95627784729004,
-      "learning_rate": 4.508369249273329e-06,
-      "loss": 2.7558,
+      "epoch": 0.97,
+      "grad_norm": 9.069717407226562,
+      "learning_rate": 1.3533866041919426e-05,
+      "loss": 0.7334,
       "step": 7729
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 13.784543991088867,
-      "learning_rate": 4.506364638668939e-06,
-      "loss": 2.2716,
+      "epoch": 0.97,
+      "grad_norm": 15.025553703308105,
+      "learning_rate": 1.3533029326862738e-05,
+      "loss": 1.7665,
       "step": 7730
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 55.046142578125,
-      "learning_rate": 4.504360028064549e-06,
-      "loss": 1.819,
+      "epoch": 0.97,
+      "grad_norm": 23.173837661743164,
+      "learning_rate": 1.353219261180605e-05,
+      "loss": 0.7893,
       "step": 7731
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 10.930252075195312,
-      "learning_rate": 4.502355417460159e-06,
-      "loss": 0.8165,
+      "epoch": 0.97,
+      "grad_norm": 17.162948608398438,
+      "learning_rate": 1.3531355896749362e-05,
+      "loss": 1.1794,
       "step": 7732
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 23.520112991333008,
-      "learning_rate": 4.500350806855769e-06,
-      "loss": 0.9347,
+      "epoch": 0.97,
+      "grad_norm": 22.472871780395508,
+      "learning_rate": 1.3530519181692676e-05,
+      "loss": 2.3618,
       "step": 7733
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 8.282185554504395,
-      "learning_rate": 4.498346196251378e-06,
-      "loss": 1.1021,
+      "epoch": 0.97,
+      "grad_norm": 31.161813735961914,
+      "learning_rate": 1.3529682466635988e-05,
+      "loss": 0.5731,
       "step": 7734
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 15.41140365600586,
-      "learning_rate": 4.496341585646988e-06,
-      "loss": 1.5547,
+      "epoch": 0.97,
+      "grad_norm": 21.318912506103516,
+      "learning_rate": 1.35288457515793e-05,
+      "loss": 0.3049,
       "step": 7735
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 26.10579490661621,
-      "learning_rate": 4.4943369750425984e-06,
-      "loss": 1.77,
+      "epoch": 0.97,
+      "grad_norm": 11.425654411315918,
+      "learning_rate": 1.3528009036522613e-05,
+      "loss": 1.5813,
       "step": 7736
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 11.937204360961914,
-      "learning_rate": 4.4923323644382085e-06,
-      "loss": 1.0362,
+      "epoch": 0.97,
+      "grad_norm": 18.119468688964844,
+      "learning_rate": 1.3527172321465927e-05,
+      "loss": 1.351,
       "step": 7737
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 21.467954635620117,
-      "learning_rate": 4.490327753833818e-06,
-      "loss": 1.802,
+      "epoch": 0.97,
+      "grad_norm": 14.75485897064209,
+      "learning_rate": 1.3526335606409237e-05,
+      "loss": 1.5202,
       "step": 7738
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 17.1977596282959,
-      "learning_rate": 4.488323143229429e-06,
-      "loss": 1.2613,
+      "epoch": 0.97,
+      "grad_norm": 22.677011489868164,
+      "learning_rate": 1.3525498891352551e-05,
+      "loss": 0.48,
       "step": 7739
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 11.602217674255371,
-      "learning_rate": 4.486318532625038e-06,
-      "loss": 1.5562,
+      "epoch": 0.97,
+      "grad_norm": 17.857650756835938,
+      "learning_rate": 1.3524662176295865e-05,
+      "loss": 1.4076,
       "step": 7740
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 50.02082061767578,
-      "learning_rate": 4.484313922020648e-06,
-      "loss": 1.2008,
+      "epoch": 0.97,
+      "grad_norm": 16.588102340698242,
+      "learning_rate": 1.3523825461239175e-05,
+      "loss": 1.1867,
       "step": 7741
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 16.229761123657227,
-      "learning_rate": 4.482309311416258e-06,
-      "loss": 1.5572,
+      "epoch": 0.97,
+      "grad_norm": 21.662322998046875,
+      "learning_rate": 1.3522988746182489e-05,
+      "loss": 1.6806,
       "step": 7742
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 16.347227096557617,
-      "learning_rate": 4.480304700811867e-06,
-      "loss": 1.4817,
+      "epoch": 0.97,
+      "grad_norm": 14.984989166259766,
+      "learning_rate": 1.3522152031125802e-05,
+      "loss": 0.6457,
       "step": 7743
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 9.144948959350586,
-      "learning_rate": 4.478300090207477e-06,
-      "loss": 0.6118,
+      "epoch": 0.97,
+      "grad_norm": 15.336214065551758,
+      "learning_rate": 1.3521315316069114e-05,
+      "loss": 1.5232,
       "step": 7744
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 45.46603775024414,
-      "learning_rate": 4.4762954796030875e-06,
-      "loss": 1.9398,
+      "epoch": 0.97,
+      "grad_norm": 7.916646480560303,
+      "learning_rate": 1.3520478601012426e-05,
+      "loss": 1.0675,
       "step": 7745
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 10.924186706542969,
-      "learning_rate": 4.474290868998698e-06,
-      "loss": 1.4317,
+      "epoch": 0.97,
+      "grad_norm": 28.336755752563477,
+      "learning_rate": 1.3519641885955738e-05,
+      "loss": 1.9443,
       "step": 7746
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 39.43347930908203,
-      "learning_rate": 4.472286258394307e-06,
-      "loss": 2.4369,
+      "epoch": 0.97,
+      "grad_norm": 8.822896003723145,
+      "learning_rate": 1.3518805170899052e-05,
+      "loss": 2.6556,
       "step": 7747
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 124.04064178466797,
-      "learning_rate": 4.470281647789917e-06,
-      "loss": 3.3272,
+      "epoch": 0.97,
+      "grad_norm": 13.60275936126709,
+      "learning_rate": 1.3517968455842364e-05,
+      "loss": 1.3155,
       "step": 7748
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 6.773767471313477,
-      "learning_rate": 4.468277037185527e-06,
-      "loss": 0.5183,
+      "epoch": 0.97,
+      "grad_norm": 11.431167602539062,
+      "learning_rate": 1.3517131740785676e-05,
+      "loss": 1.6894,
       "step": 7749
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 21.174972534179688,
-      "learning_rate": 4.466272426581137e-06,
-      "loss": 2.1152,
+      "epoch": 0.97,
+      "grad_norm": 8.643516540527344,
+      "learning_rate": 1.351629502572899e-05,
+      "loss": 0.2998,
       "step": 7750
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 15.640390396118164,
-      "learning_rate": 4.464267815976746e-06,
-      "loss": 1.0071,
+      "epoch": 0.97,
+      "grad_norm": 2.41274356842041,
+      "learning_rate": 1.3515458310672303e-05,
+      "loss": 0.0267,
       "step": 7751
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 11.41567611694336,
-      "learning_rate": 4.462263205372357e-06,
-      "loss": 0.9003,
+      "epoch": 0.97,
+      "grad_norm": 10.9503173828125,
+      "learning_rate": 1.3514621595615613e-05,
+      "loss": 1.6653,
       "step": 7752
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 13.2136869430542,
-      "learning_rate": 4.4602585947679665e-06,
-      "loss": 1.4521,
+      "epoch": 0.97,
+      "grad_norm": 8.803579330444336,
+      "learning_rate": 1.3513784880558927e-05,
+      "loss": 2.2309,
       "step": 7753
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 24.674755096435547,
-      "learning_rate": 4.4582539841635766e-06,
-      "loss": 1.1701,
+      "epoch": 0.97,
+      "grad_norm": 16.392528533935547,
+      "learning_rate": 1.351294816550224e-05,
+      "loss": 1.046,
       "step": 7754
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 27.077539443969727,
-      "learning_rate": 4.456249373559187e-06,
-      "loss": 1.3348,
+      "epoch": 0.97,
+      "grad_norm": 22.46368980407715,
+      "learning_rate": 1.351211145044555e-05,
+      "loss": 2.3814,
       "step": 7755
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 15.865095138549805,
-      "learning_rate": 4.454244762954797e-06,
-      "loss": 0.9423,
+      "epoch": 0.97,
+      "grad_norm": 18.388086318969727,
+      "learning_rate": 1.3511274735388864e-05,
+      "loss": 1.015,
       "step": 7756
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 10.586139678955078,
-      "learning_rate": 4.452240152350406e-06,
-      "loss": 0.9348,
+      "epoch": 0.97,
+      "grad_norm": 21.396652221679688,
+      "learning_rate": 1.3510438020332176e-05,
+      "loss": 1.7366,
       "step": 7757
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 14.161200523376465,
-      "learning_rate": 4.450235541746016e-06,
-      "loss": 1.5252,
+      "epoch": 0.97,
+      "grad_norm": 9.764254570007324,
+      "learning_rate": 1.350960130527549e-05,
+      "loss": 0.4437,
       "step": 7758
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 23.082622528076172,
-      "learning_rate": 4.448230931141626e-06,
-      "loss": 1.3711,
+      "epoch": 0.97,
+      "grad_norm": 4.522057056427002,
+      "learning_rate": 1.3508764590218802e-05,
+      "loss": 0.2982,
       "step": 7759
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 19.711462020874023,
-      "learning_rate": 4.446226320537235e-06,
-      "loss": 2.655,
+      "epoch": 0.97,
+      "grad_norm": 4.771112442016602,
+      "learning_rate": 1.3507927875162114e-05,
+      "loss": 0.1846,
       "step": 7760
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 24.49384880065918,
-      "learning_rate": 4.444221709932846e-06,
-      "loss": 1.0757,
+      "epoch": 0.97,
+      "grad_norm": 14.318256378173828,
+      "learning_rate": 1.3507091160105428e-05,
+      "loss": 1.8543,
       "step": 7761
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 9.763320922851562,
-      "learning_rate": 4.4422170993284555e-06,
-      "loss": 0.9065,
+      "epoch": 0.97,
+      "grad_norm": 15.801634788513184,
+      "learning_rate": 1.350625444504874e-05,
+      "loss": 1.1831,
       "step": 7762
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 8.923637390136719,
-      "learning_rate": 4.440212488724066e-06,
-      "loss": 1.1599,
+      "epoch": 0.97,
+      "grad_norm": 22.527746200561523,
+      "learning_rate": 1.3505417729992051e-05,
+      "loss": 2.1718,
       "step": 7763
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 193.84384155273438,
-      "learning_rate": 4.438207878119676e-06,
-      "loss": 1.5881,
+      "epoch": 0.97,
+      "grad_norm": 11.585763931274414,
+      "learning_rate": 1.3504581014935365e-05,
+      "loss": 1.3421,
       "step": 7764
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 4.945940971374512,
-      "learning_rate": 4.436203267515286e-06,
-      "loss": 0.698,
+      "epoch": 0.97,
+      "grad_norm": 8.993477821350098,
+      "learning_rate": 1.3503744299878679e-05,
+      "loss": 1.382,
       "step": 7765
     },
     {
-      "epoch": 2.33,
-      "grad_norm": 8.121231079101562,
-      "learning_rate": 4.434198656910895e-06,
-      "loss": 0.9452,
+      "epoch": 0.97,
+      "grad_norm": 10.9657564163208,
+      "learning_rate": 1.3502907584821989e-05,
+      "loss": 1.5614,
       "step": 7766
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 28.268239974975586,
-      "learning_rate": 4.432194046306505e-06,
-      "loss": 1.5929,
+      "epoch": 0.97,
+      "grad_norm": 22.3695125579834,
+      "learning_rate": 1.3502070869765303e-05,
+      "loss": 3.1019,
       "step": 7767
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 70.76728057861328,
-      "learning_rate": 4.430189435702115e-06,
-      "loss": 1.6534,
+      "epoch": 0.97,
+      "grad_norm": 16.434736251831055,
+      "learning_rate": 1.3501234154708616e-05,
+      "loss": 2.7474,
       "step": 7768
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 23.936372756958008,
-      "learning_rate": 4.428184825097725e-06,
-      "loss": 1.2137,
+      "epoch": 0.97,
+      "grad_norm": 9.681111335754395,
+      "learning_rate": 1.3500397439651927e-05,
+      "loss": 1.5325,
       "step": 7769
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 9.887136459350586,
-      "learning_rate": 4.426180214493335e-06,
-      "loss": 0.8383,
+      "epoch": 0.98,
+      "grad_norm": 3.7469215393066406,
+      "learning_rate": 1.349956072459524e-05,
+      "loss": 0.266,
       "step": 7770
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 14.792794227600098,
-      "learning_rate": 4.424175603888945e-06,
-      "loss": 0.9405,
+      "epoch": 0.98,
+      "grad_norm": 14.711934089660645,
+      "learning_rate": 1.3498724009538552e-05,
+      "loss": 1.5471,
       "step": 7771
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 17.71567726135254,
-      "learning_rate": 4.422170993284555e-06,
-      "loss": 1.063,
+      "epoch": 0.98,
+      "grad_norm": 43.18144226074219,
+      "learning_rate": 1.3497887294481866e-05,
+      "loss": 1.7386,
       "step": 7772
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 17.9732608795166,
-      "learning_rate": 4.420166382680165e-06,
-      "loss": 1.4261,
+      "epoch": 0.98,
+      "grad_norm": 13.835360527038574,
+      "learning_rate": 1.3497050579425178e-05,
+      "loss": 1.1166,
       "step": 7773
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 25.94655418395996,
-      "learning_rate": 4.418161772075775e-06,
-      "loss": 2.0324,
+      "epoch": 0.98,
+      "grad_norm": 14.814412117004395,
+      "learning_rate": 1.349621386436849e-05,
+      "loss": 1.0632,
       "step": 7774
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 39.89909744262695,
-      "learning_rate": 4.416157161471384e-06,
-      "loss": 1.7023,
+      "epoch": 0.98,
+      "grad_norm": 5.545921802520752,
+      "learning_rate": 1.3495377149311803e-05,
+      "loss": 0.9194,
       "step": 7775
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 23.909780502319336,
-      "learning_rate": 4.414152550866995e-06,
-      "loss": 0.9315,
+      "epoch": 0.98,
+      "grad_norm": 10.252397537231445,
+      "learning_rate": 1.3494540434255114e-05,
+      "loss": 1.2437,
       "step": 7776
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 13.213065147399902,
-      "learning_rate": 4.412147940262604e-06,
-      "loss": 0.9687,
+      "epoch": 0.98,
+      "grad_norm": 8.675298690795898,
+      "learning_rate": 1.3493703719198427e-05,
+      "loss": 0.8212,
       "step": 7777
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 45.419342041015625,
-      "learning_rate": 4.410143329658214e-06,
-      "loss": 1.2235,
+      "epoch": 0.98,
+      "grad_norm": 15.157878875732422,
+      "learning_rate": 1.3492867004141741e-05,
+      "loss": 2.6183,
       "step": 7778
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 18.985164642333984,
-      "learning_rate": 4.4081387190538244e-06,
-      "loss": 1.544,
+      "epoch": 0.98,
+      "grad_norm": 11.1221284866333,
+      "learning_rate": 1.3492030289085055e-05,
+      "loss": 1.5943,
       "step": 7779
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 9.602522850036621,
-      "learning_rate": 4.4061341084494345e-06,
-      "loss": 0.7917,
+      "epoch": 0.98,
+      "grad_norm": 46.772125244140625,
+      "learning_rate": 1.3491193574028365e-05,
+      "loss": 2.1687,
       "step": 7780
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 38.75558090209961,
-      "learning_rate": 4.404129497845044e-06,
-      "loss": 1.5868,
+      "epoch": 0.98,
+      "grad_norm": 23.514244079589844,
+      "learning_rate": 1.3490356858971679e-05,
+      "loss": 1.2343,
       "step": 7781
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 51.216766357421875,
-      "learning_rate": 4.402124887240654e-06,
-      "loss": 1.2511,
+      "epoch": 0.98,
+      "grad_norm": 11.823355674743652,
+      "learning_rate": 1.3489520143914992e-05,
+      "loss": 0.7507,
       "step": 7782
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 9.934971809387207,
-      "learning_rate": 4.400120276636264e-06,
-      "loss": 1.18,
+      "epoch": 0.98,
+      "grad_norm": 19.944049835205078,
+      "learning_rate": 1.3488683428858302e-05,
+      "loss": 1.3162,
       "step": 7783
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 10.260887145996094,
-      "learning_rate": 4.398115666031873e-06,
-      "loss": 0.5996,
+      "epoch": 0.98,
+      "grad_norm": 15.35301685333252,
+      "learning_rate": 1.3487846713801616e-05,
+      "loss": 2.3952,
       "step": 7784
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 26.906917572021484,
-      "learning_rate": 4.396111055427484e-06,
-      "loss": 1.635,
+      "epoch": 0.98,
+      "grad_norm": 12.974126815795898,
+      "learning_rate": 1.3487009998744928e-05,
+      "loss": 0.6608,
       "step": 7785
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 11.61650562286377,
-      "learning_rate": 4.394106444823093e-06,
-      "loss": 0.8312,
+      "epoch": 0.98,
+      "grad_norm": 7.92689847946167,
+      "learning_rate": 1.3486173283688242e-05,
+      "loss": 0.4993,
       "step": 7786
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 40.0263786315918,
-      "learning_rate": 4.392101834218703e-06,
-      "loss": 0.9066,
+      "epoch": 0.98,
+      "grad_norm": 12.310649871826172,
+      "learning_rate": 1.3485336568631554e-05,
+      "loss": 0.8984,
       "step": 7787
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 31.856773376464844,
-      "learning_rate": 4.390097223614313e-06,
-      "loss": 1.9564,
+      "epoch": 0.98,
+      "grad_norm": 13.593944549560547,
+      "learning_rate": 1.3484499853574866e-05,
+      "loss": 0.9544,
       "step": 7788
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 33.76646423339844,
-      "learning_rate": 4.388092613009924e-06,
-      "loss": 1.9249,
+      "epoch": 0.98,
+      "grad_norm": 10.441452026367188,
+      "learning_rate": 1.348366313851818e-05,
+      "loss": 2.1254,
       "step": 7789
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 24.660625457763672,
-      "learning_rate": 4.386088002405533e-06,
-      "loss": 2.0123,
+      "epoch": 0.98,
+      "grad_norm": 19.644514083862305,
+      "learning_rate": 1.348282642346149e-05,
+      "loss": 1.4696,
       "step": 7790
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 16.023160934448242,
-      "learning_rate": 4.384083391801143e-06,
-      "loss": 2.5144,
+      "epoch": 0.98,
+      "grad_norm": 6.888669013977051,
+      "learning_rate": 1.3481989708404803e-05,
+      "loss": 0.8234,
       "step": 7791
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 18.865354537963867,
-      "learning_rate": 4.382078781196753e-06,
-      "loss": 1.4033,
+      "epoch": 0.98,
+      "grad_norm": 6.575916767120361,
+      "learning_rate": 1.3481152993348117e-05,
+      "loss": 0.9467,
       "step": 7792
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 17.00137710571289,
-      "learning_rate": 4.380074170592363e-06,
-      "loss": 1.7571,
+      "epoch": 0.98,
+      "grad_norm": 15.154905319213867,
+      "learning_rate": 1.348031627829143e-05,
+      "loss": 1.0205,
       "step": 7793
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 22.194595336914062,
-      "learning_rate": 4.378069559987972e-06,
-      "loss": 1.5826,
+      "epoch": 0.98,
+      "grad_norm": 52.29448699951172,
+      "learning_rate": 1.347947956323474e-05,
+      "loss": 1.3675,
       "step": 7794
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 24.60336685180664,
-      "learning_rate": 4.376064949383582e-06,
-      "loss": 1.7541,
+      "epoch": 0.98,
+      "grad_norm": 25.070926666259766,
+      "learning_rate": 1.3478642848178054e-05,
+      "loss": 2.9792,
       "step": 7795
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 14.990224838256836,
-      "learning_rate": 4.3740603387791925e-06,
-      "loss": 1.1141,
+      "epoch": 0.98,
+      "grad_norm": 12.322017669677734,
+      "learning_rate": 1.3477806133121368e-05,
+      "loss": 1.9872,
       "step": 7796
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 17.370750427246094,
-      "learning_rate": 4.3720557281748026e-06,
-      "loss": 0.947,
+      "epoch": 0.98,
+      "grad_norm": 27.83733558654785,
+      "learning_rate": 1.3476969418064678e-05,
+      "loss": 0.6899,
       "step": 7797
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 32.88360595703125,
-      "learning_rate": 4.370051117570413e-06,
-      "loss": 1.035,
+      "epoch": 0.98,
+      "grad_norm": 15.455463409423828,
+      "learning_rate": 1.3476132703007992e-05,
+      "loss": 1.3138,
       "step": 7798
     },
     {
-      "epoch": 2.34,
-      "grad_norm": 13.272078514099121,
-      "learning_rate": 4.368046506966022e-06,
-      "loss": 0.8005,
+      "epoch": 0.98,
+      "grad_norm": 29.341999053955078,
+      "learning_rate": 1.3475295987951304e-05,
+      "loss": 1.3879,
       "step": 7799
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 28.69300651550293,
-      "learning_rate": 4.366041896361632e-06,
-      "loss": 1.6131,
-      "step": 7800
-    },
-    {
-      "epoch": 2.35,
-      "eval_loss": 0.17263449728488922,
-      "eval_runtime": 43.5025,
-      "eval_samples_per_second": 33.998,
-      "eval_steps_per_second": 33.998,
+      "epoch": 0.98,
+      "grad_norm": 21.08104705810547,
+      "learning_rate": 1.3474459272894618e-05,
+      "loss": 1.296,
       "step": 7800
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 21.5859317779541,
-      "learning_rate": 4.364037285757242e-06,
-      "loss": 1.961,
+      "epoch": 0.98,
+      "grad_norm": 1.2503612041473389,
+      "learning_rate": 1.347362255783793e-05,
+      "loss": 0.0344,
       "step": 7801
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 45.89167785644531,
-      "learning_rate": 4.362032675152852e-06,
-      "loss": 0.887,
+      "epoch": 0.98,
+      "grad_norm": 14.25670337677002,
+      "learning_rate": 1.3472785842781241e-05,
+      "loss": 2.0681,
       "step": 7802
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 9.14841079711914,
-      "learning_rate": 4.360028064548461e-06,
-      "loss": 0.82,
+      "epoch": 0.98,
+      "grad_norm": 8.220523834228516,
+      "learning_rate": 1.3471949127724555e-05,
+      "loss": 1.7112,
       "step": 7803
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 13.917359352111816,
-      "learning_rate": 4.358023453944072e-06,
-      "loss": 0.6227,
+      "epoch": 0.98,
+      "grad_norm": 5.848735332489014,
+      "learning_rate": 1.3471112412667865e-05,
+      "loss": 0.461,
       "step": 7804
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 10.889512062072754,
-      "learning_rate": 4.3560188433396815e-06,
-      "loss": 0.885,
+      "epoch": 0.98,
+      "grad_norm": 51.04117202758789,
+      "learning_rate": 1.3470275697611179e-05,
+      "loss": 1.8216,
       "step": 7805
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 11.779256820678711,
-      "learning_rate": 4.354014232735292e-06,
-      "loss": 1.6273,
+      "epoch": 0.98,
+      "grad_norm": 23.475095748901367,
+      "learning_rate": 1.3469438982554493e-05,
+      "loss": 2.4093,
       "step": 7806
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 9.816779136657715,
-      "learning_rate": 4.352009622130902e-06,
-      "loss": 1.2789,
+      "epoch": 0.98,
+      "grad_norm": 9.354192733764648,
+      "learning_rate": 1.3468602267497806e-05,
+      "loss": 0.8441,
       "step": 7807
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 10.623241424560547,
-      "learning_rate": 4.350005011526511e-06,
-      "loss": 0.888,
+      "epoch": 0.98,
+      "grad_norm": 7.1941728591918945,
+      "learning_rate": 1.3467765552441117e-05,
+      "loss": 0.6073,
       "step": 7808
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 15.834614753723145,
-      "learning_rate": 4.348000400922121e-06,
-      "loss": 0.8899,
+      "epoch": 0.98,
+      "grad_norm": 12.23647403717041,
+      "learning_rate": 1.346692883738443e-05,
+      "loss": 2.0373,
       "step": 7809
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 8.324065208435059,
-      "learning_rate": 4.345995790317731e-06,
-      "loss": 0.7989,
+      "epoch": 0.98,
+      "grad_norm": 14.28429889678955,
+      "learning_rate": 1.3466092122327742e-05,
+      "loss": 1.3354,
       "step": 7810
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 17.0432186126709,
-      "learning_rate": 4.343991179713341e-06,
-      "loss": 2.3516,
+      "epoch": 0.98,
+      "grad_norm": 6.032944679260254,
+      "learning_rate": 1.3465255407271054e-05,
+      "loss": 0.2188,
       "step": 7811
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 20.02086639404297,
-      "learning_rate": 4.3419865691089504e-06,
-      "loss": 0.8566,
+      "epoch": 0.98,
+      "grad_norm": 7.6251325607299805,
+      "learning_rate": 1.3464418692214368e-05,
+      "loss": 1.3673,
       "step": 7812
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 7.930890083312988,
-      "learning_rate": 4.339981958504561e-06,
-      "loss": 1.2351,
+      "epoch": 0.98,
+      "grad_norm": 38.285099029541016,
+      "learning_rate": 1.346358197715768e-05,
+      "loss": 2.4264,
       "step": 7813
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 30.14038848876953,
-      "learning_rate": 4.337977347900171e-06,
-      "loss": 1.7222,
+      "epoch": 0.98,
+      "grad_norm": 5.613325595855713,
+      "learning_rate": 1.3462745262100993e-05,
+      "loss": 1.3284,
       "step": 7814
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 12.645387649536133,
-      "learning_rate": 4.335972737295781e-06,
-      "loss": 1.3162,
+      "epoch": 0.98,
+      "grad_norm": 21.35129165649414,
+      "learning_rate": 1.3461908547044305e-05,
+      "loss": 0.6903,
       "step": 7815
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 17.415475845336914,
-      "learning_rate": 4.333968126691391e-06,
-      "loss": 1.8934,
+      "epoch": 0.98,
+      "grad_norm": 49.54762649536133,
+      "learning_rate": 1.3461071831987617e-05,
+      "loss": 2.2245,
       "step": 7816
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 13.79565143585205,
-      "learning_rate": 4.331963516087001e-06,
-      "loss": 1.4519,
+      "epoch": 0.98,
+      "grad_norm": 27.526241302490234,
+      "learning_rate": 1.3460235116930931e-05,
+      "loss": 2.7754,
       "step": 7817
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 14.080390930175781,
-      "learning_rate": 4.32995890548261e-06,
-      "loss": 0.899,
+      "epoch": 0.98,
+      "grad_norm": 120.94630432128906,
+      "learning_rate": 1.3459398401874241e-05,
+      "loss": 2.2525,
       "step": 7818
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 71.52656555175781,
-      "learning_rate": 4.32795429487822e-06,
-      "loss": 1.7353,
+      "epoch": 0.98,
+      "grad_norm": 12.913002967834473,
+      "learning_rate": 1.3458561686817555e-05,
+      "loss": 1.2333,
       "step": 7819
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 20.801849365234375,
-      "learning_rate": 4.32594968427383e-06,
-      "loss": 1.2644,
+      "epoch": 0.98,
+      "grad_norm": 16.263164520263672,
+      "learning_rate": 1.3457724971760868e-05,
+      "loss": 0.798,
       "step": 7820
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 40.78676223754883,
-      "learning_rate": 4.32394507366944e-06,
-      "loss": 2.8361,
+      "epoch": 0.98,
+      "grad_norm": 6.407655239105225,
+      "learning_rate": 1.3456888256704182e-05,
+      "loss": 0.7711,
       "step": 7821
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 14.759114265441895,
-      "learning_rate": 4.3219404630650504e-06,
-      "loss": 1.0354,
+      "epoch": 0.98,
+      "grad_norm": 17.457874298095703,
+      "learning_rate": 1.3456051541647492e-05,
+      "loss": 1.9697,
       "step": 7822
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 24.163408279418945,
-      "learning_rate": 4.31993585246066e-06,
-      "loss": 1.1916,
+      "epoch": 0.98,
+      "grad_norm": 13.347925186157227,
+      "learning_rate": 1.3455214826590806e-05,
+      "loss": 0.7008,
       "step": 7823
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 8.51475715637207,
-      "learning_rate": 4.31793124185627e-06,
-      "loss": 1.5931,
+      "epoch": 0.98,
+      "grad_norm": 26.760360717773438,
+      "learning_rate": 1.3454378111534118e-05,
+      "loss": 3.1434,
       "step": 7824
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 16.245559692382812,
-      "learning_rate": 4.31592663125188e-06,
-      "loss": 1.3994,
+      "epoch": 0.98,
+      "grad_norm": 14.519676208496094,
+      "learning_rate": 1.345354139647743e-05,
+      "loss": 0.5767,
       "step": 7825
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 78.04993438720703,
-      "learning_rate": 4.31392202064749e-06,
-      "loss": 2.4156,
+      "epoch": 0.98,
+      "grad_norm": 57.62624740600586,
+      "learning_rate": 1.3452704681420744e-05,
+      "loss": 1.4939,
       "step": 7826
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 11.318324089050293,
-      "learning_rate": 4.311917410043099e-06,
-      "loss": 1.7852,
+      "epoch": 0.98,
+      "grad_norm": 8.744379043579102,
+      "learning_rate": 1.3451867966364056e-05,
+      "loss": 1.1015,
       "step": 7827
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 9.104801177978516,
-      "learning_rate": 4.309912799438709e-06,
-      "loss": 0.7264,
+      "epoch": 0.98,
+      "grad_norm": 21.934654235839844,
+      "learning_rate": 1.345103125130737e-05,
+      "loss": 1.5129,
       "step": 7828
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 12.194317817687988,
-      "learning_rate": 4.307908188834319e-06,
-      "loss": 1.4084,
+      "epoch": 0.98,
+      "grad_norm": 10.702959060668945,
+      "learning_rate": 1.345019453625068e-05,
+      "loss": 1.1694,
       "step": 7829
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 39.487327575683594,
-      "learning_rate": 4.305903578229929e-06,
-      "loss": 1.4993,
+      "epoch": 0.98,
+      "grad_norm": 13.19983959197998,
+      "learning_rate": 1.3449357821193993e-05,
+      "loss": 1.1448,
       "step": 7830
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 24.5257511138916,
-      "learning_rate": 4.303898967625539e-06,
-      "loss": 1.7682,
+      "epoch": 0.98,
+      "grad_norm": 12.683655738830566,
+      "learning_rate": 1.3448521106137307e-05,
+      "loss": 1.0168,
       "step": 7831
     },
     {
-      "epoch": 2.35,
-      "grad_norm": 22.789091110229492,
-      "learning_rate": 4.301894357021149e-06,
-      "loss": 2.2842,
+      "epoch": 0.98,
+      "grad_norm": 44.0377197265625,
+      "learning_rate": 1.3447684391080617e-05,
+      "loss": 1.3388,
       "step": 7832
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 14.122344017028809,
-      "learning_rate": 4.299889746416759e-06,
-      "loss": 0.7481,
+      "epoch": 0.98,
+      "grad_norm": 11.622760772705078,
+      "learning_rate": 1.344684767602393e-05,
+      "loss": 0.6615,
       "step": 7833
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 9.010689735412598,
-      "learning_rate": 4.297885135812369e-06,
-      "loss": 1.0806,
+      "epoch": 0.98,
+      "grad_norm": 12.449114799499512,
+      "learning_rate": 1.3446010960967244e-05,
+      "loss": 2.0388,
       "step": 7834
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 20.649883270263672,
-      "learning_rate": 4.295880525207979e-06,
-      "loss": 1.4924,
+      "epoch": 0.98,
+      "grad_norm": 10.657881736755371,
+      "learning_rate": 1.3445174245910558e-05,
+      "loss": 0.3719,
       "step": 7835
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 24.006229400634766,
-      "learning_rate": 4.293875914603588e-06,
-      "loss": 1.4348,
+      "epoch": 0.98,
+      "grad_norm": 82.41954803466797,
+      "learning_rate": 1.3444337530853868e-05,
+      "loss": 1.0537,
       "step": 7836
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 30.37126350402832,
-      "learning_rate": 4.291871303999198e-06,
-      "loss": 1.5881,
+      "epoch": 0.98,
+      "grad_norm": 5.820470333099365,
+      "learning_rate": 1.3443500815797182e-05,
+      "loss": 0.3194,
       "step": 7837
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 10.542387962341309,
-      "learning_rate": 4.289866693394808e-06,
-      "loss": 0.7792,
+      "epoch": 0.98,
+      "grad_norm": 35.94833755493164,
+      "learning_rate": 1.3442664100740494e-05,
+      "loss": 1.8463,
       "step": 7838
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 19.010889053344727,
-      "learning_rate": 4.2878620827904185e-06,
-      "loss": 1.4441,
+      "epoch": 0.98,
+      "grad_norm": 23.072025299072266,
+      "learning_rate": 1.3441827385683806e-05,
+      "loss": 2.4423,
       "step": 7839
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 13.315712928771973,
-      "learning_rate": 4.285857472186028e-06,
-      "loss": 1.0043,
+      "epoch": 0.98,
+      "grad_norm": 26.65378761291504,
+      "learning_rate": 1.344099067062712e-05,
+      "loss": 1.1492,
       "step": 7840
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 11.758505821228027,
-      "learning_rate": 4.283852861581639e-06,
-      "loss": 1.304,
+      "epoch": 0.98,
+      "grad_norm": 44.74949264526367,
+      "learning_rate": 1.3440153955570431e-05,
+      "loss": 2.5089,
       "step": 7841
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 14.62195873260498,
-      "learning_rate": 4.281848250977248e-06,
-      "loss": 1.1863,
+      "epoch": 0.98,
+      "grad_norm": 9.149215698242188,
+      "learning_rate": 1.3439317240513745e-05,
+      "loss": 1.724,
       "step": 7842
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 17.724573135375977,
-      "learning_rate": 4.279843640372858e-06,
-      "loss": 1.0244,
+      "epoch": 0.98,
+      "grad_norm": 16.55966567993164,
+      "learning_rate": 1.3438480525457055e-05,
+      "loss": 1.4478,
       "step": 7843
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 49.88128662109375,
-      "learning_rate": 4.277839029768468e-06,
-      "loss": 1.6886,
+      "epoch": 0.98,
+      "grad_norm": 80.31683349609375,
+      "learning_rate": 1.3437643810400369e-05,
+      "loss": 2.3294,
       "step": 7844
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 28.493061065673828,
-      "learning_rate": 4.275834419164077e-06,
-      "loss": 1.7018,
+      "epoch": 0.98,
+      "grad_norm": 35.48666000366211,
+      "learning_rate": 1.3436807095343683e-05,
+      "loss": 1.6704,
       "step": 7845
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 17.226518630981445,
-      "learning_rate": 4.273829808559687e-06,
-      "loss": 1.3155,
+      "epoch": 0.98,
+      "grad_norm": 63.827598571777344,
+      "learning_rate": 1.3435970380286993e-05,
+      "loss": 2.8656,
       "step": 7846
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 14.533112525939941,
-      "learning_rate": 4.2718251979552975e-06,
-      "loss": 1.6489,
+      "epoch": 0.98,
+      "grad_norm": 21.52463722229004,
+      "learning_rate": 1.3435133665230307e-05,
+      "loss": 0.9325,
       "step": 7847
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 154.60177612304688,
-      "learning_rate": 4.2698205873509075e-06,
-      "loss": 3.9771,
+      "epoch": 0.98,
+      "grad_norm": 24.884567260742188,
+      "learning_rate": 1.343429695017362e-05,
+      "loss": 2.3111,
       "step": 7848
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 24.33751678466797,
-      "learning_rate": 4.267815976746517e-06,
-      "loss": 1.5103,
+      "epoch": 0.99,
+      "grad_norm": 13.400162696838379,
+      "learning_rate": 1.3433460235116934e-05,
+      "loss": 2.0544,
       "step": 7849
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 7.450550556182861,
-      "learning_rate": 4.265811366142128e-06,
-      "loss": 0.8299,
+      "epoch": 0.99,
+      "grad_norm": 12.546667098999023,
+      "learning_rate": 1.3432623520060244e-05,
+      "loss": 1.994,
       "step": 7850
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 11.310479164123535,
-      "learning_rate": 4.263806755537737e-06,
-      "loss": 1.0471,
+      "epoch": 0.99,
+      "grad_norm": 60.43228530883789,
+      "learning_rate": 1.3431786805003558e-05,
+      "loss": 2.4401,
       "step": 7851
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 22.815073013305664,
-      "learning_rate": 4.261802144933347e-06,
-      "loss": 2.6051,
+      "epoch": 0.99,
+      "grad_norm": 11.646256446838379,
+      "learning_rate": 1.343095008994687e-05,
+      "loss": 1.1565,
       "step": 7852
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 49.23891067504883,
-      "learning_rate": 4.259797534328957e-06,
-      "loss": 1.1229,
+      "epoch": 0.99,
+      "grad_norm": 18.161865234375,
+      "learning_rate": 1.3430113374890182e-05,
+      "loss": 1.9924,
       "step": 7853
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 12.846585273742676,
-      "learning_rate": 4.257792923724567e-06,
-      "loss": 1.0635,
+      "epoch": 0.99,
+      "grad_norm": 17.5528507232666,
+      "learning_rate": 1.3429276659833495e-05,
+      "loss": 2.2338,
       "step": 7854
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 7.844564437866211,
-      "learning_rate": 4.2557883131201764e-06,
-      "loss": 0.8797,
+      "epoch": 0.99,
+      "grad_norm": 13.696662902832031,
+      "learning_rate": 1.3428439944776807e-05,
+      "loss": 1.6114,
       "step": 7855
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 12.018916130065918,
-      "learning_rate": 4.2537837025157865e-06,
-      "loss": 1.2833,
+      "epoch": 0.99,
+      "grad_norm": 12.333256721496582,
+      "learning_rate": 1.3427603229720121e-05,
+      "loss": 1.2289,
       "step": 7856
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 15.914363861083984,
-      "learning_rate": 4.251779091911397e-06,
-      "loss": 1.2172,
+      "epoch": 0.99,
+      "grad_norm": 20.567373275756836,
+      "learning_rate": 1.3426766514663431e-05,
+      "loss": 1.2595,
       "step": 7857
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 20.0510311126709,
-      "learning_rate": 4.249774481307007e-06,
-      "loss": 0.7942,
+      "epoch": 0.99,
+      "grad_norm": 18.33525848388672,
+      "learning_rate": 1.3425929799606745e-05,
+      "loss": 1.8468,
       "step": 7858
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 45.73860549926758,
-      "learning_rate": 4.247769870702617e-06,
-      "loss": 2.2663,
+      "epoch": 0.99,
+      "grad_norm": 32.14639663696289,
+      "learning_rate": 1.3425093084550058e-05,
+      "loss": 2.8181,
       "step": 7859
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 18.577871322631836,
-      "learning_rate": 4.245765260098226e-06,
-      "loss": 1.8624,
+      "epoch": 0.99,
+      "grad_norm": 8.10889720916748,
+      "learning_rate": 1.3424256369493369e-05,
+      "loss": 1.4549,
       "step": 7860
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 14.684171676635742,
-      "learning_rate": 4.243760649493836e-06,
-      "loss": 1.2921,
+      "epoch": 0.99,
+      "grad_norm": 17.031816482543945,
+      "learning_rate": 1.3423419654436682e-05,
+      "loss": 2.38,
       "step": 7861
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 20.421449661254883,
-      "learning_rate": 4.241756038889446e-06,
-      "loss": 1.2322,
+      "epoch": 0.99,
+      "grad_norm": 16.39002227783203,
+      "learning_rate": 1.3422582939379996e-05,
+      "loss": 1.4777,
       "step": 7862
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 11.350706100463867,
-      "learning_rate": 4.239751428285056e-06,
-      "loss": 1.3288,
+      "epoch": 0.99,
+      "grad_norm": 10.644102096557617,
+      "learning_rate": 1.3421746224323308e-05,
+      "loss": 1.2411,
       "step": 7863
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 21.545028686523438,
-      "learning_rate": 4.2377468176806655e-06,
-      "loss": 1.4258,
+      "epoch": 0.99,
+      "grad_norm": 14.060710906982422,
+      "learning_rate": 1.342090950926662e-05,
+      "loss": 2.5573,
       "step": 7864
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 18.22039222717285,
-      "learning_rate": 4.2357422070762764e-06,
-      "loss": 1.4584,
+      "epoch": 0.99,
+      "grad_norm": 16.674654006958008,
+      "learning_rate": 1.3420072794209934e-05,
+      "loss": 1.6149,
       "step": 7865
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 28.657711029052734,
-      "learning_rate": 4.233737596471886e-06,
-      "loss": 2.444,
+      "epoch": 0.99,
+      "grad_norm": 16.86676025390625,
+      "learning_rate": 1.3419236079153246e-05,
+      "loss": 1.0251,
       "step": 7866
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 7.091336727142334,
-      "learning_rate": 4.231732985867496e-06,
-      "loss": 1.0055,
+      "epoch": 0.99,
+      "grad_norm": 12.13718032836914,
+      "learning_rate": 1.3418399364096557e-05,
+      "loss": 1.0789,
       "step": 7867
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 10.07259750366211,
-      "learning_rate": 4.229728375263106e-06,
-      "loss": 0.9513,
+      "epoch": 0.99,
+      "grad_norm": 21.83576202392578,
+      "learning_rate": 1.341756264903987e-05,
+      "loss": 0.7112,
       "step": 7868
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 32.32286834716797,
-      "learning_rate": 4.227723764658715e-06,
-      "loss": 1.2243,
+      "epoch": 0.99,
+      "grad_norm": 19.9146728515625,
+      "learning_rate": 1.3416725933983183e-05,
+      "loss": 3.1556,
       "step": 7869
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 14.63747787475586,
-      "learning_rate": 4.225719154054325e-06,
-      "loss": 1.3113,
+      "epoch": 0.99,
+      "grad_norm": 16.90733528137207,
+      "learning_rate": 1.3415889218926497e-05,
+      "loss": 1.7789,
       "step": 7870
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 13.216215133666992,
-      "learning_rate": 4.223714543449935e-06,
-      "loss": 1.3608,
+      "epoch": 0.99,
+      "grad_norm": 15.959580421447754,
+      "learning_rate": 1.3415052503869807e-05,
+      "loss": 1.3213,
       "step": 7871
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 22.928085327148438,
-      "learning_rate": 4.221709932845545e-06,
-      "loss": 1.5632,
+      "epoch": 0.99,
+      "grad_norm": 7.629356384277344,
+      "learning_rate": 1.341421578881312e-05,
+      "loss": 0.6035,
       "step": 7872
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 14.180635452270508,
-      "learning_rate": 4.2197053222411546e-06,
-      "loss": 1.0618,
+      "epoch": 0.99,
+      "grad_norm": 12.714489936828613,
+      "learning_rate": 1.3413379073756434e-05,
+      "loss": 1.6759,
       "step": 7873
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 13.465377807617188,
-      "learning_rate": 4.217700711636765e-06,
-      "loss": 2.2106,
+      "epoch": 0.99,
+      "grad_norm": 50.399253845214844,
+      "learning_rate": 1.3412542358699745e-05,
+      "loss": 2.6481,
       "step": 7874
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 21.83644676208496,
-      "learning_rate": 4.215696101032375e-06,
-      "loss": 1.1141,
+      "epoch": 0.99,
+      "grad_norm": 16.95625114440918,
+      "learning_rate": 1.3411705643643058e-05,
+      "loss": 1.0485,
       "step": 7875
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 9.465641021728516,
-      "learning_rate": 4.213691490427985e-06,
-      "loss": 0.9889,
+      "epoch": 0.99,
+      "grad_norm": 12.690945625305176,
+      "learning_rate": 1.3410868928586372e-05,
+      "loss": 1.2125,
       "step": 7876
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 38.847965240478516,
-      "learning_rate": 4.211686879823594e-06,
-      "loss": 2.6969,
+      "epoch": 0.99,
+      "grad_norm": 12.52241039276123,
+      "learning_rate": 1.3410032213529684e-05,
+      "loss": 2.4324,
       "step": 7877
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 12.433218002319336,
-      "learning_rate": 4.209682269219205e-06,
-      "loss": 0.5288,
+      "epoch": 0.99,
+      "grad_norm": 27.132232666015625,
+      "learning_rate": 1.3409195498472996e-05,
+      "loss": 1.5371,
       "step": 7878
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 13.307668685913086,
-      "learning_rate": 4.207677658614814e-06,
-      "loss": 0.9523,
+      "epoch": 0.99,
+      "grad_norm": 15.67525863647461,
+      "learning_rate": 1.340835878341631e-05,
+      "loss": 1.0848,
       "step": 7879
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 57.23488235473633,
-      "learning_rate": 4.205673048010424e-06,
-      "loss": 3.0293,
+      "epoch": 0.99,
+      "grad_norm": 7.077928066253662,
+      "learning_rate": 1.3407522068359621e-05,
+      "loss": 1.9246,
       "step": 7880
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 16.167064666748047,
-      "learning_rate": 4.203668437406034e-06,
-      "loss": 1.9897,
+      "epoch": 0.99,
+      "grad_norm": 13.227967262268066,
+      "learning_rate": 1.3406685353302933e-05,
+      "loss": 1.9851,
       "step": 7881
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 29.6549072265625,
-      "learning_rate": 4.2016638268016445e-06,
-      "loss": 1.3998,
+      "epoch": 0.99,
+      "grad_norm": 17.633745193481445,
+      "learning_rate": 1.3405848638246245e-05,
+      "loss": 1.5183,
       "step": 7882
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 10.331962585449219,
-      "learning_rate": 4.199659216197254e-06,
-      "loss": 0.8686,
+      "epoch": 0.99,
+      "grad_norm": 8.09128189086914,
+      "learning_rate": 1.3405011923189559e-05,
+      "loss": 2.1152,
       "step": 7883
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 16.08567237854004,
-      "learning_rate": 4.197654605592864e-06,
-      "loss": 1.3522,
+      "epoch": 0.99,
+      "grad_norm": 12.385307312011719,
+      "learning_rate": 1.3404175208132873e-05,
+      "loss": 1.4056,
       "step": 7884
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 21.387054443359375,
-      "learning_rate": 4.195649994988474e-06,
-      "loss": 1.0533,
+      "epoch": 0.99,
+      "grad_norm": 17.078733444213867,
+      "learning_rate": 1.3403338493076183e-05,
+      "loss": 1.6283,
       "step": 7885
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 18.664920806884766,
-      "learning_rate": 4.193645384384083e-06,
-      "loss": 1.7003,
+      "epoch": 0.99,
+      "grad_norm": 9.437175750732422,
+      "learning_rate": 1.3402501778019496e-05,
+      "loss": 0.3968,
       "step": 7886
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 12.59611988067627,
-      "learning_rate": 4.191640773779694e-06,
-      "loss": 1.0916,
+      "epoch": 0.99,
+      "grad_norm": 21.20670509338379,
+      "learning_rate": 1.340166506296281e-05,
+      "loss": 1.8604,
       "step": 7887
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 11.082344055175781,
-      "learning_rate": 4.189636163175303e-06,
-      "loss": 1.3448,
+      "epoch": 0.99,
+      "grad_norm": 7.615916728973389,
+      "learning_rate": 1.340082834790612e-05,
+      "loss": 0.7451,
       "step": 7888
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 15.192607879638672,
-      "learning_rate": 4.187631552570913e-06,
-      "loss": 1.1133,
+      "epoch": 0.99,
+      "grad_norm": 6.689842224121094,
+      "learning_rate": 1.3399991632849434e-05,
+      "loss": 1.3682,
       "step": 7889
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 12.27822494506836,
-      "learning_rate": 4.1856269419665235e-06,
-      "loss": 1.4396,
+      "epoch": 0.99,
+      "grad_norm": 16.554161071777344,
+      "learning_rate": 1.3399154917792748e-05,
+      "loss": 1.4628,
       "step": 7890
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 18.4125919342041,
-      "learning_rate": 4.1836223313621335e-06,
-      "loss": 0.9025,
+      "epoch": 0.99,
+      "grad_norm": 9.20697021484375,
+      "learning_rate": 1.339831820273606e-05,
+      "loss": 1.1317,
       "step": 7891
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 20.866134643554688,
-      "learning_rate": 4.181617720757743e-06,
-      "loss": 1.1648,
+      "epoch": 0.99,
+      "grad_norm": 12.420833587646484,
+      "learning_rate": 1.3397481487679372e-05,
+      "loss": 0.8767,
       "step": 7892
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 16.63862419128418,
-      "learning_rate": 4.179613110153353e-06,
-      "loss": 1.4421,
+      "epoch": 0.99,
+      "grad_norm": 9.886767387390137,
+      "learning_rate": 1.3396644772622685e-05,
+      "loss": 0.9803,
       "step": 7893
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 9.658064842224121,
-      "learning_rate": 4.177608499548963e-06,
-      "loss": 0.6854,
+      "epoch": 0.99,
+      "grad_norm": 22.698192596435547,
+      "learning_rate": 1.3395808057565997e-05,
+      "loss": 3.4859,
       "step": 7894
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 12.731464385986328,
-      "learning_rate": 4.175603888944573e-06,
-      "loss": 1.374,
+      "epoch": 0.99,
+      "grad_norm": 20.673337936401367,
+      "learning_rate": 1.3394971342509309e-05,
+      "loss": 0.9118,
       "step": 7895
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 8.585317611694336,
-      "learning_rate": 4.173599278340183e-06,
-      "loss": 0.5775,
+      "epoch": 0.99,
+      "grad_norm": 19.7998104095459,
+      "learning_rate": 1.3394134627452621e-05,
+      "loss": 0.3165,
       "step": 7896
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 24.50328826904297,
-      "learning_rate": 4.171594667735792e-06,
-      "loss": 1.5935,
+      "epoch": 0.99,
+      "grad_norm": 8.360078811645508,
+      "learning_rate": 1.3393297912395935e-05,
+      "loss": 1.5516,
       "step": 7897
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 16.08628273010254,
-      "learning_rate": 4.1695900571314024e-06,
-      "loss": 1.0627,
+      "epoch": 0.99,
+      "grad_norm": 21.86443519592285,
+      "learning_rate": 1.3392461197339248e-05,
+      "loss": 1.0845,
       "step": 7898
     },
     {
-      "epoch": 2.37,
-      "grad_norm": 15.948634147644043,
-      "learning_rate": 4.1675854465270125e-06,
-      "loss": 1.4136,
+      "epoch": 0.99,
+      "grad_norm": 14.555537223815918,
+      "learning_rate": 1.3391624482282559e-05,
+      "loss": 2.7609,
       "step": 7899
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 31.818986892700195,
-      "learning_rate": 4.165580835922623e-06,
-      "loss": 1.8128,
+      "epoch": 0.99,
+      "grad_norm": 9.91486644744873,
+      "learning_rate": 1.3390787767225872e-05,
+      "loss": 0.8442,
       "step": 7900
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 37.07505798339844,
-      "learning_rate": 4.163576225318232e-06,
-      "loss": 1.9246,
+      "epoch": 0.99,
+      "grad_norm": 16.55044174194336,
+      "learning_rate": 1.3389951052169186e-05,
+      "loss": 2.1581,
       "step": 7901
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 15.36788558959961,
-      "learning_rate": 4.161571614713843e-06,
-      "loss": 1.9555,
+      "epoch": 0.99,
+      "grad_norm": 27.388978958129883,
+      "learning_rate": 1.3389114337112496e-05,
+      "loss": 1.3599,
       "step": 7902
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 17.946441650390625,
-      "learning_rate": 4.159567004109452e-06,
-      "loss": 1.1702,
+      "epoch": 0.99,
+      "grad_norm": 5.923564910888672,
+      "learning_rate": 1.338827762205581e-05,
+      "loss": 0.4826,
       "step": 7903
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 12.651650428771973,
-      "learning_rate": 4.157562393505062e-06,
-      "loss": 1.0666,
+      "epoch": 0.99,
+      "grad_norm": 11.348058700561523,
+      "learning_rate": 1.3387440906999124e-05,
+      "loss": 1.8191,
       "step": 7904
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 31.045000076293945,
-      "learning_rate": 4.155557782900672e-06,
-      "loss": 1.553,
+      "epoch": 0.99,
+      "grad_norm": 7.931649684906006,
+      "learning_rate": 1.3386604191942435e-05,
+      "loss": 1.7959,
       "step": 7905
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 11.69579029083252,
-      "learning_rate": 4.153553172296282e-06,
-      "loss": 1.5206,
+      "epoch": 0.99,
+      "grad_norm": 9.703704833984375,
+      "learning_rate": 1.3385767476885747e-05,
+      "loss": 1.1969,
       "step": 7906
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 37.5781135559082,
-      "learning_rate": 4.1515485616918915e-06,
-      "loss": 2.1329,
+      "epoch": 0.99,
+      "grad_norm": 14.642962455749512,
+      "learning_rate": 1.3384930761829061e-05,
+      "loss": 0.6981,
       "step": 7907
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 21.697532653808594,
-      "learning_rate": 4.149543951087502e-06,
-      "loss": 1.4265,
+      "epoch": 0.99,
+      "grad_norm": 9.679839134216309,
+      "learning_rate": 1.3384094046772373e-05,
+      "loss": 1.6064,
       "step": 7908
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 9.745403289794922,
-      "learning_rate": 4.147539340483112e-06,
-      "loss": 1.1323,
+      "epoch": 0.99,
+      "grad_norm": 7.588835716247559,
+      "learning_rate": 1.3383257331715685e-05,
+      "loss": 0.8063,
       "step": 7909
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 15.652579307556152,
-      "learning_rate": 4.145534729878721e-06,
-      "loss": 1.3761,
+      "epoch": 0.99,
+      "grad_norm": 5.758579730987549,
+      "learning_rate": 1.3382420616658997e-05,
+      "loss": 0.5944,
       "step": 7910
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 12.394492149353027,
-      "learning_rate": 4.143530119274332e-06,
-      "loss": 0.9479,
+      "epoch": 0.99,
+      "grad_norm": 11.920259475708008,
+      "learning_rate": 1.338158390160231e-05,
+      "loss": 1.72,
       "step": 7911
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 19.919248580932617,
-      "learning_rate": 4.141525508669941e-06,
-      "loss": 1.2497,
+      "epoch": 0.99,
+      "grad_norm": 29.732891082763672,
+      "learning_rate": 1.3380747186545624e-05,
+      "loss": 1.4716,
       "step": 7912
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 13.425742149353027,
-      "learning_rate": 4.139520898065551e-06,
-      "loss": 1.0198,
+      "epoch": 0.99,
+      "grad_norm": 20.857481002807617,
+      "learning_rate": 1.3379910471488934e-05,
+      "loss": 2.462,
       "step": 7913
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 17.13833999633789,
-      "learning_rate": 4.13751628746116e-06,
-      "loss": 2.0541,
+      "epoch": 0.99,
+      "grad_norm": 9.390600204467773,
+      "learning_rate": 1.3379073756432248e-05,
+      "loss": 2.1183,
       "step": 7914
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 13.618558883666992,
-      "learning_rate": 4.135511676856771e-06,
-      "loss": 1.3842,
+      "epoch": 0.99,
+      "grad_norm": 16.786588668823242,
+      "learning_rate": 1.3378237041375562e-05,
+      "loss": 1.5882,
       "step": 7915
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 5.930107593536377,
-      "learning_rate": 4.1335070662523806e-06,
-      "loss": 0.517,
+      "epoch": 0.99,
+      "grad_norm": 16.013607025146484,
+      "learning_rate": 1.3377400326318872e-05,
+      "loss": 1.1315,
       "step": 7916
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 4.825802803039551,
-      "learning_rate": 4.131502455647991e-06,
-      "loss": 0.3964,
+      "epoch": 0.99,
+      "grad_norm": 13.470029830932617,
+      "learning_rate": 1.3376563611262186e-05,
+      "loss": 1.48,
       "step": 7917
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 25.973148345947266,
-      "learning_rate": 4.129497845043601e-06,
-      "loss": 0.9199,
+      "epoch": 0.99,
+      "grad_norm": 8.650081634521484,
+      "learning_rate": 1.33757268962055e-05,
+      "loss": 0.4609,
       "step": 7918
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 9.384675979614258,
-      "learning_rate": 4.127493234439211e-06,
-      "loss": 1.1662,
+      "epoch": 0.99,
+      "grad_norm": 16.66744613647461,
+      "learning_rate": 1.3374890181148811e-05,
+      "loss": 2.0696,
       "step": 7919
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 59.6272087097168,
-      "learning_rate": 4.12548862383482e-06,
-      "loss": 2.4387,
-      "step": 7920
-    },
-    {
-      "epoch": 2.38,
-      "eval_loss": 0.17472867667675018,
-      "eval_runtime": 43.5367,
-      "eval_samples_per_second": 33.971,
-      "eval_steps_per_second": 33.971,
+      "epoch": 0.99,
+      "grad_norm": 13.903091430664062,
+      "learning_rate": 1.3374053466092123e-05,
+      "loss": 1.9955,
       "step": 7920
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 17.422163009643555,
-      "learning_rate": 4.12348401323043e-06,
-      "loss": 0.9547,
+      "epoch": 0.99,
+      "grad_norm": 23.533159255981445,
+      "learning_rate": 1.3373216751035435e-05,
+      "loss": 2.6719,
       "step": 7921
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 19.566699981689453,
-      "learning_rate": 4.12147940262604e-06,
-      "loss": 1.364,
+      "epoch": 0.99,
+      "grad_norm": 28.801076889038086,
+      "learning_rate": 1.3372380035978749e-05,
+      "loss": 2.1541,
       "step": 7922
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 13.921435356140137,
-      "learning_rate": 4.11947479202165e-06,
-      "loss": 1.5498,
+      "epoch": 0.99,
+      "grad_norm": 31.399919509887695,
+      "learning_rate": 1.337154332092206e-05,
+      "loss": 2.4247,
       "step": 7923
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 46.268829345703125,
-      "learning_rate": 4.11747018141726e-06,
-      "loss": 1.7792,
+      "epoch": 0.99,
+      "grad_norm": 19.266828536987305,
+      "learning_rate": 1.3370706605865373e-05,
+      "loss": 2.8681,
       "step": 7924
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 37.72319793701172,
-      "learning_rate": 4.11546557081287e-06,
-      "loss": 1.582,
+      "epoch": 0.99,
+      "grad_norm": 8.171679496765137,
+      "learning_rate": 1.3369869890808686e-05,
+      "loss": 0.48,
       "step": 7925
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 13.244497299194336,
-      "learning_rate": 4.11346096020848e-06,
-      "loss": 1.7341,
+      "epoch": 0.99,
+      "grad_norm": 35.95570373535156,
+      "learning_rate": 1.3369033175752e-05,
+      "loss": 2.1517,
       "step": 7926
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 10.264554977416992,
-      "learning_rate": 4.11145634960409e-06,
-      "loss": 0.902,
+      "epoch": 0.99,
+      "grad_norm": 9.317283630371094,
+      "learning_rate": 1.336819646069531e-05,
+      "loss": 0.9883,
       "step": 7927
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 16.762826919555664,
-      "learning_rate": 4.1094517389997e-06,
-      "loss": 1.2989,
+      "epoch": 0.99,
+      "grad_norm": 13.376811981201172,
+      "learning_rate": 1.3367359745638624e-05,
+      "loss": 0.2741,
       "step": 7928
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 27.864768981933594,
-      "learning_rate": 4.107447128395309e-06,
-      "loss": 1.0548,
+      "epoch": 1.0,
+      "grad_norm": 27.397188186645508,
+      "learning_rate": 1.3366523030581938e-05,
+      "loss": 0.8773,
       "step": 7929
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 9.768196105957031,
-      "learning_rate": 4.105442517790919e-06,
-      "loss": 0.7929,
+      "epoch": 1.0,
+      "grad_norm": 5.359547138214111,
+      "learning_rate": 1.3365686315525248e-05,
+      "loss": 0.2089,
       "step": 7930
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 8.759325981140137,
-      "learning_rate": 4.103437907186529e-06,
-      "loss": 0.6668,
+      "epoch": 1.0,
+      "grad_norm": 11.692357063293457,
+      "learning_rate": 1.3364849600468562e-05,
+      "loss": 1.8957,
       "step": 7931
     },
     {
-      "epoch": 2.38,
-      "grad_norm": 29.228063583374023,
-      "learning_rate": 4.101433296582139e-06,
-      "loss": 1.5968,
+      "epoch": 1.0,
+      "grad_norm": 11.032751083374023,
+      "learning_rate": 1.3364012885411875e-05,
+      "loss": 1.8149,
       "step": 7932
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 15.066993713378906,
-      "learning_rate": 4.0994286859777495e-06,
-      "loss": 1.457,
+      "epoch": 1.0,
+      "grad_norm": 22.042510986328125,
+      "learning_rate": 1.3363176170355187e-05,
+      "loss": 1.264,
       "step": 7933
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 12.962151527404785,
-      "learning_rate": 4.097424075373359e-06,
-      "loss": 1.568,
+      "epoch": 1.0,
+      "grad_norm": 24.833213806152344,
+      "learning_rate": 1.3362339455298499e-05,
+      "loss": 1.1884,
       "step": 7934
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 9.651278495788574,
-      "learning_rate": 4.095419464768969e-06,
-      "loss": 0.9769,
+      "epoch": 1.0,
+      "grad_norm": 14.692005157470703,
+      "learning_rate": 1.3361502740241811e-05,
+      "loss": 1.085,
       "step": 7935
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 27.313953399658203,
-      "learning_rate": 4.093414854164579e-06,
-      "loss": 1.1726,
+      "epoch": 1.0,
+      "grad_norm": 8.472384452819824,
+      "learning_rate": 1.3360666025185125e-05,
+      "loss": 1.341,
       "step": 7936
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 30.882871627807617,
-      "learning_rate": 4.091410243560189e-06,
-      "loss": 2.3507,
+      "epoch": 1.0,
+      "grad_norm": 19.474328994750977,
+      "learning_rate": 1.3359829310128437e-05,
+      "loss": 1.6633,
       "step": 7937
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 16.88389015197754,
-      "learning_rate": 4.089405632955798e-06,
-      "loss": 0.9228,
+      "epoch": 1.0,
+      "grad_norm": 12.89220905303955,
+      "learning_rate": 1.3358992595071749e-05,
+      "loss": 0.9524,
       "step": 7938
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 54.354469299316406,
-      "learning_rate": 4.087401022351409e-06,
-      "loss": 2.5421,
+      "epoch": 1.0,
+      "grad_norm": 38.02602767944336,
+      "learning_rate": 1.3358155880015062e-05,
+      "loss": 1.6038,
       "step": 7939
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 18.750852584838867,
-      "learning_rate": 4.085396411747018e-06,
-      "loss": 1.2235,
+      "epoch": 1.0,
+      "grad_norm": 7.536783218383789,
+      "learning_rate": 1.3357319164958376e-05,
+      "loss": 0.5629,
       "step": 7940
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 7.334155559539795,
-      "learning_rate": 4.0833918011426284e-06,
-      "loss": 0.647,
+      "epoch": 1.0,
+      "grad_norm": 23.697574615478516,
+      "learning_rate": 1.3356482449901686e-05,
+      "loss": 2.2301,
       "step": 7941
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 14.79666519165039,
-      "learning_rate": 4.0813871905382385e-06,
-      "loss": 0.8448,
+      "epoch": 1.0,
+      "grad_norm": 16.0355281829834,
+      "learning_rate": 1.3355645734845e-05,
+      "loss": 1.0149,
       "step": 7942
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 35.3593864440918,
-      "learning_rate": 4.079382579933849e-06,
-      "loss": 1.4009,
+      "epoch": 1.0,
+      "grad_norm": 11.679667472839355,
+      "learning_rate": 1.3354809019788313e-05,
+      "loss": 1.4238,
       "step": 7943
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 25.9478702545166,
-      "learning_rate": 4.077377969329458e-06,
-      "loss": 1.7573,
+      "epoch": 1.0,
+      "grad_norm": 7.693532466888428,
+      "learning_rate": 1.3353972304731624e-05,
+      "loss": 1.9168,
       "step": 7944
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 13.793614387512207,
-      "learning_rate": 4.075373358725068e-06,
-      "loss": 1.5591,
+      "epoch": 1.0,
+      "grad_norm": 23.826108932495117,
+      "learning_rate": 1.3353135589674937e-05,
+      "loss": 2.9985,
       "step": 7945
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 15.494868278503418,
-      "learning_rate": 4.073368748120678e-06,
-      "loss": 1.6407,
+      "epoch": 1.0,
+      "grad_norm": 16.03788185119629,
+      "learning_rate": 1.3352298874618251e-05,
+      "loss": 0.7254,
       "step": 7946
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 26.147010803222656,
-      "learning_rate": 4.071364137516287e-06,
-      "loss": 1.6419,
+      "epoch": 1.0,
+      "grad_norm": 57.76662063598633,
+      "learning_rate": 1.3351462159561563e-05,
+      "loss": 1.2052,
       "step": 7947
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 19.630136489868164,
-      "learning_rate": 4.069359526911898e-06,
-      "loss": 1.5473,
+      "epoch": 1.0,
+      "grad_norm": 128.8223114013672,
+      "learning_rate": 1.3350625444504875e-05,
+      "loss": 2.3229,
       "step": 7948
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 75.69630432128906,
-      "learning_rate": 4.0673549163075074e-06,
-      "loss": 2.2357,
+      "epoch": 1.0,
+      "grad_norm": 12.937859535217285,
+      "learning_rate": 1.3349788729448187e-05,
+      "loss": 1.7812,
       "step": 7949
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 82.69189453125,
-      "learning_rate": 4.0653503057031175e-06,
-      "loss": 1.675,
+      "epoch": 1.0,
+      "grad_norm": 83.09746551513672,
+      "learning_rate": 1.33489520143915e-05,
+      "loss": 1.2496,
       "step": 7950
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 12.869571685791016,
-      "learning_rate": 4.063345695098728e-06,
-      "loss": 1.1058,
+      "epoch": 1.0,
+      "grad_norm": 18.56004524230957,
+      "learning_rate": 1.3348115299334812e-05,
+      "loss": 3.7962,
       "step": 7951
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 16.09258270263672,
-      "learning_rate": 4.061341084494338e-06,
-      "loss": 1.3042,
+      "epoch": 1.0,
+      "grad_norm": 7.800342559814453,
+      "learning_rate": 1.3347278584278124e-05,
+      "loss": 0.9315,
       "step": 7952
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 15.368246078491211,
-      "learning_rate": 4.059336473889947e-06,
-      "loss": 0.7744,
+      "epoch": 1.0,
+      "grad_norm": 10.73358154296875,
+      "learning_rate": 1.3346441869221438e-05,
+      "loss": 1.1961,
       "step": 7953
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 56.09071350097656,
-      "learning_rate": 4.057331863285557e-06,
-      "loss": 1.4905,
+      "epoch": 1.0,
+      "grad_norm": 23.34194564819336,
+      "learning_rate": 1.3345605154164752e-05,
+      "loss": 1.1545,
       "step": 7954
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 11.899616241455078,
-      "learning_rate": 4.055327252681167e-06,
-      "loss": 1.1252,
+      "epoch": 1.0,
+      "grad_norm": 37.151573181152344,
+      "learning_rate": 1.3344768439108062e-05,
+      "loss": 1.6355,
       "step": 7955
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 17.915721893310547,
-      "learning_rate": 4.053322642076777e-06,
-      "loss": 1.8459,
+      "epoch": 1.0,
+      "grad_norm": 29.568134307861328,
+      "learning_rate": 1.3343931724051376e-05,
+      "loss": 0.9761,
       "step": 7956
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 12.431897163391113,
-      "learning_rate": 4.051318031472386e-06,
-      "loss": 1.0242,
+      "epoch": 1.0,
+      "grad_norm": 8.467724800109863,
+      "learning_rate": 1.334309500899469e-05,
+      "loss": 1.0828,
       "step": 7957
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 15.987489700317383,
-      "learning_rate": 4.0493134208679965e-06,
-      "loss": 2.4057,
+      "epoch": 1.0,
+      "grad_norm": 12.173320770263672,
+      "learning_rate": 1.3342258293938e-05,
+      "loss": 1.5829,
       "step": 7958
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 14.070905685424805,
-      "learning_rate": 4.0473088102636066e-06,
-      "loss": 1.2005,
+      "epoch": 1.0,
+      "grad_norm": 7.051929950714111,
+      "learning_rate": 1.3341421578881313e-05,
+      "loss": 0.9812,
       "step": 7959
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 26.776086807250977,
-      "learning_rate": 4.045304199659217e-06,
-      "loss": 1.1287,
+      "epoch": 1.0,
+      "grad_norm": 10.67720890045166,
+      "learning_rate": 1.3340584863824627e-05,
+      "loss": 1.0112,
       "step": 7960
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 21.03402328491211,
-      "learning_rate": 4.043299589054827e-06,
-      "loss": 1.6414,
+      "epoch": 1.0,
+      "grad_norm": 11.94753360748291,
+      "learning_rate": 1.3339748148767939e-05,
+      "loss": 0.6997,
       "step": 7961
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 16.39931297302246,
-      "learning_rate": 4.041294978450436e-06,
-      "loss": 1.7569,
+      "epoch": 1.0,
+      "grad_norm": 39.415924072265625,
+      "learning_rate": 1.333891143371125e-05,
+      "loss": 1.9105,
       "step": 7962
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 22.343650817871094,
-      "learning_rate": 4.039290367846046e-06,
-      "loss": 1.0703,
+      "epoch": 1.0,
+      "grad_norm": 63.462162017822266,
+      "learning_rate": 1.3338074718654563e-05,
+      "loss": 2.275,
       "step": 7963
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 21.395599365234375,
-      "learning_rate": 4.037285757241656e-06,
-      "loss": 2.1687,
+      "epoch": 1.0,
+      "grad_norm": 17.261051177978516,
+      "learning_rate": 1.3337238003597876e-05,
+      "loss": 1.5312,
       "step": 7964
     },
     {
-      "epoch": 2.39,
-      "grad_norm": 53.524776458740234,
-      "learning_rate": 4.035281146637266e-06,
-      "loss": 1.6589,
+      "epoch": 1.0,
+      "grad_norm": 23.740114212036133,
+      "learning_rate": 1.3336401288541188e-05,
+      "loss": 1.1335,
       "step": 7965
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 12.918240547180176,
-      "learning_rate": 4.0332765360328755e-06,
-      "loss": 0.878,
+      "epoch": 1.0,
+      "grad_norm": 12.573298454284668,
+      "learning_rate": 1.33355645734845e-05,
+      "loss": 0.9672,
       "step": 7966
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 10.968676567077637,
-      "learning_rate": 4.031271925428486e-06,
-      "loss": 2.1511,
+      "epoch": 1.0,
+      "grad_norm": 16.43240737915039,
+      "learning_rate": 1.3334727858427814e-05,
+      "loss": 1.1471,
       "step": 7967
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 21.93174934387207,
-      "learning_rate": 4.029267314824096e-06,
-      "loss": 1.5569,
+      "epoch": 1.0,
+      "grad_norm": 25.22066879272461,
+      "learning_rate": 1.3333891143371128e-05,
+      "loss": 2.2558,
       "step": 7968
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 18.74276351928711,
-      "learning_rate": 4.027262704219706e-06,
-      "loss": 0.7945,
+      "epoch": 1.0,
+      "grad_norm": 15.474662780761719,
+      "learning_rate": 1.3333054428314438e-05,
+      "loss": 1.2835,
       "step": 7969
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 9.503168106079102,
-      "learning_rate": 4.025258093615316e-06,
-      "loss": 0.8866,
+      "epoch": 1.0,
+      "grad_norm": 7.968123435974121,
+      "learning_rate": 1.3332217713257751e-05,
+      "loss": 1.2493,
       "step": 7970
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 39.47118377685547,
-      "learning_rate": 4.023253483010925e-06,
-      "loss": 1.4191,
+      "epoch": 1.0,
+      "grad_norm": 22.447425842285156,
+      "learning_rate": 1.3331380998201065e-05,
+      "loss": 3.0713,
       "step": 7971
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 11.129521369934082,
-      "learning_rate": 4.021248872406535e-06,
-      "loss": 1.1143,
+      "epoch": 1.0,
+      "grad_norm": 23.46609878540039,
+      "learning_rate": 1.3330544283144375e-05,
+      "loss": 1.6728,
       "step": 7972
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 13.493324279785156,
-      "learning_rate": 4.019244261802145e-06,
-      "loss": 1.5284,
+      "epoch": 1.0,
+      "grad_norm": 19.523942947387695,
+      "learning_rate": 1.3329707568087689e-05,
+      "loss": 2.5787,
       "step": 7973
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 12.583949089050293,
-      "learning_rate": 4.017239651197755e-06,
-      "loss": 1.6311,
+      "epoch": 1.0,
+      "grad_norm": 18.46872329711914,
+      "learning_rate": 1.3328870853031001e-05,
+      "loss": 3.5041,
       "step": 7974
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 7.225354194641113,
-      "learning_rate": 4.0152350405933645e-06,
-      "loss": 0.5856,
+      "epoch": 1.0,
+      "grad_norm": 30.07337188720703,
+      "learning_rate": 1.3328034137974315e-05,
+      "loss": 1.852,
       "step": 7975
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 15.300594329833984,
-      "learning_rate": 4.0132304299889755e-06,
-      "loss": 1.0028,
+      "epoch": 1.0,
+      "grad_norm": 15.493572235107422,
+      "learning_rate": 1.3327197422917627e-05,
+      "loss": 2.3278,
       "step": 7976
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 16.86861228942871,
-      "learning_rate": 4.011225819384585e-06,
-      "loss": 1.7745,
+      "epoch": 1.0,
+      "grad_norm": 13.22238540649414,
+      "learning_rate": 1.3326360707860939e-05,
+      "loss": 1.1271,
       "step": 7977
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 20.883081436157227,
-      "learning_rate": 4.009221208780195e-06,
-      "loss": 1.6959,
+      "epoch": 1.0,
+      "grad_norm": 19.484376907348633,
+      "learning_rate": 1.3325523992804252e-05,
+      "loss": 0.7411,
       "step": 7978
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 120.636962890625,
-      "learning_rate": 4.007216598175805e-06,
-      "loss": 2.2987,
+      "epoch": 1.0,
+      "grad_norm": 11.625712394714355,
+      "learning_rate": 1.3324687277747562e-05,
+      "loss": 1.6891,
       "step": 7979
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 76.55606842041016,
-      "learning_rate": 4.005211987571415e-06,
-      "loss": 1.4258,
+      "epoch": 1.0,
+      "grad_norm": 28.404155731201172,
+      "learning_rate": 1.3323850562690876e-05,
+      "loss": 1.3849,
       "step": 7980
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 15.42659854888916,
-      "learning_rate": 4.003207376967024e-06,
-      "loss": 1.5873,
+      "epoch": 1.0,
+      "grad_norm": 6.87016487121582,
+      "learning_rate": 1.332301384763419e-05,
+      "loss": 0.3202,
       "step": 7981
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 9.378588676452637,
-      "learning_rate": 4.001202766362634e-06,
-      "loss": 0.9196,
+      "epoch": 1.0,
+      "grad_norm": 12.507184982299805,
+      "learning_rate": 1.3322177132577503e-05,
+      "loss": 1.6465,
       "step": 7982
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 12.483980178833008,
-      "learning_rate": 3.999198155758244e-06,
-      "loss": 0.9071,
+      "epoch": 1.0,
+      "grad_norm": 36.98790740966797,
+      "learning_rate": 1.3321340417520814e-05,
+      "loss": 2.302,
       "step": 7983
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 7.117259502410889,
-      "learning_rate": 3.9971935451538544e-06,
-      "loss": 0.6292,
+      "epoch": 1.0,
+      "grad_norm": 8.607172012329102,
+      "learning_rate": 1.3320503702464127e-05,
+      "loss": 0.8184,
       "step": 7984
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 14.306236267089844,
-      "learning_rate": 3.9951889345494645e-06,
-      "loss": 1.0767,
+      "epoch": 1.0,
+      "grad_norm": 12.417741775512695,
+      "learning_rate": 1.3319666987407441e-05,
+      "loss": 1.9958,
       "step": 7985
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 46.9137077331543,
-      "learning_rate": 3.993184323945074e-06,
-      "loss": 2.4379,
+      "epoch": 1.0,
+      "grad_norm": 17.457847595214844,
+      "learning_rate": 1.3318830272350751e-05,
+      "loss": 1.7666,
       "step": 7986
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 17.20838737487793,
-      "learning_rate": 3.991179713340684e-06,
-      "loss": 1.2174,
+      "epoch": 1.0,
+      "grad_norm": 18.212696075439453,
+      "learning_rate": 1.3317993557294065e-05,
+      "loss": 1.4751,
       "step": 7987
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 9.353306770324707,
-      "learning_rate": 3.989175102736294e-06,
-      "loss": 1.0121,
+      "epoch": 1.0,
+      "grad_norm": 9.338573455810547,
+      "learning_rate": 1.3317156842237377e-05,
+      "loss": 1.1111,
       "step": 7988
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 17.213781356811523,
-      "learning_rate": 3.987170492131904e-06,
-      "loss": 1.4443,
+      "epoch": 1.0,
+      "grad_norm": 8.514516830444336,
+      "learning_rate": 1.331632012718069e-05,
+      "loss": 0.594,
       "step": 7989
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 6.010331630706787,
-      "learning_rate": 3.985165881527513e-06,
-      "loss": 0.8314,
+      "epoch": 1.0,
+      "grad_norm": 61.38825225830078,
+      "learning_rate": 1.3315483412124002e-05,
+      "loss": 1.7547,
       "step": 7990
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 12.470282554626465,
-      "learning_rate": 3.983161270923124e-06,
-      "loss": 1.9135,
+      "epoch": 1.0,
+      "grad_norm": 9.940217018127441,
+      "learning_rate": 1.3314646697067314e-05,
+      "loss": 1.8636,
       "step": 7991
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 25.282251358032227,
-      "learning_rate": 3.9811566603187334e-06,
-      "loss": 1.9866,
+      "epoch": 1.0,
+      "grad_norm": 20.792577743530273,
+      "learning_rate": 1.3313809982010628e-05,
+      "loss": 1.5036,
       "step": 7992
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 12.670585632324219,
-      "learning_rate": 3.9791520497143435e-06,
-      "loss": 0.8544,
+      "epoch": 1.0,
+      "grad_norm": 12.03057861328125,
+      "learning_rate": 1.3312973266953938e-05,
+      "loss": 1.1097,
       "step": 7993
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 21.99867057800293,
-      "learning_rate": 3.977147439109954e-06,
-      "loss": 0.906,
+      "epoch": 1.0,
+      "grad_norm": 8.143668174743652,
+      "learning_rate": 1.3312136551897252e-05,
+      "loss": 1.62,
       "step": 7994
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 26.105510711669922,
-      "learning_rate": 3.975142828505563e-06,
-      "loss": 1.1245,
+      "epoch": 1.0,
+      "grad_norm": 22.409191131591797,
+      "learning_rate": 1.3311299836840566e-05,
+      "loss": 2.4149,
       "step": 7995
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 27.73430824279785,
-      "learning_rate": 3.973138217901173e-06,
-      "loss": 1.5627,
+      "epoch": 1.0,
+      "grad_norm": 20.784591674804688,
+      "learning_rate": 1.331046312178388e-05,
+      "loss": 1.3993,
       "step": 7996
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 31.076269149780273,
-      "learning_rate": 3.971133607296783e-06,
-      "loss": 1.5272,
+      "epoch": 1.0,
+      "grad_norm": 3.964128017425537,
+      "learning_rate": 1.330962640672719e-05,
+      "loss": 0.265,
       "step": 7997
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 20.252588272094727,
-      "learning_rate": 3.969128996692393e-06,
-      "loss": 1.3878,
+      "epoch": 1.0,
+      "grad_norm": 11.53171443939209,
+      "learning_rate": 1.3308789691670503e-05,
+      "loss": 0.9298,
       "step": 7998
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 15.601667404174805,
-      "learning_rate": 3.967124386088002e-06,
-      "loss": 1.6348,
+      "epoch": 1.0,
+      "grad_norm": 20.267475128173828,
+      "learning_rate": 1.3307952976613817e-05,
+      "loss": 2.2056,
       "step": 7999
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 15.273148536682129,
-      "learning_rate": 3.965119775483612e-06,
-      "loss": 2.6779,
+      "epoch": 1.0,
+      "grad_norm": 9.326493263244629,
+      "learning_rate": 1.3307116261557127e-05,
+      "loss": 1.0893,
       "step": 8000
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 14.738374710083008,
-      "learning_rate": 3.9631151648792225e-06,
-      "loss": 1.3222,
+      "epoch": 1.0,
+      "eval_loss": 0.10294318199157715,
+      "eval_runtime": 191.9003,
+      "eval_samples_per_second": 18.458,
+      "eval_steps_per_second": 18.458,
+      "step": 8000
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.06924057006836,
+      "learning_rate": 1.330627954650044e-05,
+      "loss": 2.4817,
       "step": 8001
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 12.968079566955566,
-      "learning_rate": 3.9611105542748326e-06,
-      "loss": 0.8718,
+      "epoch": 1.0,
+      "grad_norm": 23.34333610534668,
+      "learning_rate": 1.3305442831443753e-05,
+      "loss": 4.3463,
       "step": 8002
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 15.838987350463867,
-      "learning_rate": 3.959105943670442e-06,
-      "loss": 1.1175,
+      "epoch": 1.0,
+      "grad_norm": 42.41044235229492,
+      "learning_rate": 1.3304606116387066e-05,
+      "loss": 3.5817,
       "step": 8003
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 11.581687927246094,
-      "learning_rate": 3.957101333066053e-06,
-      "loss": 2.2458,
+      "epoch": 1.0,
+      "grad_norm": 5.411513805389404,
+      "learning_rate": 1.3303769401330378e-05,
+      "loss": 0.5194,
       "step": 8004
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 13.473747253417969,
-      "learning_rate": 3.955096722461662e-06,
-      "loss": 1.0507,
+      "epoch": 1.0,
+      "grad_norm": 7.918196678161621,
+      "learning_rate": 1.330293268627369e-05,
+      "loss": 1.5817,
       "step": 8005
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 22.461040496826172,
-      "learning_rate": 3.953092111857272e-06,
-      "loss": 1.8377,
+      "epoch": 1.0,
+      "grad_norm": 38.20075225830078,
+      "learning_rate": 1.3302095971217004e-05,
+      "loss": 3.9696,
       "step": 8006
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 9.827531814575195,
-      "learning_rate": 3.951087501252882e-06,
-      "loss": 1.3932,
+      "epoch": 1.0,
+      "grad_norm": 32.289485931396484,
+      "learning_rate": 1.3301259256160314e-05,
+      "loss": 0.9916,
       "step": 8007
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 17.85820198059082,
-      "learning_rate": 3.949082890648492e-06,
-      "loss": 1.0887,
+      "epoch": 1.0,
+      "grad_norm": 44.22406005859375,
+      "learning_rate": 1.3300422541103628e-05,
+      "loss": 2.6136,
       "step": 8008
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 10.893664360046387,
-      "learning_rate": 3.9470782800441015e-06,
-      "loss": 0.6649,
+      "epoch": 1.01,
+      "grad_norm": 23.457265853881836,
+      "learning_rate": 1.3299585826046941e-05,
+      "loss": 2.8706,
       "step": 8009
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 27.924177169799805,
-      "learning_rate": 3.9450736694397116e-06,
-      "loss": 1.7343,
+      "epoch": 1.01,
+      "grad_norm": 26.69243049621582,
+      "learning_rate": 1.3298749110990255e-05,
+      "loss": 2.6025,
       "step": 8010
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 7.996644973754883,
-      "learning_rate": 3.943069058835322e-06,
-      "loss": 1.1344,
+      "epoch": 1.01,
+      "grad_norm": 14.823225021362305,
+      "learning_rate": 1.3297912395933565e-05,
+      "loss": 0.8757,
       "step": 8011
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 29.98558807373047,
-      "learning_rate": 3.941064448230931e-06,
-      "loss": 1.4769,
+      "epoch": 1.01,
+      "grad_norm": 7.193357467651367,
+      "learning_rate": 1.3297075680876879e-05,
+      "loss": 1.0568,
       "step": 8012
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 27.507963180541992,
-      "learning_rate": 3.939059837626542e-06,
-      "loss": 0.83,
+      "epoch": 1.01,
+      "grad_norm": 29.412263870239258,
+      "learning_rate": 1.3296238965820191e-05,
+      "loss": 1.5548,
       "step": 8013
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 7.844414710998535,
-      "learning_rate": 3.937055227022151e-06,
-      "loss": 0.9977,
+      "epoch": 1.01,
+      "grad_norm": 10.851529121398926,
+      "learning_rate": 1.3295402250763503e-05,
+      "loss": 2.2944,
       "step": 8014
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 51.909061431884766,
-      "learning_rate": 3.935050616417761e-06,
-      "loss": 2.3273,
+      "epoch": 1.01,
+      "grad_norm": 18.598386764526367,
+      "learning_rate": 1.3294565535706817e-05,
+      "loss": 1.5626,
       "step": 8015
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 10.7676420211792,
-      "learning_rate": 3.933046005813371e-06,
-      "loss": 0.986,
+      "epoch": 1.01,
+      "grad_norm": 64.27809143066406,
+      "learning_rate": 1.3293728820650129e-05,
+      "loss": 1.3406,
       "step": 8016
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 14.182157516479492,
-      "learning_rate": 3.931041395208981e-06,
-      "loss": 1.3547,
+      "epoch": 1.01,
+      "grad_norm": 88.03241729736328,
+      "learning_rate": 1.3292892105593442e-05,
+      "loss": 1.6087,
       "step": 8017
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 15.017866134643555,
-      "learning_rate": 3.9290367846045905e-06,
-      "loss": 0.7605,
+      "epoch": 1.01,
+      "grad_norm": 33.14046096801758,
+      "learning_rate": 1.3292055390536754e-05,
+      "loss": 2.5191,
       "step": 8018
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 22.285579681396484,
-      "learning_rate": 3.927032174000201e-06,
-      "loss": 1.5245,
+      "epoch": 1.01,
+      "grad_norm": 10.153793334960938,
+      "learning_rate": 1.3291218675480066e-05,
+      "loss": 0.6017,
       "step": 8019
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 69.15910339355469,
-      "learning_rate": 3.925027563395811e-06,
-      "loss": 2.0588,
+      "epoch": 1.01,
+      "grad_norm": 18.304969787597656,
+      "learning_rate": 1.329038196042338e-05,
+      "loss": 1.1005,
       "step": 8020
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 57.589107513427734,
-      "learning_rate": 3.923022952791421e-06,
-      "loss": 2.0236,
+      "epoch": 1.01,
+      "grad_norm": 80.81324005126953,
+      "learning_rate": 1.328954524536669e-05,
+      "loss": 1.8401,
       "step": 8021
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 38.282440185546875,
-      "learning_rate": 3.921018342187031e-06,
-      "loss": 1.923,
+      "epoch": 1.01,
+      "grad_norm": 10.775973320007324,
+      "learning_rate": 1.3288708530310004e-05,
+      "loss": 0.8631,
       "step": 8022
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 16.257158279418945,
-      "learning_rate": 3.91901373158264e-06,
-      "loss": 1.2573,
+      "epoch": 1.01,
+      "grad_norm": 11.5303373336792,
+      "learning_rate": 1.3287871815253317e-05,
+      "loss": 1.0476,
       "step": 8023
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 13.993424415588379,
-      "learning_rate": 3.91700912097825e-06,
-      "loss": 1.1249,
+      "epoch": 1.01,
+      "grad_norm": 9.755226135253906,
+      "learning_rate": 1.3287035100196631e-05,
+      "loss": 0.5219,
       "step": 8024
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 18.90182876586914,
-      "learning_rate": 3.91500451037386e-06,
-      "loss": 1.4962,
+      "epoch": 1.01,
+      "grad_norm": 5.90783166885376,
+      "learning_rate": 1.3286198385139941e-05,
+      "loss": 0.7428,
       "step": 8025
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 19.353958129882812,
-      "learning_rate": 3.91299989976947e-06,
-      "loss": 1.0571,
+      "epoch": 1.01,
+      "grad_norm": 11.44725513458252,
+      "learning_rate": 1.3285361670083255e-05,
+      "loss": 1.4417,
       "step": 8026
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 31.27533531188965,
-      "learning_rate": 3.91099528916508e-06,
-      "loss": 1.6633,
+      "epoch": 1.01,
+      "grad_norm": 3.191082000732422,
+      "learning_rate": 1.3284524955026567e-05,
+      "loss": 0.0848,
       "step": 8027
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 15.394329071044922,
-      "learning_rate": 3.9089906785606905e-06,
-      "loss": 1.4122,
+      "epoch": 1.01,
+      "grad_norm": 30.03740119934082,
+      "learning_rate": 1.3283688239969879e-05,
+      "loss": 1.6489,
       "step": 8028
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 77.46128845214844,
-      "learning_rate": 3.9069860679563e-06,
-      "loss": 1.7462,
+      "epoch": 1.01,
+      "grad_norm": 8.293747901916504,
+      "learning_rate": 1.3282851524913192e-05,
+      "loss": 0.3589,
       "step": 8029
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 33.42092514038086,
-      "learning_rate": 3.90498145735191e-06,
-      "loss": 1.3732,
+      "epoch": 1.01,
+      "grad_norm": 5.748569965362549,
+      "learning_rate": 1.3282014809856504e-05,
+      "loss": 0.539,
       "step": 8030
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 9.593501091003418,
-      "learning_rate": 3.90297684674752e-06,
-      "loss": 0.9003,
+      "epoch": 1.01,
+      "grad_norm": 15.555586814880371,
+      "learning_rate": 1.3281178094799816e-05,
+      "loss": 0.4044,
       "step": 8031
     },
     {
-      "epoch": 2.41,
-      "grad_norm": 16.690101623535156,
-      "learning_rate": 3.900972236143129e-06,
-      "loss": 1.6157,
+      "epoch": 1.01,
+      "grad_norm": 32.48698425292969,
+      "learning_rate": 1.3280341379743128e-05,
+      "loss": 0.5353,
       "step": 8032
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 14.943684577941895,
-      "learning_rate": 3.898967625538739e-06,
-      "loss": 1.5163,
+      "epoch": 1.01,
+      "grad_norm": 13.530704498291016,
+      "learning_rate": 1.3279504664686442e-05,
+      "loss": 1.6522,
       "step": 8033
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 59.907310485839844,
-      "learning_rate": 3.896963014934349e-06,
-      "loss": 1.6031,
+      "epoch": 1.01,
+      "grad_norm": 13.967482566833496,
+      "learning_rate": 1.3278667949629756e-05,
+      "loss": 0.6138,
       "step": 8034
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 32.62445068359375,
-      "learning_rate": 3.8949584043299594e-06,
-      "loss": 1.2945,
+      "epoch": 1.01,
+      "grad_norm": 8.310638427734375,
+      "learning_rate": 1.3277831234573066e-05,
+      "loss": 1.147,
       "step": 8035
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 92.56492614746094,
-      "learning_rate": 3.892953793725569e-06,
-      "loss": 1.5538,
+      "epoch": 1.01,
+      "grad_norm": 22.596485137939453,
+      "learning_rate": 1.327699451951638e-05,
+      "loss": 1.4125,
       "step": 8036
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 11.2645902633667,
-      "learning_rate": 3.890949183121179e-06,
-      "loss": 1.8437,
+      "epoch": 1.01,
+      "grad_norm": 6.3463873863220215,
+      "learning_rate": 1.3276157804459693e-05,
+      "loss": 1.7541,
       "step": 8037
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 13.851442337036133,
-      "learning_rate": 3.888944572516789e-06,
-      "loss": 0.9473,
+      "epoch": 1.01,
+      "grad_norm": 12.85508918762207,
+      "learning_rate": 1.3275321089403003e-05,
+      "loss": 1.5066,
       "step": 8038
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 17.675752639770508,
-      "learning_rate": 3.886939961912399e-06,
-      "loss": 1.4754,
+      "epoch": 1.01,
+      "grad_norm": 14.467947006225586,
+      "learning_rate": 1.3274484374346317e-05,
+      "loss": 2.1135,
       "step": 8039
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 14.534573554992676,
-      "learning_rate": 3.884935351308008e-06,
-      "loss": 1.2699,
-      "step": 8040
-    },
-    {
-      "epoch": 2.42,
-      "eval_loss": 0.16647757589817047,
-      "eval_runtime": 43.4766,
-      "eval_samples_per_second": 34.018,
-      "eval_steps_per_second": 34.018,
+      "epoch": 1.01,
+      "grad_norm": 8.939553260803223,
+      "learning_rate": 1.327364765928963e-05,
+      "loss": 0.5817,
       "step": 8040
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 10.053756713867188,
-      "learning_rate": 3.882930740703619e-06,
-      "loss": 1.1651,
+      "epoch": 1.01,
+      "grad_norm": 4.461472511291504,
+      "learning_rate": 1.3272810944232943e-05,
+      "loss": 0.3033,
       "step": 8041
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 15.086660385131836,
-      "learning_rate": 3.880926130099228e-06,
-      "loss": 1.2307,
+      "epoch": 1.01,
+      "grad_norm": 10.413497924804688,
+      "learning_rate": 1.3271974229176255e-05,
+      "loss": 1.1546,
       "step": 8042
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 45.48273468017578,
-      "learning_rate": 3.878921519494838e-06,
-      "loss": 1.7186,
+      "epoch": 1.01,
+      "grad_norm": 42.64353942871094,
+      "learning_rate": 1.3271137514119568e-05,
+      "loss": 2.6524,
       "step": 8043
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 34.40294647216797,
-      "learning_rate": 3.8769169088904485e-06,
-      "loss": 2.7345,
+      "epoch": 1.01,
+      "grad_norm": 16.8046817779541,
+      "learning_rate": 1.327030079906288e-05,
+      "loss": 0.566,
       "step": 8044
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 14.807568550109863,
-      "learning_rate": 3.8749122982860586e-06,
-      "loss": 0.8994,
+      "epoch": 1.01,
+      "grad_norm": 27.37327003479004,
+      "learning_rate": 1.3269464084006192e-05,
+      "loss": 1.344,
       "step": 8045
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 10.855109214782715,
-      "learning_rate": 3.872907687681668e-06,
-      "loss": 1.2174,
+      "epoch": 1.01,
+      "grad_norm": 13.533464431762695,
+      "learning_rate": 1.3268627368949504e-05,
+      "loss": 1.5702,
       "step": 8046
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 36.307777404785156,
-      "learning_rate": 3.870903077077278e-06,
-      "loss": 2.1917,
+      "epoch": 1.01,
+      "grad_norm": 9.839691162109375,
+      "learning_rate": 1.3267790653892818e-05,
+      "loss": 0.8166,
       "step": 8047
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 22.572141647338867,
-      "learning_rate": 3.868898466472888e-06,
-      "loss": 0.8395,
+      "epoch": 1.01,
+      "grad_norm": 17.93497085571289,
+      "learning_rate": 1.3266953938836131e-05,
+      "loss": 0.8191,
       "step": 8048
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 17.936315536499023,
-      "learning_rate": 3.866893855868497e-06,
-      "loss": 2.4869,
+      "epoch": 1.01,
+      "grad_norm": 12.403947830200195,
+      "learning_rate": 1.3266117223779442e-05,
+      "loss": 2.1308,
       "step": 8049
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 12.054944038391113,
-      "learning_rate": 3.864889245264108e-06,
-      "loss": 1.4992,
+      "epoch": 1.01,
+      "grad_norm": 11.181997299194336,
+      "learning_rate": 1.3265280508722755e-05,
+      "loss": 1.4637,
       "step": 8050
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 10.359464645385742,
-      "learning_rate": 3.862884634659717e-06,
-      "loss": 1.1123,
+      "epoch": 1.01,
+      "grad_norm": 52.8959846496582,
+      "learning_rate": 1.3264443793666069e-05,
+      "loss": 4.37,
       "step": 8051
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 21.69719886779785,
-      "learning_rate": 3.8608800240553275e-06,
-      "loss": 1.9015,
+      "epoch": 1.01,
+      "grad_norm": 15.241668701171875,
+      "learning_rate": 1.326360707860938e-05,
+      "loss": 1.8832,
       "step": 8052
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 9.15548038482666,
-      "learning_rate": 3.8588754134509376e-06,
-      "loss": 1.1113,
+      "epoch": 1.01,
+      "grad_norm": 12.627178192138672,
+      "learning_rate": 1.3262770363552693e-05,
+      "loss": 0.7974,
       "step": 8053
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 7.168182373046875,
-      "learning_rate": 3.856870802846548e-06,
-      "loss": 0.7675,
+      "epoch": 1.01,
+      "grad_norm": 33.42836380004883,
+      "learning_rate": 1.3261933648496007e-05,
+      "loss": 1.09,
       "step": 8054
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 23.34624481201172,
-      "learning_rate": 3.854866192242157e-06,
-      "loss": 1.1852,
+      "epoch": 1.01,
+      "grad_norm": 32.938899993896484,
+      "learning_rate": 1.3261096933439318e-05,
+      "loss": 1.1665,
       "step": 8055
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 8.906826972961426,
-      "learning_rate": 3.852861581637767e-06,
-      "loss": 0.7492,
+      "epoch": 1.01,
+      "grad_norm": 3.821079730987549,
+      "learning_rate": 1.326026021838263e-05,
+      "loss": 0.2259,
       "step": 8056
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 17.47923469543457,
-      "learning_rate": 3.850856971033377e-06,
-      "loss": 1.1435,
+      "epoch": 1.01,
+      "grad_norm": 12.121703147888184,
+      "learning_rate": 1.3259423503325944e-05,
+      "loss": 1.0859,
       "step": 8057
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 8.960250854492188,
-      "learning_rate": 3.848852360428987e-06,
-      "loss": 0.9197,
+      "epoch": 1.01,
+      "grad_norm": 16.58420753479004,
+      "learning_rate": 1.3258586788269256e-05,
+      "loss": 0.6811,
       "step": 8058
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 17.509536743164062,
-      "learning_rate": 3.846847749824597e-06,
-      "loss": 1.5889,
+      "epoch": 1.01,
+      "grad_norm": 19.60185432434082,
+      "learning_rate": 1.3257750073212568e-05,
+      "loss": 1.2272,
       "step": 8059
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 14.029973030090332,
-      "learning_rate": 3.8448431392202065e-06,
-      "loss": 1.5197,
+      "epoch": 1.01,
+      "grad_norm": 17.212610244750977,
+      "learning_rate": 1.325691335815588e-05,
+      "loss": 1.0265,
       "step": 8060
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 13.85709285736084,
-      "learning_rate": 3.8428385286158165e-06,
-      "loss": 1.1813,
+      "epoch": 1.01,
+      "grad_norm": 26.601634979248047,
+      "learning_rate": 1.3256076643099194e-05,
+      "loss": 1.6516,
       "step": 8061
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 11.46907901763916,
-      "learning_rate": 3.840833918011427e-06,
-      "loss": 1.1435,
+      "epoch": 1.01,
+      "grad_norm": 9.53249454498291,
+      "learning_rate": 1.3255239928042507e-05,
+      "loss": 0.9406,
       "step": 8062
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 11.111536026000977,
-      "learning_rate": 3.838829307407037e-06,
-      "loss": 1.6378,
+      "epoch": 1.01,
+      "grad_norm": 19.139719009399414,
+      "learning_rate": 1.3254403212985817e-05,
+      "loss": 0.7048,
       "step": 8063
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 13.92691421508789,
-      "learning_rate": 3.836824696802646e-06,
-      "loss": 1.0666,
+      "epoch": 1.01,
+      "grad_norm": 24.55410385131836,
+      "learning_rate": 1.3253566497929131e-05,
+      "loss": 2.0259,
       "step": 8064
     },
     {
-      "epoch": 2.42,
-      "grad_norm": 11.448002815246582,
-      "learning_rate": 3.834820086198257e-06,
-      "loss": 1.2318,
+      "epoch": 1.01,
+      "grad_norm": 7.9865803718566895,
+      "learning_rate": 1.3252729782872445e-05,
+      "loss": 0.8302,
       "step": 8065
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 13.583879470825195,
-      "learning_rate": 3.832815475593866e-06,
-      "loss": 0.7262,
+      "epoch": 1.01,
+      "grad_norm": 12.298362731933594,
+      "learning_rate": 1.3251893067815755e-05,
+      "loss": 0.9395,
       "step": 8066
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 17.306129455566406,
-      "learning_rate": 3.830810864989476e-06,
-      "loss": 1.5805,
+      "epoch": 1.01,
+      "grad_norm": 12.418360710144043,
+      "learning_rate": 1.3251056352759069e-05,
+      "loss": 3.1056,
       "step": 8067
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 17.71708106994629,
-      "learning_rate": 3.828806254385086e-06,
-      "loss": 1.5627,
+      "epoch": 1.01,
+      "grad_norm": 16.77568244934082,
+      "learning_rate": 1.3250219637702382e-05,
+      "loss": 1.1351,
       "step": 8068
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 13.716277122497559,
-      "learning_rate": 3.826801643780696e-06,
-      "loss": 1.223,
+      "epoch": 1.01,
+      "grad_norm": 20.997798919677734,
+      "learning_rate": 1.3249382922645694e-05,
+      "loss": 1.6183,
       "step": 8069
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 9.647749900817871,
-      "learning_rate": 3.824797033176306e-06,
-      "loss": 2.1291,
+      "epoch": 1.01,
+      "grad_norm": 34.50931167602539,
+      "learning_rate": 1.3248546207589006e-05,
+      "loss": 2.1002,
       "step": 8070
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 15.30841064453125,
-      "learning_rate": 3.822792422571916e-06,
-      "loss": 1.1331,
+      "epoch": 1.01,
+      "grad_norm": 24.325166702270508,
+      "learning_rate": 1.324770949253232e-05,
+      "loss": 1.509,
       "step": 8071
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 18.855783462524414,
-      "learning_rate": 3.820787811967526e-06,
-      "loss": 1.7297,
+      "epoch": 1.01,
+      "grad_norm": 11.502999305725098,
+      "learning_rate": 1.3246872777475632e-05,
+      "loss": 1.1475,
       "step": 8072
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 35.978248596191406,
-      "learning_rate": 3.818783201363135e-06,
-      "loss": 0.7591,
+      "epoch": 1.01,
+      "grad_norm": 6.817028045654297,
+      "learning_rate": 1.3246036062418944e-05,
+      "loss": 0.7322,
       "step": 8073
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 19.79545021057129,
-      "learning_rate": 3.816778590758746e-06,
-      "loss": 1.878,
+      "epoch": 1.01,
+      "grad_norm": 23.855792999267578,
+      "learning_rate": 1.3245199347362256e-05,
+      "loss": 1.0183,
       "step": 8074
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 33.37739562988281,
-      "learning_rate": 3.814773980154355e-06,
-      "loss": 1.5872,
+      "epoch": 1.01,
+      "grad_norm": 36.64528274536133,
+      "learning_rate": 1.324436263230557e-05,
+      "loss": 2.4726,
       "step": 8075
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 19.29977035522461,
-      "learning_rate": 3.8127693695499653e-06,
-      "loss": 0.9943,
+      "epoch": 1.01,
+      "grad_norm": 15.954262733459473,
+      "learning_rate": 1.3243525917248883e-05,
+      "loss": 0.7668,
       "step": 8076
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 22.01044273376465,
-      "learning_rate": 3.8107647589455753e-06,
-      "loss": 1.9346,
+      "epoch": 1.01,
+      "grad_norm": 19.742395401000977,
+      "learning_rate": 1.3242689202192193e-05,
+      "loss": 1.0538,
       "step": 8077
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 24.812864303588867,
-      "learning_rate": 3.808760148341185e-06,
-      "loss": 1.4286,
+      "epoch": 1.01,
+      "grad_norm": 11.046028137207031,
+      "learning_rate": 1.3241852487135507e-05,
+      "loss": 1.2064,
       "step": 8078
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 16.758087158203125,
-      "learning_rate": 3.8067555377367947e-06,
-      "loss": 0.6716,
+      "epoch": 1.01,
+      "grad_norm": 16.42095947265625,
+      "learning_rate": 1.324101577207882e-05,
+      "loss": 1.3735,
       "step": 8079
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 10.366033554077148,
-      "learning_rate": 3.8047509271324047e-06,
-      "loss": 1.3467,
+      "epoch": 1.01,
+      "grad_norm": 8.204858779907227,
+      "learning_rate": 1.3240179057022131e-05,
+      "loss": 0.8107,
       "step": 8080
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 20.55314064025879,
-      "learning_rate": 3.802746316528015e-06,
-      "loss": 1.4071,
+      "epoch": 1.01,
+      "grad_norm": 19.455652236938477,
+      "learning_rate": 1.3239342341965445e-05,
+      "loss": 2.0798,
       "step": 8081
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 31.28022003173828,
-      "learning_rate": 3.8007417059236245e-06,
-      "loss": 1.7199,
+      "epoch": 1.01,
+      "grad_norm": 8.957785606384277,
+      "learning_rate": 1.3238505626908758e-05,
+      "loss": 0.5589,
       "step": 8082
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 6.907041072845459,
-      "learning_rate": 3.798737095319234e-06,
-      "loss": 0.3853,
+      "epoch": 1.01,
+      "grad_norm": 17.459102630615234,
+      "learning_rate": 1.323766891185207e-05,
+      "loss": 1.6631,
       "step": 8083
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 24.64528465270996,
-      "learning_rate": 3.7967324847148447e-06,
-      "loss": 1.4183,
+      "epoch": 1.01,
+      "grad_norm": 19.128694534301758,
+      "learning_rate": 1.3236832196795382e-05,
+      "loss": 0.8569,
       "step": 8084
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 10.540509223937988,
-      "learning_rate": 3.7947278741104543e-06,
-      "loss": 1.0129,
+      "epoch": 1.01,
+      "grad_norm": 3.326540470123291,
+      "learning_rate": 1.3235995481738694e-05,
+      "loss": 0.1619,
       "step": 8085
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 13.344951629638672,
-      "learning_rate": 3.792723263506064e-06,
-      "loss": 0.989,
+      "epoch": 1.01,
+      "grad_norm": 16.52320098876953,
+      "learning_rate": 1.3235158766682008e-05,
+      "loss": 1.2735,
       "step": 8086
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 8.0833158493042,
-      "learning_rate": 3.7907186529016745e-06,
-      "loss": 0.8597,
+      "epoch": 1.01,
+      "grad_norm": 11.69139289855957,
+      "learning_rate": 1.323432205162532e-05,
+      "loss": 0.8281,
       "step": 8087
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 19.457054138183594,
-      "learning_rate": 3.788714042297284e-06,
-      "loss": 1.3676,
+      "epoch": 1.02,
+      "grad_norm": 13.96041202545166,
+      "learning_rate": 1.3233485336568632e-05,
+      "loss": 1.0855,
       "step": 8088
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 16.89392852783203,
-      "learning_rate": 3.786709431692894e-06,
-      "loss": 1.1124,
+      "epoch": 1.02,
+      "grad_norm": 30.238624572753906,
+      "learning_rate": 1.3232648621511945e-05,
+      "loss": 2.0699,
       "step": 8089
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 9.098607063293457,
-      "learning_rate": 3.784704821088504e-06,
-      "loss": 0.5794,
+      "epoch": 1.02,
+      "grad_norm": 11.058302879333496,
+      "learning_rate": 1.3231811906455259e-05,
+      "loss": 2.3913,
       "step": 8090
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 41.21145248413086,
-      "learning_rate": 3.7827002104841136e-06,
-      "loss": 1.9724,
+      "epoch": 1.02,
+      "grad_norm": 7.368780136108398,
+      "learning_rate": 1.323097519139857e-05,
+      "loss": 1.1791,
       "step": 8091
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 5.302923679351807,
-      "learning_rate": 3.7806955998797236e-06,
-      "loss": 0.579,
+      "epoch": 1.02,
+      "grad_norm": 15.291146278381348,
+      "learning_rate": 1.3230138476341883e-05,
+      "loss": 1.3794,
       "step": 8092
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 14.981736183166504,
-      "learning_rate": 3.7786909892753337e-06,
-      "loss": 2.1375,
+      "epoch": 1.02,
+      "grad_norm": 47.3812141418457,
+      "learning_rate": 1.3229301761285196e-05,
+      "loss": 1.7557,
       "step": 8093
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 20.563865661621094,
-      "learning_rate": 3.7766863786709434e-06,
-      "loss": 1.137,
+      "epoch": 1.02,
+      "grad_norm": 24.50340461730957,
+      "learning_rate": 1.3228465046228507e-05,
+      "loss": 3.2379,
       "step": 8094
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 10.908076286315918,
-      "learning_rate": 3.774681768066553e-06,
-      "loss": 1.2199,
+      "epoch": 1.02,
+      "grad_norm": 14.086163520812988,
+      "learning_rate": 1.322762833117182e-05,
+      "loss": 0.9016,
       "step": 8095
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 79.2155532836914,
-      "learning_rate": 3.7726771574621636e-06,
-      "loss": 2.0151,
+      "epoch": 1.02,
+      "grad_norm": 14.826435089111328,
+      "learning_rate": 1.3226791616115134e-05,
+      "loss": 2.2264,
       "step": 8096
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 187.99632263183594,
-      "learning_rate": 3.7706725468577732e-06,
-      "loss": 2.0555,
+      "epoch": 1.02,
+      "grad_norm": 16.74859619140625,
+      "learning_rate": 1.3225954901058446e-05,
+      "loss": 1.3957,
       "step": 8097
     },
     {
-      "epoch": 2.43,
-      "grad_norm": 38.596275329589844,
-      "learning_rate": 3.768667936253383e-06,
-      "loss": 2.0002,
+      "epoch": 1.02,
+      "grad_norm": 12.007782936096191,
+      "learning_rate": 1.3225118186001758e-05,
+      "loss": 1.0195,
       "step": 8098
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 8.95095157623291,
-      "learning_rate": 3.766663325648993e-06,
-      "loss": 1.0331,
+      "epoch": 1.02,
+      "grad_norm": 34.34865188598633,
+      "learning_rate": 1.322428147094507e-05,
+      "loss": 2.1014,
       "step": 8099
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 17.133882522583008,
-      "learning_rate": 3.764658715044603e-06,
-      "loss": 1.5345,
+      "epoch": 1.02,
+      "grad_norm": 10.553949356079102,
+      "learning_rate": 1.3223444755888384e-05,
+      "loss": 1.3867,
       "step": 8100
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 8.687376976013184,
-      "learning_rate": 3.7626541044402127e-06,
-      "loss": 0.7787,
+      "epoch": 1.02,
+      "grad_norm": 7.074171543121338,
+      "learning_rate": 1.3222608040831695e-05,
+      "loss": 1.277,
       "step": 8101
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 69.27308654785156,
-      "learning_rate": 3.760649493835823e-06,
-      "loss": 1.252,
+      "epoch": 1.02,
+      "grad_norm": 4.532754898071289,
+      "learning_rate": 1.3221771325775007e-05,
+      "loss": 0.3835,
       "step": 8102
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 23.108619689941406,
-      "learning_rate": 3.7586448832314325e-06,
-      "loss": 1.3771,
+      "epoch": 1.02,
+      "grad_norm": 9.71251106262207,
+      "learning_rate": 1.3220934610718321e-05,
+      "loss": 2.2243,
       "step": 8103
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 33.36818313598633,
-      "learning_rate": 3.7566402726270425e-06,
-      "loss": 1.6731,
+      "epoch": 1.02,
+      "grad_norm": 12.226500511169434,
+      "learning_rate": 1.3220097895661635e-05,
+      "loss": 0.8613,
       "step": 8104
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 13.890486717224121,
-      "learning_rate": 3.7546356620226526e-06,
-      "loss": 1.39,
+      "epoch": 1.02,
+      "grad_norm": 12.565596580505371,
+      "learning_rate": 1.3219261180604945e-05,
+      "loss": 1.8522,
       "step": 8105
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 9.70341682434082,
-      "learning_rate": 3.7526310514182623e-06,
-      "loss": 0.7873,
+      "epoch": 1.02,
+      "grad_norm": 17.8659725189209,
+      "learning_rate": 1.3218424465548259e-05,
+      "loss": 0.466,
       "step": 8106
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 15.94449234008789,
-      "learning_rate": 3.750626440813872e-06,
-      "loss": 0.9446,
+      "epoch": 1.02,
+      "grad_norm": 22.846317291259766,
+      "learning_rate": 1.3217587750491572e-05,
+      "loss": 1.5893,
       "step": 8107
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 19.849061965942383,
-      "learning_rate": 3.7486218302094825e-06,
-      "loss": 1.6883,
+      "epoch": 1.02,
+      "grad_norm": 15.297237396240234,
+      "learning_rate": 1.3216751035434883e-05,
+      "loss": 0.8397,
       "step": 8108
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 19.96930503845215,
-      "learning_rate": 3.746617219605092e-06,
-      "loss": 1.2279,
+      "epoch": 1.02,
+      "grad_norm": 12.62805461883545,
+      "learning_rate": 1.3215914320378196e-05,
+      "loss": 1.6367,
       "step": 8109
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 12.898889541625977,
-      "learning_rate": 3.7446126090007018e-06,
-      "loss": 1.6165,
+      "epoch": 1.02,
+      "grad_norm": 10.7589111328125,
+      "learning_rate": 1.321507760532151e-05,
+      "loss": 0.7109,
       "step": 8110
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 12.426019668579102,
-      "learning_rate": 3.742607998396312e-06,
-      "loss": 1.4589,
+      "epoch": 1.02,
+      "grad_norm": 12.487503051757812,
+      "learning_rate": 1.3214240890264822e-05,
+      "loss": 0.972,
       "step": 8111
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 20.18886375427246,
-      "learning_rate": 3.740603387791922e-06,
-      "loss": 1.7677,
+      "epoch": 1.02,
+      "grad_norm": 13.478726387023926,
+      "learning_rate": 1.3213404175208134e-05,
+      "loss": 0.8637,
       "step": 8112
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 19.88523292541504,
-      "learning_rate": 3.7385987771875316e-06,
-      "loss": 2.5334,
+      "epoch": 1.02,
+      "grad_norm": 16.753326416015625,
+      "learning_rate": 1.3212567460151446e-05,
+      "loss": 0.9855,
       "step": 8113
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 39.80482482910156,
-      "learning_rate": 3.7365941665831417e-06,
-      "loss": 1.2999,
+      "epoch": 1.02,
+      "grad_norm": 11.019522666931152,
+      "learning_rate": 1.321173074509476e-05,
+      "loss": 1.2104,
       "step": 8114
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 20.459688186645508,
-      "learning_rate": 3.7345895559787513e-06,
-      "loss": 1.1679,
+      "epoch": 1.02,
+      "grad_norm": 9.230637550354004,
+      "learning_rate": 1.3210894030038071e-05,
+      "loss": 0.93,
       "step": 8115
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 6.700068950653076,
-      "learning_rate": 3.7325849453743614e-06,
-      "loss": 0.7833,
+      "epoch": 1.02,
+      "grad_norm": 10.129105567932129,
+      "learning_rate": 1.3210057314981383e-05,
+      "loss": 1.3537,
       "step": 8116
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 10.873579978942871,
-      "learning_rate": 3.7305803347699715e-06,
-      "loss": 1.1187,
+      "epoch": 1.02,
+      "grad_norm": 12.891251564025879,
+      "learning_rate": 1.3209220599924697e-05,
+      "loss": 0.541,
       "step": 8117
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 11.668147087097168,
-      "learning_rate": 3.728575724165581e-06,
-      "loss": 1.436,
+      "epoch": 1.02,
+      "grad_norm": 12.813945770263672,
+      "learning_rate": 1.320838388486801e-05,
+      "loss": 1.5574,
       "step": 8118
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 24.042890548706055,
-      "learning_rate": 3.726571113561191e-06,
-      "loss": 1.6814,
+      "epoch": 1.02,
+      "grad_norm": 23.467496871948242,
+      "learning_rate": 1.320754716981132e-05,
+      "loss": 2.6069,
       "step": 8119
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 7.250650882720947,
-      "learning_rate": 3.7245665029568013e-06,
-      "loss": 1.062,
+      "epoch": 1.02,
+      "grad_norm": 11.790117263793945,
+      "learning_rate": 1.3206710454754634e-05,
+      "loss": 1.5657,
       "step": 8120
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 8.547741889953613,
-      "learning_rate": 3.722561892352411e-06,
-      "loss": 1.0638,
+      "epoch": 1.02,
+      "grad_norm": 5.574734210968018,
+      "learning_rate": 1.3205873739697948e-05,
+      "loss": 1.0195,
       "step": 8121
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 9.865437507629395,
-      "learning_rate": 3.7205572817480207e-06,
-      "loss": 0.8298,
+      "epoch": 1.02,
+      "grad_norm": 12.647422790527344,
+      "learning_rate": 1.3205037024641258e-05,
+      "loss": 0.7309,
       "step": 8122
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 11.00238037109375,
-      "learning_rate": 3.7185526711436303e-06,
-      "loss": 0.7298,
+      "epoch": 1.02,
+      "grad_norm": 13.510000228881836,
+      "learning_rate": 1.3204200309584572e-05,
+      "loss": 1.8041,
       "step": 8123
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 10.733601570129395,
-      "learning_rate": 3.716548060539241e-06,
-      "loss": 0.9721,
+      "epoch": 1.02,
+      "grad_norm": 34.436222076416016,
+      "learning_rate": 1.3203363594527884e-05,
+      "loss": 1.7884,
       "step": 8124
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 32.854774475097656,
-      "learning_rate": 3.7145434499348505e-06,
-      "loss": 1.4885,
+      "epoch": 1.02,
+      "grad_norm": 39.42829895019531,
+      "learning_rate": 1.3202526879471198e-05,
+      "loss": 3.7134,
       "step": 8125
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 11.409563064575195,
-      "learning_rate": 3.71253883933046e-06,
-      "loss": 0.8726,
+      "epoch": 1.02,
+      "grad_norm": 7.951882839202881,
+      "learning_rate": 1.320169016441451e-05,
+      "loss": 0.605,
       "step": 8126
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 37.438106536865234,
-      "learning_rate": 3.7105342287260702e-06,
-      "loss": 1.3614,
+      "epoch": 1.02,
+      "grad_norm": 5.677186965942383,
+      "learning_rate": 1.3200853449357822e-05,
+      "loss": 0.1397,
       "step": 8127
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 28.72161865234375,
-      "learning_rate": 3.70852961812168e-06,
-      "loss": 1.0996,
+      "epoch": 1.02,
+      "grad_norm": 23.42045021057129,
+      "learning_rate": 1.3200016734301135e-05,
+      "loss": 1.3299,
       "step": 8128
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 10.785248756408691,
-      "learning_rate": 3.70652500751729e-06,
-      "loss": 0.7049,
+      "epoch": 1.02,
+      "grad_norm": 89.72753143310547,
+      "learning_rate": 1.3199180019244447e-05,
+      "loss": 0.9728,
       "step": 8129
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 9.802728652954102,
-      "learning_rate": 3.7045203969129e-06,
-      "loss": 1.0367,
+      "epoch": 1.02,
+      "grad_norm": 18.427221298217773,
+      "learning_rate": 1.3198343304187759e-05,
+      "loss": 1.103,
       "step": 8130
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 14.082221984863281,
-      "learning_rate": 3.7025157863085097e-06,
-      "loss": 1.0125,
+      "epoch": 1.02,
+      "grad_norm": 19.781017303466797,
+      "learning_rate": 1.3197506589131073e-05,
+      "loss": 2.0529,
       "step": 8131
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 63.74949645996094,
-      "learning_rate": 3.7005111757041194e-06,
-      "loss": 2.3996,
+      "epoch": 1.02,
+      "grad_norm": 15.670022964477539,
+      "learning_rate": 1.3196669874074386e-05,
+      "loss": 0.8893,
       "step": 8132
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 9.598031044006348,
-      "learning_rate": 3.69850656509973e-06,
-      "loss": 1.0888,
+      "epoch": 1.02,
+      "grad_norm": 57.735897064208984,
+      "learning_rate": 1.3195833159017697e-05,
+      "loss": 1.2899,
       "step": 8133
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 15.858400344848633,
-      "learning_rate": 3.6965019544953396e-06,
-      "loss": 1.055,
+      "epoch": 1.02,
+      "grad_norm": 16.303756713867188,
+      "learning_rate": 1.319499644396101e-05,
+      "loss": 1.8557,
       "step": 8134
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 10.83476734161377,
-      "learning_rate": 3.6944973438909492e-06,
-      "loss": 1.9326,
+      "epoch": 1.02,
+      "grad_norm": 12.056116104125977,
+      "learning_rate": 1.3194159728904324e-05,
+      "loss": 0.8946,
       "step": 8135
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 10.07654857635498,
-      "learning_rate": 3.6924927332865597e-06,
-      "loss": 1.5472,
+      "epoch": 1.02,
+      "grad_norm": 36.674530029296875,
+      "learning_rate": 1.3193323013847634e-05,
+      "loss": 1.7452,
       "step": 8136
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 16.537139892578125,
-      "learning_rate": 3.6904881226821694e-06,
-      "loss": 1.2108,
+      "epoch": 1.02,
+      "grad_norm": 8.627217292785645,
+      "learning_rate": 1.3192486298790948e-05,
+      "loss": 0.6619,
       "step": 8137
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 15.194998741149902,
-      "learning_rate": 3.688483512077779e-06,
-      "loss": 1.0623,
+      "epoch": 1.02,
+      "grad_norm": 14.593555450439453,
+      "learning_rate": 1.319164958373426e-05,
+      "loss": 1.1673,
       "step": 8138
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 43.36076736450195,
-      "learning_rate": 3.686478901473389e-06,
-      "loss": 1.2945,
+      "epoch": 1.02,
+      "grad_norm": 4.4681549072265625,
+      "learning_rate": 1.3190812868677573e-05,
+      "loss": 2.2662,
       "step": 8139
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 9.480472564697266,
-      "learning_rate": 3.684474290868999e-06,
-      "loss": 0.5474,
+      "epoch": 1.02,
+      "grad_norm": 14.621169090270996,
+      "learning_rate": 1.3189976153620885e-05,
+      "loss": 1.2689,
       "step": 8140
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 36.88868713378906,
-      "learning_rate": 3.682469680264609e-06,
-      "loss": 1.0721,
+      "epoch": 1.02,
+      "grad_norm": 9.740835189819336,
+      "learning_rate": 1.3189139438564197e-05,
+      "loss": 1.1295,
       "step": 8141
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 42.60389709472656,
-      "learning_rate": 3.680465069660219e-06,
-      "loss": 1.4549,
+      "epoch": 1.02,
+      "grad_norm": 21.940420150756836,
+      "learning_rate": 1.3188302723507511e-05,
+      "loss": 0.8043,
       "step": 8142
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 22.67746353149414,
-      "learning_rate": 3.6784604590558286e-06,
-      "loss": 2.1135,
+      "epoch": 1.02,
+      "grad_norm": 12.638836860656738,
+      "learning_rate": 1.3187466008450821e-05,
+      "loss": 1.6467,
       "step": 8143
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 8.67175006866455,
-      "learning_rate": 3.6764558484514383e-06,
-      "loss": 1.0506,
+      "epoch": 1.02,
+      "grad_norm": 13.792120933532715,
+      "learning_rate": 1.3186629293394135e-05,
+      "loss": 2.3832,
       "step": 8144
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 18.72547149658203,
-      "learning_rate": 3.674451237847049e-06,
-      "loss": 0.891,
+      "epoch": 1.02,
+      "grad_norm": 18.201194763183594,
+      "learning_rate": 1.3185792578337449e-05,
+      "loss": 0.1339,
       "step": 8145
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 32.578861236572266,
-      "learning_rate": 3.6724466272426585e-06,
-      "loss": 1.6625,
+      "epoch": 1.02,
+      "grad_norm": 24.331432342529297,
+      "learning_rate": 1.3184955863280762e-05,
+      "loss": 1.2141,
       "step": 8146
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 9.158719062805176,
-      "learning_rate": 3.670442016638268e-06,
-      "loss": 0.8982,
+      "epoch": 1.02,
+      "grad_norm": 16.624353408813477,
+      "learning_rate": 1.3184119148224073e-05,
+      "loss": 0.5703,
       "step": 8147
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 15.746097564697266,
-      "learning_rate": 3.6684374060338786e-06,
-      "loss": 1.349,
+      "epoch": 1.02,
+      "grad_norm": 8.646965980529785,
+      "learning_rate": 1.3183282433167386e-05,
+      "loss": 0.8436,
       "step": 8148
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 12.427379608154297,
-      "learning_rate": 3.6664327954294883e-06,
-      "loss": 1.283,
+      "epoch": 1.02,
+      "grad_norm": 9.734932899475098,
+      "learning_rate": 1.31824457181107e-05,
+      "loss": 1.2568,
       "step": 8149
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 30.601055145263672,
-      "learning_rate": 3.664428184825098e-06,
-      "loss": 0.9753,
+      "epoch": 1.02,
+      "grad_norm": 26.99791717529297,
+      "learning_rate": 1.318160900305401e-05,
+      "loss": 1.2669,
       "step": 8150
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 15.714460372924805,
-      "learning_rate": 3.662423574220708e-06,
-      "loss": 1.1643,
+      "epoch": 1.02,
+      "grad_norm": 5.206856727600098,
+      "learning_rate": 1.3180772287997324e-05,
+      "loss": 1.4424,
       "step": 8151
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 27.35734748840332,
-      "learning_rate": 3.6604189636163177e-06,
-      "loss": 1.7191,
+      "epoch": 1.02,
+      "grad_norm": 9.2411527633667,
+      "learning_rate": 1.3179935572940636e-05,
+      "loss": 0.7221,
       "step": 8152
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 16.40424919128418,
-      "learning_rate": 3.6584143530119278e-06,
-      "loss": 1.4443,
+      "epoch": 1.02,
+      "grad_norm": 28.84791374206543,
+      "learning_rate": 1.317909885788395e-05,
+      "loss": 1.5638,
       "step": 8153
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 9.086562156677246,
-      "learning_rate": 3.656409742407538e-06,
-      "loss": 1.0836,
+      "epoch": 1.02,
+      "grad_norm": 8.334434509277344,
+      "learning_rate": 1.3178262142827261e-05,
+      "loss": 0.9285,
       "step": 8154
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 16.159156799316406,
-      "learning_rate": 3.6544051318031475e-06,
-      "loss": 1.0811,
+      "epoch": 1.02,
+      "grad_norm": 6.804058074951172,
+      "learning_rate": 1.3177425427770573e-05,
+      "loss": 1.3012,
       "step": 8155
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 32.5626220703125,
-      "learning_rate": 3.652400521198757e-06,
-      "loss": 1.5753,
+      "epoch": 1.02,
+      "grad_norm": 25.955961227416992,
+      "learning_rate": 1.3176588712713887e-05,
+      "loss": 1.9186,
       "step": 8156
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 10.6674165725708,
-      "learning_rate": 3.6503959105943677e-06,
-      "loss": 0.9974,
+      "epoch": 1.02,
+      "grad_norm": 15.244351387023926,
+      "learning_rate": 1.3175751997657197e-05,
+      "loss": 1.8402,
       "step": 8157
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 11.11431884765625,
-      "learning_rate": 3.6483912999899773e-06,
-      "loss": 1.3574,
+      "epoch": 1.02,
+      "grad_norm": 13.831007957458496,
+      "learning_rate": 1.317491528260051e-05,
+      "loss": 2.3185,
       "step": 8158
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 24.535797119140625,
-      "learning_rate": 3.646386689385587e-06,
-      "loss": 1.4364,
+      "epoch": 1.02,
+      "grad_norm": 25.491474151611328,
+      "learning_rate": 1.3174078567543824e-05,
+      "loss": 1.2211,
       "step": 8159
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 49.98130416870117,
-      "learning_rate": 3.6443820787811975e-06,
-      "loss": 1.0015,
-      "step": 8160
-    },
-    {
-      "epoch": 2.45,
-      "eval_loss": 0.16650927066802979,
-      "eval_runtime": 44.1127,
-      "eval_samples_per_second": 33.528,
-      "eval_steps_per_second": 33.528,
+      "epoch": 1.02,
+      "grad_norm": 14.899972915649414,
+      "learning_rate": 1.3173241852487138e-05,
+      "loss": 0.8519,
       "step": 8160
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 9.75783634185791,
-      "learning_rate": 3.642377468176807e-06,
-      "loss": 1.1418,
+      "epoch": 1.02,
+      "grad_norm": 19.21306610107422,
+      "learning_rate": 1.3172405137430448e-05,
+      "loss": 1.4239,
       "step": 8161
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 15.587701797485352,
-      "learning_rate": 3.640372857572417e-06,
-      "loss": 0.7225,
+      "epoch": 1.02,
+      "grad_norm": 15.32010555267334,
+      "learning_rate": 1.3171568422373762e-05,
+      "loss": 1.9368,
       "step": 8162
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 17.10492706298828,
-      "learning_rate": 3.6383682469680265e-06,
-      "loss": 1.1976,
+      "epoch": 1.02,
+      "grad_norm": 35.73579788208008,
+      "learning_rate": 1.3170731707317076e-05,
+      "loss": 3.0578,
       "step": 8163
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 23.952566146850586,
-      "learning_rate": 3.6363636363636366e-06,
-      "loss": 1.2036,
+      "epoch": 1.02,
+      "grad_norm": 9.026825904846191,
+      "learning_rate": 1.3169894992260386e-05,
+      "loss": 0.4943,
       "step": 8164
     },
     {
-      "epoch": 2.45,
-      "grad_norm": 17.270137786865234,
-      "learning_rate": 3.6343590257592467e-06,
-      "loss": 1.6207,
+      "epoch": 1.02,
+      "grad_norm": 17.9897518157959,
+      "learning_rate": 1.31690582772037e-05,
+      "loss": 2.4651,
       "step": 8165
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 16.140460968017578,
-      "learning_rate": 3.6323544151548563e-06,
-      "loss": 0.9803,
+      "epoch": 1.02,
+      "grad_norm": 9.883686065673828,
+      "learning_rate": 1.3168221562147012e-05,
+      "loss": 1.4876,
       "step": 8166
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 42.045997619628906,
-      "learning_rate": 3.6303498045504664e-06,
-      "loss": 2.4721,
+      "epoch": 1.02,
+      "grad_norm": 17.007078170776367,
+      "learning_rate": 1.3167384847090325e-05,
+      "loss": 1.8312,
       "step": 8167
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 13.348779678344727,
-      "learning_rate": 3.628345193946076e-06,
-      "loss": 1.1113,
+      "epoch": 1.03,
+      "grad_norm": 26.36995506286621,
+      "learning_rate": 1.3166548132033637e-05,
+      "loss": 1.2385,
       "step": 8168
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 36.904014587402344,
-      "learning_rate": 3.6263405833416857e-06,
-      "loss": 2.5081,
+      "epoch": 1.03,
+      "grad_norm": 10.366000175476074,
+      "learning_rate": 1.3165711416976949e-05,
+      "loss": 1.5701,
       "step": 8169
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 23.20426368713379,
-      "learning_rate": 3.6243359727372962e-06,
-      "loss": 1.0172,
+      "epoch": 1.03,
+      "grad_norm": 5.277139663696289,
+      "learning_rate": 1.3164874701920263e-05,
+      "loss": 0.8169,
       "step": 8170
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 16.895856857299805,
-      "learning_rate": 3.622331362132906e-06,
-      "loss": 1.0231,
+      "epoch": 1.03,
+      "grad_norm": 59.65888977050781,
+      "learning_rate": 1.3164037986863573e-05,
+      "loss": 2.6797,
       "step": 8171
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 37.31813049316406,
-      "learning_rate": 3.6203267515285156e-06,
-      "loss": 1.3518,
+      "epoch": 1.03,
+      "grad_norm": 4.414580821990967,
+      "learning_rate": 1.3163201271806887e-05,
+      "loss": 0.1261,
       "step": 8172
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 19.724748611450195,
-      "learning_rate": 3.618322140924126e-06,
-      "loss": 1.2338,
+      "epoch": 1.03,
+      "grad_norm": 55.824485778808594,
+      "learning_rate": 1.31623645567502e-05,
+      "loss": 1.2241,
       "step": 8173
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 19.960813522338867,
-      "learning_rate": 3.6163175303197357e-06,
-      "loss": 1.4858,
+      "epoch": 1.03,
+      "grad_norm": 23.670427322387695,
+      "learning_rate": 1.3161527841693514e-05,
+      "loss": 1.7405,
       "step": 8174
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 11.546113967895508,
-      "learning_rate": 3.6143129197153454e-06,
-      "loss": 1.2553,
+      "epoch": 1.03,
+      "grad_norm": 30.104541778564453,
+      "learning_rate": 1.3160691126636824e-05,
+      "loss": 2.3148,
       "step": 8175
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 18.865678787231445,
-      "learning_rate": 3.6123083091109555e-06,
-      "loss": 1.2506,
+      "epoch": 1.03,
+      "grad_norm": 23.93840217590332,
+      "learning_rate": 1.3159854411580138e-05,
+      "loss": 1.0324,
       "step": 8176
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 22.531909942626953,
-      "learning_rate": 3.6103036985065656e-06,
-      "loss": 1.6922,
+      "epoch": 1.03,
+      "grad_norm": 7.747591972351074,
+      "learning_rate": 1.315901769652345e-05,
+      "loss": 0.6227,
       "step": 8177
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 20.21176528930664,
-      "learning_rate": 3.6082990879021752e-06,
-      "loss": 0.8544,
+      "epoch": 1.03,
+      "grad_norm": 13.66466236114502,
+      "learning_rate": 1.3158180981466762e-05,
+      "loss": 2.1594,
       "step": 8178
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 15.008464813232422,
-      "learning_rate": 3.6062944772977853e-06,
-      "loss": 1.3876,
+      "epoch": 1.03,
+      "grad_norm": 33.47099304199219,
+      "learning_rate": 1.3157344266410075e-05,
+      "loss": 1.5768,
       "step": 8179
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 24.066457748413086,
-      "learning_rate": 3.604289866693395e-06,
-      "loss": 1.4522,
+      "epoch": 1.03,
+      "grad_norm": 9.036125183105469,
+      "learning_rate": 1.3156507551353387e-05,
+      "loss": 0.7407,
       "step": 8180
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 7.842005729675293,
-      "learning_rate": 3.6022852560890046e-06,
-      "loss": 0.5896,
+      "epoch": 1.03,
+      "grad_norm": 10.413735389709473,
+      "learning_rate": 1.3155670836296701e-05,
+      "loss": 1.5001,
       "step": 8181
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 13.021004676818848,
-      "learning_rate": 3.600280645484615e-06,
-      "loss": 0.7843,
+      "epoch": 1.03,
+      "grad_norm": 10.219388008117676,
+      "learning_rate": 1.3154834121240013e-05,
+      "loss": 0.7861,
       "step": 8182
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 21.92922592163086,
-      "learning_rate": 3.598276034880225e-06,
-      "loss": 2.2506,
+      "epoch": 1.03,
+      "grad_norm": 17.917850494384766,
+      "learning_rate": 1.3153997406183325e-05,
+      "loss": 1.4693,
       "step": 8183
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 11.565603256225586,
-      "learning_rate": 3.5962714242758345e-06,
-      "loss": 1.0158,
+      "epoch": 1.03,
+      "grad_norm": 35.04172134399414,
+      "learning_rate": 1.3153160691126639e-05,
+      "loss": 0.9711,
       "step": 8184
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 18.74056053161621,
-      "learning_rate": 3.594266813671445e-06,
-      "loss": 0.9444,
+      "epoch": 1.03,
+      "grad_norm": 7.332533359527588,
+      "learning_rate": 1.3152323976069949e-05,
+      "loss": 0.7555,
       "step": 8185
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 14.562174797058105,
-      "learning_rate": 3.5922622030670546e-06,
-      "loss": 0.7885,
+      "epoch": 1.03,
+      "grad_norm": 10.041555404663086,
+      "learning_rate": 1.3151487261013262e-05,
+      "loss": 0.8399,
       "step": 8186
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 37.87458038330078,
-      "learning_rate": 3.5902575924626643e-06,
-      "loss": 1.1931,
+      "epoch": 1.03,
+      "grad_norm": 7.564031600952148,
+      "learning_rate": 1.3150650545956576e-05,
+      "loss": 0.7352,
       "step": 8187
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 14.50774097442627,
-      "learning_rate": 3.5882529818582744e-06,
-      "loss": 1.1908,
+      "epoch": 1.03,
+      "grad_norm": 12.851036071777344,
+      "learning_rate": 1.314981383089989e-05,
+      "loss": 0.5656,
       "step": 8188
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 37.05324172973633,
-      "learning_rate": 3.5862483712538845e-06,
-      "loss": 2.294,
+      "epoch": 1.03,
+      "grad_norm": 5.794101715087891,
+      "learning_rate": 1.31489771158432e-05,
+      "loss": 0.7497,
       "step": 8189
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 7.687615394592285,
-      "learning_rate": 3.584243760649494e-06,
-      "loss": 0.7946,
+      "epoch": 1.03,
+      "grad_norm": 5.937995910644531,
+      "learning_rate": 1.3148140400786514e-05,
+      "loss": 0.5418,
       "step": 8190
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 10.446063995361328,
-      "learning_rate": 3.582239150045104e-06,
-      "loss": 1.2933,
+      "epoch": 1.03,
+      "grad_norm": 3.5626814365386963,
+      "learning_rate": 1.3147303685729826e-05,
+      "loss": 0.1635,
       "step": 8191
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 25.05036735534668,
-      "learning_rate": 3.580234539440714e-06,
-      "loss": 1.8224,
+      "epoch": 1.03,
+      "grad_norm": 9.711762428283691,
+      "learning_rate": 1.3146466970673138e-05,
+      "loss": 1.463,
       "step": 8192
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 12.898480415344238,
-      "learning_rate": 3.5782299288363235e-06,
-      "loss": 1.3031,
+      "epoch": 1.03,
+      "grad_norm": 16.928871154785156,
+      "learning_rate": 1.3145630255616451e-05,
+      "loss": 2.0754,
       "step": 8193
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 23.968446731567383,
-      "learning_rate": 3.576225318231934e-06,
-      "loss": 2.3017,
+      "epoch": 1.03,
+      "grad_norm": 14.005651473999023,
+      "learning_rate": 1.3144793540559763e-05,
+      "loss": 0.7658,
       "step": 8194
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 16.672903060913086,
-      "learning_rate": 3.5742207076275437e-06,
-      "loss": 0.9025,
+      "epoch": 1.03,
+      "grad_norm": 13.096718788146973,
+      "learning_rate": 1.3143956825503077e-05,
+      "loss": 2.0003,
       "step": 8195
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 10.29025936126709,
-      "learning_rate": 3.5722160970231534e-06,
-      "loss": 1.0269,
+      "epoch": 1.03,
+      "grad_norm": 10.545746803283691,
+      "learning_rate": 1.3143120110446387e-05,
+      "loss": 1.2345,
       "step": 8196
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 11.360820770263672,
-      "learning_rate": 3.570211486418764e-06,
-      "loss": 1.7671,
+      "epoch": 1.03,
+      "grad_norm": 15.581537246704102,
+      "learning_rate": 1.31422833953897e-05,
+      "loss": 0.686,
       "step": 8197
     },
     {
-      "epoch": 2.46,
-      "grad_norm": 28.387712478637695,
-      "learning_rate": 3.5682068758143735e-06,
-      "loss": 0.9867,
+      "epoch": 1.03,
+      "grad_norm": 14.40272045135498,
+      "learning_rate": 1.3141446680333014e-05,
+      "loss": 1.2653,
       "step": 8198
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 37.20726776123047,
-      "learning_rate": 3.566202265209983e-06,
-      "loss": 2.4591,
+      "epoch": 1.03,
+      "grad_norm": 30.775413513183594,
+      "learning_rate": 1.3140609965276325e-05,
+      "loss": 2.5021,
       "step": 8199
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 8.600688934326172,
-      "learning_rate": 3.5641976546055933e-06,
-      "loss": 0.8608,
+      "epoch": 1.03,
+      "grad_norm": 9.351306915283203,
+      "learning_rate": 1.3139773250219638e-05,
+      "loss": 0.517,
       "step": 8200
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 42.92110824584961,
-      "learning_rate": 3.562193044001203e-06,
-      "loss": 1.5033,
+      "epoch": 1.03,
+      "grad_norm": 15.242072105407715,
+      "learning_rate": 1.3138936535162952e-05,
+      "loss": 0.8643,
       "step": 8201
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 9.241715431213379,
-      "learning_rate": 3.560188433396813e-06,
-      "loss": 0.8824,
+      "epoch": 1.03,
+      "grad_norm": 9.428893089294434,
+      "learning_rate": 1.3138099820106266e-05,
+      "loss": 3.8307,
       "step": 8202
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 18.05348777770996,
-      "learning_rate": 3.558183822792423e-06,
-      "loss": 0.8519,
+      "epoch": 1.03,
+      "grad_norm": 13.981907844543457,
+      "learning_rate": 1.3137263105049576e-05,
+      "loss": 1.1886,
       "step": 8203
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 13.041126251220703,
-      "learning_rate": 3.5561792121880328e-06,
-      "loss": 0.9648,
+      "epoch": 1.03,
+      "grad_norm": 12.839777946472168,
+      "learning_rate": 1.313642638999289e-05,
+      "loss": 0.8264,
       "step": 8204
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 15.101016998291016,
-      "learning_rate": 3.5541746015836424e-06,
-      "loss": 1.7081,
+      "epoch": 1.03,
+      "grad_norm": 8.989615440368652,
+      "learning_rate": 1.3135589674936201e-05,
+      "loss": 0.6943,
       "step": 8205
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 21.521345138549805,
-      "learning_rate": 3.5521699909792525e-06,
-      "loss": 2.7192,
+      "epoch": 1.03,
+      "grad_norm": 12.968503952026367,
+      "learning_rate": 1.3134752959879513e-05,
+      "loss": 0.7676,
       "step": 8206
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 27.694257736206055,
-      "learning_rate": 3.5501653803748626e-06,
-      "loss": 1.5167,
+      "epoch": 1.03,
+      "grad_norm": 10.712398529052734,
+      "learning_rate": 1.3133916244822827e-05,
+      "loss": 0.555,
       "step": 8207
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 10.351465225219727,
-      "learning_rate": 3.5481607697704722e-06,
-      "loss": 1.3463,
+      "epoch": 1.03,
+      "grad_norm": 13.51182746887207,
+      "learning_rate": 1.3133079529766139e-05,
+      "loss": 1.0422,
       "step": 8208
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 29.087156295776367,
-      "learning_rate": 3.546156159166082e-06,
-      "loss": 1.397,
+      "epoch": 1.03,
+      "grad_norm": 60.49845504760742,
+      "learning_rate": 1.3132242814709453e-05,
+      "loss": 2.5942,
       "step": 8209
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 14.024823188781738,
-      "learning_rate": 3.5441515485616924e-06,
-      "loss": 0.8097,
+      "epoch": 1.03,
+      "grad_norm": 9.75202465057373,
+      "learning_rate": 1.3131406099652763e-05,
+      "loss": 1.0263,
       "step": 8210
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 11.973261833190918,
-      "learning_rate": 3.542146937957302e-06,
-      "loss": 0.7829,
+      "epoch": 1.03,
+      "grad_norm": 7.670017242431641,
+      "learning_rate": 1.3130569384596077e-05,
+      "loss": 0.427,
       "step": 8211
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 8.828713417053223,
-      "learning_rate": 3.5401423273529117e-06,
-      "loss": 0.7569,
+      "epoch": 1.03,
+      "grad_norm": 10.970623016357422,
+      "learning_rate": 1.312973266953939e-05,
+      "loss": 0.9992,
       "step": 8212
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 32.27656173706055,
-      "learning_rate": 3.538137716748522e-06,
-      "loss": 0.9883,
+      "epoch": 1.03,
+      "grad_norm": 43.433349609375,
+      "learning_rate": 1.31288959544827e-05,
+      "loss": 2.8945,
       "step": 8213
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 54.36296463012695,
-      "learning_rate": 3.536133106144132e-06,
-      "loss": 2.0888,
+      "epoch": 1.03,
+      "grad_norm": 12.724357604980469,
+      "learning_rate": 1.3128059239426014e-05,
+      "loss": 1.0343,
       "step": 8214
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 30.008255004882812,
-      "learning_rate": 3.5341284955397416e-06,
-      "loss": 1.1675,
+      "epoch": 1.03,
+      "grad_norm": 10.736432075500488,
+      "learning_rate": 1.3127222524369328e-05,
+      "loss": 1.8907,
       "step": 8215
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 20.43587875366211,
-      "learning_rate": 3.5321238849353517e-06,
-      "loss": 1.0271,
+      "epoch": 1.03,
+      "grad_norm": 25.897153854370117,
+      "learning_rate": 1.3126385809312641e-05,
+      "loss": 1.8143,
       "step": 8216
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 9.914734840393066,
-      "learning_rate": 3.5301192743309613e-06,
-      "loss": 1.4436,
+      "epoch": 1.03,
+      "grad_norm": 10.944652557373047,
+      "learning_rate": 1.3125549094255952e-05,
+      "loss": 1.6987,
       "step": 8217
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 10.097274780273438,
-      "learning_rate": 3.5281146637265714e-06,
-      "loss": 1.2491,
+      "epoch": 1.03,
+      "grad_norm": 15.509830474853516,
+      "learning_rate": 1.3124712379199265e-05,
+      "loss": 1.0449,
       "step": 8218
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 13.336429595947266,
-      "learning_rate": 3.5261100531221815e-06,
-      "loss": 1.1802,
+      "epoch": 1.03,
+      "grad_norm": 25.927719116210938,
+      "learning_rate": 1.3123875664142577e-05,
+      "loss": 1.7837,
       "step": 8219
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 14.391281127929688,
-      "learning_rate": 3.524105442517791e-06,
-      "loss": 1.0317,
+      "epoch": 1.03,
+      "grad_norm": 16.40271759033203,
+      "learning_rate": 1.312303894908589e-05,
+      "loss": 1.5452,
       "step": 8220
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 21.83466339111328,
-      "learning_rate": 3.522100831913401e-06,
-      "loss": 2.0406,
+      "epoch": 1.03,
+      "grad_norm": 8.034306526184082,
+      "learning_rate": 1.3122202234029203e-05,
+      "loss": 0.4517,
       "step": 8221
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 13.69311809539795,
-      "learning_rate": 3.5200962213090113e-06,
-      "loss": 1.0287,
+      "epoch": 1.03,
+      "grad_norm": 18.109060287475586,
+      "learning_rate": 1.3121365518972515e-05,
+      "loss": 1.3607,
       "step": 8222
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 14.4343900680542,
-      "learning_rate": 3.518091610704621e-06,
-      "loss": 1.1169,
+      "epoch": 1.03,
+      "grad_norm": 13.58246898651123,
+      "learning_rate": 1.3120528803915829e-05,
+      "loss": 1.6415,
       "step": 8223
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 19.60431671142578,
-      "learning_rate": 3.5160870001002306e-06,
-      "loss": 1.2294,
+      "epoch": 1.03,
+      "grad_norm": 20.36124038696289,
+      "learning_rate": 1.3119692088859139e-05,
+      "loss": 1.2454,
       "step": 8224
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 8.042855262756348,
-      "learning_rate": 3.5140823894958407e-06,
-      "loss": 0.8052,
+      "epoch": 1.03,
+      "grad_norm": 32.146080017089844,
+      "learning_rate": 1.3118855373802452e-05,
+      "loss": 1.7477,
       "step": 8225
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 13.775824546813965,
-      "learning_rate": 3.512077778891451e-06,
-      "loss": 1.0722,
+      "epoch": 1.03,
+      "grad_norm": 12.584585189819336,
+      "learning_rate": 1.3118018658745766e-05,
+      "loss": 1.3979,
       "step": 8226
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 15.00515079498291,
-      "learning_rate": 3.5100731682870605e-06,
-      "loss": 1.0372,
+      "epoch": 1.03,
+      "grad_norm": 10.174505233764648,
+      "learning_rate": 1.3117181943689076e-05,
+      "loss": 0.8317,
       "step": 8227
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 52.58286666870117,
-      "learning_rate": 3.5080685576826705e-06,
-      "loss": 4.7955,
+      "epoch": 1.03,
+      "grad_norm": 39.13493728637695,
+      "learning_rate": 1.311634522863239e-05,
+      "loss": 3.3734,
       "step": 8228
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 24.154563903808594,
-      "learning_rate": 3.50606394707828e-06,
-      "loss": 1.4256,
+      "epoch": 1.03,
+      "grad_norm": 9.438246726989746,
+      "learning_rate": 1.3115508513575704e-05,
+      "loss": 0.4985,
       "step": 8229
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 13.502425193786621,
-      "learning_rate": 3.50405933647389e-06,
-      "loss": 1.381,
+      "epoch": 1.03,
+      "grad_norm": 22.061195373535156,
+      "learning_rate": 1.3114671798519016e-05,
+      "loss": 1.6837,
       "step": 8230
     },
     {
-      "epoch": 2.47,
-      "grad_norm": 9.207113265991211,
-      "learning_rate": 3.5020547258695004e-06,
-      "loss": 0.4176,
+      "epoch": 1.03,
+      "grad_norm": 11.457298278808594,
+      "learning_rate": 1.3113835083462328e-05,
+      "loss": 0.9882,
       "step": 8231
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 21.608287811279297,
-      "learning_rate": 3.50005011526511e-06,
-      "loss": 0.7372,
+      "epoch": 1.03,
+      "grad_norm": 27.55769920349121,
+      "learning_rate": 1.3112998368405641e-05,
+      "loss": 1.9697,
       "step": 8232
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 13.105839729309082,
-      "learning_rate": 3.4980455046607197e-06,
-      "loss": 0.719,
+      "epoch": 1.03,
+      "grad_norm": 61.551185607910156,
+      "learning_rate": 1.3112161653348953e-05,
+      "loss": 1.7753,
       "step": 8233
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 17.119613647460938,
-      "learning_rate": 3.49604089405633e-06,
-      "loss": 1.2089,
+      "epoch": 1.03,
+      "grad_norm": 9.996932983398438,
+      "learning_rate": 1.3111324938292265e-05,
+      "loss": 1.5993,
       "step": 8234
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 129.32228088378906,
-      "learning_rate": 3.49403628345194e-06,
-      "loss": 2.5109,
+      "epoch": 1.03,
+      "grad_norm": 10.689933776855469,
+      "learning_rate": 1.3110488223235579e-05,
+      "loss": 1.6766,
       "step": 8235
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 11.509007453918457,
-      "learning_rate": 3.4920316728475495e-06,
-      "loss": 1.1199,
+      "epoch": 1.03,
+      "grad_norm": 26.191883087158203,
+      "learning_rate": 1.310965150817889e-05,
+      "loss": 2.6339,
       "step": 8236
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 17.33429718017578,
-      "learning_rate": 3.4900270622431596e-06,
-      "loss": 1.3015,
+      "epoch": 1.03,
+      "grad_norm": 31.052392959594727,
+      "learning_rate": 1.3108814793122204e-05,
+      "loss": 4.1182,
       "step": 8237
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 28.055233001708984,
-      "learning_rate": 3.4880224516387697e-06,
-      "loss": 2.121,
+      "epoch": 1.03,
+      "grad_norm": 6.11808967590332,
+      "learning_rate": 1.3107978078065515e-05,
+      "loss": 0.3464,
       "step": 8238
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 22.95208740234375,
-      "learning_rate": 3.4860178410343794e-06,
-      "loss": 1.7331,
+      "epoch": 1.03,
+      "grad_norm": 8.085333824157715,
+      "learning_rate": 1.3107141363008828e-05,
+      "loss": 1.1866,
       "step": 8239
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 12.567434310913086,
-      "learning_rate": 3.4840132304299894e-06,
-      "loss": 1.5196,
+      "epoch": 1.03,
+      "grad_norm": 16.53407859802246,
+      "learning_rate": 1.3106304647952142e-05,
+      "loss": 3.1467,
       "step": 8240
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 40.08758544921875,
-      "learning_rate": 3.482008619825599e-06,
-      "loss": 1.1594,
+      "epoch": 1.03,
+      "grad_norm": 10.062830924987793,
+      "learning_rate": 1.3105467932895452e-05,
+      "loss": 1.7672,
       "step": 8241
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 17.78892707824707,
-      "learning_rate": 3.4800040092212088e-06,
-      "loss": 1.0566,
+      "epoch": 1.03,
+      "grad_norm": 20.216842651367188,
+      "learning_rate": 1.3104631217838766e-05,
+      "loss": 2.2723,
       "step": 8242
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 28.924680709838867,
-      "learning_rate": 3.4779993986168193e-06,
-      "loss": 1.722,
+      "epoch": 1.03,
+      "grad_norm": 17.774564743041992,
+      "learning_rate": 1.310379450278208e-05,
+      "loss": 0.7638,
       "step": 8243
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 13.244800567626953,
-      "learning_rate": 3.475994788012429e-06,
-      "loss": 1.5066,
+      "epoch": 1.03,
+      "grad_norm": 7.087859153747559,
+      "learning_rate": 1.3102957787725391e-05,
+      "loss": 0.5927,
       "step": 8244
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 46.74547576904297,
-      "learning_rate": 3.4739901774080386e-06,
-      "loss": 1.802,
+      "epoch": 1.03,
+      "grad_norm": 6.852178573608398,
+      "learning_rate": 1.3102121072668703e-05,
+      "loss": 0.5539,
       "step": 8245
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 19.008142471313477,
-      "learning_rate": 3.4719855668036482e-06,
-      "loss": 1.4982,
+      "epoch": 1.03,
+      "grad_norm": 25.83189582824707,
+      "learning_rate": 1.3101284357612017e-05,
+      "loss": 2.1372,
       "step": 8246
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 16.53547477722168,
-      "learning_rate": 3.4699809561992588e-06,
-      "loss": 1.4899,
+      "epoch": 1.03,
+      "grad_norm": 14.124688148498535,
+      "learning_rate": 1.3100447642555329e-05,
+      "loss": 1.074,
       "step": 8247
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 18.363264083862305,
-      "learning_rate": 3.4679763455948684e-06,
-      "loss": 1.022,
+      "epoch": 1.04,
+      "grad_norm": 7.1842360496521,
+      "learning_rate": 1.3099610927498641e-05,
+      "loss": 3.1602,
       "step": 8248
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 7.702117443084717,
-      "learning_rate": 3.465971734990478e-06,
-      "loss": 1.3944,
+      "epoch": 1.04,
+      "grad_norm": 16.31489372253418,
+      "learning_rate": 1.3098774212441953e-05,
+      "loss": 1.529,
       "step": 8249
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 8.90109634399414,
-      "learning_rate": 3.4639671243860886e-06,
-      "loss": 0.8569,
+      "epoch": 1.04,
+      "grad_norm": 18.872175216674805,
+      "learning_rate": 1.3097937497385267e-05,
+      "loss": 1.4403,
       "step": 8250
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 34.664485931396484,
-      "learning_rate": 3.4619625137816982e-06,
-      "loss": 1.817,
+      "epoch": 1.04,
+      "grad_norm": 24.27296257019043,
+      "learning_rate": 1.309710078232858e-05,
+      "loss": 1.354,
       "step": 8251
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 12.355849266052246,
-      "learning_rate": 3.459957903177308e-06,
-      "loss": 0.9937,
+      "epoch": 1.04,
+      "grad_norm": 19.99920082092285,
+      "learning_rate": 1.309626406727189e-05,
+      "loss": 0.6707,
       "step": 8252
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 12.493422508239746,
-      "learning_rate": 3.457953292572918e-06,
-      "loss": 1.1074,
+      "epoch": 1.04,
+      "grad_norm": 18.022979736328125,
+      "learning_rate": 1.3095427352215204e-05,
+      "loss": 0.9206,
       "step": 8253
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 20.661190032958984,
-      "learning_rate": 3.4559486819685277e-06,
-      "loss": 2.2912,
+      "epoch": 1.04,
+      "grad_norm": 8.533984184265137,
+      "learning_rate": 1.3094590637158518e-05,
+      "loss": 1.351,
       "step": 8254
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 19.867286682128906,
-      "learning_rate": 3.4539440713641377e-06,
-      "loss": 1.6991,
+      "epoch": 1.04,
+      "grad_norm": 9.833261489868164,
+      "learning_rate": 1.3093753922101828e-05,
+      "loss": 1.4907,
       "step": 8255
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 13.034092903137207,
-      "learning_rate": 3.451939460759748e-06,
-      "loss": 0.8973,
+      "epoch": 1.04,
+      "grad_norm": 21.648242950439453,
+      "learning_rate": 1.3092917207045142e-05,
+      "loss": 0.5904,
       "step": 8256
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 14.20175552368164,
-      "learning_rate": 3.4499348501553575e-06,
-      "loss": 1.1049,
+      "epoch": 1.04,
+      "grad_norm": 41.22957992553711,
+      "learning_rate": 1.3092080491988455e-05,
+      "loss": 0.6205,
       "step": 8257
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 12.550244331359863,
-      "learning_rate": 3.447930239550967e-06,
-      "loss": 1.1492,
+      "epoch": 1.04,
+      "grad_norm": 5.088689804077148,
+      "learning_rate": 1.3091243776931767e-05,
+      "loss": 0.337,
       "step": 8258
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 13.553984642028809,
-      "learning_rate": 3.4459256289465777e-06,
-      "loss": 1.1702,
+      "epoch": 1.04,
+      "grad_norm": 14.02363109588623,
+      "learning_rate": 1.309040706187508e-05,
+      "loss": 3.9714,
       "step": 8259
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 17.68621826171875,
-      "learning_rate": 3.4439210183421873e-06,
-      "loss": 1.4533,
+      "epoch": 1.04,
+      "grad_norm": 15.000255584716797,
+      "learning_rate": 1.3089570346818393e-05,
+      "loss": 1.4824,
       "step": 8260
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 19.83187484741211,
-      "learning_rate": 3.441916407737797e-06,
-      "loss": 1.4019,
+      "epoch": 1.04,
+      "grad_norm": 25.85439109802246,
+      "learning_rate": 1.3088733631761705e-05,
+      "loss": 2.2748,
       "step": 8261
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 11.274413108825684,
-      "learning_rate": 3.4399117971334075e-06,
-      "loss": 1.7506,
+      "epoch": 1.04,
+      "grad_norm": 39.99784469604492,
+      "learning_rate": 1.3087896916705017e-05,
+      "loss": 0.5131,
       "step": 8262
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 27.54680061340332,
-      "learning_rate": 3.437907186529017e-06,
-      "loss": 1.5858,
+      "epoch": 1.04,
+      "grad_norm": 6.106348514556885,
+      "learning_rate": 1.3087060201648329e-05,
+      "loss": 0.4545,
       "step": 8263
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 9.336912155151367,
-      "learning_rate": 3.435902575924627e-06,
-      "loss": 0.7806,
+      "epoch": 1.04,
+      "grad_norm": 17.503942489624023,
+      "learning_rate": 1.3086223486591642e-05,
+      "loss": 2.8439,
       "step": 8264
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 53.21189498901367,
-      "learning_rate": 3.433897965320237e-06,
-      "loss": 1.7986,
+      "epoch": 1.04,
+      "grad_norm": 7.972207069396973,
+      "learning_rate": 1.3085386771534956e-05,
+      "loss": 0.4385,
       "step": 8265
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 9.917101860046387,
-      "learning_rate": 3.4318933547158465e-06,
-      "loss": 0.8816,
+      "epoch": 1.04,
+      "grad_norm": 21.280460357666016,
+      "learning_rate": 1.3084550056478266e-05,
+      "loss": 1.4569,
       "step": 8266
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 8.793371200561523,
-      "learning_rate": 3.4298887441114566e-06,
-      "loss": 0.8211,
+      "epoch": 1.04,
+      "grad_norm": 9.055754661560059,
+      "learning_rate": 1.308371334142158e-05,
+      "loss": 1.4348,
       "step": 8267
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 6.4236626625061035,
-      "learning_rate": 3.4278841335070667e-06,
-      "loss": 0.7913,
+      "epoch": 1.04,
+      "grad_norm": 10.815393447875977,
+      "learning_rate": 1.3082876626364894e-05,
+      "loss": 0.6324,
       "step": 8268
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 22.39311408996582,
-      "learning_rate": 3.4258795229026764e-06,
-      "loss": 1.1071,
+      "epoch": 1.04,
+      "grad_norm": 4.52388858795166,
+      "learning_rate": 1.3082039911308204e-05,
+      "loss": 0.0638,
       "step": 8269
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 15.763972282409668,
-      "learning_rate": 3.423874912298286e-06,
-      "loss": 0.7142,
+      "epoch": 1.04,
+      "grad_norm": 19.21198081970215,
+      "learning_rate": 1.3081203196251517e-05,
+      "loss": 1.4016,
       "step": 8270
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 16.8847713470459,
-      "learning_rate": 3.4218703016938965e-06,
-      "loss": 1.2469,
+      "epoch": 1.04,
+      "grad_norm": 16.03714370727539,
+      "learning_rate": 1.3080366481194831e-05,
+      "loss": 1.0052,
       "step": 8271
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 9.058008193969727,
-      "learning_rate": 3.419865691089506e-06,
-      "loss": 1.3107,
+      "epoch": 1.04,
+      "grad_norm": 14.488385200500488,
+      "learning_rate": 1.3079529766138143e-05,
+      "loss": 1.3679,
       "step": 8272
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 19.072185516357422,
-      "learning_rate": 3.417861080485116e-06,
-      "loss": 1.2581,
+      "epoch": 1.04,
+      "grad_norm": 19.894723892211914,
+      "learning_rate": 1.3078693051081455e-05,
+      "loss": 1.1507,
       "step": 8273
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 30.158096313476562,
-      "learning_rate": 3.4158564698807264e-06,
-      "loss": 1.2928,
+      "epoch": 1.04,
+      "grad_norm": 5.145405292510986,
+      "learning_rate": 1.3077856336024769e-05,
+      "loss": 0.2995,
       "step": 8274
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 11.686331748962402,
-      "learning_rate": 3.413851859276336e-06,
-      "loss": 1.0867,
+      "epoch": 1.04,
+      "grad_norm": 9.702325820922852,
+      "learning_rate": 1.307701962096808e-05,
+      "loss": 1.1392,
       "step": 8275
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 13.757401466369629,
-      "learning_rate": 3.4118472486719457e-06,
-      "loss": 1.5826,
+      "epoch": 1.04,
+      "grad_norm": 14.014906883239746,
+      "learning_rate": 1.3076182905911393e-05,
+      "loss": 1.3747,
       "step": 8276
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 18.740379333496094,
-      "learning_rate": 3.4098426380675558e-06,
-      "loss": 1.2056,
+      "epoch": 1.04,
+      "grad_norm": 20.981679916381836,
+      "learning_rate": 1.3075346190854705e-05,
+      "loss": 1.193,
       "step": 8277
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 32.20988082885742,
-      "learning_rate": 3.4078380274631654e-06,
-      "loss": 1.3858,
+      "epoch": 1.04,
+      "grad_norm": 77.5040512084961,
+      "learning_rate": 1.3074509475798018e-05,
+      "loss": 2.1638,
       "step": 8278
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 13.55139446258545,
-      "learning_rate": 3.4058334168587755e-06,
-      "loss": 1.0321,
+      "epoch": 1.04,
+      "grad_norm": 6.79390811920166,
+      "learning_rate": 1.3073672760741332e-05,
+      "loss": 0.7708,
       "step": 8279
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 10.795275688171387,
-      "learning_rate": 3.4038288062543856e-06,
-      "loss": 1.2745,
-      "step": 8280
-    },
-    {
-      "epoch": 2.49,
-      "eval_loss": 0.17804668843746185,
-      "eval_runtime": 43.714,
-      "eval_samples_per_second": 33.834,
-      "eval_steps_per_second": 33.834,
+      "epoch": 1.04,
+      "grad_norm": 43.72199249267578,
+      "learning_rate": 1.3072836045684642e-05,
+      "loss": 2.7219,
       "step": 8280
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 11.243696212768555,
-      "learning_rate": 3.4018241956499953e-06,
-      "loss": 0.8101,
+      "epoch": 1.04,
+      "grad_norm": 12.676359176635742,
+      "learning_rate": 1.3071999330627956e-05,
+      "loss": 0.864,
       "step": 8281
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 10.823609352111816,
-      "learning_rate": 3.399819585045605e-06,
-      "loss": 2.0621,
+      "epoch": 1.04,
+      "grad_norm": 13.768181800842285,
+      "learning_rate": 1.307116261557127e-05,
+      "loss": 0.7752,
       "step": 8282
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 52.87884521484375,
-      "learning_rate": 3.3978149744412154e-06,
-      "loss": 2.0929,
+      "epoch": 1.04,
+      "grad_norm": 13.410317420959473,
+      "learning_rate": 1.307032590051458e-05,
+      "loss": 1.1399,
       "step": 8283
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 11.563464164733887,
-      "learning_rate": 3.395810363836825e-06,
-      "loss": 0.8609,
+      "epoch": 1.04,
+      "grad_norm": 55.05906677246094,
+      "learning_rate": 1.3069489185457893e-05,
+      "loss": 3.1471,
       "step": 8284
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 39.08991241455078,
-      "learning_rate": 3.3938057532324348e-06,
-      "loss": 1.3453,
+      "epoch": 1.04,
+      "grad_norm": 11.868683815002441,
+      "learning_rate": 1.3068652470401207e-05,
+      "loss": 0.7911,
       "step": 8285
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 72.77201843261719,
-      "learning_rate": 3.391801142628045e-06,
-      "loss": 1.5782,
+      "epoch": 1.04,
+      "grad_norm": 16.719898223876953,
+      "learning_rate": 1.3067815755344519e-05,
+      "loss": 2.0644,
       "step": 8286
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 44.493431091308594,
-      "learning_rate": 3.389796532023655e-06,
-      "loss": 1.312,
+      "epoch": 1.04,
+      "grad_norm": 9.625706672668457,
+      "learning_rate": 1.3066979040287831e-05,
+      "loss": 1.6071,
       "step": 8287
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 49.612998962402344,
-      "learning_rate": 3.3877919214192646e-06,
-      "loss": 2.2307,
+      "epoch": 1.04,
+      "grad_norm": 117.92691040039062,
+      "learning_rate": 1.3066142325231143e-05,
+      "loss": 2.3713,
       "step": 8288
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 30.34926986694336,
-      "learning_rate": 3.3857873108148743e-06,
-      "loss": 1.1781,
+      "epoch": 1.04,
+      "grad_norm": 8.23699951171875,
+      "learning_rate": 1.3065305610174456e-05,
+      "loss": 0.5104,
       "step": 8289
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 9.271419525146484,
-      "learning_rate": 3.3837827002104843e-06,
-      "loss": 0.9995,
+      "epoch": 1.04,
+      "grad_norm": 15.290740966796875,
+      "learning_rate": 1.3064468895117768e-05,
+      "loss": 0.795,
       "step": 8290
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 16.43708610534668,
-      "learning_rate": 3.3817780896060944e-06,
-      "loss": 1.193,
+      "epoch": 1.04,
+      "grad_norm": 8.10437297821045,
+      "learning_rate": 1.306363218006108e-05,
+      "loss": 0.5317,
       "step": 8291
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 15.72551441192627,
-      "learning_rate": 3.379773479001704e-06,
-      "loss": 1.3098,
+      "epoch": 1.04,
+      "grad_norm": 13.65834903717041,
+      "learning_rate": 1.3062795465004394e-05,
+      "loss": 1.32,
       "step": 8292
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 7.720631122589111,
-      "learning_rate": 3.377768868397314e-06,
-      "loss": 0.8755,
+      "epoch": 1.04,
+      "grad_norm": 38.906044006347656,
+      "learning_rate": 1.3061958749947708e-05,
+      "loss": 0.8079,
       "step": 8293
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 27.38102149963379,
-      "learning_rate": 3.375764257792924e-06,
-      "loss": 1.4355,
+      "epoch": 1.04,
+      "grad_norm": 25.86734962463379,
+      "learning_rate": 1.3061122034891018e-05,
+      "loss": 1.9084,
       "step": 8294
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 12.947731018066406,
-      "learning_rate": 3.3737596471885335e-06,
-      "loss": 1.0127,
+      "epoch": 1.04,
+      "grad_norm": 13.268917083740234,
+      "learning_rate": 1.3060285319834332e-05,
+      "loss": 1.4258,
       "step": 8295
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 18.314491271972656,
-      "learning_rate": 3.371755036584144e-06,
-      "loss": 1.1366,
+      "epoch": 1.04,
+      "grad_norm": 6.482137203216553,
+      "learning_rate": 1.3059448604777645e-05,
+      "loss": 0.7315,
       "step": 8296
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 20.400217056274414,
-      "learning_rate": 3.3697504259797537e-06,
-      "loss": 1.2425,
+      "epoch": 1.04,
+      "grad_norm": 22.378463745117188,
+      "learning_rate": 1.3058611889720956e-05,
+      "loss": 1.3847,
       "step": 8297
     },
     {
-      "epoch": 2.49,
-      "grad_norm": 21.475561141967773,
-      "learning_rate": 3.3677458153753633e-06,
-      "loss": 0.9925,
+      "epoch": 1.04,
+      "grad_norm": 5.98436975479126,
+      "learning_rate": 1.305777517466427e-05,
+      "loss": 1.3013,
       "step": 8298
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 21.971511840820312,
-      "learning_rate": 3.365741204770974e-06,
-      "loss": 1.3472,
+      "epoch": 1.04,
+      "grad_norm": 24.26239776611328,
+      "learning_rate": 1.3056938459607583e-05,
+      "loss": 2.889,
       "step": 8299
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 11.167675018310547,
-      "learning_rate": 3.3637365941665835e-06,
-      "loss": 1.0643,
+      "epoch": 1.04,
+      "grad_norm": 20.896337509155273,
+      "learning_rate": 1.3056101744550895e-05,
+      "loss": 2.7163,
       "step": 8300
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 55.813114166259766,
-      "learning_rate": 3.361731983562193e-06,
-      "loss": 2.2609,
+      "epoch": 1.04,
+      "grad_norm": 7.655635833740234,
+      "learning_rate": 1.3055265029494207e-05,
+      "loss": 0.6938,
       "step": 8301
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 14.630870819091797,
-      "learning_rate": 3.3597273729578032e-06,
-      "loss": 0.6388,
+      "epoch": 1.04,
+      "grad_norm": 8.687627792358398,
+      "learning_rate": 1.3054428314437519e-05,
+      "loss": 1.4532,
       "step": 8302
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 21.16858673095703,
-      "learning_rate": 3.357722762353413e-06,
-      "loss": 1.4621,
+      "epoch": 1.04,
+      "grad_norm": 27.534439086914062,
+      "learning_rate": 1.3053591599380832e-05,
+      "loss": 0.9914,
       "step": 8303
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 25.93169593811035,
-      "learning_rate": 3.355718151749023e-06,
-      "loss": 1.6922,
+      "epoch": 1.04,
+      "grad_norm": 11.624176025390625,
+      "learning_rate": 1.3052754884324144e-05,
+      "loss": 1.4505,
       "step": 8304
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 17.35527229309082,
-      "learning_rate": 3.353713541144633e-06,
-      "loss": 1.1722,
+      "epoch": 1.04,
+      "grad_norm": 35.57938766479492,
+      "learning_rate": 1.3051918169267456e-05,
+      "loss": 0.9582,
       "step": 8305
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 33.93075942993164,
-      "learning_rate": 3.3517089305402427e-06,
-      "loss": 1.9865,
+      "epoch": 1.04,
+      "grad_norm": 17.190563201904297,
+      "learning_rate": 1.305108145421077e-05,
+      "loss": 1.7702,
       "step": 8306
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 39.736629486083984,
-      "learning_rate": 3.3497043199358524e-06,
-      "loss": 2.1434,
+      "epoch": 1.04,
+      "grad_norm": 12.20495891571045,
+      "learning_rate": 1.3050244739154084e-05,
+      "loss": 0.6509,
       "step": 8307
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 16.60782814025879,
-      "learning_rate": 3.347699709331463e-06,
-      "loss": 1.4811,
+      "epoch": 1.04,
+      "grad_norm": 3.5402653217315674,
+      "learning_rate": 1.3049408024097394e-05,
+      "loss": 0.1798,
       "step": 8308
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 23.1959285736084,
-      "learning_rate": 3.3456950987270725e-06,
-      "loss": 1.8844,
+      "epoch": 1.04,
+      "grad_norm": 37.674285888671875,
+      "learning_rate": 1.3048571309040707e-05,
+      "loss": 2.1001,
       "step": 8309
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 15.47692584991455,
-      "learning_rate": 3.343690488122682e-06,
-      "loss": 1.1461,
+      "epoch": 1.04,
+      "grad_norm": 9.620979309082031,
+      "learning_rate": 1.3047734593984021e-05,
+      "loss": 1.6433,
       "step": 8310
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 10.405765533447266,
-      "learning_rate": 3.3416858775182927e-06,
-      "loss": 1.0609,
+      "epoch": 1.04,
+      "grad_norm": 8.073790550231934,
+      "learning_rate": 1.3046897878927331e-05,
+      "loss": 0.5259,
       "step": 8311
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 7.443643569946289,
-      "learning_rate": 3.3396812669139024e-06,
-      "loss": 0.5891,
+      "epoch": 1.04,
+      "grad_norm": 17.23143768310547,
+      "learning_rate": 1.3046061163870645e-05,
+      "loss": 0.92,
       "step": 8312
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 40.757137298583984,
-      "learning_rate": 3.337676656309512e-06,
-      "loss": 1.1556,
+      "epoch": 1.04,
+      "grad_norm": 15.197664260864258,
+      "learning_rate": 1.3045224448813959e-05,
+      "loss": 0.8573,
       "step": 8313
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 23.71717071533203,
-      "learning_rate": 3.335672045705122e-06,
-      "loss": 1.3438,
+      "epoch": 1.04,
+      "grad_norm": 40.136722564697266,
+      "learning_rate": 1.304438773375727e-05,
+      "loss": 1.027,
       "step": 8314
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 9.829071998596191,
-      "learning_rate": 3.3336674351007318e-06,
-      "loss": 1.1778,
+      "epoch": 1.04,
+      "grad_norm": 13.674126625061035,
+      "learning_rate": 1.3043551018700583e-05,
+      "loss": 1.138,
       "step": 8315
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 36.342872619628906,
-      "learning_rate": 3.331662824496342e-06,
-      "loss": 1.5935,
+      "epoch": 1.04,
+      "grad_norm": 15.043445587158203,
+      "learning_rate": 1.3042714303643895e-05,
+      "loss": 1.1383,
       "step": 8316
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 34.38883590698242,
-      "learning_rate": 3.329658213891952e-06,
-      "loss": 1.1699,
+      "epoch": 1.04,
+      "grad_norm": 17.719526290893555,
+      "learning_rate": 1.3041877588587208e-05,
+      "loss": 0.6344,
       "step": 8317
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 13.522592544555664,
-      "learning_rate": 3.3276536032875616e-06,
-      "loss": 1.3218,
+      "epoch": 1.04,
+      "grad_norm": 16.272741317749023,
+      "learning_rate": 1.304104087353052e-05,
+      "loss": 0.9304,
       "step": 8318
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 18.872207641601562,
-      "learning_rate": 3.3256489926831713e-06,
-      "loss": 1.0661,
+      "epoch": 1.04,
+      "grad_norm": 13.373222351074219,
+      "learning_rate": 1.3040204158473832e-05,
+      "loss": 1.0971,
       "step": 8319
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 48.556331634521484,
-      "learning_rate": 3.3236443820787818e-06,
-      "loss": 2.3702,
+      "epoch": 1.04,
+      "grad_norm": 15.191892623901367,
+      "learning_rate": 1.3039367443417146e-05,
+      "loss": 2.0868,
       "step": 8320
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 26.391420364379883,
-      "learning_rate": 3.3216397714743914e-06,
-      "loss": 1.0047,
+      "epoch": 1.04,
+      "grad_norm": 8.404037475585938,
+      "learning_rate": 1.303853072836046e-05,
+      "loss": 1.9416,
       "step": 8321
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 10.691487312316895,
-      "learning_rate": 3.319635160870001e-06,
-      "loss": 0.6147,
+      "epoch": 1.04,
+      "grad_norm": 26.0845890045166,
+      "learning_rate": 1.303769401330377e-05,
+      "loss": 2.1982,
       "step": 8322
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 11.963746070861816,
-      "learning_rate": 3.3176305502656116e-06,
-      "loss": 1.4303,
+      "epoch": 1.04,
+      "grad_norm": 24.468116760253906,
+      "learning_rate": 1.3036857298247083e-05,
+      "loss": 0.9575,
       "step": 8323
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 14.355230331420898,
-      "learning_rate": 3.3156259396612213e-06,
-      "loss": 1.6003,
+      "epoch": 1.04,
+      "grad_norm": 6.69379186630249,
+      "learning_rate": 1.3036020583190397e-05,
+      "loss": 0.4925,
       "step": 8324
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 14.952115058898926,
-      "learning_rate": 3.313621329056831e-06,
-      "loss": 1.2936,
+      "epoch": 1.04,
+      "grad_norm": 7.881843090057373,
+      "learning_rate": 1.3035183868133707e-05,
+      "loss": 0.6911,
       "step": 8325
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 11.36012077331543,
-      "learning_rate": 3.311616718452441e-06,
-      "loss": 1.3233,
+      "epoch": 1.04,
+      "grad_norm": 7.593258857727051,
+      "learning_rate": 1.303434715307702e-05,
+      "loss": 1.0592,
       "step": 8326
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 73.88568115234375,
-      "learning_rate": 3.3096121078480507e-06,
-      "loss": 2.0371,
+      "epoch": 1.05,
+      "grad_norm": 45.25659942626953,
+      "learning_rate": 1.3033510438020334e-05,
+      "loss": 1.0704,
       "step": 8327
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 7.392289638519287,
-      "learning_rate": 3.3076074972436608e-06,
-      "loss": 0.817,
+      "epoch": 1.05,
+      "grad_norm": 8.04725170135498,
+      "learning_rate": 1.3032673722963646e-05,
+      "loss": 0.7889,
       "step": 8328
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 26.323123931884766,
-      "learning_rate": 3.305602886639271e-06,
-      "loss": 1.2377,
+      "epoch": 1.05,
+      "grad_norm": 14.932684898376465,
+      "learning_rate": 1.3031837007906958e-05,
+      "loss": 0.8947,
       "step": 8329
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 6.049129486083984,
-      "learning_rate": 3.3035982760348805e-06,
-      "loss": 0.725,
+      "epoch": 1.05,
+      "grad_norm": 21.940521240234375,
+      "learning_rate": 1.303100029285027e-05,
+      "loss": 2.4528,
       "step": 8330
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 18.688451766967773,
-      "learning_rate": 3.30159366543049e-06,
-      "loss": 1.0841,
+      "epoch": 1.05,
+      "grad_norm": 10.622476577758789,
+      "learning_rate": 1.3030163577793584e-05,
+      "loss": 1.2091,
       "step": 8331
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 19.779504776000977,
-      "learning_rate": 3.2995890548261e-06,
-      "loss": 1.6004,
+      "epoch": 1.05,
+      "grad_norm": 5.0412774085998535,
+      "learning_rate": 1.3029326862736896e-05,
+      "loss": 0.4054,
       "step": 8332
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 37.673744201660156,
-      "learning_rate": 3.2975844442217103e-06,
-      "loss": 1.7093,
+      "epoch": 1.05,
+      "grad_norm": 4.07755708694458,
+      "learning_rate": 1.3028490147680208e-05,
+      "loss": 0.287,
       "step": 8333
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 15.679644584655762,
-      "learning_rate": 3.29557983361732e-06,
-      "loss": 1.4718,
+      "epoch": 1.05,
+      "grad_norm": 12.299552917480469,
+      "learning_rate": 1.3027653432623522e-05,
+      "loss": 1.523,
       "step": 8334
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 17.381481170654297,
-      "learning_rate": 3.2935752230129297e-06,
-      "loss": 2.5925,
+      "epoch": 1.05,
+      "grad_norm": 30.969959259033203,
+      "learning_rate": 1.3026816717566835e-05,
+      "loss": 1.1815,
       "step": 8335
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 9.106160163879395,
-      "learning_rate": 3.29157061240854e-06,
-      "loss": 0.8652,
+      "epoch": 1.05,
+      "grad_norm": 10.52627944946289,
+      "learning_rate": 1.3025980002510145e-05,
+      "loss": 0.4909,
       "step": 8336
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 25.941999435424805,
-      "learning_rate": 3.28956600180415e-06,
-      "loss": 0.8842,
+      "epoch": 1.05,
+      "grad_norm": 8.907776832580566,
+      "learning_rate": 1.3025143287453459e-05,
+      "loss": 0.4805,
       "step": 8337
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 9.563976287841797,
-      "learning_rate": 3.2875613911997595e-06,
-      "loss": 1.6659,
+      "epoch": 1.05,
+      "grad_norm": 14.624920845031738,
+      "learning_rate": 1.3024306572396773e-05,
+      "loss": 0.9164,
       "step": 8338
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 46.76715087890625,
-      "learning_rate": 3.2855567805953696e-06,
-      "loss": 1.8769,
+      "epoch": 1.05,
+      "grad_norm": 9.58926010131836,
+      "learning_rate": 1.3023469857340083e-05,
+      "loss": 1.1276,
       "step": 8339
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 20.058713912963867,
-      "learning_rate": 3.2835521699909797e-06,
-      "loss": 1.3938,
+      "epoch": 1.05,
+      "grad_norm": 2.0990917682647705,
+      "learning_rate": 1.3022633142283397e-05,
+      "loss": 0.0459,
       "step": 8340
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 41.1780891418457,
-      "learning_rate": 3.2815475593865893e-06,
-      "loss": 1.9891,
+      "epoch": 1.05,
+      "grad_norm": 13.683082580566406,
+      "learning_rate": 1.3021796427226709e-05,
+      "loss": 1.0213,
       "step": 8341
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 14.344354629516602,
-      "learning_rate": 3.2795429487821994e-06,
-      "loss": 1.4836,
+      "epoch": 1.05,
+      "grad_norm": 5.27555513381958,
+      "learning_rate": 1.3020959712170022e-05,
+      "loss": 1.4565,
       "step": 8342
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 12.271919250488281,
-      "learning_rate": 3.277538338177809e-06,
-      "loss": 1.1799,
+      "epoch": 1.05,
+      "grad_norm": 9.36648178100586,
+      "learning_rate": 1.3020122997113334e-05,
+      "loss": 1.678,
       "step": 8343
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 10.539395332336426,
-      "learning_rate": 3.2755337275734187e-06,
-      "loss": 1.0998,
+      "epoch": 1.05,
+      "grad_norm": 15.057022094726562,
+      "learning_rate": 1.3019286282056646e-05,
+      "loss": 0.7218,
       "step": 8344
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 18.392601013183594,
-      "learning_rate": 3.2735291169690292e-06,
-      "loss": 1.2702,
+      "epoch": 1.05,
+      "grad_norm": 6.689286708831787,
+      "learning_rate": 1.301844956699996e-05,
+      "loss": 0.9842,
       "step": 8345
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 9.823795318603516,
-      "learning_rate": 3.271524506364639e-06,
-      "loss": 1.4675,
+      "epoch": 1.05,
+      "grad_norm": 4.9068708419799805,
+      "learning_rate": 1.3017612851943272e-05,
+      "loss": 0.3076,
       "step": 8346
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 43.700782775878906,
-      "learning_rate": 3.2695198957602486e-06,
-      "loss": 2.5785,
+      "epoch": 1.05,
+      "grad_norm": 22.96696662902832,
+      "learning_rate": 1.3016776136886584e-05,
+      "loss": 1.6388,
       "step": 8347
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 18.977764129638672,
-      "learning_rate": 3.267515285155859e-06,
-      "loss": 2.1341,
+      "epoch": 1.05,
+      "grad_norm": 12.055654525756836,
+      "learning_rate": 1.3015939421829897e-05,
+      "loss": 3.3452,
       "step": 8348
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 11.805031776428223,
-      "learning_rate": 3.2655106745514687e-06,
-      "loss": 1.0303,
+      "epoch": 1.05,
+      "grad_norm": 21.18698501586914,
+      "learning_rate": 1.3015102706773211e-05,
+      "loss": 2.8709,
       "step": 8349
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 35.22747802734375,
-      "learning_rate": 3.2635060639470784e-06,
-      "loss": 1.0836,
+      "epoch": 1.05,
+      "grad_norm": 12.760493278503418,
+      "learning_rate": 1.3014265991716521e-05,
+      "loss": 0.861,
       "step": 8350
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 48.45168685913086,
-      "learning_rate": 3.2615014533426885e-06,
-      "loss": 1.3489,
+      "epoch": 1.05,
+      "grad_norm": 22.128196716308594,
+      "learning_rate": 1.3013429276659835e-05,
+      "loss": 2.5848,
       "step": 8351
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 22.869895935058594,
-      "learning_rate": 3.2594968427382986e-06,
-      "loss": 2.6811,
+      "epoch": 1.05,
+      "grad_norm": 91.83653259277344,
+      "learning_rate": 1.3012592561603149e-05,
+      "loss": 2.6359,
       "step": 8352
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 18.353343963623047,
-      "learning_rate": 3.257492232133908e-06,
-      "loss": 0.976,
+      "epoch": 1.05,
+      "grad_norm": 38.465980529785156,
+      "learning_rate": 1.3011755846546459e-05,
+      "loss": 0.6478,
       "step": 8353
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 29.93402862548828,
-      "learning_rate": 3.2554876215295183e-06,
-      "loss": 1.1663,
+      "epoch": 1.05,
+      "grad_norm": 11.067882537841797,
+      "learning_rate": 1.3010919131489773e-05,
+      "loss": 0.8612,
       "step": 8354
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 13.383641242980957,
-      "learning_rate": 3.253483010925128e-06,
-      "loss": 0.7903,
+      "epoch": 1.05,
+      "grad_norm": 6.735037326812744,
+      "learning_rate": 1.3010082416433084e-05,
+      "loss": 0.5627,
       "step": 8355
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 6.8472747802734375,
-      "learning_rate": 3.2514784003207376e-06,
-      "loss": 1.1989,
+      "epoch": 1.05,
+      "grad_norm": 12.847387313842773,
+      "learning_rate": 1.3009245701376398e-05,
+      "loss": 1.3777,
       "step": 8356
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 32.73374938964844,
-      "learning_rate": 3.249473789716348e-06,
-      "loss": 0.9284,
+      "epoch": 1.05,
+      "grad_norm": 19.365373611450195,
+      "learning_rate": 1.300840898631971e-05,
+      "loss": 1.3549,
       "step": 8357
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 11.246085166931152,
-      "learning_rate": 3.2474691791119578e-06,
-      "loss": 1.5323,
+      "epoch": 1.05,
+      "grad_norm": 43.83978271484375,
+      "learning_rate": 1.3007572271263022e-05,
+      "loss": 2.663,
       "step": 8358
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 8.259552955627441,
-      "learning_rate": 3.2454645685075674e-06,
-      "loss": 0.8713,
+      "epoch": 1.05,
+      "grad_norm": 7.216566562652588,
+      "learning_rate": 1.3006735556206336e-05,
+      "loss": 0.6891,
       "step": 8359
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 28.066495895385742,
-      "learning_rate": 3.243459957903178e-06,
-      "loss": 1.1289,
+      "epoch": 1.05,
+      "grad_norm": 12.197381019592285,
+      "learning_rate": 1.3005898841149646e-05,
+      "loss": 1.2583,
       "step": 8360
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 16.770734786987305,
-      "learning_rate": 3.2414553472987876e-06,
-      "loss": 1.1323,
+      "epoch": 1.05,
+      "grad_norm": 11.418231964111328,
+      "learning_rate": 1.300506212609296e-05,
+      "loss": 0.7901,
       "step": 8361
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 10.900749206542969,
-      "learning_rate": 3.2394507366943973e-06,
-      "loss": 0.7672,
+      "epoch": 1.05,
+      "grad_norm": 28.907398223876953,
+      "learning_rate": 1.3004225411036273e-05,
+      "loss": 1.1503,
       "step": 8362
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 21.28752326965332,
-      "learning_rate": 3.2374461260900074e-06,
-      "loss": 1.4755,
+      "epoch": 1.05,
+      "grad_norm": 7.6004638671875,
+      "learning_rate": 1.3003388695979587e-05,
+      "loss": 2.7435,
       "step": 8363
     },
     {
-      "epoch": 2.51,
-      "grad_norm": 15.661283493041992,
-      "learning_rate": 3.2354415154856174e-06,
-      "loss": 1.3512,
+      "epoch": 1.05,
+      "grad_norm": 6.2184906005859375,
+      "learning_rate": 1.3002551980922897e-05,
+      "loss": 0.4766,
       "step": 8364
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 16.886972427368164,
-      "learning_rate": 3.233436904881227e-06,
-      "loss": 1.4028,
+      "epoch": 1.05,
+      "grad_norm": 20.67194938659668,
+      "learning_rate": 1.300171526586621e-05,
+      "loss": 1.1035,
       "step": 8365
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 17.953754425048828,
-      "learning_rate": 3.231432294276837e-06,
-      "loss": 1.2938,
+      "epoch": 1.05,
+      "grad_norm": 17.56820297241211,
+      "learning_rate": 1.3000878550809524e-05,
+      "loss": 1.4941,
       "step": 8366
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 9.600143432617188,
-      "learning_rate": 3.229427683672447e-06,
-      "loss": 0.7052,
+      "epoch": 1.05,
+      "grad_norm": 9.857599258422852,
+      "learning_rate": 1.3000041835752835e-05,
+      "loss": 1.516,
       "step": 8367
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 19.288686752319336,
-      "learning_rate": 3.2274230730680565e-06,
-      "loss": 1.588,
+      "epoch": 1.05,
+      "grad_norm": 15.678145408630371,
+      "learning_rate": 1.2999205120696148e-05,
+      "loss": 2.4484,
       "step": 8368
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 12.002143859863281,
-      "learning_rate": 3.225418462463667e-06,
-      "loss": 0.5881,
+      "epoch": 1.05,
+      "grad_norm": 26.623945236206055,
+      "learning_rate": 1.299836840563946e-05,
+      "loss": 2.4554,
       "step": 8369
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 18.510684967041016,
-      "learning_rate": 3.2234138518592767e-06,
-      "loss": 1.543,
+      "epoch": 1.05,
+      "grad_norm": 30.240673065185547,
+      "learning_rate": 1.2997531690582774e-05,
+      "loss": 1.1681,
       "step": 8370
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 31.459373474121094,
-      "learning_rate": 3.2214092412548863e-06,
-      "loss": 1.0949,
+      "epoch": 1.05,
+      "grad_norm": 6.327847003936768,
+      "learning_rate": 1.2996694975526086e-05,
+      "loss": 1.967,
       "step": 8371
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 7.364177703857422,
-      "learning_rate": 3.219404630650496e-06,
-      "loss": 0.7282,
+      "epoch": 1.05,
+      "grad_norm": 11.448563575744629,
+      "learning_rate": 1.2995858260469398e-05,
+      "loss": 1.3557,
       "step": 8372
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 10.49648666381836,
-      "learning_rate": 3.2174000200461065e-06,
-      "loss": 0.6402,
+      "epoch": 1.05,
+      "grad_norm": 10.462690353393555,
+      "learning_rate": 1.2995021545412712e-05,
+      "loss": 0.842,
       "step": 8373
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 25.187376022338867,
-      "learning_rate": 3.215395409441716e-06,
-      "loss": 1.4291,
+      "epoch": 1.05,
+      "grad_norm": 6.635849952697754,
+      "learning_rate": 1.2994184830356022e-05,
+      "loss": 2.6515,
       "step": 8374
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 26.098617553710938,
-      "learning_rate": 3.213390798837326e-06,
-      "loss": 1.861,
+      "epoch": 1.05,
+      "grad_norm": 10.515811920166016,
+      "learning_rate": 1.2993348115299335e-05,
+      "loss": 1.0804,
       "step": 8375
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 46.278839111328125,
-      "learning_rate": 3.2113861882329363e-06,
-      "loss": 1.8303,
+      "epoch": 1.05,
+      "grad_norm": 10.180510520935059,
+      "learning_rate": 1.2992511400242649e-05,
+      "loss": 0.8655,
       "step": 8376
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 12.867042541503906,
-      "learning_rate": 3.209381577628546e-06,
-      "loss": 1.0696,
+      "epoch": 1.05,
+      "grad_norm": 12.36981201171875,
+      "learning_rate": 1.2991674685185963e-05,
+      "loss": 0.8491,
       "step": 8377
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 20.583019256591797,
-      "learning_rate": 3.2073769670241557e-06,
-      "loss": 0.992,
+      "epoch": 1.05,
+      "grad_norm": 23.697532653808594,
+      "learning_rate": 1.2990837970129273e-05,
+      "loss": 1.1596,
       "step": 8378
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 56.57286071777344,
-      "learning_rate": 3.2053723564197657e-06,
-      "loss": 1.8542,
+      "epoch": 1.05,
+      "grad_norm": 10.437012672424316,
+      "learning_rate": 1.2990001255072587e-05,
+      "loss": 1.5892,
       "step": 8379
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 13.195068359375,
-      "learning_rate": 3.2033677458153754e-06,
-      "loss": 1.3004,
+      "epoch": 1.05,
+      "grad_norm": 40.67985534667969,
+      "learning_rate": 1.29891645400159e-05,
+      "loss": 0.9195,
       "step": 8380
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 16.263944625854492,
-      "learning_rate": 3.2013631352109855e-06,
-      "loss": 1.7439,
+      "epoch": 1.05,
+      "grad_norm": 22.098230361938477,
+      "learning_rate": 1.298832782495921e-05,
+      "loss": 1.1321,
       "step": 8381
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 11.452669143676758,
-      "learning_rate": 3.1993585246065956e-06,
-      "loss": 0.7889,
+      "epoch": 1.05,
+      "grad_norm": 8.74831771850586,
+      "learning_rate": 1.2987491109902524e-05,
+      "loss": 0.8672,
       "step": 8382
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 17.5844669342041,
-      "learning_rate": 3.1973539140022052e-06,
-      "loss": 1.2866,
+      "epoch": 1.05,
+      "grad_norm": 10.82736587524414,
+      "learning_rate": 1.2986654394845836e-05,
+      "loss": 0.5641,
       "step": 8383
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 12.583413124084473,
-      "learning_rate": 3.195349303397815e-06,
-      "loss": 0.9804,
+      "epoch": 1.05,
+      "grad_norm": 11.020607948303223,
+      "learning_rate": 1.298581767978915e-05,
+      "loss": 1.3105,
       "step": 8384
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 13.724156379699707,
-      "learning_rate": 3.1933446927934254e-06,
-      "loss": 0.7936,
+      "epoch": 1.05,
+      "grad_norm": 15.373324394226074,
+      "learning_rate": 1.2984980964732462e-05,
+      "loss": 2.5396,
       "step": 8385
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 11.498400688171387,
-      "learning_rate": 3.191340082189035e-06,
-      "loss": 1.1136,
+      "epoch": 1.05,
+      "grad_norm": 5.774839878082275,
+      "learning_rate": 1.2984144249675774e-05,
+      "loss": 0.4541,
       "step": 8386
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 41.235206604003906,
-      "learning_rate": 3.1893354715846447e-06,
-      "loss": 1.1562,
+      "epoch": 1.05,
+      "grad_norm": 9.357996940612793,
+      "learning_rate": 1.2983307534619087e-05,
+      "loss": 1.65,
       "step": 8387
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 26.88475799560547,
-      "learning_rate": 3.187330860980255e-06,
-      "loss": 2.1271,
+      "epoch": 1.05,
+      "grad_norm": 27.872713088989258,
+      "learning_rate": 1.2982470819562398e-05,
+      "loss": 3.2732,
       "step": 8388
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 16.73887825012207,
-      "learning_rate": 3.185326250375865e-06,
-      "loss": 1.5196,
+      "epoch": 1.05,
+      "grad_norm": 3.760759115219116,
+      "learning_rate": 1.2981634104505711e-05,
+      "loss": 0.2349,
       "step": 8389
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 20.1945858001709,
-      "learning_rate": 3.1833216397714746e-06,
-      "loss": 0.9805,
+      "epoch": 1.05,
+      "grad_norm": 15.883977890014648,
+      "learning_rate": 1.2980797389449025e-05,
+      "loss": 1.1503,
       "step": 8390
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 34.976627349853516,
-      "learning_rate": 3.1813170291670846e-06,
-      "loss": 1.1922,
+      "epoch": 1.05,
+      "grad_norm": 4.68303918838501,
+      "learning_rate": 1.2979960674392335e-05,
+      "loss": 0.3075,
       "step": 8391
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 36.179351806640625,
-      "learning_rate": 3.1793124185626943e-06,
-      "loss": 4.2282,
+      "epoch": 1.05,
+      "grad_norm": 23.391752243041992,
+      "learning_rate": 1.2979123959335649e-05,
+      "loss": 1.0449,
       "step": 8392
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 9.375896453857422,
-      "learning_rate": 3.1773078079583044e-06,
-      "loss": 0.9379,
+      "epoch": 1.05,
+      "grad_norm": 20.98833465576172,
+      "learning_rate": 1.2978287244278962e-05,
+      "loss": 1.4492,
       "step": 8393
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 132.8348846435547,
-      "learning_rate": 3.1753031973539145e-06,
-      "loss": 3.0371,
+      "epoch": 1.05,
+      "grad_norm": 14.180381774902344,
+      "learning_rate": 1.2977450529222274e-05,
+      "loss": 1.2787,
       "step": 8394
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 30.87514877319336,
-      "learning_rate": 3.173298586749524e-06,
-      "loss": 2.5639,
+      "epoch": 1.05,
+      "grad_norm": 13.713095664978027,
+      "learning_rate": 1.2976613814165586e-05,
+      "loss": 0.6493,
       "step": 8395
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 12.229470252990723,
-      "learning_rate": 3.171293976145134e-06,
-      "loss": 0.9845,
+      "epoch": 1.05,
+      "grad_norm": 35.20022201538086,
+      "learning_rate": 1.29757770991089e-05,
+      "loss": 2.4457,
       "step": 8396
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 26.513559341430664,
-      "learning_rate": 3.1692893655407443e-06,
-      "loss": 1.1088,
+      "epoch": 1.05,
+      "grad_norm": 13.833316802978516,
+      "learning_rate": 1.2974940384052212e-05,
+      "loss": 1.1561,
       "step": 8397
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 7.9161553382873535,
-      "learning_rate": 3.167284754936354e-06,
-      "loss": 0.8013,
+      "epoch": 1.05,
+      "grad_norm": 75.72393035888672,
+      "learning_rate": 1.2974103668995524e-05,
+      "loss": 2.1973,
       "step": 8398
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 12.535621643066406,
-      "learning_rate": 3.1652801443319636e-06,
-      "loss": 1.0497,
+      "epoch": 1.05,
+      "grad_norm": 37.64180374145508,
+      "learning_rate": 1.2973266953938836e-05,
+      "loss": 1.6479,
       "step": 8399
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 9.409601211547852,
-      "learning_rate": 3.1632755337275737e-06,
-      "loss": 1.132,
+      "epoch": 1.05,
+      "grad_norm": 10.788687705993652,
+      "learning_rate": 1.297243023888215e-05,
+      "loss": 0.7525,
       "step": 8400
     },
     {
-      "epoch": 2.53,
-      "eval_loss": 0.17680850625038147,
-      "eval_runtime": 44.0695,
-      "eval_samples_per_second": 33.561,
-      "eval_steps_per_second": 33.561,
+      "epoch": 1.05,
+      "eval_loss": 0.09782509505748749,
+      "eval_runtime": 95.178,
+      "eval_samples_per_second": 37.214,
+      "eval_steps_per_second": 37.214,
       "step": 8400
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 38.185997009277344,
-      "learning_rate": 3.1612709231231838e-06,
-      "loss": 1.8298,
+      "epoch": 1.05,
+      "grad_norm": 2.3076012134552,
+      "learning_rate": 1.2971593523825463e-05,
+      "loss": 0.0305,
       "step": 8401
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 103.94436645507812,
-      "learning_rate": 3.1592663125187934e-06,
-      "loss": 0.8306,
+      "epoch": 1.05,
+      "grad_norm": 17.84105682373047,
+      "learning_rate": 1.2970756808768773e-05,
+      "loss": 2.1674,
       "step": 8402
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 105.07266998291016,
-      "learning_rate": 3.1572617019144035e-06,
-      "loss": 1.827,
+      "epoch": 1.05,
+      "grad_norm": 14.507989883422852,
+      "learning_rate": 1.2969920093712087e-05,
+      "loss": 1.7957,
       "step": 8403
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 19.526813507080078,
-      "learning_rate": 3.155257091310013e-06,
-      "loss": 1.4366,
+      "epoch": 1.05,
+      "grad_norm": 34.563480377197266,
+      "learning_rate": 1.29690833786554e-05,
+      "loss": 2.2601,
       "step": 8404
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 10.135507583618164,
-      "learning_rate": 3.153252480705623e-06,
-      "loss": 0.6992,
+      "epoch": 1.05,
+      "grad_norm": 16.744068145751953,
+      "learning_rate": 1.2968246663598711e-05,
+      "loss": 2.0247,
       "step": 8405
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 35.69138717651367,
-      "learning_rate": 3.1512478701012334e-06,
-      "loss": 1.5329,
+      "epoch": 1.05,
+      "grad_norm": 14.041572570800781,
+      "learning_rate": 1.2967409948542025e-05,
+      "loss": 1.5232,
       "step": 8406
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 74.143798828125,
-      "learning_rate": 3.149243259496843e-06,
-      "loss": 0.9668,
+      "epoch": 1.06,
+      "grad_norm": 19.138408660888672,
+      "learning_rate": 1.2966573233485338e-05,
+      "loss": 0.5881,
       "step": 8407
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 15.071893692016602,
-      "learning_rate": 3.1472386488924527e-06,
-      "loss": 1.1024,
+      "epoch": 1.06,
+      "grad_norm": 44.88977813720703,
+      "learning_rate": 1.296573651842865e-05,
+      "loss": 1.8347,
       "step": 8408
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 11.287890434265137,
-      "learning_rate": 3.145234038288063e-06,
-      "loss": 1.2887,
+      "epoch": 1.06,
+      "grad_norm": 22.761062622070312,
+      "learning_rate": 1.2964899803371962e-05,
+      "loss": 1.487,
       "step": 8409
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 63.98174285888672,
-      "learning_rate": 3.143229427683673e-06,
-      "loss": 1.8663,
+      "epoch": 1.06,
+      "grad_norm": 8.038599014282227,
+      "learning_rate": 1.2964063088315276e-05,
+      "loss": 0.7646,
       "step": 8410
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 15.564671516418457,
-      "learning_rate": 3.1412248170792825e-06,
-      "loss": 1.1972,
+      "epoch": 1.06,
+      "grad_norm": 100.69731903076172,
+      "learning_rate": 1.2963226373258588e-05,
+      "loss": 1.1985,
       "step": 8411
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 26.284500122070312,
-      "learning_rate": 3.1392202064748926e-06,
-      "loss": 1.1945,
+      "epoch": 1.06,
+      "grad_norm": 22.61496925354004,
+      "learning_rate": 1.29623896582019e-05,
+      "loss": 2.8117,
       "step": 8412
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 81.35211181640625,
-      "learning_rate": 3.1372155958705027e-06,
-      "loss": 1.7933,
+      "epoch": 1.06,
+      "grad_norm": 4.329559326171875,
+      "learning_rate": 1.2961552943145212e-05,
+      "loss": 0.4211,
       "step": 8413
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 20.28238296508789,
-      "learning_rate": 3.1352109852661123e-06,
-      "loss": 1.6545,
+      "epoch": 1.06,
+      "grad_norm": 19.117862701416016,
+      "learning_rate": 1.2960716228088525e-05,
+      "loss": 2.1339,
       "step": 8414
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 11.033011436462402,
-      "learning_rate": 3.133206374661722e-06,
-      "loss": 0.8211,
+      "epoch": 1.06,
+      "grad_norm": 11.0752534866333,
+      "learning_rate": 1.2959879513031839e-05,
+      "loss": 1.2909,
       "step": 8415
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 17.557846069335938,
-      "learning_rate": 3.131201764057332e-06,
-      "loss": 1.2009,
+      "epoch": 1.06,
+      "grad_norm": 12.766861915588379,
+      "learning_rate": 1.295904279797515e-05,
+      "loss": 1.6993,
       "step": 8416
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 17.877410888671875,
-      "learning_rate": 3.1291971534529417e-06,
-      "loss": 1.1617,
+      "epoch": 1.06,
+      "grad_norm": 15.978042602539062,
+      "learning_rate": 1.2958206082918463e-05,
+      "loss": 0.2342,
       "step": 8417
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 17.008520126342773,
-      "learning_rate": 3.127192542848552e-06,
-      "loss": 1.2213,
+      "epoch": 1.06,
+      "grad_norm": 8.060977935791016,
+      "learning_rate": 1.2957369367861777e-05,
+      "loss": 0.2633,
       "step": 8418
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 13.909770965576172,
-      "learning_rate": 3.125187932244162e-06,
-      "loss": 0.9544,
+      "epoch": 1.06,
+      "grad_norm": 5.748692035675049,
+      "learning_rate": 1.2956532652805087e-05,
+      "loss": 1.5113,
       "step": 8419
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 12.314075469970703,
-      "learning_rate": 3.1231833216397716e-06,
-      "loss": 1.1267,
+      "epoch": 1.06,
+      "grad_norm": 194.9235076904297,
+      "learning_rate": 1.29556959377484e-05,
+      "loss": 1.4548,
       "step": 8420
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 8.14324951171875,
-      "learning_rate": 3.1211787110353812e-06,
-      "loss": 1.052,
+      "epoch": 1.06,
+      "grad_norm": 26.03392219543457,
+      "learning_rate": 1.2954859222691714e-05,
+      "loss": 2.4875,
       "step": 8421
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 9.060734748840332,
-      "learning_rate": 3.1191741004309917e-06,
-      "loss": 0.7107,
+      "epoch": 1.06,
+      "grad_norm": 13.564663887023926,
+      "learning_rate": 1.2954022507635026e-05,
+      "loss": 1.1039,
       "step": 8422
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 12.2525634765625,
-      "learning_rate": 3.1171694898266014e-06,
-      "loss": 0.9515,
+      "epoch": 1.06,
+      "grad_norm": 13.226625442504883,
+      "learning_rate": 1.2953185792578338e-05,
+      "loss": 2.0588,
       "step": 8423
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 18.487070083618164,
-      "learning_rate": 3.115164879222211e-06,
-      "loss": 0.9164,
+      "epoch": 1.06,
+      "grad_norm": 17.731657028198242,
+      "learning_rate": 1.2952349077521652e-05,
+      "loss": 2.594,
       "step": 8424
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 21.11574363708496,
-      "learning_rate": 3.1131602686178216e-06,
-      "loss": 1.1415,
+      "epoch": 1.06,
+      "grad_norm": 32.340492248535156,
+      "learning_rate": 1.2951512362464964e-05,
+      "loss": 2.1051,
       "step": 8425
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 10.710432052612305,
-      "learning_rate": 3.1111556580134312e-06,
-      "loss": 0.6223,
+      "epoch": 1.06,
+      "grad_norm": 4.714109897613525,
+      "learning_rate": 1.2950675647408276e-05,
+      "loss": 0.3578,
       "step": 8426
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 12.434354782104492,
-      "learning_rate": 3.109151047409041e-06,
-      "loss": 0.7547,
+      "epoch": 1.06,
+      "grad_norm": 19.667007446289062,
+      "learning_rate": 1.2949838932351588e-05,
+      "loss": 1.5086,
       "step": 8427
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 5.420670032501221,
-      "learning_rate": 3.107146436804651e-06,
-      "loss": 0.913,
+      "epoch": 1.06,
+      "grad_norm": 22.923402786254883,
+      "learning_rate": 1.2949002217294901e-05,
+      "loss": 2.1589,
       "step": 8428
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 23.501205444335938,
-      "learning_rate": 3.1051418262002606e-06,
-      "loss": 1.7807,
+      "epoch": 1.06,
+      "grad_norm": 9.848310470581055,
+      "learning_rate": 1.2948165502238215e-05,
+      "loss": 0.8451,
       "step": 8429
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 17.01691436767578,
-      "learning_rate": 3.1031372155958707e-06,
-      "loss": 1.2807,
+      "epoch": 1.06,
+      "grad_norm": 16.107831954956055,
+      "learning_rate": 1.2947328787181525e-05,
+      "loss": 0.7639,
       "step": 8430
     },
     {
-      "epoch": 2.53,
-      "grad_norm": 15.571081161499023,
-      "learning_rate": 3.101132604991481e-06,
-      "loss": 1.397,
+      "epoch": 1.06,
+      "grad_norm": 33.23837661743164,
+      "learning_rate": 1.2946492072124839e-05,
+      "loss": 2.7027,
       "step": 8431
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 32.079612731933594,
-      "learning_rate": 3.0991279943870905e-06,
-      "loss": 1.4421,
+      "epoch": 1.06,
+      "grad_norm": 15.720340728759766,
+      "learning_rate": 1.2945655357068152e-05,
+      "loss": 1.4081,
       "step": 8432
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 13.537861824035645,
-      "learning_rate": 3.0971233837827e-06,
-      "loss": 1.4324,
+      "epoch": 1.06,
+      "grad_norm": 19.901630401611328,
+      "learning_rate": 1.2944818642011463e-05,
+      "loss": 1.5531,
       "step": 8433
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 72.99842834472656,
-      "learning_rate": 3.0951187731783106e-06,
-      "loss": 2.2397,
+      "epoch": 1.06,
+      "grad_norm": 7.82213830947876,
+      "learning_rate": 1.2943981926954776e-05,
+      "loss": 1.1551,
       "step": 8434
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 32.24034118652344,
-      "learning_rate": 3.0931141625739203e-06,
-      "loss": 1.8425,
+      "epoch": 1.06,
+      "grad_norm": 6.144089698791504,
+      "learning_rate": 1.294314521189809e-05,
+      "loss": 0.5246,
       "step": 8435
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 14.41550350189209,
-      "learning_rate": 3.09110955196953e-06,
-      "loss": 1.0483,
+      "epoch": 1.06,
+      "grad_norm": 8.631068229675293,
+      "learning_rate": 1.2942308496841402e-05,
+      "loss": 1.9636,
       "step": 8436
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 47.05146789550781,
-      "learning_rate": 3.0891049413651405e-06,
-      "loss": 1.3004,
+      "epoch": 1.06,
+      "grad_norm": 14.672770500183105,
+      "learning_rate": 1.2941471781784714e-05,
+      "loss": 1.9898,
       "step": 8437
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 25.591800689697266,
-      "learning_rate": 3.08710033076075e-06,
-      "loss": 1.1771,
+      "epoch": 1.06,
+      "grad_norm": 19.22916030883789,
+      "learning_rate": 1.2940635066728028e-05,
+      "loss": 2.5212,
       "step": 8438
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 9.942156791687012,
-      "learning_rate": 3.08509572015636e-06,
-      "loss": 0.5432,
+      "epoch": 1.06,
+      "grad_norm": 12.619750022888184,
+      "learning_rate": 1.293979835167134e-05,
+      "loss": 1.0048,
       "step": 8439
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 22.738941192626953,
-      "learning_rate": 3.08309110955197e-06,
-      "loss": 2.3535,
+      "epoch": 1.06,
+      "grad_norm": 10.999564170837402,
+      "learning_rate": 1.2938961636614651e-05,
+      "loss": 2.1711,
       "step": 8440
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 14.746498107910156,
-      "learning_rate": 3.0810864989475795e-06,
-      "loss": 1.2366,
+      "epoch": 1.06,
+      "grad_norm": 6.046539306640625,
+      "learning_rate": 1.2938124921557963e-05,
+      "loss": 0.4652,
       "step": 8441
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 98.85680389404297,
-      "learning_rate": 3.0790818883431896e-06,
-      "loss": 2.4007,
+      "epoch": 1.06,
+      "grad_norm": 31.383182525634766,
+      "learning_rate": 1.2937288206501277e-05,
+      "loss": 0.6297,
       "step": 8442
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 23.468538284301758,
-      "learning_rate": 3.0770772777387997e-06,
-      "loss": 1.0986,
+      "epoch": 1.06,
+      "grad_norm": 13.119832992553711,
+      "learning_rate": 1.293645149144459e-05,
+      "loss": 2.0921,
       "step": 8443
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 8.700865745544434,
-      "learning_rate": 3.0750726671344094e-06,
-      "loss": 1.1621,
+      "epoch": 1.06,
+      "grad_norm": 6.461184978485107,
+      "learning_rate": 1.2935614776387901e-05,
+      "loss": 0.593,
       "step": 8444
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 22.677888870239258,
-      "learning_rate": 3.073068056530019e-06,
-      "loss": 1.2522,
+      "epoch": 1.06,
+      "grad_norm": 14.403892517089844,
+      "learning_rate": 1.2934778061331215e-05,
+      "loss": 1.0602,
       "step": 8445
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 21.384296417236328,
-      "learning_rate": 3.0710634459256295e-06,
-      "loss": 1.6951,
+      "epoch": 1.06,
+      "grad_norm": 24.468730926513672,
+      "learning_rate": 1.2933941346274528e-05,
+      "loss": 1.0107,
       "step": 8446
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 30.282161712646484,
-      "learning_rate": 3.069058835321239e-06,
-      "loss": 1.8486,
+      "epoch": 1.06,
+      "grad_norm": 60.255062103271484,
+      "learning_rate": 1.2933104631217839e-05,
+      "loss": 3.9681,
       "step": 8447
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 11.835555076599121,
-      "learning_rate": 3.067054224716849e-06,
-      "loss": 1.0966,
+      "epoch": 1.06,
+      "grad_norm": 20.134521484375,
+      "learning_rate": 1.2932267916161152e-05,
+      "loss": 0.6754,
       "step": 8448
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 16.40022850036621,
-      "learning_rate": 3.0650496141124594e-06,
-      "loss": 1.55,
+      "epoch": 1.06,
+      "grad_norm": 19.462905883789062,
+      "learning_rate": 1.2931431201104466e-05,
+      "loss": 1.1152,
       "step": 8449
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 34.12657928466797,
-      "learning_rate": 3.063045003508069e-06,
-      "loss": 1.6477,
+      "epoch": 1.06,
+      "grad_norm": 9.095423698425293,
+      "learning_rate": 1.2930594486047778e-05,
+      "loss": 2.4232,
       "step": 8450
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 43.224571228027344,
-      "learning_rate": 3.0610403929036787e-06,
-      "loss": 1.4323,
+      "epoch": 1.06,
+      "grad_norm": 26.529722213745117,
+      "learning_rate": 1.292975777099109e-05,
+      "loss": 1.9776,
       "step": 8451
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 21.87476921081543,
-      "learning_rate": 3.0590357822992888e-06,
-      "loss": 1.117,
+      "epoch": 1.06,
+      "grad_norm": 9.931761741638184,
+      "learning_rate": 1.2928921055934402e-05,
+      "loss": 1.3352,
       "step": 8452
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 12.986775398254395,
-      "learning_rate": 3.0570311716948984e-06,
-      "loss": 0.865,
+      "epoch": 1.06,
+      "grad_norm": 11.543974876403809,
+      "learning_rate": 1.2928084340877715e-05,
+      "loss": 1.5399,
       "step": 8453
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 12.08977222442627,
-      "learning_rate": 3.0550265610905085e-06,
-      "loss": 0.7838,
+      "epoch": 1.06,
+      "grad_norm": 5.663609504699707,
+      "learning_rate": 1.2927247625821027e-05,
+      "loss": 0.2769,
       "step": 8454
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 50.07052230834961,
-      "learning_rate": 3.053021950486118e-06,
-      "loss": 1.4847,
+      "epoch": 1.06,
+      "grad_norm": 17.997838973999023,
+      "learning_rate": 1.292641091076434e-05,
+      "loss": 1.8294,
       "step": 8455
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 7.173743724822998,
-      "learning_rate": 3.0510173398817283e-06,
-      "loss": 0.6167,
+      "epoch": 1.06,
+      "grad_norm": 5.947362899780273,
+      "learning_rate": 1.2925574195707653e-05,
+      "loss": 1.9695,
       "step": 8456
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 38.1131591796875,
-      "learning_rate": 3.049012729277338e-06,
-      "loss": 2.517,
+      "epoch": 1.06,
+      "grad_norm": 32.4327392578125,
+      "learning_rate": 1.2924737480650967e-05,
+      "loss": 2.0827,
       "step": 8457
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 9.692411422729492,
-      "learning_rate": 3.0470081186729476e-06,
-      "loss": 1.0573,
+      "epoch": 1.06,
+      "grad_norm": 10.101252555847168,
+      "learning_rate": 1.2923900765594277e-05,
+      "loss": 1.8666,
       "step": 8458
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 41.6327018737793,
-      "learning_rate": 3.045003508068558e-06,
-      "loss": 1.4349,
+      "epoch": 1.06,
+      "grad_norm": 22.610300064086914,
+      "learning_rate": 1.292306405053759e-05,
+      "loss": 1.9778,
       "step": 8459
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 29.187610626220703,
-      "learning_rate": 3.0429988974641677e-06,
-      "loss": 1.5777,
+      "epoch": 1.06,
+      "grad_norm": 3.201634168624878,
+      "learning_rate": 1.2922227335480904e-05,
+      "loss": 0.132,
       "step": 8460
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 29.35192108154297,
-      "learning_rate": 3.0409942868597774e-06,
-      "loss": 1.8016,
+      "epoch": 1.06,
+      "grad_norm": 6.934627056121826,
+      "learning_rate": 1.2921390620424214e-05,
+      "loss": 1.0572,
       "step": 8461
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 20.445228576660156,
-      "learning_rate": 3.038989676255388e-06,
-      "loss": 0.7974,
+      "epoch": 1.06,
+      "grad_norm": 23.772220611572266,
+      "learning_rate": 1.2920553905367528e-05,
+      "loss": 1.8739,
       "step": 8462
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 18.133628845214844,
-      "learning_rate": 3.0369850656509976e-06,
-      "loss": 1.1352,
+      "epoch": 1.06,
+      "grad_norm": 11.809895515441895,
+      "learning_rate": 1.2919717190310842e-05,
+      "loss": 1.0289,
       "step": 8463
     },
     {
-      "epoch": 2.54,
-      "grad_norm": 16.889158248901367,
-      "learning_rate": 3.0349804550466072e-06,
-      "loss": 1.4672,
+      "epoch": 1.06,
+      "grad_norm": 27.51059341430664,
+      "learning_rate": 1.2918880475254154e-05,
+      "loss": 2.3416,
       "step": 8464
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 20.985919952392578,
-      "learning_rate": 3.0329758444422173e-06,
-      "loss": 1.192,
+      "epoch": 1.06,
+      "grad_norm": 56.7679328918457,
+      "learning_rate": 1.2918043760197466e-05,
+      "loss": 1.933,
       "step": 8465
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 9.734439849853516,
-      "learning_rate": 3.0309712338378274e-06,
-      "loss": 0.7422,
+      "epoch": 1.06,
+      "grad_norm": 25.842777252197266,
+      "learning_rate": 1.2917207045140778e-05,
+      "loss": 1.3563,
       "step": 8466
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 23.414813995361328,
-      "learning_rate": 3.028966623233437e-06,
-      "loss": 1.8114,
+      "epoch": 1.06,
+      "grad_norm": 16.23643684387207,
+      "learning_rate": 1.2916370330084091e-05,
+      "loss": 1.3781,
       "step": 8467
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 12.681184768676758,
-      "learning_rate": 3.026962012629047e-06,
-      "loss": 1.1328,
+      "epoch": 1.06,
+      "grad_norm": 8.489397048950195,
+      "learning_rate": 1.2915533615027403e-05,
+      "loss": 0.4774,
       "step": 8468
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 23.120763778686523,
-      "learning_rate": 3.024957402024657e-06,
-      "loss": 1.1875,
+      "epoch": 1.06,
+      "grad_norm": 98.74266052246094,
+      "learning_rate": 1.2914696899970715e-05,
+      "loss": 3.1134,
       "step": 8469
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 19.008502960205078,
-      "learning_rate": 3.0229527914202665e-06,
-      "loss": 1.2166,
+      "epoch": 1.06,
+      "grad_norm": 16.9528865814209,
+      "learning_rate": 1.2913860184914029e-05,
+      "loss": 1.0636,
       "step": 8470
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 22.20121192932129,
-      "learning_rate": 3.020948180815877e-06,
-      "loss": 1.4729,
+      "epoch": 1.06,
+      "grad_norm": 11.784737586975098,
+      "learning_rate": 1.2913023469857342e-05,
+      "loss": 0.9287,
       "step": 8471
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 16.53184700012207,
-      "learning_rate": 3.0189435702114866e-06,
-      "loss": 1.8034,
+      "epoch": 1.06,
+      "grad_norm": 31.83979034423828,
+      "learning_rate": 1.2912186754800653e-05,
+      "loss": 1.2546,
       "step": 8472
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 21.981178283691406,
-      "learning_rate": 3.0169389596070963e-06,
-      "loss": 2.2006,
+      "epoch": 1.06,
+      "grad_norm": 20.695783615112305,
+      "learning_rate": 1.2911350039743966e-05,
+      "loss": 0.9959,
       "step": 8473
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 36.195106506347656,
-      "learning_rate": 3.014934349002707e-06,
-      "loss": 2.1031,
+      "epoch": 1.06,
+      "grad_norm": 32.69658279418945,
+      "learning_rate": 1.291051332468728e-05,
+      "loss": 2.0376,
       "step": 8474
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 33.1042366027832,
-      "learning_rate": 3.0129297383983165e-06,
-      "loss": 1.5522,
+      "epoch": 1.06,
+      "grad_norm": 42.36573028564453,
+      "learning_rate": 1.290967660963059e-05,
+      "loss": 1.5448,
       "step": 8475
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 14.721949577331543,
-      "learning_rate": 3.010925127793926e-06,
-      "loss": 1.5655,
+      "epoch": 1.06,
+      "grad_norm": 17.114213943481445,
+      "learning_rate": 1.2908839894573904e-05,
+      "loss": 1.0816,
       "step": 8476
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 14.444605827331543,
-      "learning_rate": 3.0089205171895362e-06,
-      "loss": 1.2094,
+      "epoch": 1.06,
+      "grad_norm": 7.809509754180908,
+      "learning_rate": 1.2908003179517217e-05,
+      "loss": 1.6434,
       "step": 8477
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 14.284456253051758,
-      "learning_rate": 3.0069159065851463e-06,
-      "loss": 0.986,
+      "epoch": 1.06,
+      "grad_norm": 10.35164737701416,
+      "learning_rate": 1.290716646446053e-05,
+      "loss": 0.9084,
       "step": 8478
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 12.816495895385742,
-      "learning_rate": 3.004911295980756e-06,
-      "loss": 1.0418,
+      "epoch": 1.06,
+      "grad_norm": 12.399913787841797,
+      "learning_rate": 1.2906329749403841e-05,
+      "loss": 3.8119,
       "step": 8479
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 67.11431884765625,
-      "learning_rate": 3.002906685376366e-06,
-      "loss": 1.5002,
+      "epoch": 1.06,
+      "grad_norm": 23.38853645324707,
+      "learning_rate": 1.2905493034347153e-05,
+      "loss": 1.6561,
       "step": 8480
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 24.34355926513672,
-      "learning_rate": 3.0009020747719757e-06,
-      "loss": 1.329,
+      "epoch": 1.06,
+      "grad_norm": 8.756095886230469,
+      "learning_rate": 1.2904656319290467e-05,
+      "loss": 1.7878,
       "step": 8481
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 11.59719467163086,
-      "learning_rate": 2.9988974641675854e-06,
-      "loss": 1.4363,
+      "epoch": 1.06,
+      "grad_norm": 14.100663185119629,
+      "learning_rate": 1.2903819604233779e-05,
+      "loss": 0.8435,
       "step": 8482
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 32.02231216430664,
-      "learning_rate": 2.996892853563196e-06,
-      "loss": 1.2263,
+      "epoch": 1.06,
+      "grad_norm": 7.758046627044678,
+      "learning_rate": 1.2902982889177091e-05,
+      "loss": 1.9517,
       "step": 8483
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 9.977734565734863,
-      "learning_rate": 2.9948882429588055e-06,
-      "loss": 0.6075,
+      "epoch": 1.06,
+      "grad_norm": 13.198431968688965,
+      "learning_rate": 1.2902146174120405e-05,
+      "loss": 2.3742,
       "step": 8484
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 48.880313873291016,
-      "learning_rate": 2.992883632354415e-06,
-      "loss": 1.5793,
+      "epoch": 1.06,
+      "grad_norm": 18.462568283081055,
+      "learning_rate": 1.2901309459063718e-05,
+      "loss": 3.8024,
       "step": 8485
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 7.916026592254639,
-      "learning_rate": 2.9908790217500257e-06,
-      "loss": 0.7116,
+      "epoch": 1.06,
+      "grad_norm": 11.379865646362305,
+      "learning_rate": 1.2900472744007028e-05,
+      "loss": 0.5839,
       "step": 8486
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 29.04728889465332,
-      "learning_rate": 2.9888744111456354e-06,
-      "loss": 1.2325,
+      "epoch": 1.07,
+      "grad_norm": 52.05200958251953,
+      "learning_rate": 1.2899636028950342e-05,
+      "loss": 1.2563,
       "step": 8487
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 13.980238914489746,
-      "learning_rate": 2.986869800541245e-06,
-      "loss": 1.496,
+      "epoch": 1.07,
+      "grad_norm": 6.643377304077148,
+      "learning_rate": 1.2898799313893656e-05,
+      "loss": 1.4281,
       "step": 8488
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 15.269109725952148,
-      "learning_rate": 2.984865189936855e-06,
-      "loss": 1.4313,
+      "epoch": 1.07,
+      "grad_norm": 5.62060022354126,
+      "learning_rate": 1.2897962598836966e-05,
+      "loss": 0.6797,
       "step": 8489
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 20.857440948486328,
-      "learning_rate": 2.9828605793324648e-06,
-      "loss": 1.5363,
+      "epoch": 1.07,
+      "grad_norm": 16.942380905151367,
+      "learning_rate": 1.289712588378028e-05,
+      "loss": 1.6484,
       "step": 8490
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 14.627253532409668,
-      "learning_rate": 2.980855968728075e-06,
-      "loss": 1.3487,
+      "epoch": 1.07,
+      "grad_norm": 20.18743133544922,
+      "learning_rate": 1.2896289168723593e-05,
+      "loss": 2.1858,
       "step": 8491
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 7.229565620422363,
-      "learning_rate": 2.978851358123685e-06,
-      "loss": 0.8008,
+      "epoch": 1.07,
+      "grad_norm": 27.09478759765625,
+      "learning_rate": 1.2895452453666905e-05,
+      "loss": 1.9353,
       "step": 8492
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 29.895063400268555,
-      "learning_rate": 2.9768467475192946e-06,
-      "loss": 1.7014,
+      "epoch": 1.07,
+      "grad_norm": 10.600691795349121,
+      "learning_rate": 1.2894615738610217e-05,
+      "loss": 1.1805,
       "step": 8493
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 26.894834518432617,
-      "learning_rate": 2.9748421369149043e-06,
-      "loss": 1.8372,
+      "epoch": 1.07,
+      "grad_norm": 13.021747589111328,
+      "learning_rate": 1.289377902355353e-05,
+      "loss": 1.6513,
       "step": 8494
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 50.57379150390625,
-      "learning_rate": 2.9728375263105148e-06,
-      "loss": 1.3365,
+      "epoch": 1.07,
+      "grad_norm": 13.426149368286133,
+      "learning_rate": 1.2892942308496843e-05,
+      "loss": 0.9426,
       "step": 8495
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 12.997958183288574,
-      "learning_rate": 2.9708329157061244e-06,
-      "loss": 1.9539,
+      "epoch": 1.07,
+      "grad_norm": 8.105659484863281,
+      "learning_rate": 1.2892105593440155e-05,
+      "loss": 0.8835,
       "step": 8496
     },
     {
-      "epoch": 2.55,
-      "grad_norm": 11.23823356628418,
-      "learning_rate": 2.968828305101734e-06,
-      "loss": 1.2944,
+      "epoch": 1.07,
+      "grad_norm": 11.75523853302002,
+      "learning_rate": 1.2891268878383467e-05,
+      "loss": 0.6368,
       "step": 8497
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 21.072771072387695,
-      "learning_rate": 2.9668236944973438e-06,
-      "loss": 1.4985,
+      "epoch": 1.07,
+      "grad_norm": 15.636470794677734,
+      "learning_rate": 1.289043216332678e-05,
+      "loss": 1.5947,
       "step": 8498
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 61.82595443725586,
-      "learning_rate": 2.9648190838929543e-06,
-      "loss": 2.2333,
+      "epoch": 1.07,
+      "grad_norm": 17.106548309326172,
+      "learning_rate": 1.2889595448270094e-05,
+      "loss": 1.625,
       "step": 8499
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 26.898548126220703,
-      "learning_rate": 2.962814473288564e-06,
-      "loss": 1.2997,
+      "epoch": 1.07,
+      "grad_norm": 17.792394638061523,
+      "learning_rate": 1.2888758733213404e-05,
+      "loss": 2.9065,
       "step": 8500
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 14.761646270751953,
-      "learning_rate": 2.9608098626841736e-06,
-      "loss": 1.4087,
+      "epoch": 1.07,
+      "grad_norm": 13.237042427062988,
+      "learning_rate": 1.2887922018156718e-05,
+      "loss": 1.746,
       "step": 8501
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 26.489471435546875,
-      "learning_rate": 2.9588052520797837e-06,
-      "loss": 1.6874,
+      "epoch": 1.07,
+      "grad_norm": 72.85885620117188,
+      "learning_rate": 1.2887085303100032e-05,
+      "loss": 4.4072,
       "step": 8502
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 25.354101181030273,
-      "learning_rate": 2.9568006414753938e-06,
-      "loss": 0.7809,
+      "epoch": 1.07,
+      "grad_norm": 5.319523811340332,
+      "learning_rate": 1.2886248588043342e-05,
+      "loss": 0.8691,
       "step": 8503
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 8.844145774841309,
-      "learning_rate": 2.9547960308710034e-06,
-      "loss": 0.9255,
+      "epoch": 1.07,
+      "grad_norm": 4.781764984130859,
+      "learning_rate": 1.2885411872986656e-05,
+      "loss": 0.1701,
       "step": 8504
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 18.32469940185547,
-      "learning_rate": 2.9527914202666135e-06,
-      "loss": 1.0907,
+      "epoch": 1.07,
+      "grad_norm": 41.50909423828125,
+      "learning_rate": 1.2884575157929967e-05,
+      "loss": 1.0517,
       "step": 8505
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 14.326888084411621,
-      "learning_rate": 2.950786809662223e-06,
-      "loss": 1.1902,
+      "epoch": 1.07,
+      "grad_norm": 20.49448585510254,
+      "learning_rate": 1.2883738442873281e-05,
+      "loss": 1.4417,
       "step": 8506
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 19.863540649414062,
-      "learning_rate": 2.948782199057833e-06,
-      "loss": 1.5195,
+      "epoch": 1.07,
+      "grad_norm": 16.194725036621094,
+      "learning_rate": 1.2882901727816593e-05,
+      "loss": 2.1433,
       "step": 8507
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 9.05129623413086,
-      "learning_rate": 2.9467775884534433e-06,
-      "loss": 0.7649,
+      "epoch": 1.07,
+      "grad_norm": 10.775847434997559,
+      "learning_rate": 1.2882065012759905e-05,
+      "loss": 1.8694,
       "step": 8508
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 11.893647193908691,
-      "learning_rate": 2.944772977849053e-06,
-      "loss": 1.4486,
+      "epoch": 1.07,
+      "grad_norm": 12.136926651000977,
+      "learning_rate": 1.2881228297703219e-05,
+      "loss": 1.9675,
       "step": 8509
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 18.236352920532227,
-      "learning_rate": 2.9427683672446626e-06,
-      "loss": 1.0206,
+      "epoch": 1.07,
+      "grad_norm": 9.653098106384277,
+      "learning_rate": 1.2880391582646529e-05,
+      "loss": 1.2999,
       "step": 8510
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 32.056034088134766,
-      "learning_rate": 2.940763756640273e-06,
-      "loss": 2.3092,
+      "epoch": 1.07,
+      "grad_norm": 23.440593719482422,
+      "learning_rate": 1.2879554867589843e-05,
+      "loss": 2.8655,
       "step": 8511
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 72.33465576171875,
-      "learning_rate": 2.938759146035883e-06,
-      "loss": 2.1425,
+      "epoch": 1.07,
+      "grad_norm": 7.745482921600342,
+      "learning_rate": 1.2878718152533156e-05,
+      "loss": 0.3792,
       "step": 8512
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 56.66347122192383,
-      "learning_rate": 2.9367545354314925e-06,
-      "loss": 1.1673,
+      "epoch": 1.07,
+      "grad_norm": 12.097042083740234,
+      "learning_rate": 1.287788143747647e-05,
+      "loss": 0.9682,
       "step": 8513
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 14.660852432250977,
-      "learning_rate": 2.9347499248271026e-06,
-      "loss": 1.1077,
+      "epoch": 1.07,
+      "grad_norm": 13.924994468688965,
+      "learning_rate": 1.287704472241978e-05,
+      "loss": 1.8219,
       "step": 8514
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 45.072975158691406,
-      "learning_rate": 2.9327453142227126e-06,
-      "loss": 1.7196,
+      "epoch": 1.07,
+      "grad_norm": 5.2052693367004395,
+      "learning_rate": 1.2876208007363094e-05,
+      "loss": 0.1782,
       "step": 8515
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 15.052632331848145,
-      "learning_rate": 2.9307407036183223e-06,
-      "loss": 1.245,
+      "epoch": 1.07,
+      "grad_norm": 9.861640930175781,
+      "learning_rate": 1.2875371292306407e-05,
+      "loss": 1.4817,
       "step": 8516
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 17.311857223510742,
-      "learning_rate": 2.9287360930139324e-06,
-      "loss": 1.241,
+      "epoch": 1.07,
+      "grad_norm": 11.262444496154785,
+      "learning_rate": 1.2874534577249718e-05,
+      "loss": 1.083,
       "step": 8517
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 71.18388366699219,
-      "learning_rate": 2.926731482409542e-06,
-      "loss": 1.4936,
+      "epoch": 1.07,
+      "grad_norm": 14.59859848022461,
+      "learning_rate": 1.2873697862193031e-05,
+      "loss": 2.0179,
       "step": 8518
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 14.947158813476562,
-      "learning_rate": 2.9247268718051517e-06,
-      "loss": 1.164,
+      "epoch": 1.07,
+      "grad_norm": 25.852855682373047,
+      "learning_rate": 1.2872861147136343e-05,
+      "loss": 1.4012,
       "step": 8519
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 39.9619140625,
-      "learning_rate": 2.9227222612007622e-06,
-      "loss": 2.3687,
-      "step": 8520
-    },
-    {
-      "epoch": 2.56,
-      "eval_loss": 0.17769810557365417,
-      "eval_runtime": 44.5571,
-      "eval_samples_per_second": 33.193,
-      "eval_steps_per_second": 33.193,
+      "epoch": 1.07,
+      "grad_norm": 47.70848083496094,
+      "learning_rate": 1.2872024432079657e-05,
+      "loss": 1.2517,
       "step": 8520
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 17.896486282348633,
-      "learning_rate": 2.920717650596372e-06,
-      "loss": 1.4077,
+      "epoch": 1.07,
+      "grad_norm": 12.465068817138672,
+      "learning_rate": 1.2871187717022969e-05,
+      "loss": 2.0167,
       "step": 8521
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 32.224422454833984,
-      "learning_rate": 2.9187130399919815e-06,
-      "loss": 1.8564,
+      "epoch": 1.07,
+      "grad_norm": 5.884963035583496,
+      "learning_rate": 1.2870351001966281e-05,
+      "loss": 1.7394,
       "step": 8522
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 20.0201358795166,
-      "learning_rate": 2.916708429387592e-06,
-      "loss": 0.6378,
+      "epoch": 1.07,
+      "grad_norm": 24.14913558959961,
+      "learning_rate": 1.2869514286909595e-05,
+      "loss": 0.7824,
       "step": 8523
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 8.714916229248047,
-      "learning_rate": 2.9147038187832017e-06,
-      "loss": 0.8691,
+      "epoch": 1.07,
+      "grad_norm": 10.188133239746094,
+      "learning_rate": 1.2868677571852905e-05,
+      "loss": 1.2877,
       "step": 8524
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 13.311516761779785,
-      "learning_rate": 2.9126992081788114e-06,
-      "loss": 1.0312,
+      "epoch": 1.07,
+      "grad_norm": 48.296077728271484,
+      "learning_rate": 1.2867840856796218e-05,
+      "loss": 3.0401,
       "step": 8525
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 23.491960525512695,
-      "learning_rate": 2.9106945975744215e-06,
-      "loss": 1.2836,
+      "epoch": 1.07,
+      "grad_norm": 18.35318374633789,
+      "learning_rate": 1.2867004141739532e-05,
+      "loss": 2.7819,
       "step": 8526
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 9.72035026550293,
-      "learning_rate": 2.9086899869700315e-06,
-      "loss": 1.1779,
+      "epoch": 1.07,
+      "grad_norm": 6.806789875030518,
+      "learning_rate": 1.2866167426682846e-05,
+      "loss": 1.1726,
       "step": 8527
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 14.758877754211426,
-      "learning_rate": 2.906685376365641e-06,
-      "loss": 1.3364,
+      "epoch": 1.07,
+      "grad_norm": 2.1864826679229736,
+      "learning_rate": 1.2865330711626156e-05,
+      "loss": 0.0783,
       "step": 8528
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 9.038101196289062,
-      "learning_rate": 2.9046807657612513e-06,
-      "loss": 0.7359,
+      "epoch": 1.07,
+      "grad_norm": 10.152175903320312,
+      "learning_rate": 1.286449399656947e-05,
+      "loss": 2.3536,
       "step": 8529
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 9.687514305114746,
-      "learning_rate": 2.902676155156861e-06,
-      "loss": 0.6275,
+      "epoch": 1.07,
+      "grad_norm": 3.9797370433807373,
+      "learning_rate": 1.2863657281512783e-05,
+      "loss": 0.3895,
       "step": 8530
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 21.076562881469727,
-      "learning_rate": 2.9006715445524706e-06,
-      "loss": 1.3599,
+      "epoch": 1.07,
+      "grad_norm": 21.074848175048828,
+      "learning_rate": 1.2862820566456094e-05,
+      "loss": 1.5691,
       "step": 8531
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 8.555130004882812,
-      "learning_rate": 2.898666933948081e-06,
-      "loss": 0.9138,
+      "epoch": 1.07,
+      "grad_norm": 12.177873611450195,
+      "learning_rate": 1.2861983851399407e-05,
+      "loss": 1.2464,
       "step": 8532
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 30.38118553161621,
-      "learning_rate": 2.8966623233436908e-06,
-      "loss": 1.4925,
+      "epoch": 1.07,
+      "grad_norm": 15.976359367370605,
+      "learning_rate": 1.2861147136342719e-05,
+      "loss": 1.0228,
       "step": 8533
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 34.965641021728516,
-      "learning_rate": 2.8946577127393004e-06,
-      "loss": 1.3331,
+      "epoch": 1.07,
+      "grad_norm": 23.05260467529297,
+      "learning_rate": 1.2860310421286033e-05,
+      "loss": 0.9748,
       "step": 8534
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 10.746747016906738,
-      "learning_rate": 2.892653102134911e-06,
-      "loss": 1.2565,
+      "epoch": 1.07,
+      "grad_norm": 18.960296630859375,
+      "learning_rate": 1.2859473706229345e-05,
+      "loss": 2.3059,
       "step": 8535
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 62.186431884765625,
-      "learning_rate": 2.8906484915305206e-06,
-      "loss": 2.5688,
+      "epoch": 1.07,
+      "grad_norm": 14.666804313659668,
+      "learning_rate": 1.2858636991172657e-05,
+      "loss": 1.1556,
       "step": 8536
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 9.051580429077148,
-      "learning_rate": 2.8886438809261303e-06,
-      "loss": 0.7643,
+      "epoch": 1.07,
+      "grad_norm": 39.453224182128906,
+      "learning_rate": 1.285780027611597e-05,
+      "loss": 1.7697,
       "step": 8537
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 22.37142562866211,
-      "learning_rate": 2.88663927032174e-06,
-      "loss": 1.8968,
+      "epoch": 1.07,
+      "grad_norm": 5.983280181884766,
+      "learning_rate": 1.285696356105928e-05,
+      "loss": 0.5179,
       "step": 8538
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 15.524712562561035,
-      "learning_rate": 2.8846346597173504e-06,
-      "loss": 0.9424,
+      "epoch": 1.07,
+      "grad_norm": 16.707027435302734,
+      "learning_rate": 1.2856126846002594e-05,
+      "loss": 2.7405,
       "step": 8539
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 14.384515762329102,
-      "learning_rate": 2.88263004911296e-06,
-      "loss": 1.0429,
+      "epoch": 1.07,
+      "grad_norm": 9.118749618530273,
+      "learning_rate": 1.2855290130945908e-05,
+      "loss": 0.7396,
       "step": 8540
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 55.4445686340332,
-      "learning_rate": 2.8806254385085698e-06,
-      "loss": 2.7088,
+      "epoch": 1.07,
+      "grad_norm": 12.518699645996094,
+      "learning_rate": 1.2854453415889222e-05,
+      "loss": 0.8971,
       "step": 8541
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 11.720520973205566,
-      "learning_rate": 2.87862082790418e-06,
-      "loss": 0.7067,
+      "epoch": 1.07,
+      "grad_norm": 5.475255489349365,
+      "learning_rate": 1.2853616700832532e-05,
+      "loss": 0.307,
       "step": 8542
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 18.14213752746582,
-      "learning_rate": 2.8766162172997895e-06,
-      "loss": 1.0006,
+      "epoch": 1.07,
+      "grad_norm": 5.851897716522217,
+      "learning_rate": 1.2852779985775845e-05,
+      "loss": 1.6283,
       "step": 8543
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 9.360933303833008,
-      "learning_rate": 2.8746116066953996e-06,
-      "loss": 1.1814,
+      "epoch": 1.07,
+      "grad_norm": 25.698322296142578,
+      "learning_rate": 1.2851943270719157e-05,
+      "loss": 2.1417,
       "step": 8544
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 54.820091247558594,
-      "learning_rate": 2.8726069960910097e-06,
-      "loss": 1.0313,
+      "epoch": 1.07,
+      "grad_norm": 13.419102668762207,
+      "learning_rate": 1.285110655566247e-05,
+      "loss": 2.0774,
       "step": 8545
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 17.7171688079834,
-      "learning_rate": 2.8706023854866193e-06,
-      "loss": 0.5955,
+      "epoch": 1.07,
+      "grad_norm": 7.896689414978027,
+      "learning_rate": 1.2850269840605783e-05,
+      "loss": 1.5321,
       "step": 8546
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 26.265403747558594,
-      "learning_rate": 2.868597774882229e-06,
-      "loss": 1.2873,
+      "epoch": 1.07,
+      "grad_norm": 6.227831840515137,
+      "learning_rate": 1.2849433125549095e-05,
+      "loss": 1.4229,
       "step": 8547
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 51.058250427246094,
-      "learning_rate": 2.8665931642778395e-06,
-      "loss": 1.3523,
+      "epoch": 1.07,
+      "grad_norm": 12.581622123718262,
+      "learning_rate": 1.2848596410492409e-05,
+      "loss": 0.6904,
       "step": 8548
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 14.056093215942383,
-      "learning_rate": 2.864588553673449e-06,
-      "loss": 0.9431,
+      "epoch": 1.07,
+      "grad_norm": 13.476349830627441,
+      "learning_rate": 1.284775969543572e-05,
+      "loss": 1.8013,
       "step": 8549
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 55.49951171875,
-      "learning_rate": 2.862583943069059e-06,
-      "loss": 1.6338,
+      "epoch": 1.07,
+      "grad_norm": 11.182458877563477,
+      "learning_rate": 1.2846922980379033e-05,
+      "loss": 2.6402,
       "step": 8550
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 12.471390724182129,
-      "learning_rate": 2.8605793324646693e-06,
-      "loss": 1.1413,
+      "epoch": 1.07,
+      "grad_norm": 12.608802795410156,
+      "learning_rate": 1.2846086265322346e-05,
+      "loss": 1.9193,
       "step": 8551
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 19.35071563720703,
-      "learning_rate": 2.858574721860279e-06,
-      "loss": 0.9613,
+      "epoch": 1.07,
+      "grad_norm": 14.025070190429688,
+      "learning_rate": 1.2845249550265656e-05,
+      "loss": 2.8147,
       "step": 8552
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 16.208166122436523,
-      "learning_rate": 2.8565701112558886e-06,
-      "loss": 1.5786,
+      "epoch": 1.07,
+      "grad_norm": 2.8733303546905518,
+      "learning_rate": 1.284441283520897e-05,
+      "loss": 0.0776,
       "step": 8553
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 24.832500457763672,
-      "learning_rate": 2.8545655006514987e-06,
-      "loss": 1.6098,
+      "epoch": 1.07,
+      "grad_norm": 7.129153251647949,
+      "learning_rate": 1.2843576120152284e-05,
+      "loss": 3.1068,
       "step": 8554
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 51.225799560546875,
-      "learning_rate": 2.8525608900471084e-06,
-      "loss": 1.4513,
+      "epoch": 1.07,
+      "grad_norm": 6.150088787078857,
+      "learning_rate": 1.2842739405095597e-05,
+      "loss": 1.1727,
       "step": 8555
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 13.274884223937988,
-      "learning_rate": 2.8505562794427185e-06,
-      "loss": 0.8535,
+      "epoch": 1.07,
+      "grad_norm": 37.756011962890625,
+      "learning_rate": 1.2841902690038908e-05,
+      "loss": 1.289,
       "step": 8556
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 31.125516891479492,
-      "learning_rate": 2.8485516688383286e-06,
-      "loss": 1.388,
+      "epoch": 1.07,
+      "grad_norm": 19.50652313232422,
+      "learning_rate": 1.2841065974982221e-05,
+      "loss": 1.7728,
       "step": 8557
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 17.432456970214844,
-      "learning_rate": 2.8465470582339382e-06,
-      "loss": 1.5913,
+      "epoch": 1.07,
+      "grad_norm": 8.960627555847168,
+      "learning_rate": 1.2840229259925533e-05,
+      "loss": 0.5113,
       "step": 8558
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 24.500856399536133,
-      "learning_rate": 2.844542447629548e-06,
-      "loss": 2.2924,
+      "epoch": 1.07,
+      "grad_norm": 11.609451293945312,
+      "learning_rate": 1.2839392544868845e-05,
+      "loss": 1.1409,
       "step": 8559
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 14.529585838317871,
-      "learning_rate": 2.8425378370251584e-06,
-      "loss": 0.5999,
+      "epoch": 1.07,
+      "grad_norm": 83.95161437988281,
+      "learning_rate": 1.2838555829812159e-05,
+      "loss": 2.2761,
       "step": 8560
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 12.131428718566895,
-      "learning_rate": 2.840533226420768e-06,
-      "loss": 1.439,
+      "epoch": 1.07,
+      "grad_norm": 12.734169006347656,
+      "learning_rate": 1.283771911475547e-05,
+      "loss": 0.7802,
       "step": 8561
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 8.231968879699707,
-      "learning_rate": 2.8385286158163777e-06,
-      "loss": 0.9683,
+      "epoch": 1.07,
+      "grad_norm": 13.65036392211914,
+      "learning_rate": 1.2836882399698784e-05,
+      "loss": 1.8548,
       "step": 8562
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 7.67575740814209,
-      "learning_rate": 2.836524005211988e-06,
-      "loss": 0.6245,
+      "epoch": 1.07,
+      "grad_norm": 43.75178146362305,
+      "learning_rate": 1.2836045684642095e-05,
+      "loss": 1.3983,
       "step": 8563
     },
     {
-      "epoch": 2.57,
-      "grad_norm": 12.068075180053711,
-      "learning_rate": 2.834519394607598e-06,
-      "loss": 0.7527,
+      "epoch": 1.07,
+      "grad_norm": 7.9612274169921875,
+      "learning_rate": 1.2835208969585408e-05,
+      "loss": 0.3605,
       "step": 8564
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 14.110102653503418,
-      "learning_rate": 2.8325147840032075e-06,
-      "loss": 1.7017,
+      "epoch": 1.07,
+      "grad_norm": 12.827942848205566,
+      "learning_rate": 1.2834372254528722e-05,
+      "loss": 2.2101,
       "step": 8565
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 38.348731994628906,
-      "learning_rate": 2.8305101733988176e-06,
-      "loss": 2.0444,
+      "epoch": 1.08,
+      "grad_norm": 41.54198455810547,
+      "learning_rate": 1.2833535539472032e-05,
+      "loss": 1.6254,
       "step": 8566
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 9.310153007507324,
-      "learning_rate": 2.8285055627944273e-06,
-      "loss": 1.0862,
+      "epoch": 1.08,
+      "grad_norm": 16.435697555541992,
+      "learning_rate": 1.2832698824415346e-05,
+      "loss": 2.8402,
       "step": 8567
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 18.77411460876465,
-      "learning_rate": 2.8265009521900374e-06,
-      "loss": 1.0017,
+      "epoch": 1.08,
+      "grad_norm": 10.593535423278809,
+      "learning_rate": 1.283186210935866e-05,
+      "loss": 2.2697,
       "step": 8568
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 9.130143165588379,
-      "learning_rate": 2.8244963415856475e-06,
-      "loss": 1.1588,
+      "epoch": 1.08,
+      "grad_norm": 38.62551498413086,
+      "learning_rate": 1.2831025394301973e-05,
+      "loss": 2.4518,
       "step": 8569
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 37.241661071777344,
-      "learning_rate": 2.822491730981257e-06,
-      "loss": 1.1469,
+      "epoch": 1.08,
+      "grad_norm": 7.9425201416015625,
+      "learning_rate": 1.2830188679245283e-05,
+      "loss": 0.6873,
       "step": 8570
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 19.131694793701172,
-      "learning_rate": 2.8204871203768668e-06,
-      "loss": 1.8664,
+      "epoch": 1.08,
+      "grad_norm": 24.836193084716797,
+      "learning_rate": 1.2829351964188597e-05,
+      "loss": 2.4507,
       "step": 8571
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 16.91837501525879,
-      "learning_rate": 2.8184825097724773e-06,
-      "loss": 1.2044,
+      "epoch": 1.08,
+      "grad_norm": 19.25704002380371,
+      "learning_rate": 1.2828515249131909e-05,
+      "loss": 1.3152,
       "step": 8572
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 12.339491844177246,
-      "learning_rate": 2.816477899168087e-06,
-      "loss": 0.8794,
+      "epoch": 1.08,
+      "grad_norm": 38.70884323120117,
+      "learning_rate": 1.2827678534075221e-05,
+      "loss": 3.0298,
       "step": 8573
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 41.453243255615234,
-      "learning_rate": 2.8144732885636966e-06,
-      "loss": 1.4408,
+      "epoch": 1.08,
+      "grad_norm": 6.464087009429932,
+      "learning_rate": 1.2826841819018535e-05,
+      "loss": 1.7264,
       "step": 8574
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 36.32587814331055,
-      "learning_rate": 2.8124686779593067e-06,
-      "loss": 0.8383,
+      "epoch": 1.08,
+      "grad_norm": 14.138628959655762,
+      "learning_rate": 1.2826005103961847e-05,
+      "loss": 1.5077,
       "step": 8575
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 19.15566635131836,
-      "learning_rate": 2.8104640673549168e-06,
-      "loss": 1.0883,
+      "epoch": 1.08,
+      "grad_norm": 30.603025436401367,
+      "learning_rate": 1.282516838890516e-05,
+      "loss": 2.0811,
       "step": 8576
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 18.432010650634766,
-      "learning_rate": 2.8084594567505264e-06,
-      "loss": 1.314,
+      "epoch": 1.08,
+      "grad_norm": 8.994444847106934,
+      "learning_rate": 1.282433167384847e-05,
+      "loss": 0.3496,
       "step": 8577
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 51.9474983215332,
-      "learning_rate": 2.8064548461461365e-06,
-      "loss": 1.7958,
+      "epoch": 1.08,
+      "grad_norm": 7.73020601272583,
+      "learning_rate": 1.2823494958791784e-05,
+      "loss": 0.8861,
       "step": 8578
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 21.26700210571289,
-      "learning_rate": 2.804450235541746e-06,
-      "loss": 1.878,
+      "epoch": 1.08,
+      "grad_norm": 11.562694549560547,
+      "learning_rate": 1.2822658243735098e-05,
+      "loss": 0.9408,
       "step": 8579
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 23.973962783813477,
-      "learning_rate": 2.8024456249373563e-06,
-      "loss": 1.2876,
+      "epoch": 1.08,
+      "grad_norm": 19.258441925048828,
+      "learning_rate": 1.2821821528678408e-05,
+      "loss": 1.2217,
       "step": 8580
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 11.906970024108887,
-      "learning_rate": 2.800441014332966e-06,
-      "loss": 0.5872,
+      "epoch": 1.08,
+      "grad_norm": 12.606703758239746,
+      "learning_rate": 1.2820984813621722e-05,
+      "loss": 1.2354,
       "step": 8581
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 14.538961410522461,
-      "learning_rate": 2.798436403728576e-06,
-      "loss": 0.9155,
+      "epoch": 1.08,
+      "grad_norm": 8.044978141784668,
+      "learning_rate": 1.2820148098565035e-05,
+      "loss": 0.4183,
       "step": 8582
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 19.3804988861084,
-      "learning_rate": 2.7964317931241857e-06,
-      "loss": 1.2437,
+      "epoch": 1.08,
+      "grad_norm": 33.10894012451172,
+      "learning_rate": 1.2819311383508349e-05,
+      "loss": 1.6051,
       "step": 8583
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 21.384546279907227,
-      "learning_rate": 2.7944271825197953e-06,
-      "loss": 0.8738,
+      "epoch": 1.08,
+      "grad_norm": 10.770204544067383,
+      "learning_rate": 1.281847466845166e-05,
+      "loss": 0.8247,
       "step": 8584
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 5.975277423858643,
-      "learning_rate": 2.792422571915406e-06,
-      "loss": 0.4514,
+      "epoch": 1.08,
+      "grad_norm": 12.140080451965332,
+      "learning_rate": 1.2817637953394973e-05,
+      "loss": 0.8705,
       "step": 8585
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 13.506431579589844,
-      "learning_rate": 2.7904179613110155e-06,
-      "loss": 1.3215,
+      "epoch": 1.08,
+      "grad_norm": 12.246003150939941,
+      "learning_rate": 1.2816801238338285e-05,
+      "loss": 2.2221,
       "step": 8586
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 16.181753158569336,
-      "learning_rate": 2.788413350706625e-06,
-      "loss": 1.9854,
+      "epoch": 1.08,
+      "grad_norm": 16.525575637817383,
+      "learning_rate": 1.2815964523281597e-05,
+      "loss": 1.4073,
       "step": 8587
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 23.374975204467773,
-      "learning_rate": 2.7864087401022357e-06,
-      "loss": 1.9213,
+      "epoch": 1.08,
+      "grad_norm": 10.41401481628418,
+      "learning_rate": 1.281512780822491e-05,
+      "loss": 1.7817,
       "step": 8588
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 29.04400634765625,
-      "learning_rate": 2.7844041294978453e-06,
-      "loss": 1.8084,
+      "epoch": 1.08,
+      "grad_norm": 74.99308776855469,
+      "learning_rate": 1.2814291093168222e-05,
+      "loss": 2.2155,
       "step": 8589
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 9.371678352355957,
-      "learning_rate": 2.782399518893455e-06,
-      "loss": 1.0408,
+      "epoch": 1.08,
+      "grad_norm": 6.079598903656006,
+      "learning_rate": 1.2813454378111536e-05,
+      "loss": 0.2717,
       "step": 8590
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 11.659550666809082,
-      "learning_rate": 2.780394908289065e-06,
-      "loss": 1.3009,
+      "epoch": 1.08,
+      "grad_norm": 76.10659790039062,
+      "learning_rate": 1.2812617663054846e-05,
+      "loss": 1.7313,
       "step": 8591
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 27.285043716430664,
-      "learning_rate": 2.7783902976846747e-06,
-      "loss": 1.9198,
+      "epoch": 1.08,
+      "grad_norm": 31.880477905273438,
+      "learning_rate": 1.281178094799816e-05,
+      "loss": 0.3522,
       "step": 8592
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 30.421232223510742,
-      "learning_rate": 2.776385687080285e-06,
-      "loss": 1.3893,
+      "epoch": 1.08,
+      "grad_norm": 13.708595275878906,
+      "learning_rate": 1.2810944232941474e-05,
+      "loss": 2.8344,
       "step": 8593
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 14.901866912841797,
-      "learning_rate": 2.774381076475895e-06,
-      "loss": 1.0787,
+      "epoch": 1.08,
+      "grad_norm": 47.557804107666016,
+      "learning_rate": 1.2810107517884784e-05,
+      "loss": 2.2919,
       "step": 8594
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 10.58189868927002,
-      "learning_rate": 2.7723764658715046e-06,
-      "loss": 1.3617,
+      "epoch": 1.08,
+      "grad_norm": 6.2081685066223145,
+      "learning_rate": 1.2809270802828098e-05,
+      "loss": 0.2591,
       "step": 8595
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 83.2371826171875,
-      "learning_rate": 2.7703718552671142e-06,
-      "loss": 1.7557,
+      "epoch": 1.08,
+      "grad_norm": 17.9738826751709,
+      "learning_rate": 1.2808434087771411e-05,
+      "loss": 1.872,
       "step": 8596
     },
     {
-      "epoch": 2.58,
-      "grad_norm": 23.80765724182129,
-      "learning_rate": 2.7683672446627247e-06,
-      "loss": 2.3199,
+      "epoch": 1.08,
+      "grad_norm": 23.148326873779297,
+      "learning_rate": 1.2807597372714723e-05,
+      "loss": 1.6406,
       "step": 8597
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 15.422316551208496,
-      "learning_rate": 2.7663626340583344e-06,
-      "loss": 1.1275,
+      "epoch": 1.08,
+      "grad_norm": 26.999752044677734,
+      "learning_rate": 1.2806760657658035e-05,
+      "loss": 2.2209,
       "step": 8598
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 20.54278564453125,
-      "learning_rate": 2.764358023453944e-06,
-      "loss": 1.7601,
+      "epoch": 1.08,
+      "grad_norm": 8.44945240020752,
+      "learning_rate": 1.2805923942601349e-05,
+      "loss": 0.8549,
       "step": 8599
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 14.661285400390625,
-      "learning_rate": 2.7623534128495546e-06,
-      "loss": 1.0741,
+      "epoch": 1.08,
+      "grad_norm": 8.50709056854248,
+      "learning_rate": 1.280508722754466e-05,
+      "loss": 0.533,
       "step": 8600
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 33.70438766479492,
-      "learning_rate": 2.7603488022451642e-06,
-      "loss": 1.0387,
+      "epoch": 1.08,
+      "grad_norm": 11.294102668762207,
+      "learning_rate": 1.2804250512487973e-05,
+      "loss": 0.4275,
       "step": 8601
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 15.016576766967773,
-      "learning_rate": 2.758344191640774e-06,
-      "loss": 1.7023,
+      "epoch": 1.08,
+      "grad_norm": 4.108835697174072,
+      "learning_rate": 1.2803413797431286e-05,
+      "loss": 0.0644,
       "step": 8602
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 7.491636753082275,
-      "learning_rate": 2.756339581036384e-06,
-      "loss": 1.374,
+      "epoch": 1.08,
+      "grad_norm": 70.74079132080078,
+      "learning_rate": 1.2802577082374598e-05,
+      "loss": 1.4626,
       "step": 8603
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 13.837325096130371,
-      "learning_rate": 2.7543349704319936e-06,
-      "loss": 0.8613,
+      "epoch": 1.08,
+      "grad_norm": 21.69393539428711,
+      "learning_rate": 1.2801740367317912e-05,
+      "loss": 1.4618,
       "step": 8604
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 17.60556983947754,
-      "learning_rate": 2.7523303598276037e-06,
-      "loss": 1.1599,
+      "epoch": 1.08,
+      "grad_norm": 10.441219329833984,
+      "learning_rate": 1.2800903652261222e-05,
+      "loss": 1.2134,
       "step": 8605
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 13.817652702331543,
-      "learning_rate": 2.750325749223214e-06,
-      "loss": 1.1847,
+      "epoch": 1.08,
+      "grad_norm": 8.55947208404541,
+      "learning_rate": 1.2800066937204536e-05,
+      "loss": 1.9702,
       "step": 8606
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 39.440921783447266,
-      "learning_rate": 2.7483211386188235e-06,
-      "loss": 1.5714,
+      "epoch": 1.08,
+      "grad_norm": 18.24686050415039,
+      "learning_rate": 1.279923022214785e-05,
+      "loss": 1.4235,
       "step": 8607
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 13.094484329223633,
-      "learning_rate": 2.746316528014433e-06,
-      "loss": 1.4254,
+      "epoch": 1.08,
+      "grad_norm": 5.580203533172607,
+      "learning_rate": 1.279839350709116e-05,
+      "loss": 0.1612,
       "step": 8608
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 12.363558769226074,
-      "learning_rate": 2.7443119174100436e-06,
-      "loss": 1.4503,
+      "epoch": 1.08,
+      "grad_norm": 6.0034894943237305,
+      "learning_rate": 1.2797556792034473e-05,
+      "loss": 1.0321,
       "step": 8609
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 81.3821792602539,
-      "learning_rate": 2.7423073068056533e-06,
-      "loss": 1.6868,
+      "epoch": 1.08,
+      "grad_norm": 14.018782615661621,
+      "learning_rate": 1.2796720076977787e-05,
+      "loss": 1.6453,
       "step": 8610
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 28.759248733520508,
-      "learning_rate": 2.740302696201263e-06,
-      "loss": 1.2274,
+      "epoch": 1.08,
+      "grad_norm": 18.138858795166016,
+      "learning_rate": 1.2795883361921099e-05,
+      "loss": 1.3789,
       "step": 8611
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 22.243206024169922,
-      "learning_rate": 2.7382980855968735e-06,
-      "loss": 1.5054,
+      "epoch": 1.08,
+      "grad_norm": 9.55051326751709,
+      "learning_rate": 1.2795046646864411e-05,
+      "loss": 1.4569,
       "step": 8612
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 16.99969482421875,
-      "learning_rate": 2.736293474992483e-06,
-      "loss": 1.3755,
+      "epoch": 1.08,
+      "grad_norm": 8.357460975646973,
+      "learning_rate": 1.2794209931807725e-05,
+      "loss": 0.7194,
       "step": 8613
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 8.332833290100098,
-      "learning_rate": 2.7342888643880928e-06,
-      "loss": 1.0886,
+      "epoch": 1.08,
+      "grad_norm": 6.152368545532227,
+      "learning_rate": 1.2793373216751037e-05,
+      "loss": 0.9195,
       "step": 8614
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 17.875459671020508,
-      "learning_rate": 2.732284253783703e-06,
-      "loss": 1.2963,
+      "epoch": 1.08,
+      "grad_norm": 7.088659286499023,
+      "learning_rate": 1.2792536501694349e-05,
+      "loss": 2.3761,
       "step": 8615
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 17.195663452148438,
-      "learning_rate": 2.7302796431793125e-06,
-      "loss": 0.9124,
+      "epoch": 1.08,
+      "grad_norm": 16.12812614440918,
+      "learning_rate": 1.279169978663766e-05,
+      "loss": 1.8399,
       "step": 8616
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 45.631954193115234,
-      "learning_rate": 2.7282750325749226e-06,
-      "loss": 2.5502,
+      "epoch": 1.08,
+      "grad_norm": 12.287689208984375,
+      "learning_rate": 1.2790863071580974e-05,
+      "loss": 0.7975,
       "step": 8617
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 72.2505874633789,
-      "learning_rate": 2.7262704219705327e-06,
-      "loss": 1.9153,
+      "epoch": 1.08,
+      "grad_norm": 18.367534637451172,
+      "learning_rate": 1.2790026356524288e-05,
+      "loss": 0.6008,
       "step": 8618
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 19.540000915527344,
-      "learning_rate": 2.7242658113661424e-06,
-      "loss": 1.3127,
+      "epoch": 1.08,
+      "grad_norm": 9.941510200500488,
+      "learning_rate": 1.2789189641467598e-05,
+      "loss": 0.8487,
       "step": 8619
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 16.321380615234375,
-      "learning_rate": 2.722261200761752e-06,
-      "loss": 1.1336,
+      "epoch": 1.08,
+      "grad_norm": 11.001337051391602,
+      "learning_rate": 1.2788352926410912e-05,
+      "loss": 1.2659,
       "step": 8620
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 35.40462875366211,
-      "learning_rate": 2.7202565901573625e-06,
-      "loss": 1.4144,
+      "epoch": 1.08,
+      "grad_norm": 18.091001510620117,
+      "learning_rate": 1.2787516211354225e-05,
+      "loss": 1.5459,
       "step": 8621
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 18.681198120117188,
-      "learning_rate": 2.718251979552972e-06,
-      "loss": 0.8571,
+      "epoch": 1.08,
+      "grad_norm": 39.56760787963867,
+      "learning_rate": 1.2786679496297536e-05,
+      "loss": 2.3626,
       "step": 8622
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 37.2798957824707,
-      "learning_rate": 2.716247368948582e-06,
-      "loss": 1.1746,
+      "epoch": 1.08,
+      "grad_norm": 12.28984260559082,
+      "learning_rate": 1.278584278124085e-05,
+      "loss": 1.4275,
       "step": 8623
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 17.78194808959961,
-      "learning_rate": 2.7142427583441915e-06,
-      "loss": 1.0923,
+      "epoch": 1.08,
+      "grad_norm": 6.757332801818848,
+      "learning_rate": 1.2785006066184163e-05,
+      "loss": 0.2179,
       "step": 8624
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 18.684160232543945,
-      "learning_rate": 2.712238147739802e-06,
-      "loss": 1.3139,
+      "epoch": 1.08,
+      "grad_norm": 8.783282279968262,
+      "learning_rate": 1.2784169351127475e-05,
+      "loss": 0.4344,
       "step": 8625
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 25.179645538330078,
-      "learning_rate": 2.7102335371354117e-06,
-      "loss": 1.5291,
+      "epoch": 1.08,
+      "grad_norm": 19.863969802856445,
+      "learning_rate": 1.2783332636070787e-05,
+      "loss": 1.7297,
       "step": 8626
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 9.155866622924805,
-      "learning_rate": 2.7082289265310213e-06,
-      "loss": 0.6078,
+      "epoch": 1.08,
+      "grad_norm": 13.56358814239502,
+      "learning_rate": 1.27824959210141e-05,
+      "loss": 0.3086,
       "step": 8627
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 25.500167846679688,
-      "learning_rate": 2.7062243159266314e-06,
-      "loss": 1.8845,
+      "epoch": 1.08,
+      "grad_norm": 11.027216911315918,
+      "learning_rate": 1.2781659205957412e-05,
+      "loss": 0.847,
       "step": 8628
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 14.735610008239746,
-      "learning_rate": 2.7042197053222415e-06,
-      "loss": 1.3812,
+      "epoch": 1.08,
+      "grad_norm": 15.298280715942383,
+      "learning_rate": 1.2780822490900724e-05,
+      "loss": 1.68,
       "step": 8629
     },
     {
-      "epoch": 2.59,
-      "grad_norm": 13.087080001831055,
-      "learning_rate": 2.702215094717851e-06,
-      "loss": 2.3241,
+      "epoch": 1.08,
+      "grad_norm": 31.234249114990234,
+      "learning_rate": 1.2779985775844036e-05,
+      "loss": 2.733,
       "step": 8630
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 15.85035228729248,
-      "learning_rate": 2.7002104841134612e-06,
-      "loss": 2.1271,
+      "epoch": 1.08,
+      "grad_norm": 16.561674118041992,
+      "learning_rate": 1.277914906078735e-05,
+      "loss": 0.555,
       "step": 8631
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 16.836761474609375,
-      "learning_rate": 2.698205873509071e-06,
-      "loss": 0.9277,
+      "epoch": 1.08,
+      "grad_norm": 6.423410415649414,
+      "learning_rate": 1.2778312345730664e-05,
+      "loss": 1.22,
       "step": 8632
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 17.871294021606445,
-      "learning_rate": 2.6962012629046806e-06,
-      "loss": 1.3209,
+      "epoch": 1.08,
+      "grad_norm": 19.52211570739746,
+      "learning_rate": 1.2777475630673974e-05,
+      "loss": 2.2232,
       "step": 8633
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 10.736555099487305,
-      "learning_rate": 2.694196652300291e-06,
-      "loss": 0.9247,
+      "epoch": 1.08,
+      "grad_norm": 17.728485107421875,
+      "learning_rate": 1.2776638915617288e-05,
+      "loss": 2.1009,
       "step": 8634
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 4.654946804046631,
-      "learning_rate": 2.6921920416959007e-06,
-      "loss": 0.6192,
+      "epoch": 1.08,
+      "grad_norm": 20.689403533935547,
+      "learning_rate": 1.2775802200560601e-05,
+      "loss": 1.6795,
       "step": 8635
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 62.396331787109375,
-      "learning_rate": 2.6901874310915104e-06,
-      "loss": 2.6783,
+      "epoch": 1.08,
+      "grad_norm": 5.071125030517578,
+      "learning_rate": 1.2774965485503911e-05,
+      "loss": 0.1614,
       "step": 8636
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 37.43707275390625,
-      "learning_rate": 2.688182820487121e-06,
-      "loss": 0.6253,
+      "epoch": 1.08,
+      "grad_norm": 9.400809288024902,
+      "learning_rate": 1.2774128770447225e-05,
+      "loss": 1.924,
       "step": 8637
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 10.426443099975586,
-      "learning_rate": 2.6861782098827306e-06,
-      "loss": 1.1717,
+      "epoch": 1.08,
+      "grad_norm": 23.80042266845703,
+      "learning_rate": 1.2773292055390539e-05,
+      "loss": 2.4768,
       "step": 8638
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 77.32369995117188,
-      "learning_rate": 2.6841735992783402e-06,
-      "loss": 2.797,
+      "epoch": 1.08,
+      "grad_norm": 25.71880531311035,
+      "learning_rate": 1.277245534033385e-05,
+      "loss": 1.2631,
       "step": 8639
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 18.485462188720703,
-      "learning_rate": 2.6821689886739503e-06,
-      "loss": 0.9191,
-      "step": 8640
-    },
-    {
-      "epoch": 2.6,
-      "eval_loss": 0.17308995127677917,
-      "eval_runtime": 43.9129,
-      "eval_samples_per_second": 33.68,
-      "eval_steps_per_second": 33.68,
+      "epoch": 1.08,
+      "grad_norm": 9.267672538757324,
+      "learning_rate": 1.2771618625277163e-05,
+      "loss": 0.2955,
       "step": 8640
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 10.55259895324707,
-      "learning_rate": 2.6801643780695604e-06,
-      "loss": 1.4176,
+      "epoch": 1.08,
+      "grad_norm": 4.219285488128662,
+      "learning_rate": 1.2770781910220476e-05,
+      "loss": 0.3018,
       "step": 8641
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 14.04641056060791,
-      "learning_rate": 2.67815976746517e-06,
-      "loss": 1.3492,
+      "epoch": 1.08,
+      "grad_norm": 5.892913818359375,
+      "learning_rate": 1.2769945195163788e-05,
+      "loss": 0.7357,
       "step": 8642
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 11.673070907592773,
-      "learning_rate": 2.67615515686078e-06,
-      "loss": 1.3232,
+      "epoch": 1.08,
+      "grad_norm": 19.84624481201172,
+      "learning_rate": 1.27691084801071e-05,
+      "loss": 2.7648,
       "step": 8643
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 22.69013023376465,
-      "learning_rate": 2.67415054625639e-06,
-      "loss": 1.535,
+      "epoch": 1.08,
+      "grad_norm": 9.951528549194336,
+      "learning_rate": 1.2768271765050412e-05,
+      "loss": 1.8721,
       "step": 8644
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 31.60176658630371,
-      "learning_rate": 2.6721459356519995e-06,
-      "loss": 2.397,
+      "epoch": 1.08,
+      "grad_norm": 27.758649826049805,
+      "learning_rate": 1.2767435049993726e-05,
+      "loss": 1.1707,
       "step": 8645
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 72.22872161865234,
-      "learning_rate": 2.67014132504761e-06,
-      "loss": 3.1373,
+      "epoch": 1.09,
+      "grad_norm": 6.61989688873291,
+      "learning_rate": 1.276659833493704e-05,
+      "loss": 1.2057,
       "step": 8646
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 12.671222686767578,
-      "learning_rate": 2.6681367144432196e-06,
-      "loss": 0.9173,
+      "epoch": 1.09,
+      "grad_norm": 6.966522693634033,
+      "learning_rate": 1.276576161988035e-05,
+      "loss": 2.5643,
       "step": 8647
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 24.99388885498047,
-      "learning_rate": 2.6661321038388293e-06,
-      "loss": 1.3955,
+      "epoch": 1.09,
+      "grad_norm": 10.06793212890625,
+      "learning_rate": 1.2764924904823663e-05,
+      "loss": 0.8765,
       "step": 8648
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 28.980268478393555,
-      "learning_rate": 2.66412749323444e-06,
-      "loss": 1.3139,
+      "epoch": 1.09,
+      "grad_norm": 8.949012756347656,
+      "learning_rate": 1.2764088189766977e-05,
+      "loss": 2.0474,
       "step": 8649
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 14.720972061157227,
-      "learning_rate": 2.6621228826300495e-06,
-      "loss": 0.8373,
+      "epoch": 1.09,
+      "grad_norm": 17.571147918701172,
+      "learning_rate": 1.2763251474710287e-05,
+      "loss": 0.9656,
       "step": 8650
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 63.49948501586914,
-      "learning_rate": 2.660118272025659e-06,
-      "loss": 1.9721,
+      "epoch": 1.09,
+      "grad_norm": 12.04281234741211,
+      "learning_rate": 1.2762414759653601e-05,
+      "loss": 0.9934,
       "step": 8651
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 11.087202072143555,
-      "learning_rate": 2.658113661421269e-06,
-      "loss": 0.9022,
+      "epoch": 1.09,
+      "grad_norm": 13.158784866333008,
+      "learning_rate": 1.2761578044596915e-05,
+      "loss": 1.818,
       "step": 8652
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 21.9641170501709,
-      "learning_rate": 2.6561090508168793e-06,
-      "loss": 1.9667,
+      "epoch": 1.09,
+      "grad_norm": 21.233274459838867,
+      "learning_rate": 1.2760741329540227e-05,
+      "loss": 1.3968,
       "step": 8653
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 16.303483963012695,
-      "learning_rate": 2.654104440212489e-06,
-      "loss": 1.1034,
+      "epoch": 1.09,
+      "grad_norm": 10.787312507629395,
+      "learning_rate": 1.2759904614483539e-05,
+      "loss": 0.5281,
       "step": 8654
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 15.40654468536377,
-      "learning_rate": 2.652099829608099e-06,
-      "loss": 0.9308,
+      "epoch": 1.09,
+      "grad_norm": 19.136770248413086,
+      "learning_rate": 1.275906789942685e-05,
+      "loss": 0.8433,
       "step": 8655
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 37.456077575683594,
-      "learning_rate": 2.6500952190037087e-06,
-      "loss": 1.5839,
+      "epoch": 1.09,
+      "grad_norm": 12.546215057373047,
+      "learning_rate": 1.2758231184370164e-05,
+      "loss": 0.9112,
       "step": 8656
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 189.6568145751953,
-      "learning_rate": 2.6480906083993184e-06,
-      "loss": 1.1821,
+      "epoch": 1.09,
+      "grad_norm": 1347.7606201171875,
+      "learning_rate": 1.2757394469313476e-05,
+      "loss": 1.3567,
       "step": 8657
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 14.535365104675293,
-      "learning_rate": 2.646085997794929e-06,
-      "loss": 1.221,
+      "epoch": 1.09,
+      "grad_norm": 9.610301971435547,
+      "learning_rate": 1.2756557754256788e-05,
+      "loss": 0.2851,
       "step": 8658
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 15.61431884765625,
-      "learning_rate": 2.6440813871905385e-06,
-      "loss": 0.8718,
+      "epoch": 1.09,
+      "grad_norm": 14.493852615356445,
+      "learning_rate": 1.2755721039200102e-05,
+      "loss": 2.127,
       "step": 8659
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 16.955713272094727,
-      "learning_rate": 2.642076776586148e-06,
-      "loss": 1.0111,
+      "epoch": 1.09,
+      "grad_norm": 9.423722267150879,
+      "learning_rate": 1.2754884324143415e-05,
+      "loss": 1.8575,
       "step": 8660
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 17.068771362304688,
-      "learning_rate": 2.6400721659817587e-06,
-      "loss": 1.5695,
+      "epoch": 1.09,
+      "grad_norm": 21.096939086914062,
+      "learning_rate": 1.2754047609086726e-05,
+      "loss": 2.0822,
       "step": 8661
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 10.242925643920898,
-      "learning_rate": 2.6380675553773684e-06,
-      "loss": 1.4754,
+      "epoch": 1.09,
+      "grad_norm": 14.859004974365234,
+      "learning_rate": 1.275321089403004e-05,
+      "loss": 2.9309,
       "step": 8662
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 30.505434036254883,
-      "learning_rate": 2.636062944772978e-06,
-      "loss": 1.1623,
+      "epoch": 1.09,
+      "grad_norm": 12.358412742614746,
+      "learning_rate": 1.2752374178973353e-05,
+      "loss": 0.5245,
       "step": 8663
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 44.106178283691406,
-      "learning_rate": 2.6340583341685877e-06,
-      "loss": 2.1524,
+      "epoch": 1.09,
+      "grad_norm": 20.174772262573242,
+      "learning_rate": 1.2751537463916663e-05,
+      "loss": 1.7873,
       "step": 8664
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 15.738425254821777,
-      "learning_rate": 2.6320537235641978e-06,
-      "loss": 1.3509,
+      "epoch": 1.09,
+      "grad_norm": 13.297382354736328,
+      "learning_rate": 1.2750700748859977e-05,
+      "loss": 3.0238,
       "step": 8665
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 37.473567962646484,
-      "learning_rate": 2.630049112959808e-06,
-      "loss": 1.1996,
+      "epoch": 1.09,
+      "grad_norm": 15.283510208129883,
+      "learning_rate": 1.274986403380329e-05,
+      "loss": 1.0246,
       "step": 8666
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 20.072607040405273,
-      "learning_rate": 2.6280445023554175e-06,
-      "loss": 1.1615,
+      "epoch": 1.09,
+      "grad_norm": 31.507217407226562,
+      "learning_rate": 1.2749027318746602e-05,
+      "loss": 1.2635,
       "step": 8667
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 14.636398315429688,
-      "learning_rate": 2.6260398917510276e-06,
-      "loss": 0.863,
+      "epoch": 1.09,
+      "grad_norm": 11.496175765991211,
+      "learning_rate": 1.2748190603689914e-05,
+      "loss": 2.3934,
       "step": 8668
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 9.129220008850098,
-      "learning_rate": 2.6240352811466373e-06,
-      "loss": 1.157,
+      "epoch": 1.09,
+      "grad_norm": 9.032344818115234,
+      "learning_rate": 1.2747353888633226e-05,
+      "loss": 1.3367,
       "step": 8669
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 19.694377899169922,
-      "learning_rate": 2.6220306705422473e-06,
-      "loss": 0.8906,
+      "epoch": 1.09,
+      "grad_norm": 6.9487786293029785,
+      "learning_rate": 1.274651717357654e-05,
+      "loss": 1.802,
       "step": 8670
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 13.88318157196045,
-      "learning_rate": 2.6200260599378574e-06,
-      "loss": 1.343,
+      "epoch": 1.09,
+      "grad_norm": 33.17593002319336,
+      "learning_rate": 1.2745680458519852e-05,
+      "loss": 2.1467,
       "step": 8671
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 11.803475379943848,
-      "learning_rate": 2.618021449333467e-06,
-      "loss": 1.4005,
+      "epoch": 1.09,
+      "grad_norm": 13.267254829406738,
+      "learning_rate": 1.2744843743463164e-05,
+      "loss": 1.2365,
       "step": 8672
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 18.531572341918945,
-      "learning_rate": 2.6160168387290767e-06,
-      "loss": 1.3321,
+      "epoch": 1.09,
+      "grad_norm": 11.520238876342773,
+      "learning_rate": 1.2744007028406478e-05,
+      "loss": 1.512,
       "step": 8673
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 6.691076278686523,
-      "learning_rate": 2.6140122281246872e-06,
-      "loss": 0.7632,
+      "epoch": 1.09,
+      "grad_norm": 19.45894432067871,
+      "learning_rate": 1.2743170313349791e-05,
+      "loss": 2.1284,
       "step": 8674
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 8.980220794677734,
-      "learning_rate": 2.612007617520297e-06,
-      "loss": 0.7735,
+      "epoch": 1.09,
+      "grad_norm": 14.671815872192383,
+      "learning_rate": 1.2742333598293101e-05,
+      "loss": 1.9271,
       "step": 8675
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 8.376172065734863,
-      "learning_rate": 2.6100030069159066e-06,
-      "loss": 1.0655,
+      "epoch": 1.09,
+      "grad_norm": 11.830839157104492,
+      "learning_rate": 1.2741496883236415e-05,
+      "loss": 2.0808,
       "step": 8676
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 10.30313491821289,
-      "learning_rate": 2.6079983963115167e-06,
-      "loss": 1.6599,
+      "epoch": 1.09,
+      "grad_norm": 6.510457515716553,
+      "learning_rate": 1.2740660168179729e-05,
+      "loss": 1.049,
       "step": 8677
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 23.8471622467041,
-      "learning_rate": 2.6059937857071267e-06,
-      "loss": 0.579,
+      "epoch": 1.09,
+      "grad_norm": 11.446674346923828,
+      "learning_rate": 1.2739823453123039e-05,
+      "loss": 0.7896,
       "step": 8678
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 25.849637985229492,
-      "learning_rate": 2.6039891751027364e-06,
-      "loss": 1.8009,
+      "epoch": 1.09,
+      "grad_norm": 9.457968711853027,
+      "learning_rate": 1.2738986738066353e-05,
+      "loss": 1.7783,
       "step": 8679
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 26.335845947265625,
-      "learning_rate": 2.6019845644983465e-06,
-      "loss": 1.6794,
+      "epoch": 1.09,
+      "grad_norm": 13.267879486083984,
+      "learning_rate": 1.2738150023009666e-05,
+      "loss": 1.7797,
       "step": 8680
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 21.52005386352539,
-      "learning_rate": 2.599979953893956e-06,
-      "loss": 1.3504,
+      "epoch": 1.09,
+      "grad_norm": 10.053738594055176,
+      "learning_rate": 1.2737313307952978e-05,
+      "loss": 1.3612,
       "step": 8681
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 24.20670509338379,
-      "learning_rate": 2.5979753432895662e-06,
-      "loss": 0.9255,
+      "epoch": 1.09,
+      "grad_norm": 10.973382949829102,
+      "learning_rate": 1.273647659289629e-05,
+      "loss": 1.8119,
       "step": 8682
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 105.50646209716797,
-      "learning_rate": 2.5959707326851763e-06,
-      "loss": 2.7274,
+      "epoch": 1.09,
+      "grad_norm": 19.048812866210938,
+      "learning_rate": 1.2735639877839602e-05,
+      "loss": 1.8454,
       "step": 8683
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 28.476221084594727,
-      "learning_rate": 2.593966122080786e-06,
-      "loss": 2.7293,
+      "epoch": 1.09,
+      "grad_norm": 70.28934478759766,
+      "learning_rate": 1.2734803162782916e-05,
+      "loss": 0.5382,
       "step": 8684
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 23.278079986572266,
-      "learning_rate": 2.5919615114763956e-06,
-      "loss": 2.1648,
+      "epoch": 1.09,
+      "grad_norm": 30.25827980041504,
+      "learning_rate": 1.2733966447726228e-05,
+      "loss": 2.7756,
       "step": 8685
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 25.680274963378906,
-      "learning_rate": 2.589956900872006e-06,
-      "loss": 1.5944,
+      "epoch": 1.09,
+      "grad_norm": 24.014326095581055,
+      "learning_rate": 1.273312973266954e-05,
+      "loss": 0.9742,
       "step": 8686
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 16.657413482666016,
-      "learning_rate": 2.587952290267616e-06,
-      "loss": 0.8337,
+      "epoch": 1.09,
+      "grad_norm": 6.420015335083008,
+      "learning_rate": 1.2732293017612853e-05,
+      "loss": 1.4453,
       "step": 8687
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 24.16915512084961,
-      "learning_rate": 2.5859476796632255e-06,
-      "loss": 1.168,
+      "epoch": 1.09,
+      "grad_norm": 8.522665977478027,
+      "learning_rate": 1.2731456302556167e-05,
+      "loss": 1.9186,
       "step": 8688
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 13.357244491577148,
-      "learning_rate": 2.5839430690588355e-06,
-      "loss": 0.9006,
+      "epoch": 1.09,
+      "grad_norm": 18.01955223083496,
+      "learning_rate": 1.2730619587499477e-05,
+      "loss": 1.371,
       "step": 8689
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 17.03959083557129,
-      "learning_rate": 2.5819384584544456e-06,
-      "loss": 1.3754,
+      "epoch": 1.09,
+      "grad_norm": 11.156051635742188,
+      "learning_rate": 1.2729782872442791e-05,
+      "loss": 0.5492,
       "step": 8690
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 14.713468551635742,
-      "learning_rate": 2.5799338478500553e-06,
-      "loss": 1.3571,
+      "epoch": 1.09,
+      "grad_norm": 15.096097946166992,
+      "learning_rate": 1.2728946157386105e-05,
+      "loss": 1.4618,
       "step": 8691
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 48.19596481323242,
-      "learning_rate": 2.5779292372456654e-06,
-      "loss": 1.8712,
+      "epoch": 1.09,
+      "grad_norm": 15.180127143859863,
+      "learning_rate": 1.2728109442329415e-05,
+      "loss": 1.2357,
       "step": 8692
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 42.44739532470703,
-      "learning_rate": 2.575924626641275e-06,
-      "loss": 1.2746,
+      "epoch": 1.09,
+      "grad_norm": 4.636838436126709,
+      "learning_rate": 1.2727272727272728e-05,
+      "loss": 0.4605,
       "step": 8693
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 35.6419792175293,
-      "learning_rate": 2.5739200160368847e-06,
-      "loss": 1.5702,
+      "epoch": 1.09,
+      "grad_norm": 24.56222152709961,
+      "learning_rate": 1.2726436012216042e-05,
+      "loss": 2.0397,
       "step": 8694
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 19.48603630065918,
-      "learning_rate": 2.571915405432495e-06,
-      "loss": 1.9149,
+      "epoch": 1.09,
+      "grad_norm": 8.664571762084961,
+      "learning_rate": 1.2725599297159354e-05,
+      "loss": 1.411,
       "step": 8695
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 15.384188652038574,
-      "learning_rate": 2.569910794828105e-06,
-      "loss": 1.4403,
+      "epoch": 1.09,
+      "grad_norm": 9.557058334350586,
+      "learning_rate": 1.2724762582102666e-05,
+      "loss": 1.119,
       "step": 8696
     },
     {
-      "epoch": 2.61,
-      "grad_norm": 18.034732818603516,
-      "learning_rate": 2.5679061842237145e-06,
-      "loss": 1.2765,
+      "epoch": 1.09,
+      "grad_norm": 24.020355224609375,
+      "learning_rate": 1.2723925867045978e-05,
+      "loss": 3.312,
       "step": 8697
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 12.634345054626465,
-      "learning_rate": 2.565901573619325e-06,
-      "loss": 1.0044,
+      "epoch": 1.09,
+      "grad_norm": 40.546783447265625,
+      "learning_rate": 1.2723089151989292e-05,
+      "loss": 1.64,
       "step": 8698
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 12.71899127960205,
-      "learning_rate": 2.5638969630149347e-06,
-      "loss": 1.3239,
+      "epoch": 1.09,
+      "grad_norm": 7.633327960968018,
+      "learning_rate": 1.2722252436932604e-05,
+      "loss": 0.794,
       "step": 8699
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 597.7841186523438,
-      "learning_rate": 2.5618923524105444e-06,
-      "loss": 1.6701,
+      "epoch": 1.09,
+      "grad_norm": 21.85679817199707,
+      "learning_rate": 1.2721415721875916e-05,
+      "loss": 1.4378,
       "step": 8700
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 16.543411254882812,
-      "learning_rate": 2.5598877418061544e-06,
-      "loss": 1.4786,
+      "epoch": 1.09,
+      "grad_norm": 19.40522575378418,
+      "learning_rate": 1.272057900681923e-05,
+      "loss": 2.0973,
       "step": 8701
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 48.39625930786133,
-      "learning_rate": 2.5578831312017645e-06,
-      "loss": 1.538,
+      "epoch": 1.09,
+      "grad_norm": 22.19701385498047,
+      "learning_rate": 1.2719742291762543e-05,
+      "loss": 1.544,
       "step": 8702
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 10.015215873718262,
-      "learning_rate": 2.555878520597374e-06,
-      "loss": 1.4655,
+      "epoch": 1.09,
+      "grad_norm": 6.121273040771484,
+      "learning_rate": 1.2718905576705853e-05,
+      "loss": 0.5231,
       "step": 8703
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 18.111173629760742,
-      "learning_rate": 2.5538739099929843e-06,
-      "loss": 1.539,
+      "epoch": 1.09,
+      "grad_norm": 61.69236373901367,
+      "learning_rate": 1.2718068861649167e-05,
+      "loss": 1.9814,
       "step": 8704
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 34.77145767211914,
-      "learning_rate": 2.551869299388594e-06,
-      "loss": 1.5545,
+      "epoch": 1.09,
+      "grad_norm": 7.5420122146606445,
+      "learning_rate": 1.271723214659248e-05,
+      "loss": 0.546,
       "step": 8705
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 17.934062957763672,
-      "learning_rate": 2.5498646887842036e-06,
-      "loss": 1.3262,
+      "epoch": 1.09,
+      "grad_norm": 8.4906005859375,
+      "learning_rate": 1.271639543153579e-05,
+      "loss": 0.6149,
       "step": 8706
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 14.223177909851074,
-      "learning_rate": 2.5478600781798137e-06,
-      "loss": 2.0243,
+      "epoch": 1.09,
+      "grad_norm": 20.686519622802734,
+      "learning_rate": 1.2715558716479104e-05,
+      "loss": 2.3446,
       "step": 8707
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 21.776586532592773,
-      "learning_rate": 2.5458554675754238e-06,
-      "loss": 1.3405,
+      "epoch": 1.09,
+      "grad_norm": 61.66415786743164,
+      "learning_rate": 1.2714722001422416e-05,
+      "loss": 2.3224,
       "step": 8708
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 10.819275856018066,
-      "learning_rate": 2.5438508569710334e-06,
-      "loss": 1.488,
+      "epoch": 1.09,
+      "grad_norm": 6.78904390335083,
+      "learning_rate": 1.271388528636573e-05,
+      "loss": 0.6229,
       "step": 8709
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 12.452916145324707,
-      "learning_rate": 2.541846246366643e-06,
-      "loss": 1.3601,
+      "epoch": 1.09,
+      "grad_norm": 27.501174926757812,
+      "learning_rate": 1.2713048571309042e-05,
+      "loss": 1.8172,
       "step": 8710
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 13.645332336425781,
-      "learning_rate": 2.5398416357622536e-06,
-      "loss": 1.2305,
+      "epoch": 1.09,
+      "grad_norm": 16.328968048095703,
+      "learning_rate": 1.2712211856252354e-05,
+      "loss": 1.8874,
       "step": 8711
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 157.49203491210938,
-      "learning_rate": 2.5378370251578633e-06,
-      "loss": 1.5407,
+      "epoch": 1.09,
+      "grad_norm": 6.196768760681152,
+      "learning_rate": 1.2711375141195667e-05,
+      "loss": 0.6491,
       "step": 8712
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 15.182050704956055,
-      "learning_rate": 2.535832414553473e-06,
-      "loss": 0.874,
+      "epoch": 1.09,
+      "grad_norm": 9.383679389953613,
+      "learning_rate": 1.271053842613898e-05,
+      "loss": 1.6176,
       "step": 8713
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 16.068603515625,
-      "learning_rate": 2.5338278039490834e-06,
-      "loss": 1.3392,
+      "epoch": 1.09,
+      "grad_norm": 12.209012985229492,
+      "learning_rate": 1.2709701711082291e-05,
+      "loss": 1.4438,
       "step": 8714
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 11.687101364135742,
-      "learning_rate": 2.531823193344693e-06,
-      "loss": 0.9513,
+      "epoch": 1.09,
+      "grad_norm": 16.936748504638672,
+      "learning_rate": 1.2708864996025605e-05,
+      "loss": 1.1826,
       "step": 8715
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 16.209623336791992,
-      "learning_rate": 2.5298185827403027e-06,
-      "loss": 0.8897,
+      "epoch": 1.09,
+      "grad_norm": 18.694360733032227,
+      "learning_rate": 1.2708028280968919e-05,
+      "loss": 1.9717,
       "step": 8716
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 42.321136474609375,
-      "learning_rate": 2.527813972135913e-06,
-      "loss": 2.2684,
+      "epoch": 1.09,
+      "grad_norm": 8.575372695922852,
+      "learning_rate": 1.2707191565912229e-05,
+      "loss": 1.418,
       "step": 8717
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 19.06346321105957,
-      "learning_rate": 2.5258093615315225e-06,
-      "loss": 1.3514,
+      "epoch": 1.09,
+      "grad_norm": 17.063716888427734,
+      "learning_rate": 1.2706354850855543e-05,
+      "loss": 0.9269,
       "step": 8718
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 10.905227661132812,
-      "learning_rate": 2.5238047509271326e-06,
-      "loss": 0.6565,
+      "epoch": 1.09,
+      "grad_norm": 12.043959617614746,
+      "learning_rate": 1.2705518135798856e-05,
+      "loss": 1.7686,
       "step": 8719
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 17.082691192626953,
-      "learning_rate": 2.5218001403227427e-06,
-      "loss": 2.4891,
+      "epoch": 1.09,
+      "grad_norm": 34.632266998291016,
+      "learning_rate": 1.2704681420742166e-05,
+      "loss": 1.5823,
       "step": 8720
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 22.291873931884766,
-      "learning_rate": 2.5197955297183523e-06,
-      "loss": 2.0741,
+      "epoch": 1.09,
+      "grad_norm": 5.508233547210693,
+      "learning_rate": 1.270384470568548e-05,
+      "loss": 1.2093,
       "step": 8721
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 9.077491760253906,
-      "learning_rate": 2.517790919113962e-06,
-      "loss": 1.115,
+      "epoch": 1.09,
+      "grad_norm": 4.57993745803833,
+      "learning_rate": 1.2703007990628792e-05,
+      "loss": 1.6767,
       "step": 8722
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 10.916313171386719,
-      "learning_rate": 2.5157863085095725e-06,
-      "loss": 1.4171,
+      "epoch": 1.09,
+      "grad_norm": 15.940275192260742,
+      "learning_rate": 1.2702171275572106e-05,
+      "loss": 0.6244,
       "step": 8723
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 54.628211975097656,
-      "learning_rate": 2.513781697905182e-06,
-      "loss": 0.9545,
+      "epoch": 1.09,
+      "grad_norm": 8.179159164428711,
+      "learning_rate": 1.2701334560515418e-05,
+      "loss": 0.5239,
       "step": 8724
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 11.69243049621582,
-      "learning_rate": 2.511777087300792e-06,
-      "loss": 0.7757,
+      "epoch": 1.09,
+      "grad_norm": 8.999481201171875,
+      "learning_rate": 1.270049784545873e-05,
+      "loss": 1.9757,
       "step": 8725
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 23.144771575927734,
-      "learning_rate": 2.5097724766964023e-06,
-      "loss": 1.4352,
+      "epoch": 1.1,
+      "grad_norm": 9.038658142089844,
+      "learning_rate": 1.2699661130402043e-05,
+      "loss": 1.1503,
       "step": 8726
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 23.521812438964844,
-      "learning_rate": 2.507767866092012e-06,
-      "loss": 1.276,
+      "epoch": 1.1,
+      "grad_norm": 20.042102813720703,
+      "learning_rate": 1.2698824415345354e-05,
+      "loss": 2.745,
       "step": 8727
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 29.313697814941406,
-      "learning_rate": 2.5057632554876216e-06,
-      "loss": 1.4241,
+      "epoch": 1.1,
+      "grad_norm": 40.724525451660156,
+      "learning_rate": 1.2697987700288667e-05,
+      "loss": 1.8879,
       "step": 8728
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 5.5264716148376465,
-      "learning_rate": 2.5037586448832317e-06,
-      "loss": 0.6715,
+      "epoch": 1.1,
+      "grad_norm": 22.703800201416016,
+      "learning_rate": 1.2697150985231981e-05,
+      "loss": 1.8462,
       "step": 8729
     },
     {
-      "epoch": 2.62,
-      "grad_norm": 107.75334167480469,
-      "learning_rate": 2.5017540342788414e-06,
-      "loss": 2.3175,
+      "epoch": 1.1,
+      "grad_norm": 13.04196834564209,
+      "learning_rate": 1.2696314270175295e-05,
+      "loss": 1.4649,
       "step": 8730
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 16.5670108795166,
-      "learning_rate": 2.4997494236744515e-06,
-      "loss": 1.7357,
+      "epoch": 1.1,
+      "grad_norm": 6.677236080169678,
+      "learning_rate": 1.2695477555118605e-05,
+      "loss": 0.3258,
       "step": 8731
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 25.61155128479004,
-      "learning_rate": 2.497744813070061e-06,
-      "loss": 1.3919,
+      "epoch": 1.1,
+      "grad_norm": 9.752840995788574,
+      "learning_rate": 1.2694640840061918e-05,
+      "loss": 0.7642,
       "step": 8732
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 12.228818893432617,
-      "learning_rate": 2.495740202465671e-06,
-      "loss": 2.4026,
+      "epoch": 1.1,
+      "grad_norm": 7.2943196296691895,
+      "learning_rate": 1.2693804125005232e-05,
+      "loss": 0.4322,
       "step": 8733
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 16.512880325317383,
-      "learning_rate": 2.4937355918612813e-06,
-      "loss": 1.9608,
+      "epoch": 1.1,
+      "grad_norm": 21.28787612915039,
+      "learning_rate": 1.2692967409948542e-05,
+      "loss": 1.2366,
       "step": 8734
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 34.81612014770508,
-      "learning_rate": 2.491730981256891e-06,
-      "loss": 1.0885,
+      "epoch": 1.1,
+      "grad_norm": 113.95024108886719,
+      "learning_rate": 1.2692130694891856e-05,
+      "loss": 0.6453,
       "step": 8735
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 31.149940490722656,
-      "learning_rate": 2.489726370652501e-06,
-      "loss": 1.0149,
+      "epoch": 1.1,
+      "grad_norm": 17.61992835998535,
+      "learning_rate": 1.2691293979835168e-05,
+      "loss": 3.0286,
       "step": 8736
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 13.102104187011719,
-      "learning_rate": 2.4877217600481107e-06,
-      "loss": 1.5993,
+      "epoch": 1.1,
+      "grad_norm": 12.428811073303223,
+      "learning_rate": 1.2690457264778482e-05,
+      "loss": 2.1907,
       "step": 8737
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 18.59217071533203,
-      "learning_rate": 2.4857171494437208e-06,
-      "loss": 1.3812,
+      "epoch": 1.1,
+      "grad_norm": 10.521194458007812,
+      "learning_rate": 1.2689620549721794e-05,
+      "loss": 0.8512,
       "step": 8738
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 15.023920059204102,
-      "learning_rate": 2.483712538839331e-06,
-      "loss": 1.0019,
+      "epoch": 1.1,
+      "grad_norm": 17.262094497680664,
+      "learning_rate": 1.2688783834665105e-05,
+      "loss": 1.7099,
       "step": 8739
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 15.331181526184082,
-      "learning_rate": 2.4817079282349405e-06,
-      "loss": 1.5826,
+      "epoch": 1.1,
+      "grad_norm": 8.695687294006348,
+      "learning_rate": 1.2687947119608419e-05,
+      "loss": 0.7872,
       "step": 8740
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 32.79924774169922,
-      "learning_rate": 2.4797033176305506e-06,
-      "loss": 2.171,
+      "epoch": 1.1,
+      "grad_norm": 4.679690837860107,
+      "learning_rate": 1.268711040455173e-05,
+      "loss": 0.223,
       "step": 8741
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 11.717575073242188,
-      "learning_rate": 2.4776987070261603e-06,
-      "loss": 0.7835,
+      "epoch": 1.1,
+      "grad_norm": 7.748942852020264,
+      "learning_rate": 1.2686273689495043e-05,
+      "loss": 0.763,
       "step": 8742
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 19.555763244628906,
-      "learning_rate": 2.4756940964217704e-06,
-      "loss": 2.7394,
+      "epoch": 1.1,
+      "grad_norm": 10.922325134277344,
+      "learning_rate": 1.2685436974438357e-05,
+      "loss": 2.3854,
       "step": 8743
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 37.40543746948242,
-      "learning_rate": 2.47368948581738e-06,
-      "loss": 2.1193,
+      "epoch": 1.1,
+      "grad_norm": 11.161537170410156,
+      "learning_rate": 1.2684600259381667e-05,
+      "loss": 1.1113,
       "step": 8744
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 18.472028732299805,
-      "learning_rate": 2.47168487521299e-06,
-      "loss": 1.0426,
+      "epoch": 1.1,
+      "grad_norm": 16.115772247314453,
+      "learning_rate": 1.268376354432498e-05,
+      "loss": 1.7661,
       "step": 8745
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 15.350752830505371,
-      "learning_rate": 2.4696802646085998e-06,
-      "loss": 1.0137,
+      "epoch": 1.1,
+      "grad_norm": 14.176069259643555,
+      "learning_rate": 1.2682926829268294e-05,
+      "loss": 0.4978,
       "step": 8746
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 12.595955848693848,
-      "learning_rate": 2.46767565400421e-06,
-      "loss": 1.6324,
+      "epoch": 1.1,
+      "grad_norm": 8.118674278259277,
+      "learning_rate": 1.2682090114211608e-05,
+      "loss": 1.0084,
       "step": 8747
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 29.87061309814453,
-      "learning_rate": 2.46567104339982e-06,
-      "loss": 1.5554,
+      "epoch": 1.1,
+      "grad_norm": 24.346820831298828,
+      "learning_rate": 1.2681253399154918e-05,
+      "loss": 1.9808,
       "step": 8748
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 17.269176483154297,
-      "learning_rate": 2.4636664327954296e-06,
-      "loss": 1.6175,
+      "epoch": 1.1,
+      "grad_norm": 33.79582977294922,
+      "learning_rate": 1.2680416684098232e-05,
+      "loss": 1.0062,
       "step": 8749
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 16.87700653076172,
-      "learning_rate": 2.4616618221910397e-06,
-      "loss": 1.0128,
+      "epoch": 1.1,
+      "grad_norm": 17.37476921081543,
+      "learning_rate": 1.2679579969041544e-05,
+      "loss": 1.4613,
       "step": 8750
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 10.00948429107666,
-      "learning_rate": 2.4596572115866498e-06,
-      "loss": 0.9691,
+      "epoch": 1.1,
+      "grad_norm": 8.64911937713623,
+      "learning_rate": 1.2678743253984856e-05,
+      "loss": 1.0513,
       "step": 8751
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 21.35502815246582,
-      "learning_rate": 2.4576526009822594e-06,
-      "loss": 1.3124,
+      "epoch": 1.1,
+      "grad_norm": 12.104917526245117,
+      "learning_rate": 1.267790653892817e-05,
+      "loss": 1.5139,
       "step": 8752
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 11.689464569091797,
-      "learning_rate": 2.4556479903778695e-06,
-      "loss": 1.3952,
+      "epoch": 1.1,
+      "grad_norm": 27.262914657592773,
+      "learning_rate": 1.2677069823871481e-05,
+      "loss": 1.7941,
       "step": 8753
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 141.92120361328125,
-      "learning_rate": 2.453643379773479e-06,
-      "loss": 2.629,
+      "epoch": 1.1,
+      "grad_norm": 13.147500991821289,
+      "learning_rate": 1.2676233108814795e-05,
+      "loss": 1.0754,
       "step": 8754
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 26.330507278442383,
-      "learning_rate": 2.4516387691690893e-06,
-      "loss": 3.4614,
+      "epoch": 1.1,
+      "grad_norm": 8.71116828918457,
+      "learning_rate": 1.2675396393758105e-05,
+      "loss": 0.5459,
       "step": 8755
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 10.556838035583496,
-      "learning_rate": 2.449634158564699e-06,
-      "loss": 1.7251,
+      "epoch": 1.1,
+      "grad_norm": 63.87320327758789,
+      "learning_rate": 1.2674559678701419e-05,
+      "loss": 1.7561,
       "step": 8756
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 45.57450485229492,
-      "learning_rate": 2.4476295479603086e-06,
-      "loss": 1.7645,
+      "epoch": 1.1,
+      "grad_norm": 5.747971057891846,
+      "learning_rate": 1.2673722963644733e-05,
+      "loss": 0.4531,
       "step": 8757
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 11.817948341369629,
-      "learning_rate": 2.4456249373559187e-06,
-      "loss": 1.0033,
+      "epoch": 1.1,
+      "grad_norm": 11.756169319152832,
+      "learning_rate": 1.2672886248588043e-05,
+      "loss": 0.7117,
       "step": 8758
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 7.083280563354492,
-      "learning_rate": 2.4436203267515287e-06,
-      "loss": 0.7292,
+      "epoch": 1.1,
+      "grad_norm": 71.47728729248047,
+      "learning_rate": 1.2672049533531356e-05,
+      "loss": 2.1309,
       "step": 8759
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 18.485204696655273,
-      "learning_rate": 2.4416157161471384e-06,
-      "loss": 1.5034,
-      "step": 8760
-    },
-    {
-      "epoch": 2.63,
-      "eval_loss": 0.163481205701828,
-      "eval_runtime": 43.3055,
-      "eval_samples_per_second": 34.153,
-      "eval_steps_per_second": 34.153,
+      "epoch": 1.1,
+      "grad_norm": 5.626829624176025,
+      "learning_rate": 1.267121281847467e-05,
+      "loss": 0.6731,
       "step": 8760
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 21.252248764038086,
-      "learning_rate": 2.4396111055427485e-06,
-      "loss": 1.1208,
+      "epoch": 1.1,
+      "grad_norm": 11.510431289672852,
+      "learning_rate": 1.2670376103417982e-05,
+      "loss": 0.9846,
       "step": 8761
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 16.638259887695312,
-      "learning_rate": 2.4376064949383586e-06,
-      "loss": 1.8795,
+      "epoch": 1.1,
+      "grad_norm": 125.09648895263672,
+      "learning_rate": 1.2669539388361294e-05,
+      "loss": 0.6987,
       "step": 8762
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 20.396753311157227,
-      "learning_rate": 2.4356018843339682e-06,
-      "loss": 1.5712,
+      "epoch": 1.1,
+      "grad_norm": 12.258779525756836,
+      "learning_rate": 1.2668702673304608e-05,
+      "loss": 2.082,
       "step": 8763
     },
     {
-      "epoch": 2.63,
-      "grad_norm": 7.8299455642700195,
-      "learning_rate": 2.4335972737295783e-06,
-      "loss": 0.7808,
+      "epoch": 1.1,
+      "grad_norm": 13.310866355895996,
+      "learning_rate": 1.266786595824792e-05,
+      "loss": 1.4755,
       "step": 8764
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 11.673842430114746,
-      "learning_rate": 2.4315926631251884e-06,
-      "loss": 0.8884,
+      "epoch": 1.1,
+      "grad_norm": 7.459881782531738,
+      "learning_rate": 1.2667029243191232e-05,
+      "loss": 0.6761,
       "step": 8765
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 40.24734115600586,
-      "learning_rate": 2.429588052520798e-06,
-      "loss": 2.0446,
+      "epoch": 1.1,
+      "grad_norm": 18.733461380004883,
+      "learning_rate": 1.2666192528134544e-05,
+      "loss": 0.7805,
       "step": 8766
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 29.379608154296875,
-      "learning_rate": 2.4275834419164077e-06,
-      "loss": 1.6934,
+      "epoch": 1.1,
+      "grad_norm": 4.180849075317383,
+      "learning_rate": 1.2665355813077857e-05,
+      "loss": 0.3807,
       "step": 8767
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 66.81684112548828,
-      "learning_rate": 2.425578831312018e-06,
-      "loss": 1.426,
+      "epoch": 1.1,
+      "grad_norm": 10.724422454833984,
+      "learning_rate": 1.266451909802117e-05,
+      "loss": 2.1114,
       "step": 8768
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 81.16704559326172,
-      "learning_rate": 2.4235742207076275e-06,
-      "loss": 1.3078,
+      "epoch": 1.1,
+      "grad_norm": 8.880762100219727,
+      "learning_rate": 1.2663682382964481e-05,
+      "loss": 1.4725,
       "step": 8769
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 13.74372386932373,
-      "learning_rate": 2.4215696101032376e-06,
-      "loss": 1.543,
+      "epoch": 1.1,
+      "grad_norm": 10.20962142944336,
+      "learning_rate": 1.2662845667907795e-05,
+      "loss": 1.715,
       "step": 8770
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 9.779702186584473,
-      "learning_rate": 2.4195649994988476e-06,
-      "loss": 1.0268,
+      "epoch": 1.1,
+      "grad_norm": 30.03700065612793,
+      "learning_rate": 1.2662008952851108e-05,
+      "loss": 1.6719,
       "step": 8771
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 14.754984855651855,
-      "learning_rate": 2.4175603888944573e-06,
-      "loss": 1.7506,
+      "epoch": 1.1,
+      "grad_norm": 33.68233108520508,
+      "learning_rate": 1.2661172237794419e-05,
+      "loss": 2.2797,
       "step": 8772
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 29.609939575195312,
-      "learning_rate": 2.4155557782900674e-06,
-      "loss": 2.8884,
+      "epoch": 1.1,
+      "grad_norm": 14.828858375549316,
+      "learning_rate": 1.2660335522737732e-05,
+      "loss": 0.8591,
       "step": 8773
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 100.57302856445312,
-      "learning_rate": 2.4135511676856775e-06,
-      "loss": 1.1484,
+      "epoch": 1.1,
+      "grad_norm": 7.947477340698242,
+      "learning_rate": 1.2659498807681046e-05,
+      "loss": 0.9196,
       "step": 8774
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 29.211971282958984,
-      "learning_rate": 2.411546557081287e-06,
-      "loss": 2.1784,
+      "epoch": 1.1,
+      "grad_norm": 9.227246284484863,
+      "learning_rate": 1.2658662092624358e-05,
+      "loss": 2.4799,
       "step": 8775
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 31.128265380859375,
-      "learning_rate": 2.409541946476897e-06,
-      "loss": 1.6176,
+      "epoch": 1.1,
+      "grad_norm": 5.3799309730529785,
+      "learning_rate": 1.265782537756767e-05,
+      "loss": 1.7289,
       "step": 8776
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 16.750701904296875,
-      "learning_rate": 2.4075373358725073e-06,
-      "loss": 0.8013,
+      "epoch": 1.1,
+      "grad_norm": 3.43022084236145,
+      "learning_rate": 1.2656988662510983e-05,
+      "loss": 0.0955,
       "step": 8777
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 65.17308044433594,
-      "learning_rate": 2.405532725268117e-06,
-      "loss": 1.9283,
+      "epoch": 1.1,
+      "grad_norm": 15.954379081726074,
+      "learning_rate": 1.2656151947454295e-05,
+      "loss": 0.8652,
       "step": 8778
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 18.673601150512695,
-      "learning_rate": 2.4035281146637266e-06,
-      "loss": 1.3132,
+      "epoch": 1.1,
+      "grad_norm": 9.277589797973633,
+      "learning_rate": 1.2655315232397607e-05,
+      "loss": 2.1941,
       "step": 8779
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 12.052127838134766,
-      "learning_rate": 2.4015235040593367e-06,
-      "loss": 0.8893,
+      "epoch": 1.1,
+      "grad_norm": 25.019664764404297,
+      "learning_rate": 1.265447851734092e-05,
+      "loss": 0.8344,
       "step": 8780
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 33.23358917236328,
-      "learning_rate": 2.3995188934549464e-06,
-      "loss": 1.5921,
+      "epoch": 1.1,
+      "grad_norm": 10.636250495910645,
+      "learning_rate": 1.2653641802284233e-05,
+      "loss": 1.0796,
       "step": 8781
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 32.95631408691406,
-      "learning_rate": 2.3975142828505564e-06,
-      "loss": 1.5091,
+      "epoch": 1.1,
+      "grad_norm": 39.38780212402344,
+      "learning_rate": 1.2652805087227547e-05,
+      "loss": 1.5636,
       "step": 8782
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 21.939970016479492,
-      "learning_rate": 2.395509672246166e-06,
-      "loss": 1.1536,
+      "epoch": 1.1,
+      "grad_norm": 7.546035289764404,
+      "learning_rate": 1.2651968372170857e-05,
+      "loss": 0.5325,
       "step": 8783
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 19.46248435974121,
-      "learning_rate": 2.393505061641776e-06,
-      "loss": 1.7254,
+      "epoch": 1.1,
+      "grad_norm": 7.118682861328125,
+      "learning_rate": 1.265113165711417e-05,
+      "loss": 1.7248,
       "step": 8784
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 20.779407501220703,
-      "learning_rate": 2.3915004510373863e-06,
-      "loss": 0.6734,
+      "epoch": 1.1,
+      "grad_norm": 26.231958389282227,
+      "learning_rate": 1.2650294942057484e-05,
+      "loss": 2.1657,
       "step": 8785
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 18.915145874023438,
-      "learning_rate": 2.389495840432996e-06,
-      "loss": 1.4362,
+      "epoch": 1.1,
+      "grad_norm": 7.644165515899658,
+      "learning_rate": 1.2649458227000794e-05,
+      "loss": 0.5792,
       "step": 8786
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 15.002533912658691,
-      "learning_rate": 2.387491229828606e-06,
-      "loss": 1.3399,
+      "epoch": 1.1,
+      "grad_norm": 10.340425491333008,
+      "learning_rate": 1.2648621511944108e-05,
+      "loss": 1.8263,
       "step": 8787
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 205.16641235351562,
-      "learning_rate": 2.385486619224216e-06,
-      "loss": 1.7637,
+      "epoch": 1.1,
+      "grad_norm": 45.51021957397461,
+      "learning_rate": 1.2647784796887422e-05,
+      "loss": 1.2768,
       "step": 8788
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 16.24716567993164,
-      "learning_rate": 2.3834820086198258e-06,
-      "loss": 2.3097,
+      "epoch": 1.1,
+      "grad_norm": 20.326396942138672,
+      "learning_rate": 1.2646948081830734e-05,
+      "loss": 0.5046,
       "step": 8789
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 27.199005126953125,
-      "learning_rate": 2.381477398015436e-06,
-      "loss": 2.0923,
+      "epoch": 1.1,
+      "grad_norm": 14.142060279846191,
+      "learning_rate": 1.2646111366774046e-05,
+      "loss": 1.5594,
       "step": 8790
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 7.843896865844727,
-      "learning_rate": 2.3794727874110455e-06,
-      "loss": 0.556,
+      "epoch": 1.1,
+      "grad_norm": 7.47965145111084,
+      "learning_rate": 1.264527465171736e-05,
+      "loss": 0.8239,
       "step": 8791
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 15.323102951049805,
-      "learning_rate": 2.3774681768066556e-06,
-      "loss": 0.9552,
+      "epoch": 1.1,
+      "grad_norm": 16.085491180419922,
+      "learning_rate": 1.2644437936660671e-05,
+      "loss": 1.5384,
       "step": 8792
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 14.061708450317383,
-      "learning_rate": 2.3754635662022653e-06,
-      "loss": 1.4372,
+      "epoch": 1.1,
+      "grad_norm": 16.351972579956055,
+      "learning_rate": 1.2643601221603983e-05,
+      "loss": 0.9503,
       "step": 8793
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 12.933516502380371,
-      "learning_rate": 2.3734589555978753e-06,
-      "loss": 1.2338,
+      "epoch": 1.1,
+      "grad_norm": 12.335143089294434,
+      "learning_rate": 1.2642764506547295e-05,
+      "loss": 1.9275,
       "step": 8794
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 7.49314546585083,
-      "learning_rate": 2.371454344993485e-06,
-      "loss": 0.731,
+      "epoch": 1.1,
+      "grad_norm": 10.822340965270996,
+      "learning_rate": 1.2641927791490609e-05,
+      "loss": 0.9195,
       "step": 8795
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 15.913431167602539,
-      "learning_rate": 2.369449734389095e-06,
-      "loss": 1.2261,
+      "epoch": 1.1,
+      "grad_norm": 12.354034423828125,
+      "learning_rate": 1.2641091076433922e-05,
+      "loss": 2.6532,
       "step": 8796
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 12.59531307220459,
-      "learning_rate": 2.367445123784705e-06,
-      "loss": 1.0611,
+      "epoch": 1.1,
+      "grad_norm": 8.366477966308594,
+      "learning_rate": 1.2640254361377233e-05,
+      "loss": 0.6206,
       "step": 8797
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 16.93478012084961,
-      "learning_rate": 2.365440513180315e-06,
-      "loss": 1.6995,
+      "epoch": 1.1,
+      "grad_norm": 15.607980728149414,
+      "learning_rate": 1.2639417646320546e-05,
+      "loss": 1.3698,
       "step": 8798
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 11.499807357788086,
-      "learning_rate": 2.363435902575925e-06,
-      "loss": 1.3401,
+      "epoch": 1.1,
+      "grad_norm": 11.297418594360352,
+      "learning_rate": 1.263858093126386e-05,
+      "loss": 0.9642,
       "step": 8799
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 33.20011901855469,
-      "learning_rate": 2.3614312919715346e-06,
-      "loss": 2.086,
+      "epoch": 1.1,
+      "grad_norm": 14.777867317199707,
+      "learning_rate": 1.263774421620717e-05,
+      "loss": 2.1886,
       "step": 8800
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 10.58979606628418,
-      "learning_rate": 2.3594266813671447e-06,
-      "loss": 1.1518,
+      "epoch": 1.1,
+      "eval_loss": 0.08400662243366241,
+      "eval_runtime": 94.1728,
+      "eval_samples_per_second": 37.612,
+      "eval_steps_per_second": 37.612,
+      "step": 8800
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 76.77694702148438,
+      "learning_rate": 1.2636907501150484e-05,
+      "loss": 2.4871,
       "step": 8801
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 15.411652565002441,
-      "learning_rate": 2.3574220707627547e-06,
-      "loss": 0.9793,
+      "epoch": 1.1,
+      "grad_norm": 13.288673400878906,
+      "learning_rate": 1.2636070786093798e-05,
+      "loss": 0.84,
       "step": 8802
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 68.01510620117188,
-      "learning_rate": 2.3554174601583644e-06,
-      "loss": 2.0531,
+      "epoch": 1.1,
+      "grad_norm": 26.608503341674805,
+      "learning_rate": 1.263523407103711e-05,
+      "loss": 1.9318,
       "step": 8803
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 56.63874053955078,
-      "learning_rate": 2.3534128495539745e-06,
-      "loss": 1.1724,
+      "epoch": 1.1,
+      "grad_norm": 15.085040092468262,
+      "learning_rate": 1.2634397355980422e-05,
+      "loss": 1.5832,
       "step": 8804
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 60.3148193359375,
-      "learning_rate": 2.351408238949584e-06,
-      "loss": 2.7668,
+      "epoch": 1.11,
+      "grad_norm": 15.15543270111084,
+      "learning_rate": 1.2633560640923735e-05,
+      "loss": 2.2115,
       "step": 8805
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 38.45302963256836,
-      "learning_rate": 2.3494036283451942e-06,
-      "loss": 1.0094,
+      "epoch": 1.11,
+      "grad_norm": 32.635929107666016,
+      "learning_rate": 1.2632723925867047e-05,
+      "loss": 2.8508,
       "step": 8806
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 10.030008316040039,
-      "learning_rate": 2.347399017740804e-06,
-      "loss": 1.0792,
+      "epoch": 1.11,
+      "grad_norm": 6.457747936248779,
+      "learning_rate": 1.2631887210810359e-05,
+      "loss": 0.6168,
       "step": 8807
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 10.08592700958252,
-      "learning_rate": 2.345394407136414e-06,
-      "loss": 0.7964,
+      "epoch": 1.11,
+      "grad_norm": 6.022579193115234,
+      "learning_rate": 1.2631050495753671e-05,
+      "loss": 1.2132,
       "step": 8808
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 125.09164428710938,
-      "learning_rate": 2.3433897965320236e-06,
-      "loss": 3.0581,
+      "epoch": 1.11,
+      "grad_norm": 15.227607727050781,
+      "learning_rate": 1.2630213780696985e-05,
+      "loss": 1.8502,
       "step": 8809
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 19.070087432861328,
-      "learning_rate": 2.3413851859276337e-06,
-      "loss": 0.8685,
+      "epoch": 1.11,
+      "grad_norm": 25.35992431640625,
+      "learning_rate": 1.2629377065640298e-05,
+      "loss": 3.0445,
       "step": 8810
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 10.671518325805664,
-      "learning_rate": 2.339380575323244e-06,
-      "loss": 1.1835,
+      "epoch": 1.11,
+      "grad_norm": 74.23690032958984,
+      "learning_rate": 1.2628540350583609e-05,
+      "loss": 0.9138,
       "step": 8811
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 27.613374710083008,
-      "learning_rate": 2.3373759647188535e-06,
-      "loss": 1.4339,
+      "epoch": 1.11,
+      "grad_norm": 9.95718002319336,
+      "learning_rate": 1.2627703635526922e-05,
+      "loss": 1.125,
       "step": 8812
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 18.462194442749023,
-      "learning_rate": 2.3353713541144636e-06,
-      "loss": 1.0176,
+      "epoch": 1.11,
+      "grad_norm": 24.25860595703125,
+      "learning_rate": 1.2626866920470236e-05,
+      "loss": 1.1305,
       "step": 8813
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 13.578800201416016,
-      "learning_rate": 2.3333667435100736e-06,
-      "loss": 2.1367,
+      "epoch": 1.11,
+      "grad_norm": 29.47481346130371,
+      "learning_rate": 1.2626030205413546e-05,
+      "loss": 0.4841,
       "step": 8814
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 43.0266227722168,
-      "learning_rate": 2.3313621329056833e-06,
-      "loss": 2.1957,
+      "epoch": 1.11,
+      "grad_norm": 31.069408416748047,
+      "learning_rate": 1.262519349035686e-05,
+      "loss": 3.1053,
       "step": 8815
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 40.213008880615234,
-      "learning_rate": 2.3293575223012934e-06,
-      "loss": 1.9988,
+      "epoch": 1.11,
+      "grad_norm": 17.31085777282715,
+      "learning_rate": 1.2624356775300173e-05,
+      "loss": 2.2952,
       "step": 8816
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 14.245367050170898,
-      "learning_rate": 2.327352911696903e-06,
-      "loss": 1.891,
+      "epoch": 1.11,
+      "grad_norm": 15.16616153717041,
+      "learning_rate": 1.2623520060243485e-05,
+      "loss": 1.8619,
       "step": 8817
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 13.524105072021484,
-      "learning_rate": 2.3253483010925127e-06,
-      "loss": 1.4238,
+      "epoch": 1.11,
+      "grad_norm": 52.80331802368164,
+      "learning_rate": 1.2622683345186797e-05,
+      "loss": 0.5975,
       "step": 8818
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 22.680530548095703,
-      "learning_rate": 2.323343690488123e-06,
-      "loss": 1.2083,
+      "epoch": 1.11,
+      "grad_norm": 213.8785858154297,
+      "learning_rate": 1.262184663013011e-05,
+      "loss": 0.721,
       "step": 8819
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 13.493782043457031,
-      "learning_rate": 2.3213390798837325e-06,
-      "loss": 1.1022,
+      "epoch": 1.11,
+      "grad_norm": 47.809234619140625,
+      "learning_rate": 1.2621009915073423e-05,
+      "loss": 2.5473,
       "step": 8820
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 10.675830841064453,
-      "learning_rate": 2.3193344692793425e-06,
-      "loss": 0.5942,
+      "epoch": 1.11,
+      "grad_norm": 20.05143928527832,
+      "learning_rate": 1.2620173200016735e-05,
+      "loss": 0.8577,
       "step": 8821
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 12.953835487365723,
-      "learning_rate": 2.3173298586749526e-06,
-      "loss": 1.445,
+      "epoch": 1.11,
+      "grad_norm": 58.52271270751953,
+      "learning_rate": 1.2619336484960047e-05,
+      "loss": 2.2125,
       "step": 8822
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 10.076544761657715,
-      "learning_rate": 2.3153252480705623e-06,
-      "loss": 1.5124,
+      "epoch": 1.11,
+      "grad_norm": 73.26026916503906,
+      "learning_rate": 1.261849976990336e-05,
+      "loss": 2.2222,
       "step": 8823
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 21.92123031616211,
-      "learning_rate": 2.3133206374661724e-06,
-      "loss": 1.8663,
+      "epoch": 1.11,
+      "grad_norm": 13.545487403869629,
+      "learning_rate": 1.2617663054846674e-05,
+      "loss": 1.5535,
       "step": 8824
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 25.028993606567383,
-      "learning_rate": 2.3113160268617824e-06,
-      "loss": 1.0345,
+      "epoch": 1.11,
+      "grad_norm": 11.692740440368652,
+      "learning_rate": 1.2616826339789984e-05,
+      "loss": 1.13,
       "step": 8825
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 8.76278305053711,
-      "learning_rate": 2.309311416257392e-06,
-      "loss": 0.9227,
+      "epoch": 1.11,
+      "grad_norm": 6.153340816497803,
+      "learning_rate": 1.2615989624733298e-05,
+      "loss": 0.6864,
       "step": 8826
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 7.540186405181885,
-      "learning_rate": 2.307306805653002e-06,
-      "loss": 0.9597,
+      "epoch": 1.11,
+      "grad_norm": 8.990448951721191,
+      "learning_rate": 1.2615152909676612e-05,
+      "loss": 1.2014,
       "step": 8827
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 21.144914627075195,
-      "learning_rate": 2.3053021950486123e-06,
-      "loss": 1.3642,
+      "epoch": 1.11,
+      "grad_norm": 1.3864907026290894,
+      "learning_rate": 1.2614316194619922e-05,
+      "loss": 0.0544,
       "step": 8828
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 17.2004451751709,
-      "learning_rate": 2.303297584444222e-06,
-      "loss": 0.6705,
+      "epoch": 1.11,
+      "grad_norm": 20.9836368560791,
+      "learning_rate": 1.2613479479563236e-05,
+      "loss": 1.2737,
       "step": 8829
     },
     {
-      "epoch": 2.65,
-      "grad_norm": 8.921226501464844,
-      "learning_rate": 2.3012929738398316e-06,
-      "loss": 1.096,
+      "epoch": 1.11,
+      "grad_norm": 18.755577087402344,
+      "learning_rate": 1.261264276450655e-05,
+      "loss": 1.3111,
       "step": 8830
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 13.363105773925781,
-      "learning_rate": 2.2992883632354417e-06,
-      "loss": 0.9989,
+      "epoch": 1.11,
+      "grad_norm": 26.43598747253418,
+      "learning_rate": 1.2611806049449861e-05,
+      "loss": 2.1915,
       "step": 8831
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 11.100980758666992,
-      "learning_rate": 2.2972837526310513e-06,
-      "loss": 1.4,
+      "epoch": 1.11,
+      "grad_norm": 3.893955945968628,
+      "learning_rate": 1.2610969334393173e-05,
+      "loss": 0.1865,
       "step": 8832
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 18.512340545654297,
-      "learning_rate": 2.2952791420266614e-06,
-      "loss": 1.3327,
+      "epoch": 1.11,
+      "grad_norm": 8.938155174255371,
+      "learning_rate": 1.2610132619336485e-05,
+      "loss": 0.6947,
       "step": 8833
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 52.814693450927734,
-      "learning_rate": 2.2932745314222715e-06,
-      "loss": 1.4743,
+      "epoch": 1.11,
+      "grad_norm": 15.839922904968262,
+      "learning_rate": 1.2609295904279799e-05,
+      "loss": 1.2233,
       "step": 8834
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 37.01776123046875,
-      "learning_rate": 2.291269920817881e-06,
-      "loss": 1.1938,
+      "epoch": 1.11,
+      "grad_norm": 24.305334091186523,
+      "learning_rate": 1.260845918922311e-05,
+      "loss": 1.8932,
       "step": 8835
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 14.084249496459961,
-      "learning_rate": 2.2892653102134913e-06,
-      "loss": 1.4542,
+      "epoch": 1.11,
+      "grad_norm": 11.232524871826172,
+      "learning_rate": 1.2607622474166423e-05,
+      "loss": 2.2715,
       "step": 8836
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 34.67374038696289,
-      "learning_rate": 2.2872606996091013e-06,
-      "loss": 1.9464,
+      "epoch": 1.11,
+      "grad_norm": 11.493144989013672,
+      "learning_rate": 1.2606785759109736e-05,
+      "loss": 0.793,
       "step": 8837
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 16.59427833557129,
-      "learning_rate": 2.285256089004711e-06,
-      "loss": 1.5703,
+      "epoch": 1.11,
+      "grad_norm": 18.408987045288086,
+      "learning_rate": 1.260594904405305e-05,
+      "loss": 1.8757,
       "step": 8838
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 20.186216354370117,
-      "learning_rate": 2.283251478400321e-06,
-      "loss": 1.527,
+      "epoch": 1.11,
+      "grad_norm": 41.016929626464844,
+      "learning_rate": 1.260511232899636e-05,
+      "loss": 4.0113,
       "step": 8839
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 34.25483703613281,
-      "learning_rate": 2.281246867795931e-06,
-      "loss": 2.5718,
+      "epoch": 1.11,
+      "grad_norm": 27.24270248413086,
+      "learning_rate": 1.2604275613939674e-05,
+      "loss": 1.0534,
       "step": 8840
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 28.123106002807617,
-      "learning_rate": 2.279242257191541e-06,
-      "loss": 1.5148,
+      "epoch": 1.11,
+      "grad_norm": 24.53693389892578,
+      "learning_rate": 1.2603438898882988e-05,
+      "loss": 2.7761,
       "step": 8841
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 55.747276306152344,
-      "learning_rate": 2.2772376465871505e-06,
-      "loss": 1.7577,
+      "epoch": 1.11,
+      "grad_norm": 8.33435344696045,
+      "learning_rate": 1.2602602183826298e-05,
+      "loss": 0.3023,
       "step": 8842
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 21.762004852294922,
-      "learning_rate": 2.2752330359827606e-06,
-      "loss": 1.3397,
+      "epoch": 1.11,
+      "grad_norm": 7.682891845703125,
+      "learning_rate": 1.2601765468769611e-05,
+      "loss": 0.821,
       "step": 8843
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 38.36250305175781,
-      "learning_rate": 2.2732284253783702e-06,
-      "loss": 1.2055,
+      "epoch": 1.11,
+      "grad_norm": 41.553611755371094,
+      "learning_rate": 1.2600928753712925e-05,
+      "loss": 2.2671,
       "step": 8844
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 14.818611145019531,
-      "learning_rate": 2.2712238147739803e-06,
-      "loss": 1.3097,
+      "epoch": 1.11,
+      "grad_norm": 11.70155143737793,
+      "learning_rate": 1.2600092038656237e-05,
+      "loss": 2.1812,
       "step": 8845
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 7.749112129211426,
-      "learning_rate": 2.26921920416959e-06,
-      "loss": 0.4492,
+      "epoch": 1.11,
+      "grad_norm": 4.751408100128174,
+      "learning_rate": 1.2599255323599549e-05,
+      "loss": 0.4145,
       "step": 8846
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 10.357666969299316,
-      "learning_rate": 2.2672145935652e-06,
-      "loss": 0.9897,
+      "epoch": 1.11,
+      "grad_norm": 36.31829071044922,
+      "learning_rate": 1.2598418608542861e-05,
+      "loss": 2.0711,
       "step": 8847
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 15.227676391601562,
-      "learning_rate": 2.26520998296081e-06,
-      "loss": 1.3606,
+      "epoch": 1.11,
+      "grad_norm": 34.59927749633789,
+      "learning_rate": 1.2597581893486175e-05,
+      "loss": 2.4649,
       "step": 8848
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 16.850852966308594,
-      "learning_rate": 2.26320537235642e-06,
-      "loss": 1.4527,
+      "epoch": 1.11,
+      "grad_norm": 7.014703273773193,
+      "learning_rate": 1.2596745178429487e-05,
+      "loss": 0.497,
       "step": 8849
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 16.4047794342041,
-      "learning_rate": 2.26120076175203e-06,
-      "loss": 1.6789,
+      "epoch": 1.11,
+      "grad_norm": 15.280418395996094,
+      "learning_rate": 1.2595908463372799e-05,
+      "loss": 1.1994,
       "step": 8850
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 7.696395397186279,
-      "learning_rate": 2.25919615114764e-06,
-      "loss": 0.9858,
+      "epoch": 1.11,
+      "grad_norm": 13.042275428771973,
+      "learning_rate": 1.2595071748316112e-05,
+      "loss": 0.7759,
       "step": 8851
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 21.263986587524414,
-      "learning_rate": 2.2571915405432496e-06,
-      "loss": 1.7351,
+      "epoch": 1.11,
+      "grad_norm": 20.576278686523438,
+      "learning_rate": 1.2594235033259426e-05,
+      "loss": 1.3792,
       "step": 8852
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 7.717570781707764,
-      "learning_rate": 2.2551869299388597e-06,
-      "loss": 1.0206,
+      "epoch": 1.11,
+      "grad_norm": 14.835959434509277,
+      "learning_rate": 1.2593398318202736e-05,
+      "loss": 1.2731,
       "step": 8853
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 14.79688549041748,
-      "learning_rate": 2.2531823193344694e-06,
-      "loss": 1.7117,
+      "epoch": 1.11,
+      "grad_norm": 28.5734920501709,
+      "learning_rate": 1.259256160314605e-05,
+      "loss": 0.8112,
       "step": 8854
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 8.93305492401123,
-      "learning_rate": 2.2511777087300795e-06,
-      "loss": 1.839,
+      "epoch": 1.11,
+      "grad_norm": 4.923260688781738,
+      "learning_rate": 1.2591724888089363e-05,
+      "loss": 0.3191,
       "step": 8855
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 13.082503318786621,
-      "learning_rate": 2.249173098125689e-06,
-      "loss": 1.252,
+      "epoch": 1.11,
+      "grad_norm": 9.085829734802246,
+      "learning_rate": 1.2590888173032674e-05,
+      "loss": 1.3536,
       "step": 8856
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 11.024530410766602,
-      "learning_rate": 2.2471684875212992e-06,
-      "loss": 0.5396,
+      "epoch": 1.11,
+      "grad_norm": 11.32632827758789,
+      "learning_rate": 1.2590051457975987e-05,
+      "loss": 3.1705,
       "step": 8857
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 30.2106876373291,
-      "learning_rate": 2.245163876916909e-06,
-      "loss": 1.8259,
+      "epoch": 1.11,
+      "grad_norm": 59.948646545410156,
+      "learning_rate": 1.2589214742919301e-05,
+      "loss": 0.5682,
       "step": 8858
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 12.592206954956055,
-      "learning_rate": 2.243159266312519e-06,
-      "loss": 2.2187,
+      "epoch": 1.11,
+      "grad_norm": 10.617377281188965,
+      "learning_rate": 1.2588378027862613e-05,
+      "loss": 0.6431,
       "step": 8859
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 29.125699996948242,
-      "learning_rate": 2.241154655708129e-06,
-      "loss": 1.1455,
+      "epoch": 1.11,
+      "grad_norm": 18.310394287109375,
+      "learning_rate": 1.2587541312805925e-05,
+      "loss": 1.9332,
       "step": 8860
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 16.28235626220703,
-      "learning_rate": 2.2391500451037387e-06,
-      "loss": 1.9047,
+      "epoch": 1.11,
+      "grad_norm": 23.954254150390625,
+      "learning_rate": 1.2586704597749237e-05,
+      "loss": 1.5929,
       "step": 8861
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 30.546987533569336,
-      "learning_rate": 2.237145434499349e-06,
-      "loss": 1.7441,
+      "epoch": 1.11,
+      "grad_norm": 8.76931095123291,
+      "learning_rate": 1.258586788269255e-05,
+      "loss": 1.7597,
       "step": 8862
     },
     {
-      "epoch": 2.66,
-      "grad_norm": 12.745742797851562,
-      "learning_rate": 2.2351408238949585e-06,
-      "loss": 0.5604,
+      "epoch": 1.11,
+      "grad_norm": 9.1308012008667,
+      "learning_rate": 1.2585031167635862e-05,
+      "loss": 2.1952,
       "step": 8863
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 10.053805351257324,
-      "learning_rate": 2.2331362132905685e-06,
-      "loss": 1.467,
+      "epoch": 1.11,
+      "grad_norm": 67.87699890136719,
+      "learning_rate": 1.2584194452579174e-05,
+      "loss": 0.6862,
       "step": 8864
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 22.58064842224121,
-      "learning_rate": 2.2311316026861786e-06,
-      "loss": 1.3182,
+      "epoch": 1.11,
+      "grad_norm": 12.790180206298828,
+      "learning_rate": 1.2583357737522488e-05,
+      "loss": 3.0736,
       "step": 8865
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 10.636580467224121,
-      "learning_rate": 2.2291269920817883e-06,
-      "loss": 1.3195,
+      "epoch": 1.11,
+      "grad_norm": 33.39744186401367,
+      "learning_rate": 1.2582521022465802e-05,
+      "loss": 3.5402,
       "step": 8866
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 18.260143280029297,
-      "learning_rate": 2.2271223814773984e-06,
-      "loss": 1.3604,
+      "epoch": 1.11,
+      "grad_norm": 12.848188400268555,
+      "learning_rate": 1.2581684307409112e-05,
+      "loss": 1.7566,
       "step": 8867
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 8.831524848937988,
-      "learning_rate": 2.225117770873008e-06,
-      "loss": 1.1861,
+      "epoch": 1.11,
+      "grad_norm": 5.089822769165039,
+      "learning_rate": 1.2580847592352426e-05,
+      "loss": 0.2165,
       "step": 8868
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 13.567652702331543,
-      "learning_rate": 2.2231131602686177e-06,
-      "loss": 1.0606,
+      "epoch": 1.11,
+      "grad_norm": 12.391101837158203,
+      "learning_rate": 1.258001087729574e-05,
+      "loss": 0.4484,
       "step": 8869
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 31.04169273376465,
-      "learning_rate": 2.2211085496642278e-06,
-      "loss": 1.1305,
+      "epoch": 1.11,
+      "grad_norm": 28.662338256835938,
+      "learning_rate": 1.257917416223905e-05,
+      "loss": 1.7781,
       "step": 8870
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 17.62767219543457,
-      "learning_rate": 2.219103939059838e-06,
-      "loss": 1.8986,
+      "epoch": 1.11,
+      "grad_norm": 9.914118766784668,
+      "learning_rate": 1.2578337447182363e-05,
+      "loss": 3.4399,
       "step": 8871
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 31.392688751220703,
-      "learning_rate": 2.2170993284554475e-06,
-      "loss": 2.3315,
+      "epoch": 1.11,
+      "grad_norm": 5.98568868637085,
+      "learning_rate": 1.2577500732125675e-05,
+      "loss": 0.8468,
       "step": 8872
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 18.428953170776367,
-      "learning_rate": 2.2150947178510576e-06,
-      "loss": 1.559,
+      "epoch": 1.11,
+      "grad_norm": 11.589415550231934,
+      "learning_rate": 1.2576664017068989e-05,
+      "loss": 1.4256,
       "step": 8873
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 24.9808292388916,
-      "learning_rate": 2.2130901072466677e-06,
-      "loss": 1.6667,
+      "epoch": 1.11,
+      "grad_norm": 11.69733715057373,
+      "learning_rate": 1.25758273020123e-05,
+      "loss": 1.1597,
       "step": 8874
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 23.458446502685547,
-      "learning_rate": 2.2110854966422773e-06,
-      "loss": 0.8407,
+      "epoch": 1.11,
+      "grad_norm": 8.565375328063965,
+      "learning_rate": 1.2574990586955613e-05,
+      "loss": 1.5882,
       "step": 8875
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 10.7035493850708,
-      "learning_rate": 2.2090808860378874e-06,
-      "loss": 1.3342,
+      "epoch": 1.11,
+      "grad_norm": 46.239627838134766,
+      "learning_rate": 1.2574153871898926e-05,
+      "loss": 2.1625,
       "step": 8876
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 21.982637405395508,
-      "learning_rate": 2.2070762754334975e-06,
-      "loss": 1.8623,
+      "epoch": 1.11,
+      "grad_norm": 12.065559387207031,
+      "learning_rate": 1.2573317156842238e-05,
+      "loss": 2.7625,
       "step": 8877
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 10.788835525512695,
-      "learning_rate": 2.205071664829107e-06,
-      "loss": 1.14,
+      "epoch": 1.11,
+      "grad_norm": 19.050992965698242,
+      "learning_rate": 1.257248044178555e-05,
+      "loss": 0.8145,
       "step": 8878
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 12.52689266204834,
-      "learning_rate": 2.2030670542247173e-06,
-      "loss": 0.8905,
+      "epoch": 1.11,
+      "grad_norm": 23.644147872924805,
+      "learning_rate": 1.2571643726728864e-05,
+      "loss": 2.1455,
       "step": 8879
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 12.546585083007812,
-      "learning_rate": 2.201062443620327e-06,
-      "loss": 1.4123,
-      "step": 8880
-    },
-    {
-      "epoch": 2.67,
-      "eval_loss": 0.1634318232536316,
-      "eval_runtime": 43.2728,
-      "eval_samples_per_second": 34.179,
-      "eval_steps_per_second": 34.179,
+      "epoch": 1.11,
+      "grad_norm": 38.043373107910156,
+      "learning_rate": 1.2570807011672178e-05,
+      "loss": 1.5802,
       "step": 8880
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 23.065811157226562,
-      "learning_rate": 2.1990578330159366e-06,
-      "loss": 1.6073,
+      "epoch": 1.11,
+      "grad_norm": 11.92843246459961,
+      "learning_rate": 1.2569970296615488e-05,
+      "loss": 1.743,
       "step": 8881
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 8.505629539489746,
-      "learning_rate": 2.1970532224115467e-06,
-      "loss": 0.6363,
+      "epoch": 1.11,
+      "grad_norm": 15.821548461914062,
+      "learning_rate": 1.2569133581558801e-05,
+      "loss": 1.729,
       "step": 8882
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 19.605792999267578,
-      "learning_rate": 2.1950486118071563e-06,
-      "loss": 1.1445,
+      "epoch": 1.11,
+      "grad_norm": 9.177318572998047,
+      "learning_rate": 1.2568296866502115e-05,
+      "loss": 1.5359,
       "step": 8883
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 62.97642517089844,
-      "learning_rate": 2.1930440012027664e-06,
-      "loss": 3.0308,
+      "epoch": 1.11,
+      "grad_norm": 22.191919326782227,
+      "learning_rate": 1.2567460151445425e-05,
+      "loss": 1.4169,
       "step": 8884
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 34.175453186035156,
-      "learning_rate": 2.1910393905983765e-06,
-      "loss": 1.2063,
+      "epoch": 1.12,
+      "grad_norm": 16.08186149597168,
+      "learning_rate": 1.2566623436388739e-05,
+      "loss": 0.5535,
       "step": 8885
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 41.114051818847656,
-      "learning_rate": 2.189034779993986e-06,
-      "loss": 1.2233,
+      "epoch": 1.12,
+      "grad_norm": 20.439388275146484,
+      "learning_rate": 1.2565786721332051e-05,
+      "loss": 3.2665,
       "step": 8886
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 22.9063663482666,
-      "learning_rate": 2.1870301693895962e-06,
-      "loss": 1.1179,
+      "epoch": 1.12,
+      "grad_norm": 12.846830368041992,
+      "learning_rate": 1.2564950006275365e-05,
+      "loss": 1.4284,
       "step": 8887
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 12.28271770477295,
-      "learning_rate": 2.1850255587852063e-06,
-      "loss": 1.1284,
+      "epoch": 1.12,
+      "grad_norm": 36.724246978759766,
+      "learning_rate": 1.2564113291218677e-05,
+      "loss": 2.6891,
       "step": 8888
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 18.61870765686035,
-      "learning_rate": 2.183020948180816e-06,
-      "loss": 1.1018,
+      "epoch": 1.12,
+      "grad_norm": 31.265914916992188,
+      "learning_rate": 1.2563276576161988e-05,
+      "loss": 1.4702,
       "step": 8889
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 14.14230728149414,
-      "learning_rate": 2.181016337576426e-06,
-      "loss": 2.1612,
+      "epoch": 1.12,
+      "grad_norm": 19.516441345214844,
+      "learning_rate": 1.2562439861105302e-05,
+      "loss": 1.3568,
       "step": 8890
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 17.895893096923828,
-      "learning_rate": 2.179011726972036e-06,
-      "loss": 2.1376,
+      "epoch": 1.12,
+      "grad_norm": 22.5244140625,
+      "learning_rate": 1.2561603146048612e-05,
+      "loss": 1.7873,
       "step": 8891
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 10.05941390991211,
-      "learning_rate": 2.177007116367646e-06,
-      "loss": 1.9443,
+      "epoch": 1.12,
+      "grad_norm": 7.6699113845825195,
+      "learning_rate": 1.2560766430991926e-05,
+      "loss": 1.5238,
       "step": 8892
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 17.179000854492188,
-      "learning_rate": 2.1750025057632555e-06,
-      "loss": 1.3958,
+      "epoch": 1.12,
+      "grad_norm": 31.9620418548584,
+      "learning_rate": 1.255992971593524e-05,
+      "loss": 2.0262,
       "step": 8893
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 18.431400299072266,
-      "learning_rate": 2.1729978951588656e-06,
-      "loss": 1.2758,
+      "epoch": 1.12,
+      "grad_norm": 8.691798210144043,
+      "learning_rate": 1.2559093000878553e-05,
+      "loss": 0.5248,
       "step": 8894
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 18.084781646728516,
-      "learning_rate": 2.1709932845544752e-06,
-      "loss": 1.7674,
+      "epoch": 1.12,
+      "grad_norm": 9.342169761657715,
+      "learning_rate": 1.2558256285821864e-05,
+      "loss": 0.8525,
       "step": 8895
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 15.693217277526855,
-      "learning_rate": 2.1689886739500853e-06,
-      "loss": 1.0816,
+      "epoch": 1.12,
+      "grad_norm": 6.691696643829346,
+      "learning_rate": 1.2557419570765177e-05,
+      "loss": 1.1909,
       "step": 8896
     },
     {
-      "epoch": 2.67,
-      "grad_norm": 18.634521484375,
-      "learning_rate": 2.1669840633456954e-06,
-      "loss": 1.1063,
+      "epoch": 1.12,
+      "grad_norm": 9.33571720123291,
+      "learning_rate": 1.2556582855708491e-05,
+      "loss": 0.4905,
       "step": 8897
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 11.23396110534668,
-      "learning_rate": 2.164979452741305e-06,
-      "loss": 0.6603,
+      "epoch": 1.12,
+      "grad_norm": 4.31030797958374,
+      "learning_rate": 1.2555746140651801e-05,
+      "loss": 0.5247,
       "step": 8898
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 8.881089210510254,
-      "learning_rate": 2.162974842136915e-06,
-      "loss": 0.7917,
+      "epoch": 1.12,
+      "grad_norm": 9.584003448486328,
+      "learning_rate": 1.2554909425595115e-05,
+      "loss": 0.3755,
       "step": 8899
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 6.563168525695801,
-      "learning_rate": 2.1609702315325252e-06,
-      "loss": 0.7384,
+      "epoch": 1.12,
+      "grad_norm": 13.662652015686035,
+      "learning_rate": 1.2554072710538427e-05,
+      "loss": 1.4579,
       "step": 8900
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 45.68901824951172,
-      "learning_rate": 2.158965620928135e-06,
-      "loss": 2.6647,
+      "epoch": 1.12,
+      "grad_norm": 32.94351577758789,
+      "learning_rate": 1.255323599548174e-05,
+      "loss": 3.6991,
       "step": 8901
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 10.925772666931152,
-      "learning_rate": 2.156961010323745e-06,
-      "loss": 0.5719,
+      "epoch": 1.12,
+      "grad_norm": 26.51783561706543,
+      "learning_rate": 1.2552399280425052e-05,
+      "loss": 4.1407,
       "step": 8902
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 52.76468276977539,
-      "learning_rate": 2.1549563997193546e-06,
-      "loss": 1.5088,
+      "epoch": 1.12,
+      "grad_norm": 39.995025634765625,
+      "learning_rate": 1.2551562565368364e-05,
+      "loss": 1.6063,
       "step": 8903
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 17.248638153076172,
-      "learning_rate": 2.1529517891149647e-06,
-      "loss": 1.2967,
+      "epoch": 1.12,
+      "grad_norm": 13.210186004638672,
+      "learning_rate": 1.2550725850311678e-05,
+      "loss": 0.9404,
       "step": 8904
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 51.29323196411133,
-      "learning_rate": 2.1509471785105744e-06,
-      "loss": 1.6994,
+      "epoch": 1.12,
+      "grad_norm": 20.54436492919922,
+      "learning_rate": 1.2549889135254988e-05,
+      "loss": 0.942,
       "step": 8905
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 42.033302307128906,
-      "learning_rate": 2.1489425679061845e-06,
-      "loss": 1.3593,
+      "epoch": 1.12,
+      "grad_norm": 8.649977684020996,
+      "learning_rate": 1.2549052420198302e-05,
+      "loss": 1.3768,
       "step": 8906
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 9.53936767578125,
-      "learning_rate": 2.146937957301794e-06,
-      "loss": 0.7497,
+      "epoch": 1.12,
+      "grad_norm": 5.319600582122803,
+      "learning_rate": 1.2548215705141616e-05,
+      "loss": 0.2921,
       "step": 8907
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 29.297077178955078,
-      "learning_rate": 2.144933346697404e-06,
-      "loss": 1.042,
+      "epoch": 1.12,
+      "grad_norm": 19.230091094970703,
+      "learning_rate": 1.254737899008493e-05,
+      "loss": 0.9048,
       "step": 8908
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 37.91065216064453,
-      "learning_rate": 2.142928736093014e-06,
-      "loss": 2.1151,
+      "epoch": 1.12,
+      "grad_norm": 15.009637832641602,
+      "learning_rate": 1.254654227502824e-05,
+      "loss": 1.4137,
       "step": 8909
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 24.55820655822754,
-      "learning_rate": 2.140924125488624e-06,
-      "loss": 1.9639,
+      "epoch": 1.12,
+      "grad_norm": 83.87217712402344,
+      "learning_rate": 1.2545705559971553e-05,
+      "loss": 2.7705,
       "step": 8910
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 30.16676902770996,
-      "learning_rate": 2.138919514884234e-06,
-      "loss": 1.3401,
+      "epoch": 1.12,
+      "grad_norm": 5.943860054016113,
+      "learning_rate": 1.2544868844914865e-05,
+      "loss": 0.6931,
       "step": 8911
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 47.085235595703125,
-      "learning_rate": 2.1369149042798437e-06,
-      "loss": 2.4905,
+      "epoch": 1.12,
+      "grad_norm": 14.90063190460205,
+      "learning_rate": 1.2544032129858177e-05,
+      "loss": 1.1149,
       "step": 8912
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 19.413427352905273,
-      "learning_rate": 2.1349102936754538e-06,
-      "loss": 1.4043,
+      "epoch": 1.12,
+      "grad_norm": 25.426666259765625,
+      "learning_rate": 1.254319541480149e-05,
+      "loss": 0.6162,
       "step": 8913
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 31.471092224121094,
-      "learning_rate": 2.132905683071064e-06,
-      "loss": 1.9347,
+      "epoch": 1.12,
+      "grad_norm": 24.75497055053711,
+      "learning_rate": 1.2542358699744803e-05,
+      "loss": 2.9471,
       "step": 8914
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 26.705747604370117,
-      "learning_rate": 2.1309010724666735e-06,
-      "loss": 2.8265,
+      "epoch": 1.12,
+      "grad_norm": 17.341344833374023,
+      "learning_rate": 1.2541521984688116e-05,
+      "loss": 2.3469,
       "step": 8915
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 43.82736587524414,
-      "learning_rate": 2.1288964618622836e-06,
-      "loss": 1.3732,
+      "epoch": 1.12,
+      "grad_norm": 5.796677589416504,
+      "learning_rate": 1.2540685269631428e-05,
+      "loss": 0.5815,
       "step": 8916
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 12.801732063293457,
-      "learning_rate": 2.1268918512578933e-06,
-      "loss": 0.7379,
+      "epoch": 1.12,
+      "grad_norm": 10.825651168823242,
+      "learning_rate": 1.253984855457474e-05,
+      "loss": 0.6657,
       "step": 8917
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 14.50540828704834,
-      "learning_rate": 2.1248872406535033e-06,
-      "loss": 1.2645,
+      "epoch": 1.12,
+      "grad_norm": 15.243544578552246,
+      "learning_rate": 1.2539011839518054e-05,
+      "loss": 1.2522,
       "step": 8918
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 55.6925163269043,
-      "learning_rate": 2.122882630049113e-06,
-      "loss": 1.05,
+      "epoch": 1.12,
+      "grad_norm": 2.303821563720703,
+      "learning_rate": 1.2538175124461364e-05,
+      "loss": 0.0838,
       "step": 8919
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 8.958430290222168,
-      "learning_rate": 2.120878019444723e-06,
-      "loss": 0.9973,
+      "epoch": 1.12,
+      "grad_norm": 20.525466918945312,
+      "learning_rate": 1.2537338409404678e-05,
+      "loss": 0.8314,
       "step": 8920
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 12.112589836120605,
-      "learning_rate": 2.1188734088403328e-06,
-      "loss": 0.5669,
+      "epoch": 1.12,
+      "grad_norm": 6.544447898864746,
+      "learning_rate": 1.2536501694347991e-05,
+      "loss": 0.4977,
       "step": 8921
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 9.585906982421875,
-      "learning_rate": 2.116868798235943e-06,
-      "loss": 1.455,
+      "epoch": 1.12,
+      "grad_norm": 15.428367614746094,
+      "learning_rate": 1.2535664979291305e-05,
+      "loss": 0.4506,
       "step": 8922
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 10.292905807495117,
-      "learning_rate": 2.114864187631553e-06,
-      "loss": 0.822,
+      "epoch": 1.12,
+      "grad_norm": 29.606430053710938,
+      "learning_rate": 1.2534828264234615e-05,
+      "loss": 2.3348,
       "step": 8923
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 12.411725997924805,
-      "learning_rate": 2.1128595770271626e-06,
-      "loss": 0.5915,
+      "epoch": 1.12,
+      "grad_norm": 22.486295700073242,
+      "learning_rate": 1.2533991549177929e-05,
+      "loss": 1.0795,
       "step": 8924
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 95.50955200195312,
-      "learning_rate": 2.1108549664227727e-06,
-      "loss": 1.2842,
+      "epoch": 1.12,
+      "grad_norm": 12.869139671325684,
+      "learning_rate": 1.2533154834121241e-05,
+      "loss": 3.1516,
       "step": 8925
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 15.629396438598633,
-      "learning_rate": 2.1088503558183823e-06,
-      "loss": 1.6067,
+      "epoch": 1.12,
+      "grad_norm": 18.72137451171875,
+      "learning_rate": 1.2532318119064553e-05,
+      "loss": 1.5385,
       "step": 8926
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 15.7233304977417,
-      "learning_rate": 2.1068457452139924e-06,
-      "loss": 0.9173,
+      "epoch": 1.12,
+      "grad_norm": 4.132017135620117,
+      "learning_rate": 1.2531481404007866e-05,
+      "loss": 0.1411,
       "step": 8927
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 30.911314010620117,
-      "learning_rate": 2.1048411346096025e-06,
-      "loss": 2.3595,
+      "epoch": 1.12,
+      "grad_norm": 6.2769293785095215,
+      "learning_rate": 1.2530644688951178e-05,
+      "loss": 0.9071,
       "step": 8928
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 12.196179389953613,
-      "learning_rate": 2.102836524005212e-06,
-      "loss": 1.0742,
+      "epoch": 1.12,
+      "grad_norm": 9.791159629821777,
+      "learning_rate": 1.2529807973894492e-05,
+      "loss": 0.9885,
       "step": 8929
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 13.497405052185059,
-      "learning_rate": 2.1008319134008222e-06,
-      "loss": 1.1664,
+      "epoch": 1.12,
+      "grad_norm": 16.300458908081055,
+      "learning_rate": 1.2528971258837802e-05,
+      "loss": 1.1884,
       "step": 8930
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 14.204902648925781,
-      "learning_rate": 2.098827302796432e-06,
-      "loss": 1.0476,
+      "epoch": 1.12,
+      "grad_norm": 9.561453819274902,
+      "learning_rate": 1.2528134543781116e-05,
+      "loss": 1.2753,
       "step": 8931
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 21.779407501220703,
-      "learning_rate": 2.0968226921920416e-06,
-      "loss": 1.0798,
+      "epoch": 1.12,
+      "grad_norm": 68.78498077392578,
+      "learning_rate": 1.252729782872443e-05,
+      "loss": 4.0715,
       "step": 8932
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 30.279451370239258,
-      "learning_rate": 2.0948180815876516e-06,
-      "loss": 1.2667,
+      "epoch": 1.12,
+      "grad_norm": 8.061681747436523,
+      "learning_rate": 1.252646111366774e-05,
+      "loss": 0.4884,
       "step": 8933
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 13.741867065429688,
-      "learning_rate": 2.0928134709832617e-06,
-      "loss": 0.93,
+      "epoch": 1.12,
+      "grad_norm": 26.667295455932617,
+      "learning_rate": 1.2525624398611054e-05,
+      "loss": 3.0331,
       "step": 8934
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 9.646475791931152,
-      "learning_rate": 2.0908088603788714e-06,
-      "loss": 1.304,
+      "epoch": 1.12,
+      "grad_norm": 15.182604789733887,
+      "learning_rate": 1.2524787683554367e-05,
+      "loss": 0.8844,
       "step": 8935
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 13.321015357971191,
-      "learning_rate": 2.0888042497744815e-06,
-      "loss": 1.1108,
+      "epoch": 1.12,
+      "grad_norm": 3.172250270843506,
+      "learning_rate": 1.2523950968497681e-05,
+      "loss": 0.4641,
       "step": 8936
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 15.75296688079834,
-      "learning_rate": 2.0867996391700916e-06,
-      "loss": 0.9929,
+      "epoch": 1.12,
+      "grad_norm": 14.022339820861816,
+      "learning_rate": 1.2523114253440991e-05,
+      "loss": 2.2339,
       "step": 8937
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 24.70446014404297,
-      "learning_rate": 2.0847950285657012e-06,
-      "loss": 1.2706,
+      "epoch": 1.12,
+      "grad_norm": 12.469718933105469,
+      "learning_rate": 1.2522277538384305e-05,
+      "loss": 1.5523,
       "step": 8938
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 13.372381210327148,
-      "learning_rate": 2.0827904179613113e-06,
-      "loss": 0.6046,
+      "epoch": 1.12,
+      "grad_norm": 15.62543773651123,
+      "learning_rate": 1.2521440823327617e-05,
+      "loss": 2.1353,
       "step": 8939
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 40.783939361572266,
-      "learning_rate": 2.0807858073569214e-06,
-      "loss": 1.5435,
+      "epoch": 1.12,
+      "grad_norm": 17.858619689941406,
+      "learning_rate": 1.2520604108270929e-05,
+      "loss": 2.2041,
       "step": 8940
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 8.127727508544922,
-      "learning_rate": 2.078781196752531e-06,
-      "loss": 0.8914,
+      "epoch": 1.12,
+      "grad_norm": 5.224055290222168,
+      "learning_rate": 1.2519767393214242e-05,
+      "loss": 0.2525,
       "step": 8941
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 16.346284866333008,
-      "learning_rate": 2.076776586148141e-06,
-      "loss": 1.139,
+      "epoch": 1.12,
+      "grad_norm": 6.102902889251709,
+      "learning_rate": 1.2518930678157554e-05,
+      "loss": 0.4351,
       "step": 8942
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 83.08602142333984,
-      "learning_rate": 2.074771975543751e-06,
-      "loss": 2.625,
+      "epoch": 1.12,
+      "grad_norm": 17.269641876220703,
+      "learning_rate": 1.2518093963100868e-05,
+      "loss": 1.5761,
       "step": 8943
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 13.884016036987305,
-      "learning_rate": 2.0727673649393605e-06,
-      "loss": 1.0258,
+      "epoch": 1.12,
+      "grad_norm": 9.63320541381836,
+      "learning_rate": 1.2517257248044178e-05,
+      "loss": 0.6898,
       "step": 8944
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 11.885502815246582,
-      "learning_rate": 2.0707627543349705e-06,
-      "loss": 0.8579,
+      "epoch": 1.12,
+      "grad_norm": 4.770727157592773,
+      "learning_rate": 1.2516420532987492e-05,
+      "loss": 0.3419,
       "step": 8945
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 9.849443435668945,
-      "learning_rate": 2.06875814373058e-06,
-      "loss": 0.9329,
+      "epoch": 1.12,
+      "grad_norm": 7.858762264251709,
+      "learning_rate": 1.2515583817930805e-05,
+      "loss": 0.9,
       "step": 8946
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 9.109235763549805,
-      "learning_rate": 2.0667535331261903e-06,
-      "loss": 0.9577,
+      "epoch": 1.12,
+      "grad_norm": 8.816036224365234,
+      "learning_rate": 1.2514747102874116e-05,
+      "loss": 1.6838,
       "step": 8947
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 10.755121231079102,
-      "learning_rate": 2.0647489225218004e-06,
-      "loss": 0.6667,
+      "epoch": 1.12,
+      "grad_norm": 20.732986450195312,
+      "learning_rate": 1.251391038781743e-05,
+      "loss": 1.6266,
       "step": 8948
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 10.410956382751465,
-      "learning_rate": 2.06274431191741e-06,
-      "loss": 0.4745,
+      "epoch": 1.12,
+      "grad_norm": 17.330759048461914,
+      "learning_rate": 1.2513073672760743e-05,
+      "loss": 1.7112,
       "step": 8949
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 10.473151206970215,
-      "learning_rate": 2.06073970131302e-06,
-      "loss": 1.2898,
+      "epoch": 1.12,
+      "grad_norm": 18.552534103393555,
+      "learning_rate": 1.2512236957704057e-05,
+      "loss": 0.8954,
       "step": 8950
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 14.992378234863281,
-      "learning_rate": 2.05873509070863e-06,
-      "loss": 1.5699,
+      "epoch": 1.12,
+      "grad_norm": 9.809928894042969,
+      "learning_rate": 1.2511400242647367e-05,
+      "loss": 0.8482,
       "step": 8951
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 9.013774871826172,
-      "learning_rate": 2.05673048010424e-06,
-      "loss": 1.1274,
+      "epoch": 1.12,
+      "grad_norm": 14.423686027526855,
+      "learning_rate": 1.251056352759068e-05,
+      "loss": 1.1748,
       "step": 8952
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 18.34583854675293,
-      "learning_rate": 2.05472586949985e-06,
-      "loss": 1.1243,
+      "epoch": 1.12,
+      "grad_norm": 15.003713607788086,
+      "learning_rate": 1.2509726812533993e-05,
+      "loss": 0.8353,
       "step": 8953
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 9.34281063079834,
-      "learning_rate": 2.0527212588954596e-06,
-      "loss": 0.977,
+      "epoch": 1.12,
+      "grad_norm": 32.19587707519531,
+      "learning_rate": 1.2508890097477305e-05,
+      "loss": 1.1426,
       "step": 8954
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 19.33155632019043,
-      "learning_rate": 2.0507166482910697e-06,
-      "loss": 0.7991,
+      "epoch": 1.12,
+      "grad_norm": 26.24030303955078,
+      "learning_rate": 1.2508053382420618e-05,
+      "loss": 1.2817,
       "step": 8955
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 62.87168502807617,
-      "learning_rate": 2.0487120376866794e-06,
-      "loss": 1.8296,
+      "epoch": 1.12,
+      "grad_norm": 31.24359703063965,
+      "learning_rate": 1.250721666736393e-05,
+      "loss": 3.232,
       "step": 8956
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 14.451416015625,
-      "learning_rate": 2.0467074270822894e-06,
-      "loss": 0.9752,
+      "epoch": 1.12,
+      "grad_norm": 7.957561492919922,
+      "learning_rate": 1.2506379952307244e-05,
+      "loss": 0.8478,
       "step": 8957
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 13.052489280700684,
-      "learning_rate": 2.044702816477899e-06,
-      "loss": 1.0093,
+      "epoch": 1.12,
+      "grad_norm": 26.59343910217285,
+      "learning_rate": 1.2505543237250554e-05,
+      "loss": 1.7395,
       "step": 8958
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 10.156307220458984,
-      "learning_rate": 2.042698205873509e-06,
-      "loss": 1.142,
+      "epoch": 1.12,
+      "grad_norm": 216.5959014892578,
+      "learning_rate": 1.2504706522193868e-05,
+      "loss": 1.4375,
       "step": 8959
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 32.259788513183594,
-      "learning_rate": 2.0406935952691193e-06,
-      "loss": 2.6409,
+      "epoch": 1.12,
+      "grad_norm": 12.135117530822754,
+      "learning_rate": 1.2503869807137181e-05,
+      "loss": 0.4616,
       "step": 8960
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 35.24033737182617,
-      "learning_rate": 2.038688984664729e-06,
-      "loss": 1.7325,
+      "epoch": 1.12,
+      "grad_norm": 34.77619934082031,
+      "learning_rate": 1.2503033092080492e-05,
+      "loss": 1.1826,
       "step": 8961
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 15.073776245117188,
-      "learning_rate": 2.036684374060339e-06,
-      "loss": 0.7844,
+      "epoch": 1.12,
+      "grad_norm": 6.7146430015563965,
+      "learning_rate": 1.2502196377023805e-05,
+      "loss": 1.1367,
       "step": 8962
     },
     {
-      "epoch": 2.69,
-      "grad_norm": 26.254253387451172,
-      "learning_rate": 2.034679763455949e-06,
-      "loss": 0.9144,
+      "epoch": 1.12,
+      "grad_norm": 15.25345230102539,
+      "learning_rate": 1.2501359661967119e-05,
+      "loss": 1.2281,
       "step": 8963
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 10.655972480773926,
-      "learning_rate": 2.0326751528515588e-06,
-      "loss": 1.5722,
+      "epoch": 1.12,
+      "grad_norm": 20.43260383605957,
+      "learning_rate": 1.250052294691043e-05,
+      "loss": 1.3328,
       "step": 8964
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 15.8784761428833,
-      "learning_rate": 2.030670542247169e-06,
-      "loss": 1.5385,
+      "epoch": 1.13,
+      "grad_norm": 7.702591896057129,
+      "learning_rate": 1.2499686231853743e-05,
+      "loss": 0.6018,
       "step": 8965
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 75.25648498535156,
-      "learning_rate": 2.0286659316427785e-06,
-      "loss": 1.2864,
+      "epoch": 1.13,
+      "grad_norm": 13.08178424835205,
+      "learning_rate": 1.2498849516797056e-05,
+      "loss": 1.6774,
       "step": 8966
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 12.602245330810547,
-      "learning_rate": 2.0266613210383886e-06,
-      "loss": 1.6121,
+      "epoch": 1.13,
+      "grad_norm": 27.949283599853516,
+      "learning_rate": 1.2498012801740368e-05,
+      "loss": 1.9005,
       "step": 8967
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 26.045513153076172,
-      "learning_rate": 2.0246567104339982e-06,
-      "loss": 1.1818,
+      "epoch": 1.13,
+      "grad_norm": 21.547008514404297,
+      "learning_rate": 1.249717608668368e-05,
+      "loss": 2.2632,
       "step": 8968
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 15.361552238464355,
-      "learning_rate": 2.0226520998296083e-06,
-      "loss": 0.8867,
+      "epoch": 1.13,
+      "grad_norm": 28.614538192749023,
+      "learning_rate": 1.2496339371626994e-05,
+      "loss": 1.5634,
       "step": 8969
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 52.4283561706543,
-      "learning_rate": 2.020647489225218e-06,
-      "loss": 2.3197,
+      "epoch": 1.13,
+      "grad_norm": 9.390619277954102,
+      "learning_rate": 1.2495502656570306e-05,
+      "loss": 1.8472,
       "step": 8970
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 10.277100563049316,
-      "learning_rate": 2.018642878620828e-06,
-      "loss": 1.2036,
+      "epoch": 1.13,
+      "grad_norm": 68.93138885498047,
+      "learning_rate": 1.249466594151362e-05,
+      "loss": 0.6978,
       "step": 8971
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 56.391624450683594,
-      "learning_rate": 2.0166382680164377e-06,
-      "loss": 1.6374,
+      "epoch": 1.13,
+      "grad_norm": 16.693071365356445,
+      "learning_rate": 1.249382922645693e-05,
+      "loss": 0.9754,
       "step": 8972
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 16.56217384338379,
-      "learning_rate": 2.014633657412048e-06,
-      "loss": 1.2358,
+      "epoch": 1.13,
+      "grad_norm": 16.193252563476562,
+      "learning_rate": 1.2492992511400244e-05,
+      "loss": 2.1549,
       "step": 8973
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 30.0115966796875,
-      "learning_rate": 2.012629046807658e-06,
-      "loss": 1.2294,
+      "epoch": 1.13,
+      "grad_norm": 11.039328575134277,
+      "learning_rate": 1.2492155796343557e-05,
+      "loss": 1.1695,
       "step": 8974
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 10.70960807800293,
-      "learning_rate": 2.0106244362032676e-06,
-      "loss": 1.2672,
+      "epoch": 1.13,
+      "grad_norm": 15.025418281555176,
+      "learning_rate": 1.2491319081286867e-05,
+      "loss": 1.1417,
       "step": 8975
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 12.158578872680664,
-      "learning_rate": 2.0086198255988776e-06,
-      "loss": 0.794,
+      "epoch": 1.13,
+      "grad_norm": 4.358430862426758,
+      "learning_rate": 1.2490482366230181e-05,
+      "loss": 0.8119,
       "step": 8976
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 6.594705104827881,
-      "learning_rate": 2.0066152149944877e-06,
-      "loss": 0.964,
+      "epoch": 1.13,
+      "grad_norm": 12.447125434875488,
+      "learning_rate": 1.2489645651173495e-05,
+      "loss": 2.9231,
       "step": 8977
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 18.69150161743164,
-      "learning_rate": 2.0046106043900974e-06,
-      "loss": 1.1616,
+      "epoch": 1.13,
+      "grad_norm": 17.1337833404541,
+      "learning_rate": 1.2488808936116807e-05,
+      "loss": 1.0778,
       "step": 8978
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 19.576065063476562,
-      "learning_rate": 2.0026059937857075e-06,
-      "loss": 0.9053,
+      "epoch": 1.13,
+      "grad_norm": 6.489715576171875,
+      "learning_rate": 1.2487972221060119e-05,
+      "loss": 0.3476,
       "step": 8979
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 30.62327003479004,
-      "learning_rate": 2.000601383181317e-06,
-      "loss": 1.0141,
+      "epoch": 1.13,
+      "grad_norm": 7.586145877838135,
+      "learning_rate": 1.2487135506003432e-05,
+      "loss": 1.4568,
       "step": 8980
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 16.1602840423584,
-      "learning_rate": 1.9985967725769272e-06,
-      "loss": 1.0071,
+      "epoch": 1.13,
+      "grad_norm": 36.23131561279297,
+      "learning_rate": 1.2486298790946744e-05,
+      "loss": 1.7715,
       "step": 8981
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 16.035024642944336,
-      "learning_rate": 1.996592161972537e-06,
-      "loss": 1.9886,
+      "epoch": 1.13,
+      "grad_norm": 10.976786613464355,
+      "learning_rate": 1.2485462075890056e-05,
+      "loss": 0.4711,
       "step": 8982
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 32.672157287597656,
-      "learning_rate": 1.994587551368147e-06,
-      "loss": 1.9142,
+      "epoch": 1.13,
+      "grad_norm": 10.669611930847168,
+      "learning_rate": 1.2484625360833368e-05,
+      "loss": 1.7268,
       "step": 8983
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 34.7219352722168,
-      "learning_rate": 1.9925829407637566e-06,
-      "loss": 1.6279,
+      "epoch": 1.13,
+      "grad_norm": 10.771439552307129,
+      "learning_rate": 1.2483788645776682e-05,
+      "loss": 1.6016,
       "step": 8984
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 38.268131256103516,
-      "learning_rate": 1.9905783301593667e-06,
-      "loss": 1.3426,
+      "epoch": 1.13,
+      "grad_norm": 16.643007278442383,
+      "learning_rate": 1.2482951930719995e-05,
+      "loss": 1.1995,
       "step": 8985
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 10.52354621887207,
-      "learning_rate": 1.988573719554977e-06,
-      "loss": 1.3539,
+      "epoch": 1.13,
+      "grad_norm": 8.137476921081543,
+      "learning_rate": 1.2482115215663306e-05,
+      "loss": 1.8578,
       "step": 8986
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 19.19072914123535,
-      "learning_rate": 1.9865691089505865e-06,
-      "loss": 1.7263,
+      "epoch": 1.13,
+      "grad_norm": 11.106561660766602,
+      "learning_rate": 1.248127850060662e-05,
+      "loss": 0.9672,
       "step": 8987
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 35.17476272583008,
-      "learning_rate": 1.9845644983461965e-06,
-      "loss": 1.2377,
+      "epoch": 1.13,
+      "grad_norm": 56.206581115722656,
+      "learning_rate": 1.2480441785549933e-05,
+      "loss": 0.7559,
       "step": 8988
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 23.662670135498047,
-      "learning_rate": 1.982559887741806e-06,
-      "loss": 0.9037,
+      "epoch": 1.13,
+      "grad_norm": 12.278787612915039,
+      "learning_rate": 1.2479605070493243e-05,
+      "loss": 1.7761,
       "step": 8989
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 25.2396297454834,
-      "learning_rate": 1.9805552771374163e-06,
-      "loss": 1.8102,
+      "epoch": 1.13,
+      "grad_norm": 50.02405548095703,
+      "learning_rate": 1.2478768355436557e-05,
+      "loss": 2.2794,
       "step": 8990
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 8.776323318481445,
-      "learning_rate": 1.9785506665330264e-06,
-      "loss": 0.7533,
+      "epoch": 1.13,
+      "grad_norm": 58.88485336303711,
+      "learning_rate": 1.247793164037987e-05,
+      "loss": 3.1132,
       "step": 8991
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 16.68621253967285,
-      "learning_rate": 1.976546055928636e-06,
-      "loss": 1.3865,
+      "epoch": 1.13,
+      "grad_norm": 9.378216743469238,
+      "learning_rate": 1.2477094925323183e-05,
+      "loss": 1.1562,
       "step": 8992
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 17.21820640563965,
-      "learning_rate": 1.974541445324246e-06,
-      "loss": 1.0538,
+      "epoch": 1.13,
+      "grad_norm": 43.093109130859375,
+      "learning_rate": 1.2476258210266494e-05,
+      "loss": 2.3941,
       "step": 8993
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 13.057286262512207,
-      "learning_rate": 1.9725368347198558e-06,
-      "loss": 1.1173,
+      "epoch": 1.13,
+      "grad_norm": 21.89565658569336,
+      "learning_rate": 1.2475421495209808e-05,
+      "loss": 1.4894,
       "step": 8994
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 16.948637008666992,
-      "learning_rate": 1.9705322241154654e-06,
-      "loss": 0.5743,
+      "epoch": 1.13,
+      "grad_norm": 14.45287036895752,
+      "learning_rate": 1.247458478015312e-05,
+      "loss": 1.5325,
       "step": 8995
     },
     {
-      "epoch": 2.7,
-      "grad_norm": 11.57693099975586,
-      "learning_rate": 1.9685276135110755e-06,
-      "loss": 1.2016,
+      "epoch": 1.13,
+      "grad_norm": 10.38625431060791,
+      "learning_rate": 1.2473748065096432e-05,
+      "loss": 0.2503,
       "step": 8996
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 25.845434188842773,
-      "learning_rate": 1.9665230029066856e-06,
-      "loss": 1.6141,
+      "epoch": 1.13,
+      "grad_norm": 17.61884307861328,
+      "learning_rate": 1.2472911350039744e-05,
+      "loss": 1.1376,
       "step": 8997
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 11.109925270080566,
-      "learning_rate": 1.9645183923022953e-06,
-      "loss": 1.2974,
+      "epoch": 1.13,
+      "grad_norm": 17.80261993408203,
+      "learning_rate": 1.2472074634983058e-05,
+      "loss": 0.8628,
       "step": 8998
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 17.424144744873047,
-      "learning_rate": 1.9625137816979054e-06,
-      "loss": 1.0136,
+      "epoch": 1.13,
+      "grad_norm": 23.364965438842773,
+      "learning_rate": 1.2471237919926371e-05,
+      "loss": 0.5915,
       "step": 8999
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 13.590910911560059,
-      "learning_rate": 1.9605091710935154e-06,
-      "loss": 2.0274,
-      "step": 9000
-    },
-    {
-      "epoch": 2.71,
-      "eval_loss": 0.16217102110385895,
-      "eval_runtime": 43.386,
-      "eval_samples_per_second": 34.089,
-      "eval_steps_per_second": 34.089,
+      "epoch": 1.13,
+      "grad_norm": 9.120946884155273,
+      "learning_rate": 1.2470401204869682e-05,
+      "loss": 1.3486,
       "step": 9000
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 16.488697052001953,
-      "learning_rate": 1.958504560489125e-06,
-      "loss": 1.3625,
+      "epoch": 1.13,
+      "grad_norm": 14.520109176635742,
+      "learning_rate": 1.2469564489812995e-05,
+      "loss": 1.9042,
       "step": 9001
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 9.223793983459473,
-      "learning_rate": 1.956499949884735e-06,
-      "loss": 1.2811,
+      "epoch": 1.13,
+      "grad_norm": 7.26044225692749,
+      "learning_rate": 1.2468727774756309e-05,
+      "loss": 0.4613,
       "step": 9002
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 13.224336624145508,
-      "learning_rate": 1.9544953392803453e-06,
-      "loss": 1.5226,
+      "epoch": 1.13,
+      "grad_norm": 4.86530065536499,
+      "learning_rate": 1.2467891059699619e-05,
+      "loss": 0.1759,
       "step": 9003
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 39.96490478515625,
-      "learning_rate": 1.952490728675955e-06,
-      "loss": 0.9551,
+      "epoch": 1.13,
+      "grad_norm": 12.943403244018555,
+      "learning_rate": 1.2467054344642933e-05,
+      "loss": 1.1472,
       "step": 9004
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 8.030670166015625,
-      "learning_rate": 1.9504861180715646e-06,
-      "loss": 0.5237,
+      "epoch": 1.13,
+      "grad_norm": 12.199639320373535,
+      "learning_rate": 1.2466217629586246e-05,
+      "loss": 0.5435,
       "step": 9005
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 45.17795944213867,
-      "learning_rate": 1.9484815074671747e-06,
-      "loss": 1.1533,
+      "epoch": 1.13,
+      "grad_norm": 11.572731018066406,
+      "learning_rate": 1.2465380914529558e-05,
+      "loss": 2.8074,
       "step": 9006
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 13.118334770202637,
-      "learning_rate": 1.9464768968627843e-06,
-      "loss": 1.3035,
+      "epoch": 1.13,
+      "grad_norm": 14.404424667358398,
+      "learning_rate": 1.246454419947287e-05,
+      "loss": 1.2332,
       "step": 9007
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 13.178396224975586,
-      "learning_rate": 1.9444722862583944e-06,
-      "loss": 1.2787,
+      "epoch": 1.13,
+      "grad_norm": 16.319948196411133,
+      "learning_rate": 1.2463707484416184e-05,
+      "loss": 1.5866,
       "step": 9008
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 12.746395111083984,
-      "learning_rate": 1.942467675654004e-06,
-      "loss": 1.3914,
+      "epoch": 1.13,
+      "grad_norm": 8.145049095153809,
+      "learning_rate": 1.2462870769359496e-05,
+      "loss": 0.8607,
       "step": 9009
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 25.33348274230957,
-      "learning_rate": 1.940463065049614e-06,
-      "loss": 1.2344,
+      "epoch": 1.13,
+      "grad_norm": 57.9113883972168,
+      "learning_rate": 1.2462034054302808e-05,
+      "loss": 2.2547,
       "step": 9010
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 18.61649513244629,
-      "learning_rate": 1.9384584544452242e-06,
-      "loss": 0.9396,
+      "epoch": 1.13,
+      "grad_norm": 34.40373229980469,
+      "learning_rate": 1.246119733924612e-05,
+      "loss": 1.0205,
       "step": 9011
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 98.35498809814453,
-      "learning_rate": 1.936453843840834e-06,
-      "loss": 2.0006,
+      "epoch": 1.13,
+      "grad_norm": 6.570141315460205,
+      "learning_rate": 1.2460360624189433e-05,
+      "loss": 1.4808,
       "step": 9012
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 12.018935203552246,
-      "learning_rate": 1.934449233236444e-06,
-      "loss": 1.1039,
+      "epoch": 1.13,
+      "grad_norm": 22.321439743041992,
+      "learning_rate": 1.2459523909132747e-05,
+      "loss": 1.64,
       "step": 9013
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 18.714567184448242,
-      "learning_rate": 1.932444622632054e-06,
-      "loss": 1.2751,
+      "epoch": 1.13,
+      "grad_norm": 24.136550903320312,
+      "learning_rate": 1.2458687194076057e-05,
+      "loss": 1.4336,
       "step": 9014
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 41.53782653808594,
-      "learning_rate": 1.9304400120276637e-06,
-      "loss": 1.7534,
+      "epoch": 1.13,
+      "grad_norm": 21.779146194458008,
+      "learning_rate": 1.2457850479019371e-05,
+      "loss": 1.6992,
       "step": 9015
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 37.50891876220703,
-      "learning_rate": 1.928435401423274e-06,
-      "loss": 1.1007,
+      "epoch": 1.13,
+      "grad_norm": 16.919523239135742,
+      "learning_rate": 1.2457013763962685e-05,
+      "loss": 1.242,
       "step": 9016
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 14.983277320861816,
-      "learning_rate": 1.9264307908188835e-06,
-      "loss": 1.6123,
+      "epoch": 1.13,
+      "grad_norm": 22.42137336730957,
+      "learning_rate": 1.2456177048905995e-05,
+      "loss": 0.4977,
       "step": 9017
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 13.673733711242676,
-      "learning_rate": 1.9244261802144936e-06,
-      "loss": 1.0661,
+      "epoch": 1.13,
+      "grad_norm": 26.349485397338867,
+      "learning_rate": 1.2455340333849309e-05,
+      "loss": 0.5325,
       "step": 9018
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 9.97967529296875,
-      "learning_rate": 1.9224215696101032e-06,
-      "loss": 1.5572,
+      "epoch": 1.13,
+      "grad_norm": 5.285884380340576,
+      "learning_rate": 1.2454503618792622e-05,
+      "loss": 1.6786,
       "step": 9019
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 12.354124069213867,
-      "learning_rate": 1.9204169590057133e-06,
-      "loss": 1.2915,
+      "epoch": 1.13,
+      "grad_norm": 29.1854190826416,
+      "learning_rate": 1.2453666903735934e-05,
+      "loss": 1.9452,
       "step": 9020
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 11.94011116027832,
-      "learning_rate": 1.918412348401323e-06,
-      "loss": 1.3119,
+      "epoch": 1.13,
+      "grad_norm": 47.3242301940918,
+      "learning_rate": 1.2452830188679246e-05,
+      "loss": 2.2183,
       "step": 9021
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 11.489762306213379,
-      "learning_rate": 1.916407737796933e-06,
-      "loss": 1.1241,
+      "epoch": 1.13,
+      "grad_norm": 11.776655197143555,
+      "learning_rate": 1.245199347362256e-05,
+      "loss": 1.815,
       "step": 9022
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 26.69053840637207,
-      "learning_rate": 1.914403127192543e-06,
-      "loss": 1.3275,
+      "epoch": 1.13,
+      "grad_norm": 16.114843368530273,
+      "learning_rate": 1.2451156758565872e-05,
+      "loss": 1.3109,
       "step": 9023
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 15.625397682189941,
-      "learning_rate": 1.912398516588153e-06,
-      "loss": 1.2162,
+      "epoch": 1.13,
+      "grad_norm": 6.5335211753845215,
+      "learning_rate": 1.2450320043509184e-05,
+      "loss": 1.0004,
       "step": 9024
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 23.008207321166992,
-      "learning_rate": 1.910393905983763e-06,
-      "loss": 1.4855,
+      "epoch": 1.13,
+      "grad_norm": 22.013607025146484,
+      "learning_rate": 1.2449483328452496e-05,
+      "loss": 1.4589,
       "step": 9025
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 122.37059783935547,
-      "learning_rate": 1.908389295379373e-06,
-      "loss": 1.5003,
+      "epoch": 1.13,
+      "grad_norm": 11.52690601348877,
+      "learning_rate": 1.244864661339581e-05,
+      "loss": 0.9607,
       "step": 9026
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 47.32773208618164,
-      "learning_rate": 1.9063846847749826e-06,
-      "loss": 2.3885,
+      "epoch": 1.13,
+      "grad_norm": 13.754624366760254,
+      "learning_rate": 1.2447809898339123e-05,
+      "loss": 1.4942,
       "step": 9027
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 10.500248908996582,
-      "learning_rate": 1.9043800741705925e-06,
-      "loss": 1.3469,
+      "epoch": 1.13,
+      "grad_norm": 11.588741302490234,
+      "learning_rate": 1.2446973183282433e-05,
+      "loss": 1.6175,
       "step": 9028
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 12.691232681274414,
-      "learning_rate": 1.9023754635662024e-06,
-      "loss": 1.5982,
+      "epoch": 1.13,
+      "grad_norm": 12.320586204528809,
+      "learning_rate": 1.2446136468225747e-05,
+      "loss": 0.9224,
       "step": 9029
     },
     {
-      "epoch": 2.71,
-      "grad_norm": 15.978590965270996,
-      "learning_rate": 1.9003708529618122e-06,
-      "loss": 0.9491,
+      "epoch": 1.13,
+      "grad_norm": 7.940244197845459,
+      "learning_rate": 1.244529975316906e-05,
+      "loss": 1.0349,
       "step": 9030
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 22.821592330932617,
-      "learning_rate": 1.8983662423574223e-06,
-      "loss": 1.3128,
+      "epoch": 1.13,
+      "grad_norm": 12.41991901397705,
+      "learning_rate": 1.244446303811237e-05,
+      "loss": 0.7951,
       "step": 9031
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 33.413455963134766,
-      "learning_rate": 1.896361631753032e-06,
-      "loss": 2.5185,
+      "epoch": 1.13,
+      "grad_norm": 15.359567642211914,
+      "learning_rate": 1.2443626323055684e-05,
+      "loss": 0.455,
       "step": 9032
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 14.448841094970703,
-      "learning_rate": 1.894357021148642e-06,
-      "loss": 1.7291,
+      "epoch": 1.13,
+      "grad_norm": 9.817809104919434,
+      "learning_rate": 1.2442789607998998e-05,
+      "loss": 0.5092,
       "step": 9033
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 17.098773956298828,
-      "learning_rate": 1.892352410544252e-06,
-      "loss": 0.7803,
+      "epoch": 1.13,
+      "grad_norm": 15.863314628601074,
+      "learning_rate": 1.244195289294231e-05,
+      "loss": 1.0122,
       "step": 9034
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 22.17733383178711,
-      "learning_rate": 1.8903477999398618e-06,
-      "loss": 2.054,
+      "epoch": 1.13,
+      "grad_norm": 39.14425277709961,
+      "learning_rate": 1.2441116177885622e-05,
+      "loss": 3.0413,
       "step": 9035
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 18.0053768157959,
-      "learning_rate": 1.8883431893354717e-06,
-      "loss": 0.9891,
+      "epoch": 1.13,
+      "grad_norm": 24.42272186279297,
+      "learning_rate": 1.2440279462828934e-05,
+      "loss": 2.5408,
       "step": 9036
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 7.076298713684082,
-      "learning_rate": 1.8863385787310818e-06,
-      "loss": 0.6043,
+      "epoch": 1.13,
+      "grad_norm": 12.801360130310059,
+      "learning_rate": 1.2439442747772248e-05,
+      "loss": 1.0546,
       "step": 9037
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 10.31680965423584,
-      "learning_rate": 1.8843339681266914e-06,
-      "loss": 0.6896,
+      "epoch": 1.13,
+      "grad_norm": 16.35565948486328,
+      "learning_rate": 1.243860603271556e-05,
+      "loss": 0.3435,
       "step": 9038
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 34.10591125488281,
-      "learning_rate": 1.8823293575223015e-06,
-      "loss": 1.6691,
+      "epoch": 1.13,
+      "grad_norm": 29.577281951904297,
+      "learning_rate": 1.2437769317658871e-05,
+      "loss": 1.5476,
       "step": 9039
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 19.84099006652832,
-      "learning_rate": 1.8803247469179114e-06,
-      "loss": 1.7946,
+      "epoch": 1.13,
+      "grad_norm": 13.342458724975586,
+      "learning_rate": 1.2436932602602185e-05,
+      "loss": 2.3566,
       "step": 9040
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 16.761245727539062,
-      "learning_rate": 1.8783201363135213e-06,
-      "loss": 1.7972,
+      "epoch": 1.13,
+      "grad_norm": 30.523876190185547,
+      "learning_rate": 1.2436095887545499e-05,
+      "loss": 1.688,
       "step": 9041
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 19.77780532836914,
-      "learning_rate": 1.8763155257091311e-06,
-      "loss": 1.0711,
+      "epoch": 1.13,
+      "grad_norm": 23.766157150268555,
+      "learning_rate": 1.2435259172488809e-05,
+      "loss": 2.5348,
       "step": 9042
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 12.905445098876953,
-      "learning_rate": 1.8743109151047412e-06,
-      "loss": 0.79,
+      "epoch": 1.13,
+      "grad_norm": 19.39542007446289,
+      "learning_rate": 1.2434422457432123e-05,
+      "loss": 2.1838,
       "step": 9043
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 34.32182312011719,
-      "learning_rate": 1.8723063045003509e-06,
-      "loss": 1.5714,
+      "epoch": 1.14,
+      "grad_norm": 26.63437271118164,
+      "learning_rate": 1.2433585742375436e-05,
+      "loss": 2.7722,
       "step": 9044
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 15.384803771972656,
-      "learning_rate": 1.870301693895961e-06,
-      "loss": 0.9761,
+      "epoch": 1.14,
+      "grad_norm": 9.547733306884766,
+      "learning_rate": 1.2432749027318747e-05,
+      "loss": 0.6108,
       "step": 9045
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 10.162015914916992,
-      "learning_rate": 1.8682970832915708e-06,
-      "loss": 1.0334,
+      "epoch": 1.14,
+      "grad_norm": 4.224358081817627,
+      "learning_rate": 1.243191231226206e-05,
+      "loss": 0.2862,
       "step": 9046
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 21.561569213867188,
-      "learning_rate": 1.8662924726871807e-06,
-      "loss": 0.7688,
+      "epoch": 1.14,
+      "grad_norm": 13.092987060546875,
+      "learning_rate": 1.2431075597205374e-05,
+      "loss": 1.6516,
       "step": 9047
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 8.59749698638916,
-      "learning_rate": 1.8642878620827906e-06,
-      "loss": 0.533,
+      "epoch": 1.14,
+      "grad_norm": 10.587092399597168,
+      "learning_rate": 1.2430238882148686e-05,
+      "loss": 1.8526,
       "step": 9048
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 17.311405181884766,
-      "learning_rate": 1.8622832514784007e-06,
-      "loss": 0.9795,
+      "epoch": 1.14,
+      "grad_norm": 19.686052322387695,
+      "learning_rate": 1.2429402167091998e-05,
+      "loss": 1.1136,
       "step": 9049
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 67.3037109375,
-      "learning_rate": 1.8602786408740103e-06,
-      "loss": 2.6467,
+      "epoch": 1.14,
+      "grad_norm": 8.99689769744873,
+      "learning_rate": 1.242856545203531e-05,
+      "loss": 2.2862,
       "step": 9050
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 12.96740436553955,
-      "learning_rate": 1.8582740302696204e-06,
-      "loss": 1.1722,
+      "epoch": 1.14,
+      "grad_norm": 12.086695671081543,
+      "learning_rate": 1.2427728736978623e-05,
+      "loss": 0.4117,
       "step": 9051
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 16.440349578857422,
-      "learning_rate": 1.85626941966523e-06,
-      "loss": 0.6379,
+      "epoch": 1.14,
+      "grad_norm": 18.265026092529297,
+      "learning_rate": 1.2426892021921935e-05,
+      "loss": 0.6975,
       "step": 9052
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 35.48670196533203,
-      "learning_rate": 1.85426480906084e-06,
-      "loss": 1.4359,
+      "epoch": 1.14,
+      "grad_norm": 25.84575653076172,
+      "learning_rate": 1.2426055306865247e-05,
+      "loss": 0.4512,
       "step": 9053
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 55.38397216796875,
-      "learning_rate": 1.85226019845645e-06,
-      "loss": 2.2976,
+      "epoch": 1.14,
+      "grad_norm": 20.959501266479492,
+      "learning_rate": 1.2425218591808561e-05,
+      "loss": 1.0482,
       "step": 9054
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 20.49527931213379,
-      "learning_rate": 1.8502555878520597e-06,
-      "loss": 1.1748,
+      "epoch": 1.14,
+      "grad_norm": 13.340580940246582,
+      "learning_rate": 1.2424381876751875e-05,
+      "loss": 0.9483,
       "step": 9055
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 11.808248519897461,
-      "learning_rate": 1.8482509772476698e-06,
-      "loss": 1.6608,
+      "epoch": 1.14,
+      "grad_norm": 9.03878116607666,
+      "learning_rate": 1.2423545161695185e-05,
+      "loss": 0.34,
       "step": 9056
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 25.69436264038086,
-      "learning_rate": 1.8462463666432799e-06,
-      "loss": 1.0124,
+      "epoch": 1.14,
+      "grad_norm": 5.133035182952881,
+      "learning_rate": 1.2422708446638499e-05,
+      "loss": 0.5256,
       "step": 9057
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 27.099803924560547,
-      "learning_rate": 1.8442417560388895e-06,
-      "loss": 1.4334,
+      "epoch": 1.14,
+      "grad_norm": 8.312919616699219,
+      "learning_rate": 1.2421871731581812e-05,
+      "loss": 2.1823,
       "step": 9058
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 16.551921844482422,
-      "learning_rate": 1.8422371454344994e-06,
-      "loss": 1.4909,
+      "epoch": 1.14,
+      "grad_norm": 10.040514945983887,
+      "learning_rate": 1.2421035016525122e-05,
+      "loss": 0.9566,
       "step": 9059
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 14.500848770141602,
-      "learning_rate": 1.8402325348301095e-06,
-      "loss": 1.1442,
+      "epoch": 1.14,
+      "grad_norm": 26.158926010131836,
+      "learning_rate": 1.2420198301468436e-05,
+      "loss": 1.1192,
       "step": 9060
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 14.154791831970215,
-      "learning_rate": 1.8382279242257191e-06,
-      "loss": 1.8214,
+      "epoch": 1.14,
+      "grad_norm": 18.3576717376709,
+      "learning_rate": 1.241936158641175e-05,
+      "loss": 1.3846,
       "step": 9061
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 28.4224796295166,
-      "learning_rate": 1.8362233136213292e-06,
-      "loss": 1.2986,
+      "epoch": 1.14,
+      "grad_norm": 106.99333953857422,
+      "learning_rate": 1.2418524871355062e-05,
+      "loss": 1.9476,
       "step": 9062
     },
     {
-      "epoch": 2.72,
-      "grad_norm": 14.508337020874023,
-      "learning_rate": 1.8342187030169393e-06,
-      "loss": 1.1846,
+      "epoch": 1.14,
+      "grad_norm": 19.000123977661133,
+      "learning_rate": 1.2417688156298374e-05,
+      "loss": 2.4688,
       "step": 9063
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 17.251667022705078,
-      "learning_rate": 1.832214092412549e-06,
-      "loss": 1.2499,
+      "epoch": 1.14,
+      "grad_norm": 7.971542835235596,
+      "learning_rate": 1.2416851441241686e-05,
+      "loss": 1.0279,
       "step": 9064
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 37.404449462890625,
-      "learning_rate": 1.8302094818081588e-06,
-      "loss": 0.9916,
+      "epoch": 1.14,
+      "grad_norm": 33.608123779296875,
+      "learning_rate": 1.2416014726185e-05,
+      "loss": 2.7088,
       "step": 9065
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 13.81652545928955,
-      "learning_rate": 1.828204871203769e-06,
-      "loss": 1.2755,
+      "epoch": 1.14,
+      "grad_norm": 38.017791748046875,
+      "learning_rate": 1.2415178011128311e-05,
+      "loss": 2.9421,
       "step": 9066
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 21.725704193115234,
-      "learning_rate": 1.8262002605993786e-06,
-      "loss": 1.1399,
+      "epoch": 1.14,
+      "grad_norm": 17.621917724609375,
+      "learning_rate": 1.2414341296071623e-05,
+      "loss": 1.0552,
       "step": 9067
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 10.298942565917969,
-      "learning_rate": 1.8241956499949887e-06,
-      "loss": 0.7941,
+      "epoch": 1.14,
+      "grad_norm": 11.231819152832031,
+      "learning_rate": 1.2413504581014937e-05,
+      "loss": 1.1976,
       "step": 9068
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 25.29086685180664,
-      "learning_rate": 1.8221910393905988e-06,
-      "loss": 1.3804,
+      "epoch": 1.14,
+      "grad_norm": 17.117918014526367,
+      "learning_rate": 1.241266786595825e-05,
+      "loss": 1.6858,
       "step": 9069
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 18.70489501953125,
-      "learning_rate": 1.8201864287862084e-06,
-      "loss": 1.1317,
+      "epoch": 1.14,
+      "grad_norm": 16.66362953186035,
+      "learning_rate": 1.241183115090156e-05,
+      "loss": 0.8505,
       "step": 9070
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 11.406014442443848,
-      "learning_rate": 1.8181818181818183e-06,
-      "loss": 0.4713,
+      "epoch": 1.14,
+      "grad_norm": 13.61717700958252,
+      "learning_rate": 1.2410994435844874e-05,
+      "loss": 3.3554,
       "step": 9071
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 22.935609817504883,
-      "learning_rate": 1.8161772075774282e-06,
-      "loss": 1.6272,
+      "epoch": 1.14,
+      "grad_norm": 8.80383586883545,
+      "learning_rate": 1.2410157720788188e-05,
+      "loss": 0.3408,
       "step": 9072
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 33.32746124267578,
-      "learning_rate": 1.814172596973038e-06,
-      "loss": 1.3237,
+      "epoch": 1.14,
+      "grad_norm": 9.178657531738281,
+      "learning_rate": 1.2409321005731498e-05,
+      "loss": 0.6548,
       "step": 9073
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 8.96728801727295,
-      "learning_rate": 1.8121679863686481e-06,
-      "loss": 0.556,
+      "epoch": 1.14,
+      "grad_norm": 50.2661247253418,
+      "learning_rate": 1.2408484290674812e-05,
+      "loss": 2.264,
       "step": 9074
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 23.04669761657715,
-      "learning_rate": 1.8101633757642578e-06,
-      "loss": 1.5039,
+      "epoch": 1.14,
+      "grad_norm": 17.373971939086914,
+      "learning_rate": 1.2407647575618124e-05,
+      "loss": 1.2542,
       "step": 9075
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 10.999143600463867,
-      "learning_rate": 1.8081587651598679e-06,
-      "loss": 1.7855,
+      "epoch": 1.14,
+      "grad_norm": 9.856574058532715,
+      "learning_rate": 1.2406810860561438e-05,
+      "loss": 0.8595,
       "step": 9076
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 21.91204261779785,
-      "learning_rate": 1.8061541545554777e-06,
-      "loss": 1.4705,
+      "epoch": 1.14,
+      "grad_norm": 27.904253005981445,
+      "learning_rate": 1.240597414550475e-05,
+      "loss": 2.2831,
       "step": 9077
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 31.69646644592285,
-      "learning_rate": 1.8041495439510876e-06,
-      "loss": 1.5004,
+      "epoch": 1.14,
+      "grad_norm": 30.6309814453125,
+      "learning_rate": 1.2405137430448061e-05,
+      "loss": 1.7967,
       "step": 9078
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 8.526796340942383,
-      "learning_rate": 1.8021449333466975e-06,
-      "loss": 0.8363,
+      "epoch": 1.14,
+      "grad_norm": 5.129134178161621,
+      "learning_rate": 1.2404300715391375e-05,
+      "loss": 1.6209,
       "step": 9079
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 91.82820129394531,
-      "learning_rate": 1.8001403227423076e-06,
-      "loss": 2.5966,
+      "epoch": 1.14,
+      "grad_norm": 15.763965606689453,
+      "learning_rate": 1.2403464000334687e-05,
+      "loss": 3.2273,
       "step": 9080
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 62.776432037353516,
-      "learning_rate": 1.7981357121379172e-06,
-      "loss": 2.0551,
+      "epoch": 1.14,
+      "grad_norm": 7.9168267250061035,
+      "learning_rate": 1.2402627285277999e-05,
+      "loss": 1.4169,
       "step": 9081
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 7.502096652984619,
-      "learning_rate": 1.7961311015335273e-06,
-      "loss": 0.9674,
+      "epoch": 1.14,
+      "grad_norm": 9.040215492248535,
+      "learning_rate": 1.2401790570221313e-05,
+      "loss": 2.3169,
       "step": 9082
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 13.519929885864258,
-      "learning_rate": 1.7941264909291372e-06,
-      "loss": 0.9669,
+      "epoch": 1.14,
+      "grad_norm": 12.345662117004395,
+      "learning_rate": 1.2400953855164626e-05,
+      "loss": 0.7656,
       "step": 9083
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 18.463733673095703,
-      "learning_rate": 1.792121880324747e-06,
-      "loss": 1.557,
+      "epoch": 1.14,
+      "grad_norm": 32.8636360168457,
+      "learning_rate": 1.2400117140107937e-05,
+      "loss": 3.5922,
       "step": 9084
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 18.185571670532227,
-      "learning_rate": 1.790117269720357e-06,
-      "loss": 1.9059,
+      "epoch": 1.14,
+      "grad_norm": 31.467174530029297,
+      "learning_rate": 1.239928042505125e-05,
+      "loss": 4.4755,
       "step": 9085
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 11.766922950744629,
-      "learning_rate": 1.788112659115967e-06,
-      "loss": 1.3468,
+      "epoch": 1.14,
+      "grad_norm": 16.0479679107666,
+      "learning_rate": 1.2398443709994564e-05,
+      "loss": 1.2256,
       "step": 9086
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 22.52054786682129,
-      "learning_rate": 1.7861080485115767e-06,
-      "loss": 0.9551,
+      "epoch": 1.14,
+      "grad_norm": 15.858236312866211,
+      "learning_rate": 1.2397606994937874e-05,
+      "loss": 1.4377,
       "step": 9087
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 8.950528144836426,
-      "learning_rate": 1.7841034379071868e-06,
-      "loss": 0.6039,
+      "epoch": 1.14,
+      "grad_norm": 19.002588272094727,
+      "learning_rate": 1.2396770279881188e-05,
+      "loss": 1.7684,
       "step": 9088
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 19.103931427001953,
-      "learning_rate": 1.7820988273027966e-06,
-      "loss": 1.0202,
+      "epoch": 1.14,
+      "grad_norm": 5.8403215408325195,
+      "learning_rate": 1.23959335648245e-05,
+      "loss": 0.4953,
       "step": 9089
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 14.955410957336426,
-      "learning_rate": 1.7800942166984065e-06,
-      "loss": 1.4588,
+      "epoch": 1.14,
+      "grad_norm": 10.064393043518066,
+      "learning_rate": 1.2395096849767813e-05,
+      "loss": 0.698,
       "step": 9090
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 49.647361755371094,
-      "learning_rate": 1.7780896060940164e-06,
-      "loss": 3.6037,
+      "epoch": 1.14,
+      "grad_norm": 14.260061264038086,
+      "learning_rate": 1.2394260134711125e-05,
+      "loss": 1.5792,
       "step": 9091
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 111.04601287841797,
-      "learning_rate": 1.7760849954896263e-06,
-      "loss": 1.5603,
+      "epoch": 1.14,
+      "grad_norm": 24.716506958007812,
+      "learning_rate": 1.2393423419654437e-05,
+      "loss": 1.5605,
       "step": 9092
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 49.5386962890625,
-      "learning_rate": 1.7740803848852361e-06,
-      "loss": 2.5773,
+      "epoch": 1.14,
+      "grad_norm": 8.207243919372559,
+      "learning_rate": 1.2392586704597751e-05,
+      "loss": 0.3092,
       "step": 9093
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 12.374490737915039,
-      "learning_rate": 1.7720757742808462e-06,
-      "loss": 1.1015,
+      "epoch": 1.14,
+      "grad_norm": 12.12687873840332,
+      "learning_rate": 1.2391749989541061e-05,
+      "loss": 1.008,
       "step": 9094
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 9.451532363891602,
-      "learning_rate": 1.7700711636764559e-06,
-      "loss": 0.4628,
+      "epoch": 1.14,
+      "grad_norm": 13.592269897460938,
+      "learning_rate": 1.2390913274484375e-05,
+      "loss": 1.4859,
       "step": 9095
     },
     {
-      "epoch": 2.73,
-      "grad_norm": 13.187180519104004,
-      "learning_rate": 1.768066553072066e-06,
-      "loss": 0.8442,
+      "epoch": 1.14,
+      "grad_norm": 27.075054168701172,
+      "learning_rate": 1.2390076559427688e-05,
+      "loss": 2.4585,
       "step": 9096
     },
     {
-      "epoch": 2.74,
-      "grad_norm": 12.320247650146484,
-      "learning_rate": 1.7660619424676758e-06,
-      "loss": 0.88,
+      "epoch": 1.14,
+      "grad_norm": 13.686470031738281,
+      "learning_rate": 1.2389239844371002e-05,
+      "loss": 1.3728,
       "step": 9097
     },
     {
-      "epoch": 2.74,
-      "grad_norm": 8.388055801391602,
-      "learning_rate": 1.7640573318632857e-06,
-      "loss": 0.7241,
+      "epoch": 1.14,
+      "grad_norm": 16.724834442138672,
+      "learning_rate": 1.2388403129314312e-05,
+      "loss": 1.2046,
       "step": 9098
     },
     {
-      "epoch": 2.74,
-      "grad_norm": 11.43036937713623,
-      "learning_rate": 1.7620527212588956e-06,
-      "loss": 0.8333,
+      "epoch": 1.14,
+      "grad_norm": 53.69195556640625,
+      "learning_rate": 1.2387566414257626e-05,
+      "loss": 1.6093,
       "step": 9099
     },
     {
-      "epoch": 2.74,
-      "grad_norm": 8.495409965515137,
-      "learning_rate": 1.7600481106545057e-06,
-      "loss": 1.1459,
+      "epoch": 1.14,
+      "grad_norm": 6.855299949645996,
+      "learning_rate": 1.238672969920094e-05,
+      "loss": 0.5229,
       "step": 9100
     },
     {
-      "epoch": 2.74,
-      "grad_norm": 11.421164512634277,
-      "learning_rate": 1.7580435000501153e-06,
-      "loss": 1.0815,
+      "epoch": 1.14,
+      "grad_norm": 19.45718002319336,
+      "learning_rate": 1.238589298414425e-05,
+      "loss": 1.5701,
       "step": 9101
     },
     {
-      "epoch": 2.74,
-      "grad_norm": 17.927734375,
-      "learning_rate": 1.7560388894457254e-06,
-      "loss": 1.1274,
+      "epoch": 1.14,
+      "grad_norm": 19.386138916015625,
+      "learning_rate": 1.2385056269087564e-05,
+      "loss": 0.9297,
       "step": 9102
     },
     {
-      "epoch": 2.74,
-      "grad_norm": 22.090795516967773,
-      "learning_rate": 1.7540342788413353e-06,
-      "loss": 0.8337,
+      "epoch": 1.14,
+      "grad_norm": 17.744094848632812,
+      "learning_rate": 1.2384219554030876e-05,
+      "loss": 1.6128,
       "step": 9103
     },
     {
-      "epoch": 2.74,
-      "grad_norm": 11.036092758178711,
-      "learning_rate": 1.752029668236945e-06,
-      "loss": 1.3825,
+      "epoch": 1.14,
+      "grad_norm": 8.07080364227295,
+      "learning_rate": 1.2383382838974188e-05,
+      "loss": 0.6567,
       "step": 9104
     },
     {
-      "epoch": 2.74,
-      "grad_norm": 10.002640724182129,
-      "learning_rate": 1.750025057632555e-06,
-      "loss": 1.174,
+      "epoch": 1.14,
+      "grad_norm": 23.605592727661133,
+      "learning_rate": 1.2382546123917501e-05,
+      "loss": 3.0387,
       "step": 9105
     },
     {
-      "epoch": 2.74,
-      "grad_norm": 13.688326835632324,
-      "learning_rate": 1.748020447028165e-06,
-      "loss": 1.0106,
+      "epoch": 1.14,
+      "grad_norm": 10.705403327941895,
+      "learning_rate": 1.2381709408860813e-05,
+      "loss": 0.4005,
       "step": 9106
     },
     {
-      "epoch": 2.74,
-      "grad_norm": 18.911972045898438,
-      "learning_rate": 1.7460158364237748e-06,
-      "loss": 1.3007,
+      "epoch": 1.14,
+      "grad_norm": 9.318358421325684,
+      "learning_rate": 1.2380872693804127e-05,
+      "loss": 0.8697,
       "step": 9107
     },
     {
-      "epoch": 2.74,
-      "grad_norm": 28.944271087646484,
-      "learning_rate": 1.7440112258193848e-06,
-      "loss": 1.5335,
+      "epoch": 1.14,
+      "grad_norm": 12.265621185302734,
+      "learning_rate": 1.2380035978747437e-05,
+      "loss": 0.9076,
       "step": 9108
     },
     {
-      "epoch": 2.74,
-      "grad_norm": 17.66625213623047,
-      "learning_rate": 1.7420066152149947e-06,
-      "loss": 0.7189,
+      "epoch": 1.14,
+      "grad_norm": 8.256604194641113,
+      "learning_rate": 1.237919926369075e-05,
+      "loss": 2.0401,
       "step": 9109
     },
     {
-      "epoch": 2.74,
-      "grad_norm": 28.595500946044922,
-      "learning_rate": 1.7400020046106044e-06,
-      "loss": 1.4251,
+      "epoch": 1.14,
+      "grad_norm": 12.32669734954834,
+      "learning_rate": 1.2378362548634064e-05,
+      "loss": 1.6128,
       "step": 9110
     },
     {
-      "epoch": 2.74,
-      "grad_norm": 11.355489730834961,
-      "learning_rate": 1.7379973940062145e-06,
-      "loss": 0.8031,
+      "epoch": 1.14,
+      "grad_norm": 10.947827339172363,
+      "learning_rate": 1.2377525833577375e-05,
+      "loss": 1.6705,
       "step": 9111
     },
     {
-      "epoch": 2.74,
-      "grad_norm": 20.97258949279785,
-      "learning_rate": 1.7359927834018241e-06,
-      "loss": 0.8911,
+      "epoch": 1.14,
+      "grad_norm": 29.249637603759766,
+      "learning_rate": 1.2376689118520688e-05,
+      "loss": 1.42,
       "step": 9112
     },
     {
-      "epoch": 2.74,
-      "grad_norm": 21.422672271728516,
-      "learning_rate": 1.7339881727974342e-06,
-      "loss": 1.5909,
+      "epoch": 1.14,
+      "grad_norm": 5.854145526885986,
+      "learning_rate": 1.2375852403464002e-05,
+      "loss": 0.3727,
       "step": 9113
     },
     {
-      "epoch": 2.74,
-      "grad_norm": 16.778446197509766,
-      "learning_rate": 1.7319835621930443e-06,
-      "loss": 0.5871,
+      "epoch": 1.14,
+      "grad_norm": 8.388299942016602,
+      "learning_rate": 1.2375015688407316e-05,
+      "loss": 0.7125,
       "step": 9114
     },
     {
-      "epoch": 2.74,
-      "grad_norm": 14.981059074401855,
-      "learning_rate": 1.729978951588654e-06,
-      "loss": 0.938,
+      "epoch": 1.14,
+      "grad_norm": 5.731067180633545,
+      "learning_rate": 1.2374178973350626e-05,
+      "loss": 0.5314,
       "step": 9115
     },
     {
-      "epoch": 2.74,
-      "grad_norm": 17.02187156677246,
-      "learning_rate": 1.7279743409842638e-06,
-      "loss": 1.4252,
+      "epoch": 1.14,
+      "grad_norm": 7.892755031585693,
+      "learning_rate": 1.237334225829394e-05,
+      "loss": 0.8244,
       "step": 9116
     },
     {
-      "epoch": 2.74,
-      "grad_norm": 20.162086486816406,
-      "learning_rate": 1.725969730379874e-06,
-      "loss": 0.9187,
+      "epoch": 1.14,
+      "grad_norm": 20.525279998779297,
+      "learning_rate": 1.2372505543237251e-05,
+      "loss": 2.1919,
       "step": 9117
     },
     {
-      "epoch": 2.74,
-      "grad_norm": 22.232948303222656,
-      "learning_rate": 1.7239651197754836e-06,
-      "loss": 1.1912,
+      "epoch": 1.14,
+      "grad_norm": 13.459932327270508,
+      "learning_rate": 1.2371668828180563e-05,
+      "loss": 1.0175,
       "step": 9118
     },
     {
-      "epoch": 2.74,
-      "grad_norm": 21.19387435913086,
-      "learning_rate": 1.7219605091710937e-06,
-      "loss": 0.7006,
+      "epoch": 1.14,
+      "grad_norm": 6.225473880767822,
+      "learning_rate": 1.2370832113123877e-05,
+      "loss": 0.7936,
       "step": 9119
     },
     {
-      "epoch": 2.74,
-      "grad_norm": 11.553994178771973,
-      "learning_rate": 1.7199558985667037e-06,
-      "loss": 0.7409,
+      "epoch": 1.14,
+      "grad_norm": 19.290508270263672,
+      "learning_rate": 1.2369995398067189e-05,
+      "loss": 2.8995,
       "step": 9120
     },
     {
-      "epoch": 2.74,
-      "eval_loss": 0.16245336830615997,
-      "eval_runtime": 43.6231,
-      "eval_samples_per_second": 33.904,
-      "eval_steps_per_second": 33.904,
-      "step": 9120
+      "epoch": 1.14,
+      "grad_norm": 19.17864227294922,
+      "learning_rate": 1.2369158683010503e-05,
+      "loss": 0.7154,
+      "step": 9121
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 6.110776424407959,
+      "learning_rate": 1.2368321967953813e-05,
+      "loss": 0.2638,
+      "step": 9122
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 13.87114143371582,
+      "learning_rate": 1.2367485252897127e-05,
+      "loss": 1.0967,
+      "step": 9123
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.471000671386719,
+      "learning_rate": 1.236664853784044e-05,
+      "loss": 1.3324,
+      "step": 9124
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.963351726531982,
+      "learning_rate": 1.236581182278375e-05,
+      "loss": 1.545,
+      "step": 9125
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.431926727294922,
+      "learning_rate": 1.2364975107727064e-05,
+      "loss": 0.3971,
+      "step": 9126
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 26.25963592529297,
+      "learning_rate": 1.2364138392670378e-05,
+      "loss": 2.6638,
+      "step": 9127
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 6.95812463760376,
+      "learning_rate": 1.236330167761369e-05,
+      "loss": 1.4096,
+      "step": 9128
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 24.770366668701172,
+      "learning_rate": 1.2362464962557002e-05,
+      "loss": 2.2656,
+      "step": 9129
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.823984146118164,
+      "learning_rate": 1.2361628247500315e-05,
+      "loss": 1.6969,
+      "step": 9130
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.5049519538879395,
+      "learning_rate": 1.2360791532443627e-05,
+      "loss": 2.208,
+      "step": 9131
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 18.726926803588867,
+      "learning_rate": 1.235995481738694e-05,
+      "loss": 1.0615,
+      "step": 9132
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 18.43231964111328,
+      "learning_rate": 1.2359118102330253e-05,
+      "loss": 1.283,
+      "step": 9133
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 20.206207275390625,
+      "learning_rate": 1.2358281387273565e-05,
+      "loss": 1.5557,
+      "step": 9134
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.573138236999512,
+      "learning_rate": 1.2357444672216878e-05,
+      "loss": 0.3381,
+      "step": 9135
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.645369529724121,
+      "learning_rate": 1.2356607957160189e-05,
+      "loss": 1.2922,
+      "step": 9136
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 49.07716369628906,
+      "learning_rate": 1.2355771242103502e-05,
+      "loss": 3.4945,
+      "step": 9137
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.765970706939697,
+      "learning_rate": 1.2354934527046816e-05,
+      "loss": 0.4129,
+      "step": 9138
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 34.59571075439453,
+      "learning_rate": 1.2354097811990126e-05,
+      "loss": 2.1463,
+      "step": 9139
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 25.068452835083008,
+      "learning_rate": 1.235326109693344e-05,
+      "loss": 1.5953,
+      "step": 9140
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 18.312746047973633,
+      "learning_rate": 1.2352424381876754e-05,
+      "loss": 2.3889,
+      "step": 9141
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.369312286376953,
+      "learning_rate": 1.2351587666820066e-05,
+      "loss": 0.1455,
+      "step": 9142
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 13.723098754882812,
+      "learning_rate": 1.2350750951763377e-05,
+      "loss": 0.9427,
+      "step": 9143
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 40.40791320800781,
+      "learning_rate": 1.2349914236706691e-05,
+      "loss": 1.8195,
+      "step": 9144
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 20.710609436035156,
+      "learning_rate": 1.2349077521650003e-05,
+      "loss": 2.6318,
+      "step": 9145
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.522031307220459,
+      "learning_rate": 1.2348240806593315e-05,
+      "loss": 0.2665,
+      "step": 9146
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.169207572937012,
+      "learning_rate": 1.2347404091536627e-05,
+      "loss": 0.6525,
+      "step": 9147
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 14.694474220275879,
+      "learning_rate": 1.234656737647994e-05,
+      "loss": 0.767,
+      "step": 9148
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 19.8188533782959,
+      "learning_rate": 1.2345730661423254e-05,
+      "loss": 2.2416,
+      "step": 9149
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 24.8552188873291,
+      "learning_rate": 1.2344893946366565e-05,
+      "loss": 1.8361,
+      "step": 9150
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 22.294574737548828,
+      "learning_rate": 1.2344057231309878e-05,
+      "loss": 1.0202,
+      "step": 9151
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 23.81131362915039,
+      "learning_rate": 1.2343220516253192e-05,
+      "loss": 2.1269,
+      "step": 9152
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 19.063304901123047,
+      "learning_rate": 1.2342383801196502e-05,
+      "loss": 1.2329,
+      "step": 9153
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.666409492492676,
+      "learning_rate": 1.2341547086139816e-05,
+      "loss": 1.3701,
+      "step": 9154
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.337384223937988,
+      "learning_rate": 1.234071037108313e-05,
+      "loss": 0.4143,
+      "step": 9155
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 39.719398498535156,
+      "learning_rate": 1.2339873656026441e-05,
+      "loss": 1.3722,
+      "step": 9156
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 13.493542671203613,
+      "learning_rate": 1.2339036940969753e-05,
+      "loss": 1.4881,
+      "step": 9157
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.449769973754883,
+      "learning_rate": 1.2338200225913067e-05,
+      "loss": 0.882,
+      "step": 9158
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 14.406185150146484,
+      "learning_rate": 1.2337363510856379e-05,
+      "loss": 1.1041,
+      "step": 9159
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.488336563110352,
+      "learning_rate": 1.2336526795799691e-05,
+      "loss": 0.9631,
+      "step": 9160
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 21.735170364379883,
+      "learning_rate": 1.2335690080743003e-05,
+      "loss": 2.287,
+      "step": 9161
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 19.89383316040039,
+      "learning_rate": 1.2334853365686316e-05,
+      "loss": 1.5426,
+      "step": 9162
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 142.1265106201172,
+      "learning_rate": 1.233401665062963e-05,
+      "loss": 2.0707,
+      "step": 9163
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.750329971313477,
+      "learning_rate": 1.233317993557294e-05,
+      "loss": 1.1023,
+      "step": 9164
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.105993270874023,
+      "learning_rate": 1.2332343220516254e-05,
+      "loss": 1.6093,
+      "step": 9165
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 13.682340621948242,
+      "learning_rate": 1.2331506505459568e-05,
+      "loss": 0.7277,
+      "step": 9166
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 14.506525993347168,
+      "learning_rate": 1.2330669790402878e-05,
+      "loss": 1.3144,
+      "step": 9167
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.809659004211426,
+      "learning_rate": 1.2329833075346192e-05,
+      "loss": 0.4419,
+      "step": 9168
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 35.33649826049805,
+      "learning_rate": 1.2328996360289505e-05,
+      "loss": 0.847,
+      "step": 9169
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 14.611030578613281,
+      "learning_rate": 1.2328159645232817e-05,
+      "loss": 1.426,
+      "step": 9170
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 22.434511184692383,
+      "learning_rate": 1.2327322930176129e-05,
+      "loss": 1.5308,
+      "step": 9171
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 6.115286827087402,
+      "learning_rate": 1.2326486215119443e-05,
+      "loss": 0.9465,
+      "step": 9172
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.794254302978516,
+      "learning_rate": 1.2325649500062755e-05,
+      "loss": 1.4792,
+      "step": 9173
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.999887466430664,
+      "learning_rate": 1.2324812785006067e-05,
+      "loss": 1.8432,
+      "step": 9174
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.971545696258545,
+      "learning_rate": 1.2323976069949379e-05,
+      "loss": 0.3954,
+      "step": 9175
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.400793075561523,
+      "learning_rate": 1.2323139354892692e-05,
+      "loss": 1.5969,
+      "step": 9176
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.92943000793457,
+      "learning_rate": 1.2322302639836006e-05,
+      "loss": 2.1288,
+      "step": 9177
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 6.143886089324951,
+      "learning_rate": 1.2321465924779316e-05,
+      "loss": 1.2216,
+      "step": 9178
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.842630386352539,
+      "learning_rate": 1.232062920972263e-05,
+      "loss": 0.4645,
+      "step": 9179
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 27.85214614868164,
+      "learning_rate": 1.2319792494665944e-05,
+      "loss": 1.5454,
+      "step": 9180
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 14.218128204345703,
+      "learning_rate": 1.2318955779609254e-05,
+      "loss": 2.8992,
+      "step": 9181
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 14.0691556930542,
+      "learning_rate": 1.2318119064552567e-05,
+      "loss": 1.0916,
+      "step": 9182
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 24.324790954589844,
+      "learning_rate": 1.2317282349495881e-05,
+      "loss": 2.3248,
+      "step": 9183
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 33.42656707763672,
+      "learning_rate": 1.2316445634439193e-05,
+      "loss": 3.127,
+      "step": 9184
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 6.54421329498291,
+      "learning_rate": 1.2315608919382505e-05,
+      "loss": 2.6659,
+      "step": 9185
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.281428337097168,
+      "learning_rate": 1.2314772204325817e-05,
+      "loss": 1.0513,
+      "step": 9186
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.191452980041504,
+      "learning_rate": 1.231393548926913e-05,
+      "loss": 0.6355,
+      "step": 9187
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 16.978662490844727,
+      "learning_rate": 1.2313098774212443e-05,
+      "loss": 1.0362,
+      "step": 9188
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 40.00162887573242,
+      "learning_rate": 1.2312262059155754e-05,
+      "loss": 1.1716,
+      "step": 9189
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 14.924356460571289,
+      "learning_rate": 1.2311425344099068e-05,
+      "loss": 1.7509,
+      "step": 9190
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 9.85426139831543,
+      "learning_rate": 1.2310588629042382e-05,
+      "loss": 0.3192,
+      "step": 9191
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.907785415649414,
+      "learning_rate": 1.2309751913985692e-05,
+      "loss": 1.013,
+      "step": 9192
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.675817489624023,
+      "learning_rate": 1.2308915198929006e-05,
+      "loss": 0.9179,
+      "step": 9193
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 42.699310302734375,
+      "learning_rate": 1.230807848387232e-05,
+      "loss": 2.2475,
+      "step": 9194
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 20.156429290771484,
+      "learning_rate": 1.230724176881563e-05,
+      "loss": 2.1368,
+      "step": 9195
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.853020668029785,
+      "learning_rate": 1.2306405053758943e-05,
+      "loss": 0.9512,
+      "step": 9196
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.198304176330566,
+      "learning_rate": 1.2305568338702257e-05,
+      "loss": 0.6033,
+      "step": 9197
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.125846862792969,
+      "learning_rate": 1.2304731623645569e-05,
+      "loss": 0.7922,
+      "step": 9198
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 34.10848617553711,
+      "learning_rate": 1.230389490858888e-05,
+      "loss": 1.4218,
+      "step": 9199
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 11.90444564819336,
+      "learning_rate": 1.2303058193532193e-05,
+      "loss": 1.9948,
+      "step": 9200
+    },
+    {
+      "epoch": 1.15,
+      "eval_loss": 0.09518160670995712,
+      "eval_runtime": 94.306,
+      "eval_samples_per_second": 37.559,
+      "eval_steps_per_second": 37.559,
+      "step": 9200
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.9429755210876465,
+      "learning_rate": 1.2302221478475506e-05,
+      "loss": 0.9086,
+      "step": 9201
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 23.948022842407227,
+      "learning_rate": 1.2301384763418818e-05,
+      "loss": 1.653,
+      "step": 9202
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 27.833898544311523,
+      "learning_rate": 1.230054804836213e-05,
+      "loss": 1.4532,
+      "step": 9203
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.013521194458008,
+      "learning_rate": 1.2299711333305444e-05,
+      "loss": 0.9502,
+      "step": 9204
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 6.075160503387451,
+      "learning_rate": 1.2298874618248758e-05,
+      "loss": 1.1342,
+      "step": 9205
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 30.68280792236328,
+      "learning_rate": 1.2298037903192068e-05,
+      "loss": 1.3133,
+      "step": 9206
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.82780933380127,
+      "learning_rate": 1.2297201188135382e-05,
+      "loss": 1.5006,
+      "step": 9207
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 5.6558427810668945,
+      "learning_rate": 1.2296364473078695e-05,
+      "loss": 0.924,
+      "step": 9208
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 5.863631725311279,
+      "learning_rate": 1.2295527758022005e-05,
+      "loss": 0.4676,
+      "step": 9209
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 19.110694885253906,
+      "learning_rate": 1.2294691042965319e-05,
+      "loss": 1.1357,
+      "step": 9210
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 66.21870422363281,
+      "learning_rate": 1.2293854327908633e-05,
+      "loss": 1.5988,
+      "step": 9211
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.78620719909668,
+      "learning_rate": 1.2293017612851945e-05,
+      "loss": 1.1191,
+      "step": 9212
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 12.453399658203125,
+      "learning_rate": 1.2292180897795257e-05,
+      "loss": 0.5343,
+      "step": 9213
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 19.970155715942383,
+      "learning_rate": 1.2291344182738569e-05,
+      "loss": 0.6287,
+      "step": 9214
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.808201789855957,
+      "learning_rate": 1.2290507467681882e-05,
+      "loss": 1.2625,
+      "step": 9215
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 16.423688888549805,
+      "learning_rate": 1.2289670752625194e-05,
+      "loss": 0.4894,
+      "step": 9216
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 14.505757331848145,
+      "learning_rate": 1.2288834037568506e-05,
+      "loss": 1.3869,
+      "step": 9217
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.26084303855896,
+      "learning_rate": 1.228799732251182e-05,
+      "loss": 0.1848,
+      "step": 9218
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 96.6875228881836,
+      "learning_rate": 1.2287160607455133e-05,
+      "loss": 1.6699,
+      "step": 9219
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 131.69342041015625,
+      "learning_rate": 1.2286323892398444e-05,
+      "loss": 2.1287,
+      "step": 9220
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.78143310546875,
+      "learning_rate": 1.2285487177341757e-05,
+      "loss": 0.4735,
+      "step": 9221
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 20.10757827758789,
+      "learning_rate": 1.2284650462285071e-05,
+      "loss": 2.2554,
+      "step": 9222
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.9812092781066895,
+      "learning_rate": 1.2283813747228381e-05,
+      "loss": 0.4739,
+      "step": 9223
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.661271810531616,
+      "learning_rate": 1.2282977032171695e-05,
+      "loss": 0.2297,
+      "step": 9224
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 16.200122833251953,
+      "learning_rate": 1.2282140317115009e-05,
+      "loss": 3.4471,
+      "step": 9225
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.316689491271973,
+      "learning_rate": 1.228130360205832e-05,
+      "loss": 0.4403,
+      "step": 9226
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.576048374176025,
+      "learning_rate": 1.2280466887001632e-05,
+      "loss": 0.1168,
+      "step": 9227
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 6.368162155151367,
+      "learning_rate": 1.2279630171944944e-05,
+      "loss": 0.5202,
+      "step": 9228
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 102.35724639892578,
+      "learning_rate": 1.2278793456888258e-05,
+      "loss": 1.4843,
+      "step": 9229
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 12.44650650024414,
+      "learning_rate": 1.227795674183157e-05,
+      "loss": 2.0575,
+      "step": 9230
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.62086296081543,
+      "learning_rate": 1.2277120026774882e-05,
+      "loss": 1.6411,
+      "step": 9231
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 35.608333587646484,
+      "learning_rate": 1.2276283311718196e-05,
+      "loss": 0.4515,
+      "step": 9232
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 29.228580474853516,
+      "learning_rate": 1.227544659666151e-05,
+      "loss": 2.278,
+      "step": 9233
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 14.064435005187988,
+      "learning_rate": 1.227460988160482e-05,
+      "loss": 2.7028,
+      "step": 9234
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.022710800170898,
+      "learning_rate": 1.2273773166548133e-05,
+      "loss": 0.677,
+      "step": 9235
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.585491180419922,
+      "learning_rate": 1.2272936451491447e-05,
+      "loss": 1.2076,
+      "step": 9236
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 12.864282608032227,
+      "learning_rate": 1.2272099736434757e-05,
+      "loss": 1.8201,
+      "step": 9237
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 35.674781799316406,
+      "learning_rate": 1.227126302137807e-05,
+      "loss": 1.6032,
+      "step": 9238
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 20.760175704956055,
+      "learning_rate": 1.2270426306321383e-05,
+      "loss": 2.4856,
+      "step": 9239
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.651836395263672,
+      "learning_rate": 1.2269589591264696e-05,
+      "loss": 0.824,
+      "step": 9240
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 15.389017105102539,
+      "learning_rate": 1.2268752876208008e-05,
+      "loss": 1.4099,
+      "step": 9241
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 15.191420555114746,
+      "learning_rate": 1.226791616115132e-05,
+      "loss": 2.6674,
+      "step": 9242
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 23.26456069946289,
+      "learning_rate": 1.2267079446094634e-05,
+      "loss": 2.9811,
+      "step": 9243
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.608368873596191,
+      "learning_rate": 1.2266242731037946e-05,
+      "loss": 1.2524,
+      "step": 9244
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 14.599865913391113,
+      "learning_rate": 1.2265406015981258e-05,
+      "loss": 0.48,
+      "step": 9245
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.090410232543945,
+      "learning_rate": 1.2264569300924571e-05,
+      "loss": 0.8899,
+      "step": 9246
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 23.541912078857422,
+      "learning_rate": 1.2263732585867885e-05,
+      "loss": 1.5031,
+      "step": 9247
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 17.355106353759766,
+      "learning_rate": 1.2262895870811195e-05,
+      "loss": 1.131,
+      "step": 9248
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.4488019943237305,
+      "learning_rate": 1.2262059155754509e-05,
+      "loss": 0.4096,
+      "step": 9249
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 37.859893798828125,
+      "learning_rate": 1.2261222440697823e-05,
+      "loss": 1.2356,
+      "step": 9250
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.11927604675293,
+      "learning_rate": 1.2260385725641133e-05,
+      "loss": 0.4152,
+      "step": 9251
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 107.45669555664062,
+      "learning_rate": 1.2259549010584447e-05,
+      "loss": 2.2546,
+      "step": 9252
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 19.212947845458984,
+      "learning_rate": 1.2258712295527759e-05,
+      "loss": 1.3481,
+      "step": 9253
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 15.844950675964355,
+      "learning_rate": 1.2257875580471072e-05,
+      "loss": 0.9506,
+      "step": 9254
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 32.015132904052734,
+      "learning_rate": 1.2257038865414384e-05,
+      "loss": 1.5036,
+      "step": 9255
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 27.190282821655273,
+      "learning_rate": 1.2256202150357696e-05,
+      "loss": 2.8504,
+      "step": 9256
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.14805793762207,
+      "learning_rate": 1.225536543530101e-05,
+      "loss": 0.7194,
+      "step": 9257
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 15.691932678222656,
+      "learning_rate": 1.225452872024432e-05,
+      "loss": 1.5595,
+      "step": 9258
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 15.650479316711426,
+      "learning_rate": 1.2253692005187634e-05,
+      "loss": 0.4911,
+      "step": 9259
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.42494535446167,
+      "learning_rate": 1.2252855290130947e-05,
+      "loss": 0.4337,
+      "step": 9260
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 19.028564453125,
+      "learning_rate": 1.2252018575074261e-05,
+      "loss": 0.6831,
+      "step": 9261
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.2690110206604,
+      "learning_rate": 1.2251181860017571e-05,
+      "loss": 0.4737,
+      "step": 9262
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 26.352842330932617,
+      "learning_rate": 1.2250345144960885e-05,
+      "loss": 1.4589,
+      "step": 9263
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 31.72899627685547,
+      "learning_rate": 1.2249508429904199e-05,
+      "loss": 1.9815,
+      "step": 9264
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 33.27836608886719,
+      "learning_rate": 1.2248671714847509e-05,
+      "loss": 2.9694,
+      "step": 9265
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 17.309558868408203,
+      "learning_rate": 1.2247834999790822e-05,
+      "loss": 1.6879,
+      "step": 9266
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.14188003540039,
+      "learning_rate": 1.2246998284734134e-05,
+      "loss": 1.1734,
+      "step": 9267
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 36.28446578979492,
+      "learning_rate": 1.2246161569677448e-05,
+      "loss": 2.6889,
+      "step": 9268
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 8.997748374938965,
+      "learning_rate": 1.224532485462076e-05,
+      "loss": 1.4469,
+      "step": 9269
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 5.327282428741455,
+      "learning_rate": 1.2244488139564072e-05,
+      "loss": 0.3503,
+      "step": 9270
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.748549461364746,
+      "learning_rate": 1.2243651424507386e-05,
+      "loss": 0.467,
+      "step": 9271
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 46.473114013671875,
+      "learning_rate": 1.2242814709450696e-05,
+      "loss": 1.4334,
+      "step": 9272
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 15.982047080993652,
+      "learning_rate": 1.224197799439401e-05,
+      "loss": 2.7342,
+      "step": 9273
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.6621732711792,
+      "learning_rate": 1.2241141279337323e-05,
+      "loss": 1.6504,
+      "step": 9274
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 63.71792984008789,
+      "learning_rate": 1.2240304564280637e-05,
+      "loss": 1.4364,
+      "step": 9275
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.43286418914795,
+      "learning_rate": 1.2239467849223947e-05,
+      "loss": 0.2358,
+      "step": 9276
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 88.21047973632812,
+      "learning_rate": 1.223863113416726e-05,
+      "loss": 0.829,
+      "step": 9277
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 11.336211204528809,
+      "learning_rate": 1.2237794419110574e-05,
+      "loss": 0.9357,
+      "step": 9278
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.616588592529297,
+      "learning_rate": 1.2236957704053885e-05,
+      "loss": 0.3243,
+      "step": 9279
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 9.622396469116211,
+      "learning_rate": 1.2236120988997198e-05,
+      "loss": 0.981,
+      "step": 9280
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 10.186108589172363,
+      "learning_rate": 1.223528427394051e-05,
+      "loss": 1.0945,
+      "step": 9281
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.330602169036865,
+      "learning_rate": 1.2234447558883824e-05,
+      "loss": 0.5163,
+      "step": 9282
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.188134670257568,
+      "learning_rate": 1.2233610843827136e-05,
+      "loss": 0.4402,
+      "step": 9283
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 51.132850646972656,
+      "learning_rate": 1.2232774128770448e-05,
+      "loss": 3.2306,
+      "step": 9284
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 5.105656623840332,
+      "learning_rate": 1.2231937413713761e-05,
+      "loss": 0.5543,
+      "step": 9285
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 17.768465042114258,
+      "learning_rate": 1.2231100698657072e-05,
+      "loss": 2.3156,
+      "step": 9286
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 19.271150588989258,
+      "learning_rate": 1.2230263983600385e-05,
+      "loss": 0.8452,
+      "step": 9287
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 134.21510314941406,
+      "learning_rate": 1.2229427268543699e-05,
+      "loss": 3.0031,
+      "step": 9288
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.793252944946289,
+      "learning_rate": 1.2228590553487013e-05,
+      "loss": 2.1216,
+      "step": 9289
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 42.16161346435547,
+      "learning_rate": 1.2227753838430323e-05,
+      "loss": 0.4732,
+      "step": 9290
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.013446807861328,
+      "learning_rate": 1.2226917123373637e-05,
+      "loss": 1.3425,
+      "step": 9291
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 24.325328826904297,
+      "learning_rate": 1.2226080408316949e-05,
+      "loss": 2.1346,
+      "step": 9292
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 15.697874069213867,
+      "learning_rate": 1.222524369326026e-05,
+      "loss": 1.4914,
+      "step": 9293
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.737473487854004,
+      "learning_rate": 1.2224406978203574e-05,
+      "loss": 1.0438,
+      "step": 9294
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 35.751625061035156,
+      "learning_rate": 1.2223570263146886e-05,
+      "loss": 2.2221,
+      "step": 9295
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 120.206787109375,
+      "learning_rate": 1.22227335480902e-05,
+      "loss": 3.6724,
+      "step": 9296
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 36.27464294433594,
+      "learning_rate": 1.222189683303351e-05,
+      "loss": 1.2908,
+      "step": 9297
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.949959754943848,
+      "learning_rate": 1.2221060117976824e-05,
+      "loss": 1.3326,
+      "step": 9298
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 15.542511940002441,
+      "learning_rate": 1.2220223402920137e-05,
+      "loss": 1.6399,
+      "step": 9299
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 31.601863861083984,
+      "learning_rate": 1.2219386687863448e-05,
+      "loss": 1.898,
+      "step": 9300
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 5.258601665496826,
+      "learning_rate": 1.2218549972806761e-05,
+      "loss": 0.6207,
+      "step": 9301
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 7.814626693725586,
+      "learning_rate": 1.2217713257750075e-05,
+      "loss": 1.6168,
+      "step": 9302
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 16.59807777404785,
+      "learning_rate": 1.2216876542693388e-05,
+      "loss": 1.2187,
+      "step": 9303
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 21.79674530029297,
+      "learning_rate": 1.2216039827636699e-05,
+      "loss": 2.087,
+      "step": 9304
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.025087356567383,
+      "learning_rate": 1.2215203112580012e-05,
+      "loss": 0.156,
+      "step": 9305
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 16.624496459960938,
+      "learning_rate": 1.2214366397523324e-05,
+      "loss": 1.5409,
+      "step": 9306
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 46.0156135559082,
+      "learning_rate": 1.2213529682466636e-05,
+      "loss": 0.845,
+      "step": 9307
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 6.28947114944458,
+      "learning_rate": 1.221269296740995e-05,
+      "loss": 1.1145,
+      "step": 9308
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 7.662294864654541,
+      "learning_rate": 1.2211856252353262e-05,
+      "loss": 0.4594,
+      "step": 9309
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.82614517211914,
+      "learning_rate": 1.2211019537296576e-05,
+      "loss": 0.4761,
+      "step": 9310
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.320185661315918,
+      "learning_rate": 1.2210182822239886e-05,
+      "loss": 1.7961,
+      "step": 9311
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 24.989471435546875,
+      "learning_rate": 1.22093461071832e-05,
+      "loss": 1.0871,
+      "step": 9312
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 70.48097229003906,
+      "learning_rate": 1.2208509392126513e-05,
+      "loss": 1.5549,
+      "step": 9313
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.001704216003418,
+      "learning_rate": 1.2207672677069823e-05,
+      "loss": 1.3008,
+      "step": 9314
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 38.65049743652344,
+      "learning_rate": 1.2206835962013137e-05,
+      "loss": 1.9781,
+      "step": 9315
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 22.112384796142578,
+      "learning_rate": 1.220599924695645e-05,
+      "loss": 0.8403,
+      "step": 9316
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.990418434143066,
+      "learning_rate": 1.2205162531899764e-05,
+      "loss": 1.2436,
+      "step": 9317
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 23.103830337524414,
+      "learning_rate": 1.2204325816843075e-05,
+      "loss": 1.6826,
+      "step": 9318
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 30.30733871459961,
+      "learning_rate": 1.2203489101786388e-05,
+      "loss": 2.2576,
+      "step": 9319
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 14.125082015991211,
+      "learning_rate": 1.22026523867297e-05,
+      "loss": 2.0127,
+      "step": 9320
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 21.839176177978516,
+      "learning_rate": 1.2201815671673012e-05,
+      "loss": 1.1863,
+      "step": 9321
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.785724639892578,
+      "learning_rate": 1.2200978956616326e-05,
+      "loss": 0.6927,
+      "step": 9322
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 11.992521286010742,
+      "learning_rate": 1.2200142241559638e-05,
+      "loss": 1.8727,
+      "step": 9323
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.777341842651367,
+      "learning_rate": 1.2199305526502951e-05,
+      "loss": 0.4468,
+      "step": 9324
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.365225791931152,
+      "learning_rate": 1.2198468811446262e-05,
+      "loss": 0.5203,
+      "step": 9325
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.874039649963379,
+      "learning_rate": 1.2197632096389575e-05,
+      "loss": 1.1379,
+      "step": 9326
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 27.9597225189209,
+      "learning_rate": 1.2196795381332889e-05,
+      "loss": 1.5369,
+      "step": 9327
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 16.303752899169922,
+      "learning_rate": 1.21959586662762e-05,
+      "loss": 1.3029,
+      "step": 9328
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 14.0924072265625,
+      "learning_rate": 1.2195121951219513e-05,
+      "loss": 2.3521,
+      "step": 9329
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 21.292030334472656,
+      "learning_rate": 1.2194285236162827e-05,
+      "loss": 1.3894,
+      "step": 9330
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 14.060272216796875,
+      "learning_rate": 1.2193448521106138e-05,
+      "loss": 1.4786,
+      "step": 9331
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 19.819843292236328,
+      "learning_rate": 1.219261180604945e-05,
+      "loss": 1.2166,
+      "step": 9332
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.6164567470550537,
+      "learning_rate": 1.2191775090992764e-05,
+      "loss": 0.1673,
+      "step": 9333
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 36.086036682128906,
+      "learning_rate": 1.2190938375936076e-05,
+      "loss": 0.7871,
+      "step": 9334
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 34.218055725097656,
+      "learning_rate": 1.2190101660879388e-05,
+      "loss": 1.5674,
+      "step": 9335
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 13.336847305297852,
+      "learning_rate": 1.2189264945822702e-05,
+      "loss": 0.7522,
+      "step": 9336
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 7.8779616355896,
+      "learning_rate": 1.2188428230766014e-05,
+      "loss": 0.9476,
+      "step": 9337
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 16.02096176147461,
+      "learning_rate": 1.2187591515709327e-05,
+      "loss": 1.2687,
+      "step": 9338
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 7.479913234710693,
+      "learning_rate": 1.2186754800652637e-05,
+      "loss": 0.2904,
+      "step": 9339
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 13.346236228942871,
+      "learning_rate": 1.2185918085595951e-05,
+      "loss": 1.0085,
+      "step": 9340
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 20.374624252319336,
+      "learning_rate": 1.2185081370539265e-05,
+      "loss": 1.7728,
+      "step": 9341
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 23.75489044189453,
+      "learning_rate": 1.2184244655482575e-05,
+      "loss": 0.9994,
+      "step": 9342
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 15.159979820251465,
+      "learning_rate": 1.2183407940425889e-05,
+      "loss": 0.9954,
+      "step": 9343
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 18.356822967529297,
+      "learning_rate": 1.2182571225369202e-05,
+      "loss": 0.8046,
+      "step": 9344
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.173781394958496,
+      "learning_rate": 1.2181734510312514e-05,
+      "loss": 1.5786,
+      "step": 9345
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.518787384033203,
+      "learning_rate": 1.2180897795255826e-05,
+      "loss": 2.3228,
+      "step": 9346
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 5.2768473625183105,
+      "learning_rate": 1.218006108019914e-05,
+      "loss": 0.2989,
+      "step": 9347
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 20.644575119018555,
+      "learning_rate": 1.2179224365142452e-05,
+      "loss": 1.0783,
+      "step": 9348
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 7.835660934448242,
+      "learning_rate": 1.2178387650085764e-05,
+      "loss": 1.0289,
+      "step": 9349
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 15.82724666595459,
+      "learning_rate": 1.2177550935029076e-05,
+      "loss": 1.1451,
+      "step": 9350
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 24.29384994506836,
+      "learning_rate": 1.217671421997239e-05,
+      "loss": 1.8794,
+      "step": 9351
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 10.641297340393066,
+      "learning_rate": 1.2175877504915703e-05,
+      "loss": 2.2985,
+      "step": 9352
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 9.606209754943848,
+      "learning_rate": 1.2175040789859013e-05,
+      "loss": 0.808,
+      "step": 9353
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 23.063278198242188,
+      "learning_rate": 1.2174204074802327e-05,
+      "loss": 1.5276,
+      "step": 9354
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 13.640690803527832,
+      "learning_rate": 1.217336735974564e-05,
+      "loss": 0.839,
+      "step": 9355
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 16.292888641357422,
+      "learning_rate": 1.2172530644688951e-05,
+      "loss": 2.0711,
+      "step": 9356
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 15.702163696289062,
+      "learning_rate": 1.2171693929632265e-05,
+      "loss": 0.8678,
+      "step": 9357
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.12215518951416,
+      "learning_rate": 1.2170857214575578e-05,
+      "loss": 0.5314,
+      "step": 9358
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 30.536754608154297,
+      "learning_rate": 1.217002049951889e-05,
+      "loss": 3.6003,
+      "step": 9359
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.415929794311523,
+      "learning_rate": 1.2169183784462202e-05,
+      "loss": 1.9438,
+      "step": 9360
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 76.1987075805664,
+      "learning_rate": 1.2168347069405516e-05,
+      "loss": 2.6336,
+      "step": 9361
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 12.82883071899414,
+      "learning_rate": 1.2167510354348828e-05,
+      "loss": 2.2039,
+      "step": 9362
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.220084190368652,
+      "learning_rate": 1.216667363929214e-05,
+      "loss": 0.2783,
+      "step": 9363
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.6379334926605225,
+      "learning_rate": 1.2165836924235452e-05,
+      "loss": 0.4021,
+      "step": 9364
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 6.193840026855469,
+      "learning_rate": 1.2165000209178765e-05,
+      "loss": 1.1253,
+      "step": 9365
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 22.884403228759766,
+      "learning_rate": 1.2164163494122079e-05,
+      "loss": 2.1672,
+      "step": 9366
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 22.401716232299805,
+      "learning_rate": 1.216332677906539e-05,
+      "loss": 1.6772,
+      "step": 9367
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 27.59799575805664,
+      "learning_rate": 1.2162490064008703e-05,
+      "loss": 1.5655,
+      "step": 9368
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 18.193872451782227,
+      "learning_rate": 1.2161653348952016e-05,
+      "loss": 1.5153,
+      "step": 9369
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 18.48906135559082,
+      "learning_rate": 1.2160816633895327e-05,
+      "loss": 1.3137,
+      "step": 9370
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.570785522460938,
+      "learning_rate": 1.215997991883864e-05,
+      "loss": 0.4115,
+      "step": 9371
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 29.38730239868164,
+      "learning_rate": 1.2159143203781954e-05,
+      "loss": 0.9199,
+      "step": 9372
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 34.051998138427734,
+      "learning_rate": 1.2158306488725266e-05,
+      "loss": 1.2682,
+      "step": 9373
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 11.954904556274414,
+      "learning_rate": 1.2157469773668578e-05,
+      "loss": 1.2436,
+      "step": 9374
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 16.466739654541016,
+      "learning_rate": 1.2156633058611892e-05,
+      "loss": 0.7362,
+      "step": 9375
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 20.73897933959961,
+      "learning_rate": 1.2155796343555204e-05,
+      "loss": 1.3595,
+      "step": 9376
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 23.741695404052734,
+      "learning_rate": 1.2154959628498515e-05,
+      "loss": 2.058,
+      "step": 9377
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 21.302831649780273,
+      "learning_rate": 1.2154122913441827e-05,
+      "loss": 0.7095,
+      "step": 9378
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 6.31294584274292,
+      "learning_rate": 1.2153286198385141e-05,
+      "loss": 0.6791,
+      "step": 9379
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 33.14387130737305,
+      "learning_rate": 1.2152449483328455e-05,
+      "loss": 1.9076,
+      "step": 9380
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 33.726531982421875,
+      "learning_rate": 1.2151612768271765e-05,
+      "loss": 2.0859,
+      "step": 9381
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 19.91851234436035,
+      "learning_rate": 1.2150776053215079e-05,
+      "loss": 1.0878,
+      "step": 9382
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 6.069289684295654,
+      "learning_rate": 1.2149939338158392e-05,
+      "loss": 0.5945,
+      "step": 9383
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 24.550518035888672,
+      "learning_rate": 1.2149102623101703e-05,
+      "loss": 1.9977,
+      "step": 9384
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 23.480838775634766,
+      "learning_rate": 1.2148265908045016e-05,
+      "loss": 1.4497,
+      "step": 9385
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.521613121032715,
+      "learning_rate": 1.214742919298833e-05,
+      "loss": 0.4875,
+      "step": 9386
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 30.82222557067871,
+      "learning_rate": 1.2146592477931642e-05,
+      "loss": 1.4363,
+      "step": 9387
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.662206649780273,
+      "learning_rate": 1.2145755762874954e-05,
+      "loss": 1.2419,
+      "step": 9388
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.719494342803955,
+      "learning_rate": 1.2144919047818267e-05,
+      "loss": 0.6059,
+      "step": 9389
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.777727127075195,
+      "learning_rate": 1.214408233276158e-05,
+      "loss": 1.2383,
+      "step": 9390
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 24.78661346435547,
+      "learning_rate": 1.2143245617704891e-05,
+      "loss": 1.2971,
+      "step": 9391
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.432952880859375,
+      "learning_rate": 1.2142408902648203e-05,
+      "loss": 1.4945,
+      "step": 9392
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 86.90373229980469,
+      "learning_rate": 1.2141572187591517e-05,
+      "loss": 2.8384,
+      "step": 9393
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 16.666542053222656,
+      "learning_rate": 1.214073547253483e-05,
+      "loss": 0.9968,
+      "step": 9394
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 17.281185150146484,
+      "learning_rate": 1.213989875747814e-05,
+      "loss": 1.9701,
+      "step": 9395
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 14.964309692382812,
+      "learning_rate": 1.2139062042421454e-05,
+      "loss": 1.5485,
+      "step": 9396
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 12.134135246276855,
+      "learning_rate": 1.2138225327364768e-05,
+      "loss": 1.7547,
+      "step": 9397
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 48.08967971801758,
+      "learning_rate": 1.2137388612308078e-05,
+      "loss": 1.4255,
+      "step": 9398
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 50.927547454833984,
+      "learning_rate": 1.2136551897251392e-05,
+      "loss": 2.8541,
+      "step": 9399
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 28.605714797973633,
+      "learning_rate": 1.2135715182194706e-05,
+      "loss": 2.351,
+      "step": 9400
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.128934860229492,
+      "learning_rate": 1.2134878467138018e-05,
+      "loss": 1.8754,
+      "step": 9401
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 7.692806243896484,
+      "learning_rate": 1.213404175208133e-05,
+      "loss": 0.3688,
+      "step": 9402
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.0777764320373535,
+      "learning_rate": 1.2133205037024642e-05,
+      "loss": 0.4806,
+      "step": 9403
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 7.016156196594238,
+      "learning_rate": 1.2132368321967955e-05,
+      "loss": 1.9269,
+      "step": 9404
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 11.855449676513672,
+      "learning_rate": 1.2131531606911267e-05,
+      "loss": 1.3325,
+      "step": 9405
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 33.3604736328125,
+      "learning_rate": 1.2130694891854579e-05,
+      "loss": 1.5821,
+      "step": 9406
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 15.44910717010498,
+      "learning_rate": 1.2129858176797893e-05,
+      "loss": 1.9013,
+      "step": 9407
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 14.292071342468262,
+      "learning_rate": 1.2129021461741206e-05,
+      "loss": 0.676,
+      "step": 9408
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 17.352420806884766,
+      "learning_rate": 1.2128184746684517e-05,
+      "loss": 2.2619,
+      "step": 9409
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 15.946002006530762,
+      "learning_rate": 1.212734803162783e-05,
+      "loss": 1.85,
+      "step": 9410
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 20.615232467651367,
+      "learning_rate": 1.2126511316571144e-05,
+      "loss": 1.3353,
+      "step": 9411
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 17.23834800720215,
+      "learning_rate": 1.2125674601514454e-05,
+      "loss": 1.5281,
+      "step": 9412
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 35.31224822998047,
+      "learning_rate": 1.2124837886457768e-05,
+      "loss": 1.5222,
+      "step": 9413
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.358646392822266,
+      "learning_rate": 1.2124001171401082e-05,
+      "loss": 1.0086,
+      "step": 9414
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 16.674345016479492,
+      "learning_rate": 1.2123164456344393e-05,
+      "loss": 1.2741,
+      "step": 9415
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 24.218284606933594,
+      "learning_rate": 1.2122327741287705e-05,
+      "loss": 1.7174,
+      "step": 9416
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 16.73403549194336,
+      "learning_rate": 1.2121491026231017e-05,
+      "loss": 0.8684,
+      "step": 9417
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 16.773893356323242,
+      "learning_rate": 1.2120654311174331e-05,
+      "loss": 2.454,
+      "step": 9418
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 11.357705116271973,
+      "learning_rate": 1.2119817596117643e-05,
+      "loss": 1.5169,
+      "step": 9419
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.991296768188477,
+      "learning_rate": 1.2118980881060955e-05,
+      "loss": 1.1927,
+      "step": 9420
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 33.26502990722656,
+      "learning_rate": 1.2118144166004269e-05,
+      "loss": 1.6418,
+      "step": 9421
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.168608665466309,
+      "learning_rate": 1.2117307450947582e-05,
+      "loss": 1.3493,
+      "step": 9422
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 7.570466041564941,
+      "learning_rate": 1.2116470735890893e-05,
+      "loss": 0.6953,
+      "step": 9423
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 14.572734832763672,
+      "learning_rate": 1.2115634020834206e-05,
+      "loss": 1.4042,
+      "step": 9424
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 7.8805694580078125,
+      "learning_rate": 1.211479730577752e-05,
+      "loss": 0.6732,
+      "step": 9425
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 15.376143455505371,
+      "learning_rate": 1.211396059072083e-05,
+      "loss": 1.7732,
+      "step": 9426
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 24.803756713867188,
+      "learning_rate": 1.2113123875664144e-05,
+      "loss": 1.6964,
+      "step": 9427
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 8.024435997009277,
+      "learning_rate": 1.2112287160607457e-05,
+      "loss": 1.6161,
+      "step": 9428
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 7.506372928619385,
+      "learning_rate": 1.211145044555077e-05,
+      "loss": 0.4802,
+      "step": 9429
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.939638137817383,
+      "learning_rate": 1.2110613730494081e-05,
+      "loss": 1.49,
+      "step": 9430
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 14.319657325744629,
+      "learning_rate": 1.2109777015437393e-05,
+      "loss": 3.0862,
+      "step": 9431
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 17.68967628479004,
+      "learning_rate": 1.2108940300380707e-05,
+      "loss": 3.3128,
+      "step": 9432
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 67.42790222167969,
+      "learning_rate": 1.2108103585324019e-05,
+      "loss": 1.6133,
+      "step": 9433
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 13.851888656616211,
+      "learning_rate": 1.210726687026733e-05,
+      "loss": 1.8041,
+      "step": 9434
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 17.65105438232422,
+      "learning_rate": 1.2106430155210644e-05,
+      "loss": 1.6052,
+      "step": 9435
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 21.24810791015625,
+      "learning_rate": 1.2105593440153958e-05,
+      "loss": 2.0339,
+      "step": 9436
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 17.26531410217285,
+      "learning_rate": 1.2104756725097268e-05,
+      "loss": 1.6868,
+      "step": 9437
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 31.289440155029297,
+      "learning_rate": 1.2103920010040582e-05,
+      "loss": 2.7377,
+      "step": 9438
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.300957679748535,
+      "learning_rate": 1.2103083294983896e-05,
+      "loss": 1.3814,
+      "step": 9439
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 9.7875394821167,
+      "learning_rate": 1.2102246579927206e-05,
+      "loss": 0.6515,
+      "step": 9440
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 25.404930114746094,
+      "learning_rate": 1.210140986487052e-05,
+      "loss": 0.9794,
+      "step": 9441
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 44.835391998291016,
+      "learning_rate": 1.2100573149813832e-05,
+      "loss": 2.3566,
+      "step": 9442
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 23.619457244873047,
+      "learning_rate": 1.2099736434757145e-05,
+      "loss": 1.4021,
+      "step": 9443
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 11.631579399108887,
+      "learning_rate": 1.2098899719700457e-05,
+      "loss": 0.956,
+      "step": 9444
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 11.371979713439941,
+      "learning_rate": 1.2098063004643769e-05,
+      "loss": 1.1925,
+      "step": 9445
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.25599479675293,
+      "learning_rate": 1.2097226289587083e-05,
+      "loss": 0.718,
+      "step": 9446
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 16.738479614257812,
+      "learning_rate": 1.2096389574530395e-05,
+      "loss": 1.8042,
+      "step": 9447
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.86751651763916,
+      "learning_rate": 1.2095552859473707e-05,
+      "loss": 0.4073,
+      "step": 9448
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.155218124389648,
+      "learning_rate": 1.209471614441702e-05,
+      "loss": 0.9582,
+      "step": 9449
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.982494354248047,
+      "learning_rate": 1.2093879429360334e-05,
+      "loss": 0.7085,
+      "step": 9450
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 11.960037231445312,
+      "learning_rate": 1.2093042714303644e-05,
+      "loss": 0.6946,
+      "step": 9451
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 13.039807319641113,
+      "learning_rate": 1.2092205999246958e-05,
+      "loss": 2.2661,
+      "step": 9452
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 19.76038932800293,
+      "learning_rate": 1.2091369284190271e-05,
+      "loss": 1.2509,
+      "step": 9453
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 41.41776657104492,
+      "learning_rate": 1.2090532569133582e-05,
+      "loss": 2.9388,
+      "step": 9454
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 29.097850799560547,
+      "learning_rate": 1.2089695854076895e-05,
+      "loss": 1.1569,
+      "step": 9455
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 24.557373046875,
+      "learning_rate": 1.2088859139020207e-05,
+      "loss": 2.6758,
+      "step": 9456
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 17.72818946838379,
+      "learning_rate": 1.208802242396352e-05,
+      "loss": 1.5893,
+      "step": 9457
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.774362564086914,
+      "learning_rate": 1.2087185708906833e-05,
+      "loss": 0.3926,
+      "step": 9458
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 27.049278259277344,
+      "learning_rate": 1.2086348993850145e-05,
+      "loss": 2.1748,
+      "step": 9459
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.9520184993743896,
+      "learning_rate": 1.2085512278793459e-05,
+      "loss": 0.1981,
+      "step": 9460
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 13.119494438171387,
+      "learning_rate": 1.2084675563736769e-05,
+      "loss": 0.9479,
+      "step": 9461
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 62.712379455566406,
+      "learning_rate": 1.2083838848680082e-05,
+      "loss": 1.5795,
+      "step": 9462
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 21.442840576171875,
+      "learning_rate": 1.2083002133623396e-05,
+      "loss": 0.9436,
+      "step": 9463
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 16.34206771850586,
+      "learning_rate": 1.2082165418566706e-05,
+      "loss": 1.7083,
+      "step": 9464
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.9206340312957764,
+      "learning_rate": 1.208132870351002e-05,
+      "loss": 0.1217,
+      "step": 9465
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 12.184715270996094,
+      "learning_rate": 1.2080491988453334e-05,
+      "loss": 1.1073,
+      "step": 9466
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 14.369974136352539,
+      "learning_rate": 1.2079655273396647e-05,
+      "loss": 2.6136,
+      "step": 9467
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 11.854850769042969,
+      "learning_rate": 1.2078818558339958e-05,
+      "loss": 2.6624,
+      "step": 9468
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 13.329398155212402,
+      "learning_rate": 1.2077981843283271e-05,
+      "loss": 1.7601,
+      "step": 9469
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 15.483675956726074,
+      "learning_rate": 1.2077145128226583e-05,
+      "loss": 1.6243,
+      "step": 9470
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.290410995483398,
+      "learning_rate": 1.2076308413169895e-05,
+      "loss": 1.1646,
+      "step": 9471
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 14.221891403198242,
+      "learning_rate": 1.2075471698113209e-05,
+      "loss": 1.1798,
+      "step": 9472
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 13.933663368225098,
+      "learning_rate": 1.207463498305652e-05,
+      "loss": 1.2359,
+      "step": 9473
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 64.45641326904297,
+      "learning_rate": 1.2073798267999834e-05,
+      "loss": 2.1569,
+      "step": 9474
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.408763885498047,
+      "learning_rate": 1.2072961552943145e-05,
+      "loss": 1.7103,
+      "step": 9475
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 12.310800552368164,
+      "learning_rate": 1.2072124837886458e-05,
+      "loss": 1.4251,
+      "step": 9476
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 12.391608238220215,
+      "learning_rate": 1.2071288122829772e-05,
+      "loss": 1.4087,
+      "step": 9477
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.71947193145752,
+      "learning_rate": 1.2070451407773082e-05,
+      "loss": 2.111,
+      "step": 9478
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 14.64475154876709,
+      "learning_rate": 1.2069614692716396e-05,
+      "loss": 1.4027,
+      "step": 9479
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 24.743968963623047,
+      "learning_rate": 1.206877797765971e-05,
+      "loss": 2.4821,
+      "step": 9480
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 26.19554901123047,
+      "learning_rate": 1.2067941262603023e-05,
+      "loss": 1.9297,
+      "step": 9481
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.869488716125488,
+      "learning_rate": 1.2067104547546333e-05,
+      "loss": 0.6276,
+      "step": 9482
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.796403884887695,
+      "learning_rate": 1.2066267832489647e-05,
+      "loss": 1.121,
+      "step": 9483
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 16.084341049194336,
+      "learning_rate": 1.2065431117432959e-05,
+      "loss": 1.5531,
+      "step": 9484
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 16.72888946533203,
+      "learning_rate": 1.2064594402376271e-05,
+      "loss": 1.2476,
+      "step": 9485
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 16.48817253112793,
+      "learning_rate": 1.2063757687319585e-05,
+      "loss": 2.8725,
+      "step": 9486
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 16.455909729003906,
+      "learning_rate": 1.2062920972262897e-05,
+      "loss": 2.6381,
+      "step": 9487
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 12.815619468688965,
+      "learning_rate": 1.206208425720621e-05,
+      "loss": 0.9785,
+      "step": 9488
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.762529373168945,
+      "learning_rate": 1.206124754214952e-05,
+      "loss": 0.1354,
+      "step": 9489
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 30.014179229736328,
+      "learning_rate": 1.2060410827092834e-05,
+      "loss": 1.2246,
+      "step": 9490
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 19.356752395629883,
+      "learning_rate": 1.2059574112036148e-05,
+      "loss": 0.9066,
+      "step": 9491
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.490389823913574,
+      "learning_rate": 1.2058737396979458e-05,
+      "loss": 1.1231,
+      "step": 9492
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 17.957721710205078,
+      "learning_rate": 1.2057900681922772e-05,
+      "loss": 1.0549,
+      "step": 9493
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.60168170928955,
+      "learning_rate": 1.2057063966866085e-05,
+      "loss": 1.1987,
+      "step": 9494
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.99045467376709,
+      "learning_rate": 1.2056227251809397e-05,
+      "loss": 0.9482,
+      "step": 9495
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 15.194043159484863,
+      "learning_rate": 1.205539053675271e-05,
+      "loss": 2.5112,
+      "step": 9496
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 24.683975219726562,
+      "learning_rate": 1.2054553821696023e-05,
+      "loss": 2.1607,
+      "step": 9497
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.09386157989502,
+      "learning_rate": 1.2053717106639335e-05,
+      "loss": 0.9821,
+      "step": 9498
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 11.077011108398438,
+      "learning_rate": 1.2052880391582647e-05,
+      "loss": 1.0397,
+      "step": 9499
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 15.114931106567383,
+      "learning_rate": 1.205204367652596e-05,
+      "loss": 1.6384,
+      "step": 9500
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 22.820463180541992,
+      "learning_rate": 1.2051206961469272e-05,
+      "loss": 1.9424,
+      "step": 9501
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.954093933105469,
+      "learning_rate": 1.2050370246412586e-05,
+      "loss": 1.0824,
+      "step": 9502
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 22.414491653442383,
+      "learning_rate": 1.2049533531355896e-05,
+      "loss": 1.0721,
+      "step": 9503
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.525824069976807,
+      "learning_rate": 1.204869681629921e-05,
+      "loss": 0.3603,
+      "step": 9504
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 30.8877010345459,
+      "learning_rate": 1.2047860101242524e-05,
+      "loss": 2.2824,
+      "step": 9505
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 24.046707153320312,
+      "learning_rate": 1.2047023386185834e-05,
+      "loss": 0.8705,
+      "step": 9506
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.529458045959473,
+      "learning_rate": 1.2046186671129148e-05,
+      "loss": 0.9239,
+      "step": 9507
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 32.48001480102539,
+      "learning_rate": 1.2045349956072461e-05,
+      "loss": 2.0995,
+      "step": 9508
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 16.195293426513672,
+      "learning_rate": 1.2044513241015773e-05,
+      "loss": 1.5304,
+      "step": 9509
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.123017311096191,
+      "learning_rate": 1.2043676525959085e-05,
+      "loss": 0.7504,
+      "step": 9510
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 24.254032135009766,
+      "learning_rate": 1.2042839810902399e-05,
+      "loss": 2.4331,
+      "step": 9511
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 16.45452117919922,
+      "learning_rate": 1.204200309584571e-05,
+      "loss": 2.5155,
+      "step": 9512
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 21.005456924438477,
+      "learning_rate": 1.2041166380789023e-05,
+      "loss": 1.4441,
+      "step": 9513
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 20.228805541992188,
+      "learning_rate": 1.2040329665732335e-05,
+      "loss": 0.5575,
+      "step": 9514
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 64.70474243164062,
+      "learning_rate": 1.2039492950675648e-05,
+      "loss": 2.7917,
+      "step": 9515
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 38.86469650268555,
+      "learning_rate": 1.2038656235618962e-05,
+      "loss": 2.3465,
+      "step": 9516
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 9.966197967529297,
+      "learning_rate": 1.2037819520562272e-05,
+      "loss": 0.9642,
+      "step": 9517
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 18.563018798828125,
+      "learning_rate": 1.2036982805505586e-05,
+      "loss": 2.0094,
+      "step": 9518
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 17.217992782592773,
+      "learning_rate": 1.20361460904489e-05,
+      "loss": 2.0467,
+      "step": 9519
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 11.091466903686523,
+      "learning_rate": 1.203530937539221e-05,
+      "loss": 2.0162,
+      "step": 9520
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 35.912391662597656,
+      "learning_rate": 1.2034472660335523e-05,
+      "loss": 2.7718,
+      "step": 9521
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 7.927547454833984,
+      "learning_rate": 1.2033635945278837e-05,
+      "loss": 0.4048,
+      "step": 9522
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.039243698120117,
+      "learning_rate": 1.2032799230222149e-05,
+      "loss": 1.4761,
+      "step": 9523
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.982276916503906,
+      "learning_rate": 1.2031962515165461e-05,
+      "loss": 1.8997,
+      "step": 9524
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.009096145629883,
+      "learning_rate": 1.2031125800108775e-05,
+      "loss": 1.811,
+      "step": 9525
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.981189727783203,
+      "learning_rate": 1.2030289085052087e-05,
+      "loss": 2.2421,
+      "step": 9526
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 7.837049961090088,
+      "learning_rate": 1.2029452369995398e-05,
+      "loss": 1.7047,
+      "step": 9527
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.705050468444824,
+      "learning_rate": 1.202861565493871e-05,
+      "loss": 0.3916,
+      "step": 9528
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 18.591602325439453,
+      "learning_rate": 1.2027778939882024e-05,
+      "loss": 0.877,
+      "step": 9529
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.062577247619629,
+      "learning_rate": 1.2026942224825338e-05,
+      "loss": 0.8878,
+      "step": 9530
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 18.63389015197754,
+      "learning_rate": 1.2026105509768648e-05,
+      "loss": 1.3153,
+      "step": 9531
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.728708267211914,
+      "learning_rate": 1.2025268794711962e-05,
+      "loss": 2.1566,
+      "step": 9532
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.1477296352386475,
+      "learning_rate": 1.2024432079655275e-05,
+      "loss": 0.1559,
+      "step": 9533
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.642887592315674,
+      "learning_rate": 1.2023595364598586e-05,
+      "loss": 0.4317,
+      "step": 9534
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 17.34537124633789,
+      "learning_rate": 1.20227586495419e-05,
+      "loss": 1.5061,
+      "step": 9535
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 14.58328914642334,
+      "learning_rate": 1.2021921934485213e-05,
+      "loss": 1.1147,
+      "step": 9536
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.474074363708496,
+      "learning_rate": 1.2021085219428525e-05,
+      "loss": 0.676,
+      "step": 9537
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 26.583112716674805,
+      "learning_rate": 1.2020248504371837e-05,
+      "loss": 2.1511,
+      "step": 9538
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 15.964679718017578,
+      "learning_rate": 1.201941178931515e-05,
+      "loss": 0.8863,
+      "step": 9539
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 22.612957000732422,
+      "learning_rate": 1.2018575074258462e-05,
+      "loss": 3.0547,
+      "step": 9540
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.257503509521484,
+      "learning_rate": 1.2017738359201774e-05,
+      "loss": 0.7493,
+      "step": 9541
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 63.64929962158203,
+      "learning_rate": 1.2016901644145086e-05,
+      "loss": 0.9087,
+      "step": 9542
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.743460655212402,
+      "learning_rate": 1.20160649290884e-05,
+      "loss": 1.8593,
+      "step": 9543
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 16.666603088378906,
+      "learning_rate": 1.2015228214031714e-05,
+      "loss": 1.6155,
+      "step": 9544
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 13.350123405456543,
+      "learning_rate": 1.2014391498975024e-05,
+      "loss": 2.1969,
+      "step": 9545
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.722788333892822,
+      "learning_rate": 1.2013554783918337e-05,
+      "loss": 0.7194,
+      "step": 9546
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 18.07695198059082,
+      "learning_rate": 1.2012718068861651e-05,
+      "loss": 0.9043,
+      "step": 9547
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.8741455078125,
+      "learning_rate": 1.2011881353804961e-05,
+      "loss": 1.1142,
+      "step": 9548
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 27.6228084564209,
+      "learning_rate": 1.2011044638748275e-05,
+      "loss": 2.4636,
+      "step": 9549
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 21.675600051879883,
+      "learning_rate": 1.2010207923691589e-05,
+      "loss": 1.4064,
+      "step": 9550
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 42.42505645751953,
+      "learning_rate": 1.20093712086349e-05,
+      "loss": 1.4687,
+      "step": 9551
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 24.894182205200195,
+      "learning_rate": 1.2008534493578213e-05,
+      "loss": 0.8802,
+      "step": 9552
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 6.725581169128418,
+      "learning_rate": 1.2007697778521525e-05,
+      "loss": 0.6847,
+      "step": 9553
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 18.131277084350586,
+      "learning_rate": 1.2006861063464838e-05,
+      "loss": 0.911,
+      "step": 9554
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.308471202850342,
+      "learning_rate": 1.200602434840815e-05,
+      "loss": 0.1425,
+      "step": 9555
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.859646797180176,
+      "learning_rate": 1.2005187633351462e-05,
+      "loss": 1.3277,
+      "step": 9556
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.082441329956055,
+      "learning_rate": 1.2004350918294776e-05,
+      "loss": 2.6566,
+      "step": 9557
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 14.841496467590332,
+      "learning_rate": 1.200351420323809e-05,
+      "loss": 2.0174,
+      "step": 9558
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.623128890991211,
+      "learning_rate": 1.20026774881814e-05,
+      "loss": 0.8134,
+      "step": 9559
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 16.343841552734375,
+      "learning_rate": 1.2001840773124713e-05,
+      "loss": 1.8796,
+      "step": 9560
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.812308311462402,
+      "learning_rate": 1.2001004058068027e-05,
+      "loss": 0.5763,
+      "step": 9561
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.021820068359375,
+      "learning_rate": 1.2000167343011337e-05,
+      "loss": 0.3513,
+      "step": 9562
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.692317008972168,
+      "learning_rate": 1.1999330627954651e-05,
+      "loss": 0.9401,
+      "step": 9563
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.400107383728027,
+      "learning_rate": 1.1998493912897965e-05,
+      "loss": 0.2327,
+      "step": 9564
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 25.09235382080078,
+      "learning_rate": 1.1997657197841276e-05,
+      "loss": 2.2308,
+      "step": 9565
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 9.318791389465332,
+      "learning_rate": 1.1996820482784588e-05,
+      "loss": 0.8867,
+      "step": 9566
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 21.22977066040039,
+      "learning_rate": 1.19959837677279e-05,
+      "loss": 1.4477,
+      "step": 9567
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.80361270904541,
+      "learning_rate": 1.1995147052671214e-05,
+      "loss": 0.7557,
+      "step": 9568
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 20.52949333190918,
+      "learning_rate": 1.1994310337614526e-05,
+      "loss": 1.7729,
+      "step": 9569
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.310013771057129,
+      "learning_rate": 1.1993473622557838e-05,
+      "loss": 0.2901,
+      "step": 9570
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 27.86255645751953,
+      "learning_rate": 1.1992636907501152e-05,
+      "loss": 0.8871,
+      "step": 9571
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 11.100020408630371,
+      "learning_rate": 1.1991800192444465e-05,
+      "loss": 0.7504,
+      "step": 9572
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 38.98814010620117,
+      "learning_rate": 1.1990963477387776e-05,
+      "loss": 1.8995,
+      "step": 9573
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 7.79847526550293,
+      "learning_rate": 1.199012676233109e-05,
+      "loss": 0.8019,
+      "step": 9574
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 14.081271171569824,
+      "learning_rate": 1.1989290047274403e-05,
+      "loss": 0.744,
+      "step": 9575
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 64.49124145507812,
+      "learning_rate": 1.1988453332217713e-05,
+      "loss": 1.0467,
+      "step": 9576
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.7524309158325195,
+      "learning_rate": 1.1987616617161027e-05,
+      "loss": 1.1582,
+      "step": 9577
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 7.193423271179199,
+      "learning_rate": 1.198677990210434e-05,
+      "loss": 1.817,
+      "step": 9578
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.854019165039062,
+      "learning_rate": 1.1985943187047652e-05,
+      "loss": 1.3564,
+      "step": 9579
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.136621475219727,
+      "learning_rate": 1.1985106471990964e-05,
+      "loss": 1.3315,
+      "step": 9580
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 21.077999114990234,
+      "learning_rate": 1.1984269756934276e-05,
+      "loss": 1.4171,
+      "step": 9581
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.653505325317383,
+      "learning_rate": 1.198343304187759e-05,
+      "loss": 0.7738,
+      "step": 9582
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 17.65804100036621,
+      "learning_rate": 1.1982596326820902e-05,
+      "loss": 0.7794,
+      "step": 9583
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 20.522994995117188,
+      "learning_rate": 1.1981759611764214e-05,
+      "loss": 4.257,
+      "step": 9584
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 42.09373092651367,
+      "learning_rate": 1.1980922896707527e-05,
+      "loss": 2.0501,
+      "step": 9585
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.918661117553711,
+      "learning_rate": 1.1980086181650841e-05,
+      "loss": 1.1037,
+      "step": 9586
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 6.394575595855713,
+      "learning_rate": 1.1979249466594151e-05,
+      "loss": 0.4025,
+      "step": 9587
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 42.787078857421875,
+      "learning_rate": 1.1978412751537465e-05,
+      "loss": 2.5623,
+      "step": 9588
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 21.253686904907227,
+      "learning_rate": 1.1977576036480779e-05,
+      "loss": 1.4255,
+      "step": 9589
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 18.87506866455078,
+      "learning_rate": 1.1976739321424089e-05,
+      "loss": 1.337,
+      "step": 9590
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.737640380859375,
+      "learning_rate": 1.1975902606367403e-05,
+      "loss": 0.3225,
+      "step": 9591
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 32.349571228027344,
+      "learning_rate": 1.1975065891310716e-05,
+      "loss": 2.7442,
+      "step": 9592
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 67.66178894042969,
+      "learning_rate": 1.1974229176254028e-05,
+      "loss": 1.4558,
+      "step": 9593
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 75.0477066040039,
+      "learning_rate": 1.197339246119734e-05,
+      "loss": 1.3897,
+      "step": 9594
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 15.68415641784668,
+      "learning_rate": 1.1972555746140652e-05,
+      "loss": 1.2878,
+      "step": 9595
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 17.703025817871094,
+      "learning_rate": 1.1971719031083966e-05,
+      "loss": 0.7005,
+      "step": 9596
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 15.474891662597656,
+      "learning_rate": 1.1970882316027278e-05,
+      "loss": 1.1582,
+      "step": 9597
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 40.227596282958984,
+      "learning_rate": 1.197004560097059e-05,
+      "loss": 2.1653,
+      "step": 9598
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.242630958557129,
+      "learning_rate": 1.1969208885913903e-05,
+      "loss": 0.6583,
+      "step": 9599
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.278730392456055,
+      "learning_rate": 1.1968372170857217e-05,
+      "loss": 0.7933,
+      "step": 9600
+    },
+    {
+      "epoch": 1.2,
+      "eval_loss": 0.08709228038787842,
+      "eval_runtime": 94.6545,
+      "eval_samples_per_second": 37.42,
+      "eval_steps_per_second": 37.42,
+      "step": 9600
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 10.507102966308594,
+      "learning_rate": 1.1967535455800527e-05,
+      "loss": 1.25,
+      "step": 9601
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 44.48877716064453,
+      "learning_rate": 1.196669874074384e-05,
+      "loss": 1.0483,
+      "step": 9602
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.914356231689453,
+      "learning_rate": 1.1965862025687154e-05,
+      "loss": 1.2343,
+      "step": 9603
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 14.555281639099121,
+      "learning_rate": 1.1965025310630465e-05,
+      "loss": 2.6171,
+      "step": 9604
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 40.05111312866211,
+      "learning_rate": 1.1964188595573778e-05,
+      "loss": 1.6383,
+      "step": 9605
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 13.92563533782959,
+      "learning_rate": 1.196335188051709e-05,
+      "loss": 0.8991,
+      "step": 9606
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 6.110177516937256,
+      "learning_rate": 1.1962515165460404e-05,
+      "loss": 0.4676,
+      "step": 9607
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.3621063232421875,
+      "learning_rate": 1.1961678450403716e-05,
+      "loss": 0.866,
+      "step": 9608
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 27.311689376831055,
+      "learning_rate": 1.1960841735347028e-05,
+      "loss": 0.966,
+      "step": 9609
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 21.225568771362305,
+      "learning_rate": 1.1960005020290342e-05,
+      "loss": 0.9207,
+      "step": 9610
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 5.978994369506836,
+      "learning_rate": 1.1959168305233654e-05,
+      "loss": 1.9674,
+      "step": 9611
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.083782196044922,
+      "learning_rate": 1.1958331590176965e-05,
+      "loss": 1.1243,
+      "step": 9612
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 16.697175979614258,
+      "learning_rate": 1.1957494875120279e-05,
+      "loss": 1.5605,
+      "step": 9613
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.813374519348145,
+      "learning_rate": 1.1956658160063593e-05,
+      "loss": 1.3633,
+      "step": 9614
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 19.021514892578125,
+      "learning_rate": 1.1955821445006903e-05,
+      "loss": 1.9111,
+      "step": 9615
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 17.797969818115234,
+      "learning_rate": 1.1954984729950217e-05,
+      "loss": 1.2812,
+      "step": 9616
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 11.785213470458984,
+      "learning_rate": 1.195414801489353e-05,
+      "loss": 1.0591,
+      "step": 9617
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.536766052246094,
+      "learning_rate": 1.195331129983684e-05,
+      "loss": 0.3705,
+      "step": 9618
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.87598180770874,
+      "learning_rate": 1.1952474584780154e-05,
+      "loss": 0.8564,
+      "step": 9619
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 5.448538780212402,
+      "learning_rate": 1.1951637869723466e-05,
+      "loss": 0.4751,
+      "step": 9620
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.328154563903809,
+      "learning_rate": 1.195080115466678e-05,
+      "loss": 0.3251,
+      "step": 9621
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.848305702209473,
+      "learning_rate": 1.1949964439610092e-05,
+      "loss": 0.9825,
+      "step": 9622
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 33.84737777709961,
+      "learning_rate": 1.1949127724553404e-05,
+      "loss": 1.1589,
+      "step": 9623
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 54.30733871459961,
+      "learning_rate": 1.1948291009496717e-05,
+      "loss": 1.9604,
+      "step": 9624
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 13.447796821594238,
+      "learning_rate": 1.1947454294440028e-05,
+      "loss": 0.7447,
+      "step": 9625
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 46.959598541259766,
+      "learning_rate": 1.1946617579383341e-05,
+      "loss": 1.8551,
+      "step": 9626
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 23.30669403076172,
+      "learning_rate": 1.1945780864326655e-05,
+      "loss": 2.3335,
+      "step": 9627
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.455014228820801,
+      "learning_rate": 1.1944944149269969e-05,
+      "loss": 1.0215,
+      "step": 9628
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 35.093746185302734,
+      "learning_rate": 1.1944107434213279e-05,
+      "loss": 1.5814,
+      "step": 9629
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.487492084503174,
+      "learning_rate": 1.1943270719156593e-05,
+      "loss": 1.1403,
+      "step": 9630
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 87.34281158447266,
+      "learning_rate": 1.1942434004099906e-05,
+      "loss": 2.2956,
+      "step": 9631
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.11362361907959,
+      "learning_rate": 1.1941597289043216e-05,
+      "loss": 0.8279,
+      "step": 9632
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.813751697540283,
+      "learning_rate": 1.194076057398653e-05,
+      "loss": 1.2104,
+      "step": 9633
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 18.270851135253906,
+      "learning_rate": 1.1939923858929842e-05,
+      "loss": 0.8104,
+      "step": 9634
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 19.307849884033203,
+      "learning_rate": 1.1939087143873156e-05,
+      "loss": 0.422,
+      "step": 9635
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 15.76440143585205,
+      "learning_rate": 1.1938250428816468e-05,
+      "loss": 1.7508,
+      "step": 9636
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 43.16948318481445,
+      "learning_rate": 1.193741371375978e-05,
+      "loss": 0.1653,
+      "step": 9637
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 21.53644371032715,
+      "learning_rate": 1.1936576998703093e-05,
+      "loss": 1.9636,
+      "step": 9638
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 5.95522403717041,
+      "learning_rate": 1.1935740283646403e-05,
+      "loss": 0.4349,
+      "step": 9639
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 11.78096866607666,
+      "learning_rate": 1.1934903568589717e-05,
+      "loss": 1.6425,
+      "step": 9640
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 49.823890686035156,
+      "learning_rate": 1.193406685353303e-05,
+      "loss": 2.4893,
+      "step": 9641
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 12.305930137634277,
+      "learning_rate": 1.1933230138476344e-05,
+      "loss": 0.9311,
+      "step": 9642
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 17.963987350463867,
+      "learning_rate": 1.1932393423419655e-05,
+      "loss": 0.686,
+      "step": 9643
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.458902359008789,
+      "learning_rate": 1.1931556708362968e-05,
+      "loss": 0.6735,
+      "step": 9644
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 26.11444091796875,
+      "learning_rate": 1.1930719993306282e-05,
+      "loss": 2.1362,
+      "step": 9645
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 39.13798141479492,
+      "learning_rate": 1.1929883278249592e-05,
+      "loss": 2.6185,
+      "step": 9646
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 11.085432052612305,
+      "learning_rate": 1.1929046563192906e-05,
+      "loss": 0.6792,
+      "step": 9647
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 25.43505096435547,
+      "learning_rate": 1.1928209848136218e-05,
+      "loss": 0.9783,
+      "step": 9648
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 11.517626762390137,
+      "learning_rate": 1.1927373133079532e-05,
+      "loss": 1.6657,
+      "step": 9649
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.83224105834961,
+      "learning_rate": 1.1926536418022843e-05,
+      "loss": 2.2246,
+      "step": 9650
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 33.41084671020508,
+      "learning_rate": 1.1925699702966155e-05,
+      "loss": 1.7027,
+      "step": 9651
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.236109733581543,
+      "learning_rate": 1.1924862987909469e-05,
+      "loss": 0.8989,
+      "step": 9652
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 32.25490188598633,
+      "learning_rate": 1.192402627285278e-05,
+      "loss": 0.4262,
+      "step": 9653
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 22.86430549621582,
+      "learning_rate": 1.1923189557796093e-05,
+      "loss": 1.3529,
+      "step": 9654
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 135.71849060058594,
+      "learning_rate": 1.1922352842739407e-05,
+      "loss": 3.4075,
+      "step": 9655
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.792234420776367,
+      "learning_rate": 1.192151612768272e-05,
+      "loss": 0.2315,
+      "step": 9656
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 18.394742965698242,
+      "learning_rate": 1.192067941262603e-05,
+      "loss": 1.0349,
+      "step": 9657
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 6.221388816833496,
+      "learning_rate": 1.1919842697569344e-05,
+      "loss": 0.3315,
+      "step": 9658
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 8.869363784790039,
+      "learning_rate": 1.1919005982512656e-05,
+      "loss": 0.4505,
+      "step": 9659
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.412805557250977,
+      "learning_rate": 1.1918169267455968e-05,
+      "loss": 0.6926,
+      "step": 9660
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.7438879013061523,
+      "learning_rate": 1.1917332552399282e-05,
+      "loss": 0.237,
+      "step": 9661
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 19.818506240844727,
+      "learning_rate": 1.1916495837342594e-05,
+      "loss": 0.9026,
+      "step": 9662
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.5379445552825928,
+      "learning_rate": 1.1915659122285907e-05,
+      "loss": 0.0703,
+      "step": 9663
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 100.18061065673828,
+      "learning_rate": 1.191482240722922e-05,
+      "loss": 1.19,
+      "step": 9664
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.609733581542969,
+      "learning_rate": 1.1913985692172531e-05,
+      "loss": 0.2924,
+      "step": 9665
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 89.28193664550781,
+      "learning_rate": 1.1913148977115845e-05,
+      "loss": 2.1219,
+      "step": 9666
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 44.11702346801758,
+      "learning_rate": 1.1912312262059155e-05,
+      "loss": 1.3472,
+      "step": 9667
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.148381233215332,
+      "learning_rate": 1.1911475547002469e-05,
+      "loss": 1.6529,
+      "step": 9668
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 60.04631042480469,
+      "learning_rate": 1.1910638831945782e-05,
+      "loss": 0.3966,
+      "step": 9669
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 9.514388084411621,
+      "learning_rate": 1.1909802116889096e-05,
+      "loss": 0.8149,
+      "step": 9670
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 13.282306671142578,
+      "learning_rate": 1.1908965401832406e-05,
+      "loss": 1.6097,
+      "step": 9671
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 93.24066925048828,
+      "learning_rate": 1.190812868677572e-05,
+      "loss": 1.8783,
+      "step": 9672
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 24.716209411621094,
+      "learning_rate": 1.1907291971719032e-05,
+      "loss": 3.0982,
+      "step": 9673
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 8.042412757873535,
+      "learning_rate": 1.1906455256662344e-05,
+      "loss": 1.1715,
+      "step": 9674
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 47.73003387451172,
+      "learning_rate": 1.1905618541605658e-05,
+      "loss": 1.9625,
+      "step": 9675
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 19.045705795288086,
+      "learning_rate": 1.190478182654897e-05,
+      "loss": 1.1437,
+      "step": 9676
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 13.466931343078613,
+      "learning_rate": 1.1903945111492283e-05,
+      "loss": 1.1626,
+      "step": 9677
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 11.230758666992188,
+      "learning_rate": 1.1903108396435593e-05,
+      "loss": 1.8554,
+      "step": 9678
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 28.8801326751709,
+      "learning_rate": 1.1902271681378907e-05,
+      "loss": 2.4584,
+      "step": 9679
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.975315809249878,
+      "learning_rate": 1.190143496632222e-05,
+      "loss": 0.378,
+      "step": 9680
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 8.131440162658691,
+      "learning_rate": 1.1900598251265531e-05,
+      "loss": 0.597,
+      "step": 9681
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.377288818359375,
+      "learning_rate": 1.1899761536208845e-05,
+      "loss": 1.2148,
+      "step": 9682
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.312132358551025,
+      "learning_rate": 1.1898924821152158e-05,
+      "loss": 0.8062,
+      "step": 9683
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.498202323913574,
+      "learning_rate": 1.1898088106095472e-05,
+      "loss": 0.9118,
+      "step": 9684
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 4.082993507385254,
+      "learning_rate": 1.1897251391038782e-05,
+      "loss": 1.4373,
+      "step": 9685
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 24.017648696899414,
+      "learning_rate": 1.1896414675982096e-05,
+      "loss": 2.074,
+      "step": 9686
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 25.38010025024414,
+      "learning_rate": 1.1895577960925408e-05,
+      "loss": 1.515,
+      "step": 9687
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 13.79157543182373,
+      "learning_rate": 1.189474124586872e-05,
+      "loss": 0.6778,
+      "step": 9688
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 30.164793014526367,
+      "learning_rate": 1.1893904530812033e-05,
+      "loss": 1.5578,
+      "step": 9689
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 16.952709197998047,
+      "learning_rate": 1.1893067815755345e-05,
+      "loss": 1.8537,
+      "step": 9690
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.253092765808105,
+      "learning_rate": 1.1892231100698659e-05,
+      "loss": 0.723,
+      "step": 9691
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 59.83599853515625,
+      "learning_rate": 1.189139438564197e-05,
+      "loss": 2.5,
+      "step": 9692
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 61.06582260131836,
+      "learning_rate": 1.1890557670585283e-05,
+      "loss": 1.3422,
+      "step": 9693
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 11.885011672973633,
+      "learning_rate": 1.1889720955528597e-05,
+      "loss": 1.3652,
+      "step": 9694
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 15.825852394104004,
+      "learning_rate": 1.1888884240471907e-05,
+      "loss": 3.3345,
+      "step": 9695
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 19.11424446105957,
+      "learning_rate": 1.188804752541522e-05,
+      "loss": 1.8801,
+      "step": 9696
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 36.316688537597656,
+      "learning_rate": 1.1887210810358534e-05,
+      "loss": 2.3157,
+      "step": 9697
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.54317569732666,
+      "learning_rate": 1.1886374095301846e-05,
+      "loss": 1.6094,
+      "step": 9698
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 66.97698211669922,
+      "learning_rate": 1.1885537380245158e-05,
+      "loss": 1.1063,
+      "step": 9699
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 5.200563430786133,
+      "learning_rate": 1.1884700665188472e-05,
+      "loss": 0.6615,
+      "step": 9700
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 18.189468383789062,
+      "learning_rate": 1.1883863950131784e-05,
+      "loss": 1.4355,
+      "step": 9701
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 11.666252136230469,
+      "learning_rate": 1.1883027235075096e-05,
+      "loss": 0.8702,
+      "step": 9702
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.471773147583008,
+      "learning_rate": 1.188219052001841e-05,
+      "loss": 0.8736,
+      "step": 9703
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 15.201849937438965,
+      "learning_rate": 1.1881353804961721e-05,
+      "loss": 1.2114,
+      "step": 9704
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 14.885287284851074,
+      "learning_rate": 1.1880517089905035e-05,
+      "loss": 0.916,
+      "step": 9705
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 5.266726016998291,
+      "learning_rate": 1.1879680374848345e-05,
+      "loss": 0.4309,
+      "step": 9706
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.320878505706787,
+      "learning_rate": 1.1878843659791659e-05,
+      "loss": 0.4919,
+      "step": 9707
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.768665313720703,
+      "learning_rate": 1.1878006944734972e-05,
+      "loss": 1.098,
+      "step": 9708
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.963107109069824,
+      "learning_rate": 1.1877170229678283e-05,
+      "loss": 0.6641,
+      "step": 9709
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 14.743289947509766,
+      "learning_rate": 1.1876333514621596e-05,
+      "loss": 1.0066,
+      "step": 9710
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.966519355773926,
+      "learning_rate": 1.187549679956491e-05,
+      "loss": 0.6277,
+      "step": 9711
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 17.250471115112305,
+      "learning_rate": 1.1874660084508222e-05,
+      "loss": 1.2056,
+      "step": 9712
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 18.81276512145996,
+      "learning_rate": 1.1873823369451534e-05,
+      "loss": 1.4703,
+      "step": 9713
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 14.694525718688965,
+      "learning_rate": 1.1872986654394848e-05,
+      "loss": 1.6066,
+      "step": 9714
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 24.63129425048828,
+      "learning_rate": 1.187214993933816e-05,
+      "loss": 3.3113,
+      "step": 9715
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 16.88232421875,
+      "learning_rate": 1.1871313224281471e-05,
+      "loss": 0.742,
+      "step": 9716
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.315510749816895,
+      "learning_rate": 1.1870476509224783e-05,
+      "loss": 0.6042,
+      "step": 9717
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.096471786499023,
+      "learning_rate": 1.1869639794168097e-05,
+      "loss": 0.448,
+      "step": 9718
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 16.892091751098633,
+      "learning_rate": 1.186880307911141e-05,
+      "loss": 1.3073,
+      "step": 9719
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.081926345825195,
+      "learning_rate": 1.1867966364054721e-05,
+      "loss": 0.9993,
+      "step": 9720
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.064104080200195,
+      "learning_rate": 1.1867129648998035e-05,
+      "loss": 0.731,
+      "step": 9721
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 50.32485580444336,
+      "learning_rate": 1.1866292933941348e-05,
+      "loss": 1.3959,
+      "step": 9722
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 26.517179489135742,
+      "learning_rate": 1.1865456218884659e-05,
+      "loss": 4.9971,
+      "step": 9723
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 98.24976348876953,
+      "learning_rate": 1.1864619503827972e-05,
+      "loss": 1.5253,
+      "step": 9724
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 36.602752685546875,
+      "learning_rate": 1.1863782788771286e-05,
+      "loss": 1.2071,
+      "step": 9725
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.339238166809082,
+      "learning_rate": 1.1862946073714598e-05,
+      "loss": 1.1399,
+      "step": 9726
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 17.146467208862305,
+      "learning_rate": 1.186210935865791e-05,
+      "loss": 0.4849,
+      "step": 9727
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.974125862121582,
+      "learning_rate": 1.1861272643601223e-05,
+      "loss": 1.5309,
+      "step": 9728
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 13.77196979522705,
+      "learning_rate": 1.1860435928544535e-05,
+      "loss": 0.5355,
+      "step": 9729
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.349488735198975,
+      "learning_rate": 1.1859599213487847e-05,
+      "loss": 1.2569,
+      "step": 9730
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 20.587677001953125,
+      "learning_rate": 1.185876249843116e-05,
+      "loss": 1.1589,
+      "step": 9731
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 20.375568389892578,
+      "learning_rate": 1.1857925783374473e-05,
+      "loss": 2.3431,
+      "step": 9732
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.66171932220459,
+      "learning_rate": 1.1857089068317787e-05,
+      "loss": 1.5323,
+      "step": 9733
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 14.362919807434082,
+      "learning_rate": 1.1856252353261097e-05,
+      "loss": 1.5258,
+      "step": 9734
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 18.112083435058594,
+      "learning_rate": 1.185541563820441e-05,
+      "loss": 2.2156,
+      "step": 9735
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.53700590133667,
+      "learning_rate": 1.1854578923147724e-05,
+      "loss": 1.7609,
+      "step": 9736
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.45809268951416,
+      "learning_rate": 1.1853742208091034e-05,
+      "loss": 1.2105,
+      "step": 9737
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 99.39317321777344,
+      "learning_rate": 1.1852905493034348e-05,
+      "loss": 2.9131,
+      "step": 9738
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 14.499650955200195,
+      "learning_rate": 1.1852068777977662e-05,
+      "loss": 1.1414,
+      "step": 9739
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 9.993605613708496,
+      "learning_rate": 1.1851232062920974e-05,
+      "loss": 2.3243,
+      "step": 9740
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 32.130714416503906,
+      "learning_rate": 1.1850395347864286e-05,
+      "loss": 0.9311,
+      "step": 9741
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.0457258224487305,
+      "learning_rate": 1.18495586328076e-05,
+      "loss": 0.4676,
+      "step": 9742
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 56.041038513183594,
+      "learning_rate": 1.1848721917750911e-05,
+      "loss": 1.9098,
+      "step": 9743
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 19.361499786376953,
+      "learning_rate": 1.1847885202694223e-05,
+      "loss": 1.6205,
+      "step": 9744
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 21.862525939941406,
+      "learning_rate": 1.1847048487637535e-05,
+      "loss": 1.8168,
+      "step": 9745
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.574731826782227,
+      "learning_rate": 1.1846211772580849e-05,
+      "loss": 0.8181,
+      "step": 9746
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 21.914243698120117,
+      "learning_rate": 1.1845375057524162e-05,
+      "loss": 1.3609,
+      "step": 9747
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 10.41545581817627,
+      "learning_rate": 1.1844538342467473e-05,
+      "loss": 0.5635,
+      "step": 9748
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.6245574951171875,
+      "learning_rate": 1.1843701627410786e-05,
+      "loss": 1.0122,
+      "step": 9749
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 19.98027992248535,
+      "learning_rate": 1.18428649123541e-05,
+      "loss": 1.5541,
+      "step": 9750
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 6.076691627502441,
+      "learning_rate": 1.184202819729741e-05,
+      "loss": 0.6404,
+      "step": 9751
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.035378456115723,
+      "learning_rate": 1.1841191482240724e-05,
+      "loss": 0.839,
+      "step": 9752
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.503005981445312,
+      "learning_rate": 1.1840354767184037e-05,
+      "loss": 1.9198,
+      "step": 9753
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 19.694664001464844,
+      "learning_rate": 1.183951805212735e-05,
+      "loss": 1.6603,
+      "step": 9754
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.645819664001465,
+      "learning_rate": 1.1838681337070661e-05,
+      "loss": 0.6592,
+      "step": 9755
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 5.133186340332031,
+      "learning_rate": 1.1837844622013975e-05,
+      "loss": 0.3353,
+      "step": 9756
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 21.087446212768555,
+      "learning_rate": 1.1837007906957287e-05,
+      "loss": 3.1026,
+      "step": 9757
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 19.393712997436523,
+      "learning_rate": 1.1836171191900599e-05,
+      "loss": 1.9517,
+      "step": 9758
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 77.39122772216797,
+      "learning_rate": 1.1835334476843911e-05,
+      "loss": 1.8786,
+      "step": 9759
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 22.556842803955078,
+      "learning_rate": 1.1834497761787225e-05,
+      "loss": 0.9785,
+      "step": 9760
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 12.212501525878906,
+      "learning_rate": 1.1833661046730538e-05,
+      "loss": 1.126,
+      "step": 9761
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.766206741333008,
+      "learning_rate": 1.1832824331673848e-05,
+      "loss": 1.5055,
+      "step": 9762
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 17.554147720336914,
+      "learning_rate": 1.1831987616617162e-05,
+      "loss": 0.9103,
+      "step": 9763
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 23.220369338989258,
+      "learning_rate": 1.1831150901560476e-05,
+      "loss": 3.0983,
+      "step": 9764
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 22.07280158996582,
+      "learning_rate": 1.1830314186503786e-05,
+      "loss": 1.9888,
+      "step": 9765
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 8.112464904785156,
+      "learning_rate": 1.18294774714471e-05,
+      "loss": 0.3538,
+      "step": 9766
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.294885635375977,
+      "learning_rate": 1.1828640756390413e-05,
+      "loss": 1.4769,
+      "step": 9767
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.150555610656738,
+      "learning_rate": 1.1827804041333725e-05,
+      "loss": 2.2846,
+      "step": 9768
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 5.326368808746338,
+      "learning_rate": 1.1826967326277037e-05,
+      "loss": 0.5095,
+      "step": 9769
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 13.025184631347656,
+      "learning_rate": 1.182613061122035e-05,
+      "loss": 1.3693,
+      "step": 9770
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.305625915527344,
+      "learning_rate": 1.1825293896163663e-05,
+      "loss": 1.7269,
+      "step": 9771
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 16.152267456054688,
+      "learning_rate": 1.1824457181106975e-05,
+      "loss": 1.6834,
+      "step": 9772
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 17.717496871948242,
+      "learning_rate": 1.1823620466050287e-05,
+      "loss": 1.5123,
+      "step": 9773
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 6.071959972381592,
+      "learning_rate": 1.18227837509936e-05,
+      "loss": 0.5821,
+      "step": 9774
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 35.208396911621094,
+      "learning_rate": 1.1821947035936914e-05,
+      "loss": 0.7074,
+      "step": 9775
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 21.743438720703125,
+      "learning_rate": 1.1821110320880224e-05,
+      "loss": 1.6365,
+      "step": 9776
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 15.401273727416992,
+      "learning_rate": 1.1820273605823538e-05,
+      "loss": 2.3279,
+      "step": 9777
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.5009899139404297,
+      "learning_rate": 1.1819436890766852e-05,
+      "loss": 0.0616,
+      "step": 9778
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 8.305074691772461,
+      "learning_rate": 1.1818600175710162e-05,
+      "loss": 0.8248,
+      "step": 9779
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 8.066091537475586,
+      "learning_rate": 1.1817763460653476e-05,
+      "loss": 0.7442,
+      "step": 9780
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 7.680723667144775,
+      "learning_rate": 1.181692674559679e-05,
+      "loss": 1.5972,
+      "step": 9781
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 21.709930419921875,
+      "learning_rate": 1.1816090030540101e-05,
+      "loss": 1.9213,
+      "step": 9782
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 8.652266502380371,
+      "learning_rate": 1.1815253315483413e-05,
+      "loss": 1.3114,
+      "step": 9783
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.539308547973633,
+      "learning_rate": 1.1814416600426725e-05,
+      "loss": 1.4542,
+      "step": 9784
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 6.910467147827148,
+      "learning_rate": 1.1813579885370039e-05,
+      "loss": 0.9087,
+      "step": 9785
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 4.772772312164307,
+      "learning_rate": 1.181274317031335e-05,
+      "loss": 0.1523,
+      "step": 9786
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 13.30754280090332,
+      "learning_rate": 1.1811906455256663e-05,
+      "loss": 1.1662,
+      "step": 9787
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 17.31911277770996,
+      "learning_rate": 1.1811069740199976e-05,
+      "loss": 1.2639,
+      "step": 9788
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.7666053771972656,
+      "learning_rate": 1.181023302514329e-05,
+      "loss": 0.0387,
+      "step": 9789
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.537734985351562,
+      "learning_rate": 1.18093963100866e-05,
+      "loss": 1.752,
+      "step": 9790
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 7.068645477294922,
+      "learning_rate": 1.1808559595029914e-05,
+      "loss": 1.4841,
+      "step": 9791
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 14.287550926208496,
+      "learning_rate": 1.1807722879973227e-05,
+      "loss": 2.7825,
+      "step": 9792
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 52.4526252746582,
+      "learning_rate": 1.1806886164916538e-05,
+      "loss": 2.5035,
+      "step": 9793
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.312798500061035,
+      "learning_rate": 1.1806049449859851e-05,
+      "loss": 1.8516,
+      "step": 9794
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 31.503602981567383,
+      "learning_rate": 1.1805212734803165e-05,
+      "loss": 1.1349,
+      "step": 9795
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.658240556716919,
+      "learning_rate": 1.1804376019746477e-05,
+      "loss": 0.3557,
+      "step": 9796
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 18.69940185546875,
+      "learning_rate": 1.1803539304689789e-05,
+      "loss": 1.3528,
+      "step": 9797
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 17.545040130615234,
+      "learning_rate": 1.1802702589633101e-05,
+      "loss": 0.8369,
+      "step": 9798
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 5.3122944831848145,
+      "learning_rate": 1.1801865874576415e-05,
+      "loss": 1.507,
+      "step": 9799
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 17.907398223876953,
+      "learning_rate": 1.1801029159519726e-05,
+      "loss": 1.8676,
+      "step": 9800
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 24.098581314086914,
+      "learning_rate": 1.1800192444463038e-05,
+      "loss": 2.6623,
+      "step": 9801
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 15.955728530883789,
+      "learning_rate": 1.1799355729406352e-05,
+      "loss": 1.4016,
+      "step": 9802
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 4.137416839599609,
+      "learning_rate": 1.1798519014349666e-05,
+      "loss": 0.114,
+      "step": 9803
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 20.56151580810547,
+      "learning_rate": 1.1797682299292976e-05,
+      "loss": 1.1279,
+      "step": 9804
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 18.528879165649414,
+      "learning_rate": 1.179684558423629e-05,
+      "loss": 1.7007,
+      "step": 9805
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.145833969116211,
+      "learning_rate": 1.1796008869179603e-05,
+      "loss": 2.0885,
+      "step": 9806
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 4.693984031677246,
+      "learning_rate": 1.1795172154122914e-05,
+      "loss": 1.038,
+      "step": 9807
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 16.135169982910156,
+      "learning_rate": 1.1794335439066227e-05,
+      "loss": 1.1995,
+      "step": 9808
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 13.0604248046875,
+      "learning_rate": 1.179349872400954e-05,
+      "loss": 3.9162,
+      "step": 9809
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 7.696951389312744,
+      "learning_rate": 1.1792662008952853e-05,
+      "loss": 0.8307,
+      "step": 9810
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.879096031188965,
+      "learning_rate": 1.1791825293896165e-05,
+      "loss": 1.2128,
+      "step": 9811
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.9865870475769043,
+      "learning_rate": 1.1790988578839477e-05,
+      "loss": 0.3161,
+      "step": 9812
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 14.160734176635742,
+      "learning_rate": 1.179015186378279e-05,
+      "loss": 2.3232,
+      "step": 9813
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 14.625443458557129,
+      "learning_rate": 1.1789315148726102e-05,
+      "loss": 2.0368,
+      "step": 9814
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 17.27570915222168,
+      "learning_rate": 1.1788478433669414e-05,
+      "loss": 0.7004,
+      "step": 9815
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 5.550584316253662,
+      "learning_rate": 1.1787641718612728e-05,
+      "loss": 0.354,
+      "step": 9816
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 6.084341049194336,
+      "learning_rate": 1.1786805003556038e-05,
+      "loss": 1.1003,
+      "step": 9817
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 21.174999237060547,
+      "learning_rate": 1.1785968288499352e-05,
+      "loss": 1.3518,
+      "step": 9818
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 12.584552764892578,
+      "learning_rate": 1.1785131573442665e-05,
+      "loss": 2.6013,
+      "step": 9819
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 21.88260269165039,
+      "learning_rate": 1.1784294858385979e-05,
+      "loss": 1.9915,
+      "step": 9820
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.529659271240234,
+      "learning_rate": 1.178345814332929e-05,
+      "loss": 0.7695,
+      "step": 9821
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 7.979567527770996,
+      "learning_rate": 1.1782621428272603e-05,
+      "loss": 0.9663,
+      "step": 9822
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 18.05103874206543,
+      "learning_rate": 1.1781784713215915e-05,
+      "loss": 3.1226,
+      "step": 9823
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 6.140756607055664,
+      "learning_rate": 1.1780947998159227e-05,
+      "loss": 0.2484,
+      "step": 9824
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 7.758366107940674,
+      "learning_rate": 1.178011128310254e-05,
+      "loss": 0.21,
+      "step": 9825
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 36.641845703125,
+      "learning_rate": 1.1779274568045853e-05,
+      "loss": 2.524,
+      "step": 9826
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 5.3544158935546875,
+      "learning_rate": 1.1778437852989166e-05,
+      "loss": 0.1326,
+      "step": 9827
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 90.64929962158203,
+      "learning_rate": 1.1777601137932476e-05,
+      "loss": 2.0421,
+      "step": 9828
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 14.497895240783691,
+      "learning_rate": 1.177676442287579e-05,
+      "loss": 1.7791,
+      "step": 9829
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 36.353668212890625,
+      "learning_rate": 1.1775927707819104e-05,
+      "loss": 0.3409,
+      "step": 9830
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 18.138263702392578,
+      "learning_rate": 1.1775090992762414e-05,
+      "loss": 0.6608,
+      "step": 9831
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.4441068768501282,
+      "learning_rate": 1.1774254277705728e-05,
+      "loss": 0.0099,
+      "step": 9832
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 16.02644157409668,
+      "learning_rate": 1.1773417562649041e-05,
+      "loss": 1.1821,
+      "step": 9833
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.014534950256348,
+      "learning_rate": 1.1772580847592355e-05,
+      "loss": 1.8367,
+      "step": 9834
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 13.180492401123047,
+      "learning_rate": 1.1771744132535665e-05,
+      "loss": 0.5815,
+      "step": 9835
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 55.19274139404297,
+      "learning_rate": 1.1770907417478979e-05,
+      "loss": 1.3104,
+      "step": 9836
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.048669815063477,
+      "learning_rate": 1.177007070242229e-05,
+      "loss": 1.456,
+      "step": 9837
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 40.642486572265625,
+      "learning_rate": 1.1769233987365603e-05,
+      "loss": 1.2127,
+      "step": 9838
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 17.61302375793457,
+      "learning_rate": 1.1768397272308916e-05,
+      "loss": 1.1644,
+      "step": 9839
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.714911460876465,
+      "learning_rate": 1.1767560557252228e-05,
+      "loss": 1.2779,
+      "step": 9840
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.274693489074707,
+      "learning_rate": 1.1766723842195542e-05,
+      "loss": 2.0046,
+      "step": 9841
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 4.059805393218994,
+      "learning_rate": 1.1765887127138852e-05,
+      "loss": 0.1009,
+      "step": 9842
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 20.27713394165039,
+      "learning_rate": 1.1765050412082166e-05,
+      "loss": 1.6268,
+      "step": 9843
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 82.43860626220703,
+      "learning_rate": 1.176421369702548e-05,
+      "loss": 2.0341,
+      "step": 9844
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.47548770904541,
+      "learning_rate": 1.176337698196879e-05,
+      "loss": 1.4052,
+      "step": 9845
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.602141380310059,
+      "learning_rate": 1.1762540266912103e-05,
+      "loss": 0.8891,
+      "step": 9846
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 16.42512321472168,
+      "learning_rate": 1.1761703551855417e-05,
+      "loss": 0.9752,
+      "step": 9847
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 21.03297233581543,
+      "learning_rate": 1.176086683679873e-05,
+      "loss": 1.2994,
+      "step": 9848
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.0785598754882812,
+      "learning_rate": 1.1760030121742041e-05,
+      "loss": 0.1167,
+      "step": 9849
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 17.921354293823242,
+      "learning_rate": 1.1759193406685355e-05,
+      "loss": 1.7992,
+      "step": 9850
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 18.078763961791992,
+      "learning_rate": 1.1758356691628667e-05,
+      "loss": 1.1432,
+      "step": 9851
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.481141090393066,
+      "learning_rate": 1.1757519976571979e-05,
+      "loss": 0.5549,
+      "step": 9852
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 21.629512786865234,
+      "learning_rate": 1.1756683261515292e-05,
+      "loss": 1.5444,
+      "step": 9853
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 11.057011604309082,
+      "learning_rate": 1.1755846546458604e-05,
+      "loss": 0.4012,
+      "step": 9854
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 16.92287826538086,
+      "learning_rate": 1.1755009831401918e-05,
+      "loss": 1.2856,
+      "step": 9855
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 18.003637313842773,
+      "learning_rate": 1.1754173116345228e-05,
+      "loss": 0.7065,
+      "step": 9856
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 13.426071166992188,
+      "learning_rate": 1.1753336401288542e-05,
+      "loss": 2.3196,
+      "step": 9857
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 14.647819519042969,
+      "learning_rate": 1.1752499686231855e-05,
+      "loss": 0.7164,
+      "step": 9858
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 13.148072242736816,
+      "learning_rate": 1.1751662971175166e-05,
+      "loss": 0.8424,
+      "step": 9859
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 11.636898040771484,
+      "learning_rate": 1.175082625611848e-05,
+      "loss": 0.9816,
+      "step": 9860
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.009572982788086,
+      "learning_rate": 1.1749989541061793e-05,
+      "loss": 1.0581,
+      "step": 9861
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 14.94993782043457,
+      "learning_rate": 1.1749152826005105e-05,
+      "loss": 1.0659,
+      "step": 9862
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 14.876943588256836,
+      "learning_rate": 1.1748316110948417e-05,
+      "loss": 1.4571,
+      "step": 9863
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 28.33251953125,
+      "learning_rate": 1.174747939589173e-05,
+      "loss": 1.6219,
+      "step": 9864
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.158002853393555,
+      "learning_rate": 1.1746642680835042e-05,
+      "loss": 0.3761,
+      "step": 9865
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 16.329578399658203,
+      "learning_rate": 1.1745805965778354e-05,
+      "loss": 0.9413,
+      "step": 9866
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.986481666564941,
+      "learning_rate": 1.1744969250721668e-05,
+      "loss": 0.9236,
+      "step": 9867
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.439972400665283,
+      "learning_rate": 1.174413253566498e-05,
+      "loss": 0.0313,
+      "step": 9868
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.496933937072754,
+      "learning_rate": 1.1743295820608294e-05,
+      "loss": 0.9859,
+      "step": 9869
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 13.342422485351562,
+      "learning_rate": 1.1742459105551604e-05,
+      "loss": 0.8337,
+      "step": 9870
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 28.25615119934082,
+      "learning_rate": 1.1741622390494918e-05,
+      "loss": 2.9762,
+      "step": 9871
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.167847633361816,
+      "learning_rate": 1.1740785675438231e-05,
+      "loss": 0.8082,
+      "step": 9872
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 17.874998092651367,
+      "learning_rate": 1.1739948960381542e-05,
+      "loss": 2.2444,
+      "step": 9873
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 13.657346725463867,
+      "learning_rate": 1.1739112245324855e-05,
+      "loss": 1.3361,
+      "step": 9874
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 26.667905807495117,
+      "learning_rate": 1.1738275530268169e-05,
+      "loss": 1.6523,
+      "step": 9875
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 23.977022171020508,
+      "learning_rate": 1.173743881521148e-05,
+      "loss": 2.1226,
+      "step": 9876
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.629140853881836,
+      "learning_rate": 1.1736602100154793e-05,
+      "loss": 1.351,
+      "step": 9877
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.725996971130371,
+      "learning_rate": 1.1735765385098106e-05,
+      "loss": 0.6322,
+      "step": 9878
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.691288948059082,
+      "learning_rate": 1.1734928670041418e-05,
+      "loss": 0.3133,
+      "step": 9879
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.4821033477783203,
+      "learning_rate": 1.173409195498473e-05,
+      "loss": 0.0902,
+      "step": 9880
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 13.584442138671875,
+      "learning_rate": 1.1733255239928042e-05,
+      "loss": 2.056,
+      "step": 9881
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 22.304651260375977,
+      "learning_rate": 1.1732418524871356e-05,
+      "loss": 1.782,
+      "step": 9882
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 15.612174987792969,
+      "learning_rate": 1.173158180981467e-05,
+      "loss": 1.2614,
+      "step": 9883
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 16.01333999633789,
+      "learning_rate": 1.173074509475798e-05,
+      "loss": 1.5578,
+      "step": 9884
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.732401847839355,
+      "learning_rate": 1.1729908379701293e-05,
+      "loss": 1.0354,
+      "step": 9885
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 30.70211410522461,
+      "learning_rate": 1.1729071664644607e-05,
+      "loss": 1.0566,
+      "step": 9886
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 9.202712059020996,
+      "learning_rate": 1.1728234949587917e-05,
+      "loss": 0.4007,
+      "step": 9887
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 68.00823211669922,
+      "learning_rate": 1.1727398234531231e-05,
+      "loss": 1.4828,
+      "step": 9888
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 11.506952285766602,
+      "learning_rate": 1.1726561519474545e-05,
+      "loss": 1.145,
+      "step": 9889
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 21.97418975830078,
+      "learning_rate": 1.1725724804417857e-05,
+      "loss": 1.2828,
+      "step": 9890
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 13.470939636230469,
+      "learning_rate": 1.1724888089361169e-05,
+      "loss": 1.4388,
+      "step": 9891
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 15.934314727783203,
+      "learning_rate": 1.1724051374304482e-05,
+      "loss": 1.5466,
+      "step": 9892
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 16.9025936126709,
+      "learning_rate": 1.1723214659247794e-05,
+      "loss": 1.1422,
+      "step": 9893
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 11.071962356567383,
+      "learning_rate": 1.1722377944191106e-05,
+      "loss": 2.2181,
+      "step": 9894
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.60844612121582,
+      "learning_rate": 1.1721541229134418e-05,
+      "loss": 1.6378,
+      "step": 9895
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 19.60494041442871,
+      "learning_rate": 1.1720704514077732e-05,
+      "loss": 0.4927,
+      "step": 9896
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 6.539211273193359,
+      "learning_rate": 1.1719867799021045e-05,
+      "loss": 0.5612,
+      "step": 9897
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 83.48982238769531,
+      "learning_rate": 1.1719031083964356e-05,
+      "loss": 1.712,
+      "step": 9898
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 11.819092750549316,
+      "learning_rate": 1.171819436890767e-05,
+      "loss": 1.1643,
+      "step": 9899
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 40.66287612915039,
+      "learning_rate": 1.1717357653850983e-05,
+      "loss": 1.8852,
+      "step": 9900
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 4.276702880859375,
+      "learning_rate": 1.1716520938794293e-05,
+      "loss": 0.2434,
+      "step": 9901
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 12.493239402770996,
+      "learning_rate": 1.1715684223737607e-05,
+      "loss": 0.4886,
+      "step": 9902
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.208495140075684,
+      "learning_rate": 1.171484750868092e-05,
+      "loss": 2.4325,
+      "step": 9903
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 12.189680099487305,
+      "learning_rate": 1.1714010793624232e-05,
+      "loss": 1.207,
+      "step": 9904
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.993623733520508,
+      "learning_rate": 1.1713174078567544e-05,
+      "loss": 1.8157,
+      "step": 9905
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 7.953149318695068,
+      "learning_rate": 1.1712337363510858e-05,
+      "loss": 0.8497,
+      "step": 9906
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 14.442801475524902,
+      "learning_rate": 1.171150064845417e-05,
+      "loss": 0.9703,
+      "step": 9907
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 27.209766387939453,
+      "learning_rate": 1.1710663933397482e-05,
+      "loss": 2.2167,
+      "step": 9908
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 34.28467559814453,
+      "learning_rate": 1.1709827218340794e-05,
+      "loss": 1.4908,
+      "step": 9909
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 101.06977844238281,
+      "learning_rate": 1.1708990503284108e-05,
+      "loss": 2.5449,
+      "step": 9910
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 16.86500358581543,
+      "learning_rate": 1.1708153788227421e-05,
+      "loss": 1.1078,
+      "step": 9911
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.253978729248047,
+      "learning_rate": 1.1707317073170731e-05,
+      "loss": 1.1973,
+      "step": 9912
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 4.450164794921875,
+      "learning_rate": 1.1706480358114045e-05,
+      "loss": 0.4521,
+      "step": 9913
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 19.197463989257812,
+      "learning_rate": 1.1705643643057359e-05,
+      "loss": 2.4202,
+      "step": 9914
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 12.841304779052734,
+      "learning_rate": 1.1704806928000669e-05,
+      "loss": 1.0224,
+      "step": 9915
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 24.760807037353516,
+      "learning_rate": 1.1703970212943983e-05,
+      "loss": 1.556,
+      "step": 9916
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.061955451965332,
+      "learning_rate": 1.1703133497887296e-05,
+      "loss": 0.5102,
+      "step": 9917
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 13.719026565551758,
+      "learning_rate": 1.1702296782830608e-05,
+      "loss": 0.7795,
+      "step": 9918
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 8.608823776245117,
+      "learning_rate": 1.170146006777392e-05,
+      "loss": 2.1716,
+      "step": 9919
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 65.23345184326172,
+      "learning_rate": 1.1700623352717234e-05,
+      "loss": 1.4773,
+      "step": 9920
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.659074783325195,
+      "learning_rate": 1.1699786637660546e-05,
+      "loss": 0.7318,
+      "step": 9921
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.6577730178833,
+      "learning_rate": 1.1698949922603858e-05,
+      "loss": 0.9231,
+      "step": 9922
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 5.324812889099121,
+      "learning_rate": 1.169811320754717e-05,
+      "loss": 0.6034,
+      "step": 9923
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 14.057004928588867,
+      "learning_rate": 1.1697276492490483e-05,
+      "loss": 0.9089,
+      "step": 9924
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 43.46983337402344,
+      "learning_rate": 1.1696439777433797e-05,
+      "loss": 1.6469,
+      "step": 9925
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 18.37807273864746,
+      "learning_rate": 1.1695603062377107e-05,
+      "loss": 0.5716,
+      "step": 9926
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.710125923156738,
+      "learning_rate": 1.1694766347320421e-05,
+      "loss": 2.0483,
+      "step": 9927
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 6.957853317260742,
+      "learning_rate": 1.1693929632263735e-05,
+      "loss": 0.4949,
+      "step": 9928
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 122.19803619384766,
+      "learning_rate": 1.1693092917207045e-05,
+      "loss": 2.0142,
+      "step": 9929
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 4.989986419677734,
+      "learning_rate": 1.1692256202150359e-05,
+      "loss": 0.5258,
+      "step": 9930
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 17.40357208251953,
+      "learning_rate": 1.1691419487093672e-05,
+      "loss": 0.9121,
+      "step": 9931
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 14.683014869689941,
+      "learning_rate": 1.1690582772036984e-05,
+      "loss": 1.2121,
+      "step": 9932
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 42.671024322509766,
+      "learning_rate": 1.1689746056980296e-05,
+      "loss": 2.0181,
+      "step": 9933
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 13.469338417053223,
+      "learning_rate": 1.1688909341923608e-05,
+      "loss": 4.0186,
+      "step": 9934
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 28.573440551757812,
+      "learning_rate": 1.1688072626866922e-05,
+      "loss": 1.1826,
+      "step": 9935
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 70.24644470214844,
+      "learning_rate": 1.1687235911810234e-05,
+      "loss": 2.1922,
+      "step": 9936
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 21.14872169494629,
+      "learning_rate": 1.1686399196753546e-05,
+      "loss": 1.451,
+      "step": 9937
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 6.562117576599121,
+      "learning_rate": 1.168556248169686e-05,
+      "loss": 0.2988,
+      "step": 9938
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 31.07131004333496,
+      "learning_rate": 1.1684725766640173e-05,
+      "loss": 1.3988,
+      "step": 9939
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 12.469810485839844,
+      "learning_rate": 1.1683889051583483e-05,
+      "loss": 1.2987,
+      "step": 9940
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 16.9166202545166,
+      "learning_rate": 1.1683052336526797e-05,
+      "loss": 3.1104,
+      "step": 9941
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 12.717508316040039,
+      "learning_rate": 1.168221562147011e-05,
+      "loss": 2.7321,
+      "step": 9942
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 17.788097381591797,
+      "learning_rate": 1.168137890641342e-05,
+      "loss": 1.7442,
+      "step": 9943
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.754151344299316,
+      "learning_rate": 1.1680542191356734e-05,
+      "loss": 0.6006,
+      "step": 9944
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 18.577787399291992,
+      "learning_rate": 1.1679705476300048e-05,
+      "loss": 2.614,
+      "step": 9945
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 48.640342712402344,
+      "learning_rate": 1.167886876124336e-05,
+      "loss": 1.2527,
+      "step": 9946
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 59.63102340698242,
+      "learning_rate": 1.1678032046186672e-05,
+      "loss": 1.4984,
+      "step": 9947
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 22.24765396118164,
+      "learning_rate": 1.1677195331129984e-05,
+      "loss": 1.0165,
+      "step": 9948
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 20.635051727294922,
+      "learning_rate": 1.1676358616073298e-05,
+      "loss": 1.4636,
+      "step": 9949
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 14.577654838562012,
+      "learning_rate": 1.167552190101661e-05,
+      "loss": 1.8911,
+      "step": 9950
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 15.311452865600586,
+      "learning_rate": 1.1674685185959921e-05,
+      "loss": 0.9604,
+      "step": 9951
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 23.743816375732422,
+      "learning_rate": 1.1673848470903235e-05,
+      "loss": 2.5244,
+      "step": 9952
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 22.989370346069336,
+      "learning_rate": 1.1673011755846549e-05,
+      "loss": 2.0244,
+      "step": 9953
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 14.441717147827148,
+      "learning_rate": 1.1672175040789859e-05,
+      "loss": 0.8623,
+      "step": 9954
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.20020580291748,
+      "learning_rate": 1.1671338325733173e-05,
+      "loss": 1.2268,
+      "step": 9955
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 16.0340576171875,
+      "learning_rate": 1.1670501610676486e-05,
+      "loss": 1.6768,
+      "step": 9956
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 5.675755977630615,
+      "learning_rate": 1.1669664895619797e-05,
+      "loss": 1.0931,
+      "step": 9957
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 5.870267391204834,
+      "learning_rate": 1.166882818056311e-05,
+      "loss": 0.6885,
+      "step": 9958
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 15.182581901550293,
+      "learning_rate": 1.1667991465506424e-05,
+      "loss": 1.7034,
+      "step": 9959
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 41.7207145690918,
+      "learning_rate": 1.1667154750449736e-05,
+      "loss": 1.8537,
+      "step": 9960
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.536687850952148,
+      "learning_rate": 1.1666318035393048e-05,
+      "loss": 0.9629,
+      "step": 9961
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 5.596810817718506,
+      "learning_rate": 1.166548132033636e-05,
+      "loss": 0.3131,
+      "step": 9962
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 24.205638885498047,
+      "learning_rate": 1.1664644605279673e-05,
+      "loss": 1.0994,
+      "step": 9963
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 80.2908706665039,
+      "learning_rate": 1.1663807890222985e-05,
+      "loss": 0.9488,
+      "step": 9964
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.021303176879883,
+      "learning_rate": 1.1662971175166297e-05,
+      "loss": 1.2746,
+      "step": 9965
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.356732368469238,
+      "learning_rate": 1.1662134460109611e-05,
+      "loss": 1.9948,
+      "step": 9966
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0912623405456543,
+      "learning_rate": 1.1661297745052925e-05,
+      "loss": 0.0345,
+      "step": 9967
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 31.7636661529541,
+      "learning_rate": 1.1660461029996235e-05,
+      "loss": 1.817,
+      "step": 9968
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 6.6061553955078125,
+      "learning_rate": 1.1659624314939548e-05,
+      "loss": 1.1319,
+      "step": 9969
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 13.415298461914062,
+      "learning_rate": 1.1658787599882862e-05,
+      "loss": 0.6593,
+      "step": 9970
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 14.584321022033691,
+      "learning_rate": 1.1657950884826172e-05,
+      "loss": 1.9673,
+      "step": 9971
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 14.611739158630371,
+      "learning_rate": 1.1657114169769486e-05,
+      "loss": 0.9902,
+      "step": 9972
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.319719314575195,
+      "learning_rate": 1.1656277454712798e-05,
+      "loss": 1.1964,
+      "step": 9973
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 20.697874069213867,
+      "learning_rate": 1.1655440739656112e-05,
+      "loss": 1.9791,
+      "step": 9974
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 7.637447834014893,
+      "learning_rate": 1.1654604024599424e-05,
+      "loss": 1.1018,
+      "step": 9975
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.925323486328125,
+      "learning_rate": 1.1653767309542736e-05,
+      "loss": 0.5566,
+      "step": 9976
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 80.9381332397461,
+      "learning_rate": 1.165293059448605e-05,
+      "loss": 1.4074,
+      "step": 9977
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 14.44371223449707,
+      "learning_rate": 1.1652093879429361e-05,
+      "loss": 1.5644,
+      "step": 9978
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.81545639038086,
+      "learning_rate": 1.1651257164372673e-05,
+      "loss": 1.211,
+      "step": 9979
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 23.71071434020996,
+      "learning_rate": 1.1650420449315987e-05,
+      "loss": 1.4741,
+      "step": 9980
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 9.054380416870117,
+      "learning_rate": 1.16495837342593e-05,
+      "loss": 2.1966,
+      "step": 9981
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 13.376519203186035,
+      "learning_rate": 1.164874701920261e-05,
+      "loss": 3.4557,
+      "step": 9982
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 22.065784454345703,
+      "learning_rate": 1.1647910304145924e-05,
+      "loss": 2.3747,
+      "step": 9983
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.869088172912598,
+      "learning_rate": 1.1647073589089238e-05,
+      "loss": 1.893,
+      "step": 9984
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 12.72069263458252,
+      "learning_rate": 1.1646236874032548e-05,
+      "loss": 1.6795,
+      "step": 9985
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.286107063293457,
+      "learning_rate": 1.1645400158975862e-05,
+      "loss": 1.1739,
+      "step": 9986
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 28.606319427490234,
+      "learning_rate": 1.1644563443919174e-05,
+      "loss": 1.7524,
+      "step": 9987
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.5166430473327637,
+      "learning_rate": 1.1643726728862487e-05,
+      "loss": 0.0647,
+      "step": 9988
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 21.91144561767578,
+      "learning_rate": 1.16428900138058e-05,
+      "loss": 1.3016,
+      "step": 9989
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 23.434961318969727,
+      "learning_rate": 1.1642053298749111e-05,
+      "loss": 0.9999,
+      "step": 9990
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.6171557903289795,
+      "learning_rate": 1.1641216583692425e-05,
+      "loss": 0.3816,
+      "step": 9991
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 25.213613510131836,
+      "learning_rate": 1.1640379868635735e-05,
+      "loss": 2.231,
+      "step": 9992
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.4147605895996094,
+      "learning_rate": 1.1639543153579049e-05,
+      "loss": 0.1599,
+      "step": 9993
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.12276840209961,
+      "learning_rate": 1.1638706438522363e-05,
+      "loss": 0.9217,
+      "step": 9994
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 7.428321838378906,
+      "learning_rate": 1.1637869723465676e-05,
+      "loss": 1.4817,
+      "step": 9995
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.9801888465881348,
+      "learning_rate": 1.1637033008408986e-05,
+      "loss": 0.0822,
+      "step": 9996
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 15.356182098388672,
+      "learning_rate": 1.16361962933523e-05,
+      "loss": 1.5217,
+      "step": 9997
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.371597290039062,
+      "learning_rate": 1.1635359578295614e-05,
+      "loss": 2.9021,
+      "step": 9998
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 18.860326766967773,
+      "learning_rate": 1.1634522863238924e-05,
+      "loss": 1.3427,
+      "step": 9999
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 8.820758819580078,
+      "learning_rate": 1.1633686148182238e-05,
+      "loss": 2.0757,
+      "step": 10000
+    },
+    {
+      "epoch": 1.25,
+      "eval_loss": 0.08529716730117798,
+      "eval_runtime": 95.5005,
+      "eval_samples_per_second": 37.089,
+      "eval_steps_per_second": 37.089,
+      "step": 10000
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 18.028676986694336,
+      "learning_rate": 1.163284943312555e-05,
+      "loss": 2.2488,
+      "step": 10001
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 23.154457092285156,
+      "learning_rate": 1.1632012718068863e-05,
+      "loss": 0.8694,
+      "step": 10002
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 10.192426681518555,
+      "learning_rate": 1.1631176003012175e-05,
+      "loss": 0.8179,
+      "step": 10003
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 37.10844802856445,
+      "learning_rate": 1.1630339287955487e-05,
+      "loss": 1.7208,
+      "step": 10004
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 44.653175354003906,
+      "learning_rate": 1.1629502572898801e-05,
+      "loss": 0.8287,
+      "step": 10005
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 4.824666976928711,
+      "learning_rate": 1.1628665857842111e-05,
+      "loss": 0.2451,
+      "step": 10006
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 10.745501518249512,
+      "learning_rate": 1.1627829142785425e-05,
+      "loss": 1.5447,
+      "step": 10007
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 17.71950912475586,
+      "learning_rate": 1.1626992427728738e-05,
+      "loss": 2.7385,
+      "step": 10008
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 4.815904140472412,
+      "learning_rate": 1.1626155712672052e-05,
+      "loss": 0.3241,
+      "step": 10009
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.331512451171875,
+      "learning_rate": 1.1625318997615362e-05,
+      "loss": 0.8059,
+      "step": 10010
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 11.508204460144043,
+      "learning_rate": 1.1624482282558676e-05,
+      "loss": 1.0577,
+      "step": 10011
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 38.958702087402344,
+      "learning_rate": 1.162364556750199e-05,
+      "loss": 1.4877,
+      "step": 10012
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 12.809269905090332,
+      "learning_rate": 1.16228088524453e-05,
+      "loss": 0.8781,
+      "step": 10013
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 21.30847930908203,
+      "learning_rate": 1.1621972137388614e-05,
+      "loss": 0.6281,
+      "step": 10014
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 12.957433700561523,
+      "learning_rate": 1.1621135422331925e-05,
+      "loss": 0.8955,
+      "step": 10015
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 13.342180252075195,
+      "learning_rate": 1.1620298707275239e-05,
+      "loss": 1.9799,
+      "step": 10016
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 27.924131393432617,
+      "learning_rate": 1.1619461992218551e-05,
+      "loss": 3.0026,
+      "step": 10017
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 20.1268253326416,
+      "learning_rate": 1.1618625277161863e-05,
+      "loss": 0.5763,
+      "step": 10018
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 6.630133152008057,
+      "learning_rate": 1.1617788562105177e-05,
+      "loss": 0.4575,
+      "step": 10019
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 19.63210678100586,
+      "learning_rate": 1.1616951847048487e-05,
+      "loss": 1.0933,
+      "step": 10020
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 14.773037910461426,
+      "learning_rate": 1.16161151319918e-05,
+      "loss": 1.8225,
+      "step": 10021
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 30.617250442504883,
+      "learning_rate": 1.1615278416935114e-05,
+      "loss": 1.1239,
+      "step": 10022
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 10.700183868408203,
+      "learning_rate": 1.1614441701878428e-05,
+      "loss": 0.7783,
+      "step": 10023
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 14.855048179626465,
+      "learning_rate": 1.1613604986821738e-05,
+      "loss": 2.5965,
+      "step": 10024
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 15.018572807312012,
+      "learning_rate": 1.1612768271765052e-05,
+      "loss": 1.3255,
+      "step": 10025
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 6.425628662109375,
+      "learning_rate": 1.1611931556708364e-05,
+      "loss": 0.6021,
+      "step": 10026
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 104.21813201904297,
+      "learning_rate": 1.1611094841651676e-05,
+      "loss": 1.1573,
+      "step": 10027
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.305505752563477,
+      "learning_rate": 1.161025812659499e-05,
+      "loss": 1.0939,
+      "step": 10028
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 11.391011238098145,
+      "learning_rate": 1.1609421411538301e-05,
+      "loss": 1.2346,
+      "step": 10029
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 39.1953125,
+      "learning_rate": 1.1608584696481615e-05,
+      "loss": 3.9511,
+      "step": 10030
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 28.51256561279297,
+      "learning_rate": 1.1607747981424927e-05,
+      "loss": 1.0599,
+      "step": 10031
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.189704895019531,
+      "learning_rate": 1.1606911266368239e-05,
+      "loss": 1.0351,
+      "step": 10032
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 12.981938362121582,
+      "learning_rate": 1.1606074551311553e-05,
+      "loss": 1.1728,
+      "step": 10033
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 56.41666030883789,
+      "learning_rate": 1.1605237836254863e-05,
+      "loss": 0.7182,
+      "step": 10034
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 14.686779975891113,
+      "learning_rate": 1.1604401121198176e-05,
+      "loss": 2.1056,
+      "step": 10035
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 18.131771087646484,
+      "learning_rate": 1.160356440614149e-05,
+      "loss": 2.0395,
+      "step": 10036
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.616911888122559,
+      "learning_rate": 1.1602727691084804e-05,
+      "loss": 1.0235,
+      "step": 10037
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 23.04092025756836,
+      "learning_rate": 1.1601890976028114e-05,
+      "loss": 1.743,
+      "step": 10038
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 48.13896560668945,
+      "learning_rate": 1.1601054260971428e-05,
+      "loss": 2.179,
+      "step": 10039
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 19.76043701171875,
+      "learning_rate": 1.160021754591474e-05,
+      "loss": 1.943,
+      "step": 10040
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 10.525476455688477,
+      "learning_rate": 1.1599380830858052e-05,
+      "loss": 3.165,
+      "step": 10041
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 16.542749404907227,
+      "learning_rate": 1.1598544115801365e-05,
+      "loss": 2.4273,
+      "step": 10042
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.2783255577087402,
+      "learning_rate": 1.1597707400744677e-05,
+      "loss": 0.0453,
+      "step": 10043
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 7.882564544677734,
+      "learning_rate": 1.159687068568799e-05,
+      "loss": 0.4292,
+      "step": 10044
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 30.36086082458496,
+      "learning_rate": 1.1596033970631301e-05,
+      "loss": 0.7863,
+      "step": 10045
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 16.81833839416504,
+      "learning_rate": 1.1595197255574615e-05,
+      "loss": 2.1633,
+      "step": 10046
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 16.21168327331543,
+      "learning_rate": 1.1594360540517928e-05,
+      "loss": 0.6202,
+      "step": 10047
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 12.192089080810547,
+      "learning_rate": 1.1593523825461239e-05,
+      "loss": 0.6212,
+      "step": 10048
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 15.387441635131836,
+      "learning_rate": 1.1592687110404552e-05,
+      "loss": 0.7201,
+      "step": 10049
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 15.565264701843262,
+      "learning_rate": 1.1591850395347866e-05,
+      "loss": 0.8358,
+      "step": 10050
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 12.644769668579102,
+      "learning_rate": 1.159101368029118e-05,
+      "loss": 2.3882,
+      "step": 10051
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 10.12132453918457,
+      "learning_rate": 1.159017696523449e-05,
+      "loss": 1.5997,
+      "step": 10052
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 133.52984619140625,
+      "learning_rate": 1.1589340250177803e-05,
+      "loss": 2.4285,
+      "step": 10053
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 13.751230239868164,
+      "learning_rate": 1.1588503535121115e-05,
+      "loss": 0.8718,
+      "step": 10054
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 10.26984977722168,
+      "learning_rate": 1.1587666820064427e-05,
+      "loss": 1.512,
+      "step": 10055
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 6.855957984924316,
+      "learning_rate": 1.1586830105007741e-05,
+      "loss": 0.2865,
+      "step": 10056
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 14.386962890625,
+      "learning_rate": 1.1585993389951053e-05,
+      "loss": 0.7904,
+      "step": 10057
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 11.048480033874512,
+      "learning_rate": 1.1585156674894367e-05,
+      "loss": 1.1862,
+      "step": 10058
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.805039882659912,
+      "learning_rate": 1.1584319959837677e-05,
+      "loss": 1.1205,
+      "step": 10059
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 52.64353561401367,
+      "learning_rate": 1.158348324478099e-05,
+      "loss": 1.2169,
+      "step": 10060
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.508051872253418,
+      "learning_rate": 1.1582646529724304e-05,
+      "loss": 0.441,
+      "step": 10061
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 10.83204460144043,
+      "learning_rate": 1.1581809814667614e-05,
+      "loss": 0.7029,
+      "step": 10062
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.536264419555664,
+      "learning_rate": 1.1580973099610928e-05,
+      "loss": 0.5355,
+      "step": 10063
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 18.436201095581055,
+      "learning_rate": 1.1580136384554242e-05,
+      "loss": 2.426,
+      "step": 10064
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 4.101004123687744,
+      "learning_rate": 1.1579299669497555e-05,
+      "loss": 0.3596,
+      "step": 10065
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 61.37518310546875,
+      "learning_rate": 1.1578462954440866e-05,
+      "loss": 1.6832,
+      "step": 10066
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 9.92491340637207,
+      "learning_rate": 1.157762623938418e-05,
+      "loss": 2.079,
+      "step": 10067
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 35.844871520996094,
+      "learning_rate": 1.1576789524327491e-05,
+      "loss": 2.0752,
+      "step": 10068
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 4.12769079208374,
+      "learning_rate": 1.1575952809270803e-05,
+      "loss": 0.2739,
+      "step": 10069
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 11.006604194641113,
+      "learning_rate": 1.1575116094214117e-05,
+      "loss": 1.53,
+      "step": 10070
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 13.370402336120605,
+      "learning_rate": 1.1574279379157429e-05,
+      "loss": 1.1828,
+      "step": 10071
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 20.020004272460938,
+      "learning_rate": 1.1573442664100742e-05,
+      "loss": 1.3519,
+      "step": 10072
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 20.790355682373047,
+      "learning_rate": 1.1572605949044053e-05,
+      "loss": 1.2607,
+      "step": 10073
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 13.527242660522461,
+      "learning_rate": 1.1571769233987366e-05,
+      "loss": 0.8212,
+      "step": 10074
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 11.991751670837402,
+      "learning_rate": 1.157093251893068e-05,
+      "loss": 0.7136,
+      "step": 10075
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.771205186843872,
+      "learning_rate": 1.157009580387399e-05,
+      "loss": 0.0826,
+      "step": 10076
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 7.323941707611084,
+      "learning_rate": 1.1569259088817304e-05,
+      "loss": 0.6736,
+      "step": 10077
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 25.943374633789062,
+      "learning_rate": 1.1568422373760618e-05,
+      "loss": 3.555,
+      "step": 10078
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 17.632131576538086,
+      "learning_rate": 1.156758565870393e-05,
+      "loss": 2.1017,
+      "step": 10079
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 27.554595947265625,
+      "learning_rate": 1.1566748943647242e-05,
+      "loss": 2.1187,
+      "step": 10080
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 11.076910018920898,
+      "learning_rate": 1.1565912228590555e-05,
+      "loss": 2.0688,
+      "step": 10081
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 27.039587020874023,
+      "learning_rate": 1.1565075513533867e-05,
+      "loss": 1.2906,
+      "step": 10082
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 57.26781463623047,
+      "learning_rate": 1.1564238798477179e-05,
+      "loss": 0.9118,
+      "step": 10083
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.590510368347168,
+      "learning_rate": 1.1563402083420491e-05,
+      "loss": 0.9175,
+      "step": 10084
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 7.459293842315674,
+      "learning_rate": 1.1562565368363805e-05,
+      "loss": 0.862,
+      "step": 10085
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 60.84637451171875,
+      "learning_rate": 1.1561728653307118e-05,
+      "loss": 2.0098,
+      "step": 10086
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 5.666141986846924,
+      "learning_rate": 1.1560891938250429e-05,
+      "loss": 0.5453,
+      "step": 10087
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.006087303161621,
+      "learning_rate": 1.1560055223193742e-05,
+      "loss": 2.3826,
+      "step": 10088
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 13.376801490783691,
+      "learning_rate": 1.1559218508137056e-05,
+      "loss": 2.2379,
+      "step": 10089
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 18.88728141784668,
+      "learning_rate": 1.1558381793080366e-05,
+      "loss": 3.1723,
+      "step": 10090
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 15.521059036254883,
+      "learning_rate": 1.155754507802368e-05,
+      "loss": 2.7883,
+      "step": 10091
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 24.361106872558594,
+      "learning_rate": 1.1556708362966993e-05,
+      "loss": 1.4142,
+      "step": 10092
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 5.3813347816467285,
+      "learning_rate": 1.1555871647910305e-05,
+      "loss": 0.3879,
+      "step": 10093
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.4180572032928467,
+      "learning_rate": 1.1555034932853617e-05,
+      "loss": 0.1363,
+      "step": 10094
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 18.977209091186523,
+      "learning_rate": 1.1554198217796931e-05,
+      "loss": 1.1672,
+      "step": 10095
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.979006767272949,
+      "learning_rate": 1.1553361502740243e-05,
+      "loss": 0.4125,
+      "step": 10096
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 7.230010986328125,
+      "learning_rate": 1.1552524787683555e-05,
+      "loss": 0.8564,
+      "step": 10097
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 31.010128021240234,
+      "learning_rate": 1.1551688072626867e-05,
+      "loss": 1.5528,
+      "step": 10098
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 13.065135955810547,
+      "learning_rate": 1.155085135757018e-05,
+      "loss": 0.4326,
+      "step": 10099
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 37.013404846191406,
+      "learning_rate": 1.1550014642513494e-05,
+      "loss": 1.5276,
+      "step": 10100
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 77.3665542602539,
+      "learning_rate": 1.1549177927456804e-05,
+      "loss": 2.7431,
+      "step": 10101
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 18.20163345336914,
+      "learning_rate": 1.1548341212400118e-05,
+      "loss": 1.9462,
+      "step": 10102
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 11.497478485107422,
+      "learning_rate": 1.1547504497343432e-05,
+      "loss": 1.6377,
+      "step": 10103
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 5.81686544418335,
+      "learning_rate": 1.1546667782286742e-05,
+      "loss": 0.2629,
+      "step": 10104
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.53093147277832,
+      "learning_rate": 1.1545831067230056e-05,
+      "loss": 0.8507,
+      "step": 10105
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 25.47751235961914,
+      "learning_rate": 1.154499435217337e-05,
+      "loss": 1.11,
+      "step": 10106
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 23.080636978149414,
+      "learning_rate": 1.1544157637116681e-05,
+      "loss": 1.3284,
+      "step": 10107
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.221174240112305,
+      "learning_rate": 1.1543320922059993e-05,
+      "loss": 0.8157,
+      "step": 10108
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 26.856416702270508,
+      "learning_rate": 1.1542484207003307e-05,
+      "loss": 2.0644,
+      "step": 10109
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.893000602722168,
+      "learning_rate": 1.1541647491946619e-05,
+      "loss": 1.6379,
+      "step": 10110
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.401641845703125,
+      "learning_rate": 1.154081077688993e-05,
+      "loss": 2.956,
+      "step": 10111
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 14.263781547546387,
+      "learning_rate": 1.1539974061833243e-05,
+      "loss": 1.6514,
+      "step": 10112
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 19.887052536010742,
+      "learning_rate": 1.1539137346776556e-05,
+      "loss": 0.8531,
+      "step": 10113
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.784709930419922,
+      "learning_rate": 1.153830063171987e-05,
+      "loss": 1.4221,
+      "step": 10114
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 6.730569839477539,
+      "learning_rate": 1.153746391666318e-05,
+      "loss": 0.5405,
+      "step": 10115
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 12.676787376403809,
+      "learning_rate": 1.1536627201606494e-05,
+      "loss": 0.8175,
+      "step": 10116
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.611346244812012,
+      "learning_rate": 1.1535790486549808e-05,
+      "loss": 0.5565,
+      "step": 10117
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 12.043750762939453,
+      "learning_rate": 1.1534953771493118e-05,
+      "loss": 0.3946,
+      "step": 10118
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 13.019969940185547,
+      "learning_rate": 1.1534117056436431e-05,
+      "loss": 1.1617,
+      "step": 10119
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 13.252829551696777,
+      "learning_rate": 1.1533280341379745e-05,
+      "loss": 1.5584,
+      "step": 10120
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 16.159860610961914,
+      "learning_rate": 1.1532443626323057e-05,
+      "loss": 1.3501,
+      "step": 10121
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 21.49095344543457,
+      "learning_rate": 1.1531606911266369e-05,
+      "loss": 0.746,
+      "step": 10122
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 12.386560440063477,
+      "learning_rate": 1.1530770196209683e-05,
+      "loss": 0.5928,
+      "step": 10123
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 19.16834831237793,
+      "learning_rate": 1.1529933481152995e-05,
+      "loss": 1.1127,
+      "step": 10124
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 7.827854633331299,
+      "learning_rate": 1.1529096766096307e-05,
+      "loss": 1.0985,
+      "step": 10125
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 7.0746541023254395,
+      "learning_rate": 1.1528260051039619e-05,
+      "loss": 0.8428,
+      "step": 10126
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 74.46416473388672,
+      "learning_rate": 1.1527423335982932e-05,
+      "loss": 3.9228,
+      "step": 10127
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 44.8482666015625,
+      "learning_rate": 1.1526586620926246e-05,
+      "loss": 1.3062,
+      "step": 10128
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 24.330764770507812,
+      "learning_rate": 1.1525749905869556e-05,
+      "loss": 1.4034,
+      "step": 10129
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 17.377822875976562,
+      "learning_rate": 1.152491319081287e-05,
+      "loss": 2.4979,
+      "step": 10130
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 6.2692646980285645,
+      "learning_rate": 1.1524076475756183e-05,
+      "loss": 1.3149,
+      "step": 10131
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 15.297842979431152,
+      "learning_rate": 1.1523239760699494e-05,
+      "loss": 1.3075,
+      "step": 10132
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 12.967483520507812,
+      "learning_rate": 1.1522403045642807e-05,
+      "loss": 0.7149,
+      "step": 10133
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 20.41457176208496,
+      "learning_rate": 1.1521566330586121e-05,
+      "loss": 1.4163,
+      "step": 10134
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.600605964660645,
+      "learning_rate": 1.1520729615529433e-05,
+      "loss": 1.5241,
+      "step": 10135
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.931024551391602,
+      "learning_rate": 1.1519892900472745e-05,
+      "loss": 0.6916,
+      "step": 10136
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 18.53843116760254,
+      "learning_rate": 1.1519056185416057e-05,
+      "loss": 0.78,
+      "step": 10137
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.421236991882324,
+      "learning_rate": 1.151821947035937e-05,
+      "loss": 0.7191,
+      "step": 10138
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 6.311631679534912,
+      "learning_rate": 1.1517382755302682e-05,
+      "loss": 0.4033,
+      "step": 10139
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.313081741333008,
+      "learning_rate": 1.1516546040245994e-05,
+      "loss": 0.8096,
+      "step": 10140
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 22.80303955078125,
+      "learning_rate": 1.1515709325189308e-05,
+      "loss": 1.9656,
+      "step": 10141
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 14.31672191619873,
+      "learning_rate": 1.1514872610132622e-05,
+      "loss": 0.536,
+      "step": 10142
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.868819236755371,
+      "learning_rate": 1.1514035895075932e-05,
+      "loss": 0.5144,
+      "step": 10143
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 14.120258331298828,
+      "learning_rate": 1.1513199180019246e-05,
+      "loss": 1.4992,
+      "step": 10144
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.662248611450195,
+      "learning_rate": 1.151236246496256e-05,
+      "loss": 0.5324,
+      "step": 10145
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 13.44205093383789,
+      "learning_rate": 1.151152574990587e-05,
+      "loss": 0.8438,
+      "step": 10146
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 6.933963775634766,
+      "learning_rate": 1.1510689034849183e-05,
+      "loss": 1.4326,
+      "step": 10147
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 16.528087615966797,
+      "learning_rate": 1.1509852319792497e-05,
+      "loss": 3.5467,
+      "step": 10148
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 12.353951454162598,
+      "learning_rate": 1.1509015604735809e-05,
+      "loss": 0.9497,
+      "step": 10149
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 19.795032501220703,
+      "learning_rate": 1.150817888967912e-05,
+      "loss": 2.5206,
+      "step": 10150
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 7.626039505004883,
+      "learning_rate": 1.1507342174622433e-05,
+      "loss": 0.5983,
+      "step": 10151
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 16.426559448242188,
+      "learning_rate": 1.1506505459565746e-05,
+      "loss": 1.5464,
+      "step": 10152
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.337430000305176,
+      "learning_rate": 1.1505668744509058e-05,
+      "loss": 0.5066,
+      "step": 10153
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 7.870647430419922,
+      "learning_rate": 1.150483202945237e-05,
+      "loss": 0.7625,
+      "step": 10154
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 13.128697395324707,
+      "learning_rate": 1.1503995314395684e-05,
+      "loss": 2.1313,
+      "step": 10155
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 29.010080337524414,
+      "learning_rate": 1.1503158599338998e-05,
+      "loss": 0.5929,
+      "step": 10156
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 26.012073516845703,
+      "learning_rate": 1.1502321884282308e-05,
+      "loss": 1.9542,
+      "step": 10157
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 11.357194900512695,
+      "learning_rate": 1.1501485169225621e-05,
+      "loss": 0.6633,
+      "step": 10158
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 9.94775390625,
+      "learning_rate": 1.1500648454168935e-05,
+      "loss": 1.4055,
+      "step": 10159
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 15.528973579406738,
+      "learning_rate": 1.1499811739112245e-05,
+      "loss": 2.0949,
+      "step": 10160
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 19.442302703857422,
+      "learning_rate": 1.1498975024055559e-05,
+      "loss": 1.303,
+      "step": 10161
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.6374619007110596,
+      "learning_rate": 1.1498138308998873e-05,
+      "loss": 0.3394,
+      "step": 10162
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 39.10694885253906,
+      "learning_rate": 1.1497301593942185e-05,
+      "loss": 3.5757,
+      "step": 10163
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 5.301551342010498,
+      "learning_rate": 1.1496464878885497e-05,
+      "loss": 0.4299,
+      "step": 10164
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 14.305914878845215,
+      "learning_rate": 1.1495628163828808e-05,
+      "loss": 2.2254,
+      "step": 10165
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 10.280922889709473,
+      "learning_rate": 1.1494791448772122e-05,
+      "loss": 0.5056,
+      "step": 10166
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 19.907215118408203,
+      "learning_rate": 1.1493954733715434e-05,
+      "loss": 0.862,
+      "step": 10167
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 13.114398956298828,
+      "learning_rate": 1.1493118018658746e-05,
+      "loss": 1.274,
+      "step": 10168
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.666918754577637,
+      "learning_rate": 1.149228130360206e-05,
+      "loss": 1.2034,
+      "step": 10169
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 16.747922897338867,
+      "learning_rate": 1.149144458854537e-05,
+      "loss": 1.3977,
+      "step": 10170
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.491082191467285,
+      "learning_rate": 1.1490607873488684e-05,
+      "loss": 1.5995,
+      "step": 10171
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 109.70430755615234,
+      "learning_rate": 1.1489771158431997e-05,
+      "loss": 2.0431,
+      "step": 10172
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 19.952125549316406,
+      "learning_rate": 1.1488934443375311e-05,
+      "loss": 2.6242,
+      "step": 10173
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 14.38827896118164,
+      "learning_rate": 1.1488097728318621e-05,
+      "loss": 1.0839,
+      "step": 10174
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.935357093811035,
+      "learning_rate": 1.1487261013261935e-05,
+      "loss": 0.4499,
+      "step": 10175
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.330789566040039,
+      "learning_rate": 1.1486424298205248e-05,
+      "loss": 1.8868,
+      "step": 10176
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 5.454981327056885,
+      "learning_rate": 1.1485587583148559e-05,
+      "loss": 0.5025,
+      "step": 10177
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.598280906677246,
+      "learning_rate": 1.1484750868091872e-05,
+      "loss": 0.4013,
+      "step": 10178
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.526079177856445,
+      "learning_rate": 1.1483914153035184e-05,
+      "loss": 0.933,
+      "step": 10179
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 15.199457168579102,
+      "learning_rate": 1.1483077437978498e-05,
+      "loss": 1.4231,
+      "step": 10180
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 78.98918151855469,
+      "learning_rate": 1.148224072292181e-05,
+      "loss": 2.0349,
+      "step": 10181
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 32.13898849487305,
+      "learning_rate": 1.1481404007865122e-05,
+      "loss": 1.6432,
+      "step": 10182
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 38.61654281616211,
+      "learning_rate": 1.1480567292808436e-05,
+      "loss": 2.6025,
+      "step": 10183
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.5785717964172363,
+      "learning_rate": 1.1479730577751746e-05,
+      "loss": 0.1609,
+      "step": 10184
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 4.924627304077148,
+      "learning_rate": 1.147889386269506e-05,
+      "loss": 0.2792,
+      "step": 10185
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 61.50282287597656,
+      "learning_rate": 1.1478057147638373e-05,
+      "loss": 2.2775,
+      "step": 10186
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 17.36076545715332,
+      "learning_rate": 1.1477220432581687e-05,
+      "loss": 0.7657,
+      "step": 10187
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.714561462402344,
+      "learning_rate": 1.1476383717524997e-05,
+      "loss": 1.351,
+      "step": 10188
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 16.25249481201172,
+      "learning_rate": 1.147554700246831e-05,
+      "loss": 1.2711,
+      "step": 10189
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 22.99129867553711,
+      "learning_rate": 1.1474710287411623e-05,
+      "loss": 1.2257,
+      "step": 10190
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 50.55567169189453,
+      "learning_rate": 1.1473873572354935e-05,
+      "loss": 3.1723,
+      "step": 10191
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 29.830339431762695,
+      "learning_rate": 1.1473036857298248e-05,
+      "loss": 2.6605,
+      "step": 10192
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 11.058066368103027,
+      "learning_rate": 1.147220014224156e-05,
+      "loss": 0.4693,
+      "step": 10193
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 27.12210464477539,
+      "learning_rate": 1.1471363427184874e-05,
+      "loss": 1.9006,
+      "step": 10194
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 15.129981994628906,
+      "learning_rate": 1.1470526712128184e-05,
+      "loss": 0.7715,
+      "step": 10195
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 11.411121368408203,
+      "learning_rate": 1.1469689997071498e-05,
+      "loss": 1.2326,
+      "step": 10196
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 14.986672401428223,
+      "learning_rate": 1.1468853282014811e-05,
+      "loss": 1.0021,
+      "step": 10197
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 7.28682279586792,
+      "learning_rate": 1.1468016566958122e-05,
+      "loss": 1.6545,
+      "step": 10198
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.118386268615723,
+      "learning_rate": 1.1467179851901435e-05,
+      "loss": 0.605,
+      "step": 10199
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 11.546998977661133,
+      "learning_rate": 1.1466343136844749e-05,
+      "loss": 0.5888,
+      "step": 10200
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.022869110107422,
+      "learning_rate": 1.1465506421788063e-05,
+      "loss": 0.7267,
+      "step": 10201
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 11.095171928405762,
+      "learning_rate": 1.1464669706731373e-05,
+      "loss": 2.0642,
+      "step": 10202
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 10.303417205810547,
+      "learning_rate": 1.1463832991674686e-05,
+      "loss": 0.5233,
+      "step": 10203
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 11.605952262878418,
+      "learning_rate": 1.1462996276617998e-05,
+      "loss": 2.1517,
+      "step": 10204
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 15.570901870727539,
+      "learning_rate": 1.146215956156131e-05,
+      "loss": 0.9086,
+      "step": 10205
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 81.48786926269531,
+      "learning_rate": 1.1461322846504624e-05,
+      "loss": 2.5043,
+      "step": 10206
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 38.58003616333008,
+      "learning_rate": 1.1460486131447936e-05,
+      "loss": 0.7259,
+      "step": 10207
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 18.89350128173828,
+      "learning_rate": 1.145964941639125e-05,
+      "loss": 2.7088,
+      "step": 10208
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 11.42757511138916,
+      "learning_rate": 1.145881270133456e-05,
+      "loss": 1.7884,
+      "step": 10209
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 14.500580787658691,
+      "learning_rate": 1.1457975986277874e-05,
+      "loss": 1.8369,
+      "step": 10210
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 10.310991287231445,
+      "learning_rate": 1.1457139271221187e-05,
+      "loss": 0.2204,
+      "step": 10211
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 7.678720951080322,
+      "learning_rate": 1.1456302556164497e-05,
+      "loss": 0.7853,
+      "step": 10212
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 23.72821617126465,
+      "learning_rate": 1.1455465841107811e-05,
+      "loss": 1.7404,
+      "step": 10213
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 11.855911254882812,
+      "learning_rate": 1.1454629126051125e-05,
+      "loss": 1.0215,
+      "step": 10214
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 5.976694583892822,
+      "learning_rate": 1.1453792410994438e-05,
+      "loss": 0.2622,
+      "step": 10215
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 23.440793991088867,
+      "learning_rate": 1.1452955695937749e-05,
+      "loss": 1.4206,
+      "step": 10216
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 13.946830749511719,
+      "learning_rate": 1.1452118980881062e-05,
+      "loss": 2.2023,
+      "step": 10217
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 88.03887939453125,
+      "learning_rate": 1.1451282265824374e-05,
+      "loss": 3.502,
+      "step": 10218
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 19.027698516845703,
+      "learning_rate": 1.1450445550767686e-05,
+      "loss": 0.7178,
+      "step": 10219
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 56.99128723144531,
+      "learning_rate": 1.1449608835711e-05,
+      "loss": 0.9991,
+      "step": 10220
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 13.91539192199707,
+      "learning_rate": 1.1448772120654312e-05,
+      "loss": 1.4028,
+      "step": 10221
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 34.86525344848633,
+      "learning_rate": 1.1447935405597625e-05,
+      "loss": 1.5506,
+      "step": 10222
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 18.596843719482422,
+      "learning_rate": 1.1447098690540936e-05,
+      "loss": 1.7121,
+      "step": 10223
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.9318630695343018,
+      "learning_rate": 1.144626197548425e-05,
+      "loss": 0.2067,
+      "step": 10224
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 22.74734878540039,
+      "learning_rate": 1.1445425260427563e-05,
+      "loss": 1.2098,
+      "step": 10225
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 13.152312278747559,
+      "learning_rate": 1.1444588545370873e-05,
+      "loss": 0.6433,
+      "step": 10226
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 13.77457046508789,
+      "learning_rate": 1.1443751830314187e-05,
+      "loss": 1.9025,
+      "step": 10227
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 8.574159622192383,
+      "learning_rate": 1.14429151152575e-05,
+      "loss": 0.8724,
+      "step": 10228
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 21.462326049804688,
+      "learning_rate": 1.1442078400200813e-05,
+      "loss": 1.4886,
+      "step": 10229
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 19.787872314453125,
+      "learning_rate": 1.1441241685144125e-05,
+      "loss": 2.7115,
+      "step": 10230
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 14.572142601013184,
+      "learning_rate": 1.1440404970087438e-05,
+      "loss": 1.9277,
+      "step": 10231
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 67.16240692138672,
+      "learning_rate": 1.143956825503075e-05,
+      "loss": 1.009,
+      "step": 10232
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 15.596375465393066,
+      "learning_rate": 1.1438731539974062e-05,
+      "loss": 1.0441,
+      "step": 10233
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 5.920174598693848,
+      "learning_rate": 1.1437894824917376e-05,
+      "loss": 0.291,
+      "step": 10234
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 14.65095329284668,
+      "learning_rate": 1.1437058109860688e-05,
+      "loss": 0.7002,
+      "step": 10235
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 16.111230850219727,
+      "learning_rate": 1.1436221394804001e-05,
+      "loss": 1.2901,
+      "step": 10236
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 52.37575149536133,
+      "learning_rate": 1.1435384679747312e-05,
+      "loss": 0.932,
+      "step": 10237
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 16.191383361816406,
+      "learning_rate": 1.1434547964690625e-05,
+      "loss": 0.8914,
+      "step": 10238
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 39.67690658569336,
+      "learning_rate": 1.1433711249633939e-05,
+      "loss": 1.0724,
+      "step": 10239
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 17.87415885925293,
+      "learning_rate": 1.1432874534577249e-05,
+      "loss": 2.0775,
+      "step": 10240
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 17.937177658081055,
+      "learning_rate": 1.1432037819520563e-05,
+      "loss": 1.4651,
+      "step": 10241
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.352197647094727,
+      "learning_rate": 1.1431201104463876e-05,
+      "loss": 0.971,
+      "step": 10242
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 18.424087524414062,
+      "learning_rate": 1.1430364389407188e-05,
+      "loss": 0.9773,
+      "step": 10243
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 38.48517990112305,
+      "learning_rate": 1.14295276743505e-05,
+      "loss": 1.7147,
+      "step": 10244
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 43.285911560058594,
+      "learning_rate": 1.1428690959293814e-05,
+      "loss": 1.2354,
+      "step": 10245
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 20.401620864868164,
+      "learning_rate": 1.1427854244237126e-05,
+      "loss": 1.2132,
+      "step": 10246
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.969087600708008,
+      "learning_rate": 1.1427017529180438e-05,
+      "loss": 0.4129,
+      "step": 10247
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.487325668334961,
+      "learning_rate": 1.142618081412375e-05,
+      "loss": 0.5936,
+      "step": 10248
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 15.239189147949219,
+      "learning_rate": 1.1425344099067064e-05,
+      "loss": 1.5616,
+      "step": 10249
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 6.267992973327637,
+      "learning_rate": 1.1424507384010377e-05,
+      "loss": 0.5374,
+      "step": 10250
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 6.666161060333252,
+      "learning_rate": 1.1423670668953687e-05,
+      "loss": 0.2554,
+      "step": 10251
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.305466175079346,
+      "learning_rate": 1.1422833953897001e-05,
+      "loss": 0.8353,
+      "step": 10252
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 4.593288421630859,
+      "learning_rate": 1.1421997238840315e-05,
+      "loss": 0.2813,
+      "step": 10253
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 6.5083909034729,
+      "learning_rate": 1.1421160523783625e-05,
+      "loss": 0.8337,
+      "step": 10254
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.180171012878418,
+      "learning_rate": 1.1420323808726939e-05,
+      "loss": 0.6955,
+      "step": 10255
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 16.141521453857422,
+      "learning_rate": 1.1419487093670252e-05,
+      "loss": 2.3699,
+      "step": 10256
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 16.75731086730957,
+      "learning_rate": 1.1418650378613564e-05,
+      "loss": 1.3129,
+      "step": 10257
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 16.07633399963379,
+      "learning_rate": 1.1417813663556876e-05,
+      "loss": 1.6392,
+      "step": 10258
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 18.466657638549805,
+      "learning_rate": 1.141697694850019e-05,
+      "loss": 1.7056,
+      "step": 10259
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.80053997039795,
+      "learning_rate": 1.1416140233443502e-05,
+      "loss": 1.4135,
+      "step": 10260
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 28.12500762939453,
+      "learning_rate": 1.1415303518386814e-05,
+      "loss": 1.8611,
+      "step": 10261
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 12.086811065673828,
+      "learning_rate": 1.1414466803330126e-05,
+      "loss": 1.3992,
+      "step": 10262
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 30.865243911743164,
+      "learning_rate": 1.141363008827344e-05,
+      "loss": 1.8881,
+      "step": 10263
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 11.483710289001465,
+      "learning_rate": 1.1412793373216753e-05,
+      "loss": 1.1132,
+      "step": 10264
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 26.61813735961914,
+      "learning_rate": 1.1411956658160063e-05,
+      "loss": 1.5824,
+      "step": 10265
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 28.661653518676758,
+      "learning_rate": 1.1411119943103377e-05,
+      "loss": 2.1407,
+      "step": 10266
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.828034400939941,
+      "learning_rate": 1.141028322804669e-05,
+      "loss": 1.0855,
+      "step": 10267
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 15.128324508666992,
+      "learning_rate": 1.140944651299e-05,
+      "loss": 2.3863,
+      "step": 10268
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.595596313476562,
+      "learning_rate": 1.1408609797933314e-05,
+      "loss": 1.7249,
+      "step": 10269
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 16.92155647277832,
+      "learning_rate": 1.1407773082876628e-05,
+      "loss": 1.4368,
+      "step": 10270
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 6.1554646492004395,
+      "learning_rate": 1.140693636781994e-05,
+      "loss": 1.1061,
+      "step": 10271
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.05190372467041,
+      "learning_rate": 1.1406099652763252e-05,
+      "loss": 0.6833,
+      "step": 10272
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 32.9219856262207,
+      "learning_rate": 1.1405262937706566e-05,
+      "loss": 1.4377,
+      "step": 10273
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 16.694856643676758,
+      "learning_rate": 1.1404426222649878e-05,
+      "loss": 0.8126,
+      "step": 10274
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 34.169456481933594,
+      "learning_rate": 1.140358950759319e-05,
+      "loss": 1.5287,
+      "step": 10275
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 4.406486511230469,
+      "learning_rate": 1.1402752792536502e-05,
+      "loss": 0.2657,
+      "step": 10276
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 256.61065673828125,
+      "learning_rate": 1.1401916077479815e-05,
+      "loss": 0.7833,
+      "step": 10277
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 23.040742874145508,
+      "learning_rate": 1.1401079362423129e-05,
+      "loss": 1.6952,
+      "step": 10278
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 45.938045501708984,
+      "learning_rate": 1.1400242647366439e-05,
+      "loss": 1.9528,
+      "step": 10279
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 19.950336456298828,
+      "learning_rate": 1.1399405932309753e-05,
+      "loss": 1.8527,
+      "step": 10280
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 31.180498123168945,
+      "learning_rate": 1.1398569217253066e-05,
+      "loss": 3.2534,
+      "step": 10281
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 16.345396041870117,
+      "learning_rate": 1.1397732502196377e-05,
+      "loss": 1.582,
+      "step": 10282
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 11.217645645141602,
+      "learning_rate": 1.139689578713969e-05,
+      "loss": 0.863,
+      "step": 10283
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 22.57853126525879,
+      "learning_rate": 1.1396059072083004e-05,
+      "loss": 1.0994,
+      "step": 10284
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 16.58045768737793,
+      "learning_rate": 1.1395222357026316e-05,
+      "loss": 1.5042,
+      "step": 10285
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 13.019248962402344,
+      "learning_rate": 1.1394385641969628e-05,
+      "loss": 1.068,
+      "step": 10286
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 66.55805969238281,
+      "learning_rate": 1.1393548926912942e-05,
+      "loss": 2.7568,
+      "step": 10287
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 3.6205310821533203,
+      "learning_rate": 1.1392712211856253e-05,
+      "loss": 0.3069,
+      "step": 10288
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.0787482261657715,
+      "learning_rate": 1.1391875496799565e-05,
+      "loss": 0.6211,
+      "step": 10289
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 22.246379852294922,
+      "learning_rate": 1.1391038781742877e-05,
+      "loss": 1.4295,
+      "step": 10290
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 36.28958511352539,
+      "learning_rate": 1.1390202066686191e-05,
+      "loss": 2.3918,
+      "step": 10291
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.485498905181885,
+      "learning_rate": 1.1389365351629505e-05,
+      "loss": 0.6775,
+      "step": 10292
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 15.13989543914795,
+      "learning_rate": 1.1388528636572815e-05,
+      "loss": 0.5463,
+      "step": 10293
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.350567817687988,
+      "learning_rate": 1.1387691921516129e-05,
+      "loss": 1.4391,
+      "step": 10294
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 5.292941570281982,
+      "learning_rate": 1.1386855206459442e-05,
+      "loss": 0.5071,
+      "step": 10295
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 14.215531349182129,
+      "learning_rate": 1.1386018491402752e-05,
+      "loss": 1.0031,
+      "step": 10296
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 24.693675994873047,
+      "learning_rate": 1.1385181776346066e-05,
+      "loss": 2.4584,
+      "step": 10297
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 5.77023983001709,
+      "learning_rate": 1.138434506128938e-05,
+      "loss": 0.4245,
+      "step": 10298
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 18.520322799682617,
+      "learning_rate": 1.1383508346232692e-05,
+      "loss": 1.4537,
+      "step": 10299
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 35.155845642089844,
+      "learning_rate": 1.1382671631176004e-05,
+      "loss": 1.9235,
+      "step": 10300
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 6.172835350036621,
+      "learning_rate": 1.1381834916119316e-05,
+      "loss": 0.4536,
+      "step": 10301
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.883145332336426,
+      "learning_rate": 1.138099820106263e-05,
+      "loss": 1.2697,
+      "step": 10302
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 11.846651077270508,
+      "learning_rate": 1.1380161486005941e-05,
+      "loss": 0.9475,
+      "step": 10303
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 13.66583251953125,
+      "learning_rate": 1.1379324770949253e-05,
+      "loss": 2.1352,
+      "step": 10304
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 88.45223999023438,
+      "learning_rate": 1.1378488055892567e-05,
+      "loss": 3.0673,
+      "step": 10305
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.919356346130371,
+      "learning_rate": 1.137765134083588e-05,
+      "loss": 0.8511,
+      "step": 10306
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 34.995784759521484,
+      "learning_rate": 1.137681462577919e-05,
+      "loss": 1.4538,
+      "step": 10307
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 58.23386001586914,
+      "learning_rate": 1.1375977910722504e-05,
+      "loss": 2.2188,
+      "step": 10308
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 5.6721272468566895,
+      "learning_rate": 1.1375141195665818e-05,
+      "loss": 0.5302,
+      "step": 10309
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 6.369467258453369,
+      "learning_rate": 1.1374304480609128e-05,
+      "loss": 1.5864,
+      "step": 10310
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.525195121765137,
+      "learning_rate": 1.1373467765552442e-05,
+      "loss": 0.5,
+      "step": 10311
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.663617134094238,
+      "learning_rate": 1.1372631050495756e-05,
+      "loss": 1.0208,
+      "step": 10312
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.464286804199219,
+      "learning_rate": 1.1371794335439068e-05,
+      "loss": 1.6635,
+      "step": 10313
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 29.7788143157959,
+      "learning_rate": 1.137095762038238e-05,
+      "loss": 0.7489,
+      "step": 10314
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 27.088542938232422,
+      "learning_rate": 1.1370120905325691e-05,
+      "loss": 1.816,
+      "step": 10315
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 13.230093002319336,
+      "learning_rate": 1.1369284190269005e-05,
+      "loss": 1.5894,
+      "step": 10316
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 11.129487991333008,
+      "learning_rate": 1.1368447475212317e-05,
+      "loss": 0.9775,
+      "step": 10317
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 10.884594917297363,
+      "learning_rate": 1.1367610760155629e-05,
+      "loss": 1.1296,
+      "step": 10318
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 32.00385665893555,
+      "learning_rate": 1.1366774045098943e-05,
+      "loss": 2.2619,
+      "step": 10319
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 79.75541687011719,
+      "learning_rate": 1.1365937330042256e-05,
+      "loss": 1.1296,
+      "step": 10320
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 25.60192108154297,
+      "learning_rate": 1.1365100614985567e-05,
+      "loss": 2.0423,
+      "step": 10321
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.8787202835083,
+      "learning_rate": 1.136426389992888e-05,
+      "loss": 1.4763,
+      "step": 10322
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 13.253055572509766,
+      "learning_rate": 1.1363427184872194e-05,
+      "loss": 1.5555,
+      "step": 10323
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 16.878116607666016,
+      "learning_rate": 1.1362590469815504e-05,
+      "loss": 1.4943,
+      "step": 10324
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 19.132755279541016,
+      "learning_rate": 1.1361753754758818e-05,
+      "loss": 1.8916,
+      "step": 10325
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 7.025834083557129,
+      "learning_rate": 1.1360917039702131e-05,
+      "loss": 0.3252,
+      "step": 10326
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 27.166078567504883,
+      "learning_rate": 1.1360080324645443e-05,
+      "loss": 2.4286,
+      "step": 10327
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.396801948547363,
+      "learning_rate": 1.1359243609588755e-05,
+      "loss": 1.6724,
+      "step": 10328
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 7.133232116699219,
+      "learning_rate": 1.1358406894532067e-05,
+      "loss": 0.3781,
+      "step": 10329
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 18.406600952148438,
+      "learning_rate": 1.1357570179475381e-05,
+      "loss": 2.059,
+      "step": 10330
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 37.831607818603516,
+      "learning_rate": 1.1356733464418693e-05,
+      "loss": 1.8232,
+      "step": 10331
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 19.915050506591797,
+      "learning_rate": 1.1355896749362005e-05,
+      "loss": 1.4515,
+      "step": 10332
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 21.95366668701172,
+      "learning_rate": 1.1355060034305319e-05,
+      "loss": 1.7563,
+      "step": 10333
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 12.465929985046387,
+      "learning_rate": 1.1354223319248632e-05,
+      "loss": 0.2971,
+      "step": 10334
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 5.080347537994385,
+      "learning_rate": 1.1353386604191942e-05,
+      "loss": 0.3639,
+      "step": 10335
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.8196351528167725,
+      "learning_rate": 1.1352549889135256e-05,
+      "loss": 0.4015,
+      "step": 10336
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 7.730131149291992,
+      "learning_rate": 1.135171317407857e-05,
+      "loss": 0.5276,
+      "step": 10337
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 30.008684158325195,
+      "learning_rate": 1.135087645902188e-05,
+      "loss": 2.2609,
+      "step": 10338
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.635030746459961,
+      "learning_rate": 1.1350039743965194e-05,
+      "loss": 0.962,
+      "step": 10339
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 5.797532558441162,
+      "learning_rate": 1.1349203028908506e-05,
+      "loss": 0.7564,
+      "step": 10340
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.949183464050293,
+      "learning_rate": 1.134836631385182e-05,
+      "loss": 0.9836,
+      "step": 10341
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 4.199320316314697,
+      "learning_rate": 1.1347529598795131e-05,
+      "loss": 1.0451,
+      "step": 10342
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 38.582027435302734,
+      "learning_rate": 1.1346692883738443e-05,
+      "loss": 2.0152,
+      "step": 10343
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 19.874881744384766,
+      "learning_rate": 1.1345856168681757e-05,
+      "loss": 1.53,
+      "step": 10344
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.34642219543457,
+      "learning_rate": 1.1345019453625069e-05,
+      "loss": 1.1512,
+      "step": 10345
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 22.101593017578125,
+      "learning_rate": 1.134418273856838e-05,
+      "loss": 0.8094,
+      "step": 10346
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 13.938207626342773,
+      "learning_rate": 1.1343346023511694e-05,
+      "loss": 0.6246,
+      "step": 10347
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 10.098628044128418,
+      "learning_rate": 1.1342509308455008e-05,
+      "loss": 0.4555,
+      "step": 10348
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 16.11585807800293,
+      "learning_rate": 1.1341672593398318e-05,
+      "loss": 2.9272,
+      "step": 10349
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 15.630293846130371,
+      "learning_rate": 1.1340835878341632e-05,
+      "loss": 1.1305,
+      "step": 10350
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 17.844301223754883,
+      "learning_rate": 1.1339999163284946e-05,
+      "loss": 1.4772,
+      "step": 10351
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.894214630126953,
+      "learning_rate": 1.1339162448228256e-05,
+      "loss": 0.5381,
+      "step": 10352
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 12.17094898223877,
+      "learning_rate": 1.133832573317157e-05,
+      "loss": 0.7044,
+      "step": 10353
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 50.2935905456543,
+      "learning_rate": 1.1337489018114881e-05,
+      "loss": 3.1994,
+      "step": 10354
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 106.10897064208984,
+      "learning_rate": 1.1336652303058195e-05,
+      "loss": 1.2581,
+      "step": 10355
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 29.674652099609375,
+      "learning_rate": 1.1335815588001507e-05,
+      "loss": 2.473,
+      "step": 10356
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 16.675195693969727,
+      "learning_rate": 1.1334978872944819e-05,
+      "loss": 3.5494,
+      "step": 10357
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 14.187880516052246,
+      "learning_rate": 1.1334142157888133e-05,
+      "loss": 1.0302,
+      "step": 10358
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 16.538631439208984,
+      "learning_rate": 1.1333305442831443e-05,
+      "loss": 3.3621,
+      "step": 10359
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 10.103822708129883,
+      "learning_rate": 1.1332468727774757e-05,
+      "loss": 2.0432,
+      "step": 10360
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 74.78076171875,
+      "learning_rate": 1.133163201271807e-05,
+      "loss": 1.4483,
+      "step": 10361
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 19.359525680541992,
+      "learning_rate": 1.1330795297661384e-05,
+      "loss": 2.5388,
+      "step": 10362
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 12.176506996154785,
+      "learning_rate": 1.1329958582604694e-05,
+      "loss": 1.2399,
+      "step": 10363
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 16.648141860961914,
+      "learning_rate": 1.1329121867548008e-05,
+      "loss": 0.4769,
+      "step": 10364
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 4.545150279998779,
+      "learning_rate": 1.1328285152491321e-05,
+      "loss": 0.4344,
+      "step": 10365
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 7.920763969421387,
+      "learning_rate": 1.1327448437434632e-05,
+      "loss": 1.3722,
+      "step": 10366
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 8.198457717895508,
+      "learning_rate": 1.1326611722377945e-05,
+      "loss": 0.9668,
+      "step": 10367
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 5.3731913566589355,
+      "learning_rate": 1.1325775007321257e-05,
+      "loss": 0.2896,
+      "step": 10368
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.775003433227539,
+      "learning_rate": 1.1324938292264571e-05,
+      "loss": 2.7216,
+      "step": 10369
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.60084342956543,
+      "learning_rate": 1.1324101577207883e-05,
+      "loss": 1.4263,
+      "step": 10370
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 6.094202041625977,
+      "learning_rate": 1.1323264862151195e-05,
+      "loss": 0.6605,
+      "step": 10371
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 6.6341471672058105,
+      "learning_rate": 1.1322428147094508e-05,
+      "loss": 0.8613,
+      "step": 10372
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.536539077758789,
+      "learning_rate": 1.1321591432037819e-05,
+      "loss": 2.1967,
+      "step": 10373
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 17.777511596679688,
+      "learning_rate": 1.1320754716981132e-05,
+      "loss": 1.1906,
+      "step": 10374
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 18.944923400878906,
+      "learning_rate": 1.1319918001924446e-05,
+      "loss": 1.1539,
+      "step": 10375
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 10.789403915405273,
+      "learning_rate": 1.131908128686776e-05,
+      "loss": 2.2528,
+      "step": 10376
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.346477508544922,
+      "learning_rate": 1.131824457181107e-05,
+      "loss": 2.3022,
+      "step": 10377
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.01475715637207,
+      "learning_rate": 1.1317407856754384e-05,
+      "loss": 1.7589,
+      "step": 10378
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 12.307167053222656,
+      "learning_rate": 1.1316571141697697e-05,
+      "loss": 2.5049,
+      "step": 10379
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 6.579891681671143,
+      "learning_rate": 1.1315734426641008e-05,
+      "loss": 0.5182,
+      "step": 10380
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 16.720767974853516,
+      "learning_rate": 1.1314897711584321e-05,
+      "loss": 1.0419,
+      "step": 10381
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 28.597412109375,
+      "learning_rate": 1.1314060996527633e-05,
+      "loss": 1.2098,
+      "step": 10382
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 16.08039665222168,
+      "learning_rate": 1.1313224281470947e-05,
+      "loss": 2.1011,
+      "step": 10383
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.145201683044434,
+      "learning_rate": 1.1312387566414259e-05,
+      "loss": 0.54,
+      "step": 10384
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 10.521228790283203,
+      "learning_rate": 1.131155085135757e-05,
+      "loss": 0.9657,
+      "step": 10385
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 12.017659187316895,
+      "learning_rate": 1.1310714136300884e-05,
+      "loss": 0.163,
+      "step": 10386
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 16.294795989990234,
+      "learning_rate": 1.1309877421244195e-05,
+      "loss": 0.9599,
+      "step": 10387
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 9.123615264892578,
+      "learning_rate": 1.1309040706187508e-05,
+      "loss": 0.214,
+      "step": 10388
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 21.298059463500977,
+      "learning_rate": 1.1308203991130822e-05,
+      "loss": 1.9316,
+      "step": 10389
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 15.311811447143555,
+      "learning_rate": 1.1307367276074136e-05,
+      "loss": 1.2914,
+      "step": 10390
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 19.213844299316406,
+      "learning_rate": 1.1306530561017446e-05,
+      "loss": 1.6291,
+      "step": 10391
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 12.955446243286133,
+      "learning_rate": 1.130569384596076e-05,
+      "loss": 1.1187,
+      "step": 10392
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 12.431927680969238,
+      "learning_rate": 1.1304857130904071e-05,
+      "loss": 0.8056,
+      "step": 10393
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 18.048500061035156,
+      "learning_rate": 1.1304020415847383e-05,
+      "loss": 1.8015,
+      "step": 10394
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 52.673152923583984,
+      "learning_rate": 1.1303183700790697e-05,
+      "loss": 2.2139,
+      "step": 10395
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 28.4801082611084,
+      "learning_rate": 1.1302346985734009e-05,
+      "loss": 2.603,
+      "step": 10396
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 13.845949172973633,
+      "learning_rate": 1.1301510270677323e-05,
+      "loss": 0.7731,
+      "step": 10397
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 19.7300968170166,
+      "learning_rate": 1.1300673555620635e-05,
+      "loss": 2.4962,
+      "step": 10398
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 18.611753463745117,
+      "learning_rate": 1.1299836840563947e-05,
+      "loss": 2.0268,
+      "step": 10399
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 13.834794998168945,
+      "learning_rate": 1.129900012550726e-05,
+      "loss": 0.6129,
+      "step": 10400
+    },
+    {
+      "epoch": 1.31,
+      "eval_loss": 0.08567207306623459,
+      "eval_runtime": 97.5514,
+      "eval_samples_per_second": 36.309,
+      "eval_steps_per_second": 36.309,
+      "step": 10400
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.377630233764648,
+      "learning_rate": 1.129816341045057e-05,
+      "loss": 2.2945,
+      "step": 10401
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 14.499702453613281,
+      "learning_rate": 1.1297326695393884e-05,
+      "loss": 1.7374,
+      "step": 10402
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 6.5097575187683105,
+      "learning_rate": 1.1296489980337198e-05,
+      "loss": 0.7856,
+      "step": 10403
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 7.979915142059326,
+      "learning_rate": 1.1295653265280511e-05,
+      "loss": 1.0574,
+      "step": 10404
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.252052307128906,
+      "learning_rate": 1.1294816550223822e-05,
+      "loss": 1.1272,
+      "step": 10405
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.146780967712402,
+      "learning_rate": 1.1293979835167135e-05,
+      "loss": 1.1183,
+      "step": 10406
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 38.974308013916016,
+      "learning_rate": 1.1293143120110447e-05,
+      "loss": 3.3828,
+      "step": 10407
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 20.843727111816406,
+      "learning_rate": 1.129230640505376e-05,
+      "loss": 2.2591,
+      "step": 10408
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 36.327911376953125,
+      "learning_rate": 1.1291469689997073e-05,
+      "loss": 2.9167,
+      "step": 10409
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 5.336467742919922,
+      "learning_rate": 1.1290632974940385e-05,
+      "loss": 0.3331,
+      "step": 10410
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 17.777746200561523,
+      "learning_rate": 1.1289796259883698e-05,
+      "loss": 1.5454,
+      "step": 10411
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 6.859896659851074,
+      "learning_rate": 1.1288959544827009e-05,
+      "loss": 0.3444,
+      "step": 10412
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.759364128112793,
+      "learning_rate": 1.1288122829770322e-05,
+      "loss": 0.8933,
+      "step": 10413
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 7.504409313201904,
+      "learning_rate": 1.1287286114713636e-05,
+      "loss": 0.4092,
+      "step": 10414
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 20.213590621948242,
+      "learning_rate": 1.1286449399656946e-05,
+      "loss": 1.075,
+      "step": 10415
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 15.260299682617188,
+      "learning_rate": 1.128561268460026e-05,
+      "loss": 1.079,
+      "step": 10416
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 29.809040069580078,
+      "learning_rate": 1.1284775969543574e-05,
+      "loss": 1.168,
+      "step": 10417
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 15.28903865814209,
+      "learning_rate": 1.1283939254486887e-05,
+      "loss": 1.3027,
+      "step": 10418
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 6.6696858406066895,
+      "learning_rate": 1.1283102539430197e-05,
+      "loss": 1.1304,
+      "step": 10419
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.441073417663574,
+      "learning_rate": 1.1282265824373511e-05,
+      "loss": 1.3255,
+      "step": 10420
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 3.9966790676116943,
+      "learning_rate": 1.1281429109316823e-05,
+      "loss": 0.7552,
+      "step": 10421
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.323063850402832,
+      "learning_rate": 1.1280592394260135e-05,
+      "loss": 1.3449,
+      "step": 10422
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 5.556552886962891,
+      "learning_rate": 1.1279755679203449e-05,
+      "loss": 0.4441,
+      "step": 10423
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 7.900179386138916,
+      "learning_rate": 1.127891896414676e-05,
+      "loss": 0.2652,
+      "step": 10424
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 21.98341941833496,
+      "learning_rate": 1.1278082249090074e-05,
+      "loss": 2.6199,
+      "step": 10425
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 38.724266052246094,
+      "learning_rate": 1.1277245534033385e-05,
+      "loss": 1.2014,
+      "step": 10426
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 4.7703938484191895,
+      "learning_rate": 1.1276408818976698e-05,
+      "loss": 0.4701,
+      "step": 10427
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.077507972717285,
+      "learning_rate": 1.1275572103920012e-05,
+      "loss": 1.7238,
+      "step": 10428
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 15.9498929977417,
+      "learning_rate": 1.1274735388863322e-05,
+      "loss": 2.0018,
+      "step": 10429
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 6.577724933624268,
+      "learning_rate": 1.1273898673806636e-05,
+      "loss": 0.5287,
+      "step": 10430
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 14.119277000427246,
+      "learning_rate": 1.127306195874995e-05,
+      "loss": 1.7477,
+      "step": 10431
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 78.08936309814453,
+      "learning_rate": 1.1272225243693263e-05,
+      "loss": 1.2009,
+      "step": 10432
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 5.330587387084961,
+      "learning_rate": 1.1271388528636573e-05,
+      "loss": 0.5086,
+      "step": 10433
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.60766887664795,
+      "learning_rate": 1.1270551813579887e-05,
+      "loss": 0.2803,
+      "step": 10434
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 22.309850692749023,
+      "learning_rate": 1.1269715098523199e-05,
+      "loss": 1.0715,
+      "step": 10435
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.775199890136719,
+      "learning_rate": 1.1268878383466511e-05,
+      "loss": 0.8933,
+      "step": 10436
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 22.4728946685791,
+      "learning_rate": 1.1268041668409825e-05,
+      "loss": 2.7902,
+      "step": 10437
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.569796562194824,
+      "learning_rate": 1.1267204953353136e-05,
+      "loss": 1.4144,
+      "step": 10438
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 20.00957489013672,
+      "learning_rate": 1.126636823829645e-05,
+      "loss": 1.4074,
+      "step": 10439
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.141829490661621,
+      "learning_rate": 1.126553152323976e-05,
+      "loss": 0.4434,
+      "step": 10440
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 18.492023468017578,
+      "learning_rate": 1.1264694808183074e-05,
+      "loss": 1.4251,
+      "step": 10441
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 44.78361511230469,
+      "learning_rate": 1.1263858093126388e-05,
+      "loss": 3.159,
+      "step": 10442
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 4.946948051452637,
+      "learning_rate": 1.1263021378069698e-05,
+      "loss": 0.1349,
+      "step": 10443
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 24.42609977722168,
+      "learning_rate": 1.1262184663013012e-05,
+      "loss": 2.367,
+      "step": 10444
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 12.066911697387695,
+      "learning_rate": 1.1261347947956325e-05,
+      "loss": 1.1732,
+      "step": 10445
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 21.528846740722656,
+      "learning_rate": 1.1260511232899637e-05,
+      "loss": 0.433,
+      "step": 10446
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 20.1120662689209,
+      "learning_rate": 1.1259674517842949e-05,
+      "loss": 0.6791,
+      "step": 10447
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 12.262086868286133,
+      "learning_rate": 1.1258837802786263e-05,
+      "loss": 1.2916,
+      "step": 10448
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 22.04985809326172,
+      "learning_rate": 1.1258001087729575e-05,
+      "loss": 1.5374,
+      "step": 10449
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 7.273778915405273,
+      "learning_rate": 1.1257164372672887e-05,
+      "loss": 2.0661,
+      "step": 10450
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 26.157926559448242,
+      "learning_rate": 1.12563276576162e-05,
+      "loss": 0.8122,
+      "step": 10451
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 3.5167133808135986,
+      "learning_rate": 1.1255490942559512e-05,
+      "loss": 0.2834,
+      "step": 10452
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 26.909486770629883,
+      "learning_rate": 1.1254654227502826e-05,
+      "loss": 2.0141,
+      "step": 10453
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 15.352017402648926,
+      "learning_rate": 1.1253817512446136e-05,
+      "loss": 2.5512,
+      "step": 10454
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 13.897164344787598,
+      "learning_rate": 1.125298079738945e-05,
+      "loss": 1.4139,
+      "step": 10455
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 49.78348159790039,
+      "learning_rate": 1.1252144082332764e-05,
+      "loss": 2.1975,
+      "step": 10456
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 38.473941802978516,
+      "learning_rate": 1.1251307367276074e-05,
+      "loss": 3.2748,
+      "step": 10457
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 11.407572746276855,
+      "learning_rate": 1.1250470652219387e-05,
+      "loss": 0.8564,
+      "step": 10458
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 35.14699172973633,
+      "learning_rate": 1.1249633937162701e-05,
+      "loss": 1.3608,
+      "step": 10459
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 24.109121322631836,
+      "learning_rate": 1.1248797222106013e-05,
+      "loss": 1.8741,
+      "step": 10460
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 47.487579345703125,
+      "learning_rate": 1.1247960507049325e-05,
+      "loss": 1.6083,
+      "step": 10461
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 7.928770542144775,
+      "learning_rate": 1.1247123791992639e-05,
+      "loss": 0.6158,
+      "step": 10462
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 21.42852783203125,
+      "learning_rate": 1.124628707693595e-05,
+      "loss": 0.7657,
+      "step": 10463
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 15.748943328857422,
+      "learning_rate": 1.1245450361879263e-05,
+      "loss": 1.8107,
+      "step": 10464
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 100.10545349121094,
+      "learning_rate": 1.1244613646822574e-05,
+      "loss": 2.828,
+      "step": 10465
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 17.006912231445312,
+      "learning_rate": 1.1243776931765888e-05,
+      "loss": 1.855,
+      "step": 10466
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 18.44497299194336,
+      "learning_rate": 1.1242940216709202e-05,
+      "loss": 2.5444,
+      "step": 10467
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 12.248494148254395,
+      "learning_rate": 1.1242103501652512e-05,
+      "loss": 1.3102,
+      "step": 10468
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 9.847447395324707,
+      "learning_rate": 1.1241266786595826e-05,
+      "loss": 0.9991,
+      "step": 10469
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 16.546358108520508,
+      "learning_rate": 1.124043007153914e-05,
+      "loss": 1.3043,
+      "step": 10470
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 15.240071296691895,
+      "learning_rate": 1.123959335648245e-05,
+      "loss": 2.4128,
+      "step": 10471
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.652076721191406,
+      "learning_rate": 1.1238756641425763e-05,
+      "loss": 0.7773,
+      "step": 10472
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 16.13754653930664,
+      "learning_rate": 1.1237919926369077e-05,
+      "loss": 1.7918,
+      "step": 10473
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 12.626401901245117,
+      "learning_rate": 1.1237083211312389e-05,
+      "loss": 1.8125,
+      "step": 10474
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 14.035736083984375,
+      "learning_rate": 1.12362464962557e-05,
+      "loss": 0.7317,
+      "step": 10475
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 21.7222957611084,
+      "learning_rate": 1.1235409781199014e-05,
+      "loss": 2.3465,
+      "step": 10476
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.77615737915039,
+      "learning_rate": 1.1234573066142326e-05,
+      "loss": 0.7189,
+      "step": 10477
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 24.431882858276367,
+      "learning_rate": 1.1233736351085638e-05,
+      "loss": 1.5463,
+      "step": 10478
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 19.643756866455078,
+      "learning_rate": 1.123289963602895e-05,
+      "loss": 1.9038,
+      "step": 10479
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 10.335158348083496,
+      "learning_rate": 1.1232062920972264e-05,
+      "loss": 0.5165,
+      "step": 10480
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 31.67599105834961,
+      "learning_rate": 1.1231226205915578e-05,
+      "loss": 1.2921,
+      "step": 10481
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 20.917774200439453,
+      "learning_rate": 1.1230389490858888e-05,
+      "loss": 2.4767,
+      "step": 10482
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.009300231933594,
+      "learning_rate": 1.1229552775802202e-05,
+      "loss": 0.771,
+      "step": 10483
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 6.347224235534668,
+      "learning_rate": 1.1228716060745515e-05,
+      "loss": 1.6127,
+      "step": 10484
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 31.478605270385742,
+      "learning_rate": 1.1227879345688825e-05,
+      "loss": 1.0239,
+      "step": 10485
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.87148380279541,
+      "learning_rate": 1.1227042630632139e-05,
+      "loss": 0.6186,
+      "step": 10486
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 135.9146270751953,
+      "learning_rate": 1.1226205915575453e-05,
+      "loss": 0.802,
+      "step": 10487
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 20.080284118652344,
+      "learning_rate": 1.1225369200518765e-05,
+      "loss": 1.8482,
+      "step": 10488
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 47.902992248535156,
+      "learning_rate": 1.1224532485462077e-05,
+      "loss": 3.7796,
+      "step": 10489
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 24.935144424438477,
+      "learning_rate": 1.122369577040539e-05,
+      "loss": 2.8729,
+      "step": 10490
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 19.10072898864746,
+      "learning_rate": 1.1222859055348702e-05,
+      "loss": 1.1724,
+      "step": 10491
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 20.33785057067871,
+      "learning_rate": 1.1222022340292014e-05,
+      "loss": 1.2364,
+      "step": 10492
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 7.977625370025635,
+      "learning_rate": 1.1221185625235326e-05,
+      "loss": 1.8418,
+      "step": 10493
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.695846557617188,
+      "learning_rate": 1.122034891017864e-05,
+      "loss": 1.9317,
+      "step": 10494
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.307378768920898,
+      "learning_rate": 1.1219512195121953e-05,
+      "loss": 1.586,
+      "step": 10495
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 17.317626953125,
+      "learning_rate": 1.1218675480065264e-05,
+      "loss": 1.4528,
+      "step": 10496
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 18.528837203979492,
+      "learning_rate": 1.1217838765008577e-05,
+      "loss": 1.5133,
+      "step": 10497
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 19.242237091064453,
+      "learning_rate": 1.1217002049951891e-05,
+      "loss": 1.4202,
+      "step": 10498
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 7.817399501800537,
+      "learning_rate": 1.1216165334895201e-05,
+      "loss": 0.3527,
+      "step": 10499
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.068084716796875,
+      "learning_rate": 1.1215328619838515e-05,
+      "loss": 0.566,
+      "step": 10500
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 21.281558990478516,
+      "learning_rate": 1.1214491904781829e-05,
+      "loss": 1.6802,
+      "step": 10501
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 12.254424095153809,
+      "learning_rate": 1.121365518972514e-05,
+      "loss": 0.5506,
+      "step": 10502
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 16.627653121948242,
+      "learning_rate": 1.1212818474668452e-05,
+      "loss": 1.5935,
+      "step": 10503
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.4081315994262695,
+      "learning_rate": 1.1211981759611764e-05,
+      "loss": 0.3894,
+      "step": 10504
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 14.867437362670898,
+      "learning_rate": 1.1211145044555078e-05,
+      "loss": 1.5122,
+      "step": 10505
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 46.8331184387207,
+      "learning_rate": 1.121030832949839e-05,
+      "loss": 1.1616,
+      "step": 10506
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.698062896728516,
+      "learning_rate": 1.1209471614441702e-05,
+      "loss": 0.7611,
+      "step": 10507
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.20580005645752,
+      "learning_rate": 1.1208634899385016e-05,
+      "loss": 0.2818,
+      "step": 10508
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.287290573120117,
+      "learning_rate": 1.120779818432833e-05,
+      "loss": 0.8224,
+      "step": 10509
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 16.682458877563477,
+      "learning_rate": 1.120696146927164e-05,
+      "loss": 1.1507,
+      "step": 10510
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 15.22974967956543,
+      "learning_rate": 1.1206124754214953e-05,
+      "loss": 0.662,
+      "step": 10511
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 10.623566627502441,
+      "learning_rate": 1.1205288039158267e-05,
+      "loss": 0.5222,
+      "step": 10512
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.88694953918457,
+      "learning_rate": 1.1204451324101577e-05,
+      "loss": 1.2391,
+      "step": 10513
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 27.594900131225586,
+      "learning_rate": 1.120361460904489e-05,
+      "loss": 1.774,
+      "step": 10514
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 15.125885009765625,
+      "learning_rate": 1.1202777893988204e-05,
+      "loss": 1.3167,
+      "step": 10515
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 27.385690689086914,
+      "learning_rate": 1.1201941178931516e-05,
+      "loss": 2.3734,
+      "step": 10516
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 15.248473167419434,
+      "learning_rate": 1.1201104463874828e-05,
+      "loss": 1.4631,
+      "step": 10517
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 13.878178596496582,
+      "learning_rate": 1.120026774881814e-05,
+      "loss": 1.1966,
+      "step": 10518
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 13.134208679199219,
+      "learning_rate": 1.1199431033761454e-05,
+      "loss": 2.1645,
+      "step": 10519
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 17.41299057006836,
+      "learning_rate": 1.1198594318704766e-05,
+      "loss": 1.3558,
+      "step": 10520
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 7.347055912017822,
+      "learning_rate": 1.1197757603648078e-05,
+      "loss": 0.6305,
+      "step": 10521
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 4.793478488922119,
+      "learning_rate": 1.1196920888591391e-05,
+      "loss": 0.4807,
+      "step": 10522
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 15.79783821105957,
+      "learning_rate": 1.1196084173534705e-05,
+      "loss": 1.2236,
+      "step": 10523
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9226435422897339,
+      "learning_rate": 1.1195247458478015e-05,
+      "loss": 0.0487,
+      "step": 10524
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 58.216766357421875,
+      "learning_rate": 1.1194410743421329e-05,
+      "loss": 2.5017,
+      "step": 10525
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 12.894635200500488,
+      "learning_rate": 1.1193574028364643e-05,
+      "loss": 2.6699,
+      "step": 10526
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 60.26153564453125,
+      "learning_rate": 1.1192737313307953e-05,
+      "loss": 1.6459,
+      "step": 10527
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.879268646240234,
+      "learning_rate": 1.1191900598251267e-05,
+      "loss": 1.5969,
+      "step": 10528
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.770275115966797,
+      "learning_rate": 1.119106388319458e-05,
+      "loss": 1.3921,
+      "step": 10529
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 22.978065490722656,
+      "learning_rate": 1.119022716813789e-05,
+      "loss": 1.5411,
+      "step": 10530
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.687366485595703,
+      "learning_rate": 1.1189390453081204e-05,
+      "loss": 0.6153,
+      "step": 10531
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 6.889460563659668,
+      "learning_rate": 1.1188553738024516e-05,
+      "loss": 0.4407,
+      "step": 10532
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 7.504435062408447,
+      "learning_rate": 1.118771702296783e-05,
+      "loss": 0.3158,
+      "step": 10533
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.856164932250977,
+      "learning_rate": 1.1186880307911142e-05,
+      "loss": 0.6373,
+      "step": 10534
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 10.153422355651855,
+      "learning_rate": 1.1186043592854454e-05,
+      "loss": 0.352,
+      "step": 10535
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 14.861977577209473,
+      "learning_rate": 1.1185206877797767e-05,
+      "loss": 2.5537,
+      "step": 10536
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.03333854675293,
+      "learning_rate": 1.1184370162741078e-05,
+      "loss": 0.7913,
+      "step": 10537
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 25.498279571533203,
+      "learning_rate": 1.1183533447684391e-05,
+      "loss": 1.5244,
+      "step": 10538
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 13.05024528503418,
+      "learning_rate": 1.1182696732627705e-05,
+      "loss": 1.8159,
+      "step": 10539
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 17.579965591430664,
+      "learning_rate": 1.1181860017571019e-05,
+      "loss": 1.3664,
+      "step": 10540
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 7.962263584136963,
+      "learning_rate": 1.1181023302514329e-05,
+      "loss": 1.0635,
+      "step": 10541
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 7.066946506500244,
+      "learning_rate": 1.1180186587457642e-05,
+      "loss": 1.743,
+      "step": 10542
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.174063682556152,
+      "learning_rate": 1.1179349872400956e-05,
+      "loss": 1.1828,
+      "step": 10543
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 33.12421417236328,
+      "learning_rate": 1.1178513157344266e-05,
+      "loss": 2.0509,
+      "step": 10544
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 9.112988471984863,
+      "learning_rate": 1.117767644228758e-05,
+      "loss": 0.6407,
+      "step": 10545
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 74.02664184570312,
+      "learning_rate": 1.1176839727230892e-05,
+      "loss": 1.7778,
+      "step": 10546
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 13.02247428894043,
+      "learning_rate": 1.1176003012174206e-05,
+      "loss": 1.609,
+      "step": 10547
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 27.995586395263672,
+      "learning_rate": 1.1175166297117518e-05,
+      "loss": 3.5742,
+      "step": 10548
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.910315990447998,
+      "learning_rate": 1.117432958206083e-05,
+      "loss": 0.4359,
+      "step": 10549
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 22.431060791015625,
+      "learning_rate": 1.1173492867004143e-05,
+      "loss": 2.211,
+      "step": 10550
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 13.337568283081055,
+      "learning_rate": 1.1172656151947453e-05,
+      "loss": 1.2511,
+      "step": 10551
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 59.052276611328125,
+      "learning_rate": 1.1171819436890767e-05,
+      "loss": 1.3315,
+      "step": 10552
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 6.686931133270264,
+      "learning_rate": 1.117098272183408e-05,
+      "loss": 0.3952,
+      "step": 10553
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 8.727797508239746,
+      "learning_rate": 1.1170146006777394e-05,
+      "loss": 1.8726,
+      "step": 10554
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 16.314666748046875,
+      "learning_rate": 1.1169309291720705e-05,
+      "loss": 2.3723,
+      "step": 10555
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 24.2736873626709,
+      "learning_rate": 1.1168472576664018e-05,
+      "loss": 1.8258,
+      "step": 10556
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 17.27292251586914,
+      "learning_rate": 1.116763586160733e-05,
+      "loss": 2.3978,
+      "step": 10557
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.159758567810059,
+      "learning_rate": 1.1166799146550642e-05,
+      "loss": 1.688,
+      "step": 10558
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 95.5091781616211,
+      "learning_rate": 1.1165962431493956e-05,
+      "loss": 2.191,
+      "step": 10559
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 18.80675506591797,
+      "learning_rate": 1.1165125716437268e-05,
+      "loss": 1.5938,
+      "step": 10560
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 6.572784423828125,
+      "learning_rate": 1.1164289001380581e-05,
+      "loss": 0.4247,
+      "step": 10561
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 15.58597183227539,
+      "learning_rate": 1.1163452286323893e-05,
+      "loss": 1.5509,
+      "step": 10562
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 34.90790557861328,
+      "learning_rate": 1.1162615571267205e-05,
+      "loss": 2.4227,
+      "step": 10563
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 13.600082397460938,
+      "learning_rate": 1.1161778856210519e-05,
+      "loss": 1.3753,
+      "step": 10564
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 18.139291763305664,
+      "learning_rate": 1.116094214115383e-05,
+      "loss": 0.3906,
+      "step": 10565
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.340546607971191,
+      "learning_rate": 1.1160105426097143e-05,
+      "loss": 1.2416,
+      "step": 10566
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.842409133911133,
+      "learning_rate": 1.1159268711040457e-05,
+      "loss": 0.9771,
+      "step": 10567
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.24379825592041,
+      "learning_rate": 1.115843199598377e-05,
+      "loss": 0.9641,
+      "step": 10568
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 13.406096458435059,
+      "learning_rate": 1.115759528092708e-05,
+      "loss": 1.8182,
+      "step": 10569
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 25.022401809692383,
+      "learning_rate": 1.1156758565870394e-05,
+      "loss": 2.7239,
+      "step": 10570
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.70371150970459,
+      "learning_rate": 1.1155921850813706e-05,
+      "loss": 1.6387,
+      "step": 10571
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.708685874938965,
+      "learning_rate": 1.1155085135757018e-05,
+      "loss": 1.5522,
+      "step": 10572
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 16.037275314331055,
+      "learning_rate": 1.1154248420700332e-05,
+      "loss": 1.4569,
+      "step": 10573
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 13.24316692352295,
+      "learning_rate": 1.1153411705643644e-05,
+      "loss": 1.8347,
+      "step": 10574
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 24.94005012512207,
+      "learning_rate": 1.1152574990586957e-05,
+      "loss": 1.1633,
+      "step": 10575
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 40.46428680419922,
+      "learning_rate": 1.1151738275530268e-05,
+      "loss": 1.9529,
+      "step": 10576
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.654131889343262,
+      "learning_rate": 1.1150901560473581e-05,
+      "loss": 0.9671,
+      "step": 10577
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 13.445412635803223,
+      "learning_rate": 1.1150064845416895e-05,
+      "loss": 1.0693,
+      "step": 10578
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 6.598649501800537,
+      "learning_rate": 1.1149228130360205e-05,
+      "loss": 0.256,
+      "step": 10579
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.3994269371032715,
+      "learning_rate": 1.1148391415303519e-05,
+      "loss": 0.4632,
+      "step": 10580
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 6.5095343589782715,
+      "learning_rate": 1.1147554700246832e-05,
+      "loss": 1.4937,
+      "step": 10581
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 6.868313789367676,
+      "learning_rate": 1.1146717985190146e-05,
+      "loss": 1.0561,
+      "step": 10582
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 14.947495460510254,
+      "learning_rate": 1.1145881270133456e-05,
+      "loss": 1.4978,
+      "step": 10583
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.344189643859863,
+      "learning_rate": 1.114504455507677e-05,
+      "loss": 0.6595,
+      "step": 10584
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 20.680248260498047,
+      "learning_rate": 1.1144207840020082e-05,
+      "loss": 1.0697,
+      "step": 10585
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 6.246676445007324,
+      "learning_rate": 1.1143371124963394e-05,
+      "loss": 1.0448,
+      "step": 10586
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.066667556762695,
+      "learning_rate": 1.1142534409906708e-05,
+      "loss": 1.8635,
+      "step": 10587
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.610118389129639,
+      "learning_rate": 1.114169769485002e-05,
+      "loss": 1.1762,
+      "step": 10588
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 38.041683197021484,
+      "learning_rate": 1.1140860979793333e-05,
+      "loss": 1.7216,
+      "step": 10589
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 19.77991485595703,
+      "learning_rate": 1.1140024264736643e-05,
+      "loss": 0.9676,
+      "step": 10590
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 14.37332534790039,
+      "learning_rate": 1.1139187549679957e-05,
+      "loss": 1.7085,
+      "step": 10591
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.033374786376953,
+      "learning_rate": 1.113835083462327e-05,
+      "loss": 1.5882,
+      "step": 10592
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.050271511077881,
+      "learning_rate": 1.1137514119566581e-05,
+      "loss": 0.2143,
+      "step": 10593
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.486701488494873,
+      "learning_rate": 1.1136677404509895e-05,
+      "loss": 0.5395,
+      "step": 10594
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.2949371337890625,
+      "learning_rate": 1.1135840689453208e-05,
+      "loss": 0.5285,
+      "step": 10595
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.485579490661621,
+      "learning_rate": 1.1135003974396522e-05,
+      "loss": 1.1012,
+      "step": 10596
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 5.428603649139404,
+      "learning_rate": 1.1134167259339832e-05,
+      "loss": 1.3909,
+      "step": 10597
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.4245710372924805,
+      "learning_rate": 1.1133330544283146e-05,
+      "loss": 0.6164,
+      "step": 10598
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 44.259098052978516,
+      "learning_rate": 1.1132493829226458e-05,
+      "loss": 1.6817,
+      "step": 10599
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 12.517207145690918,
+      "learning_rate": 1.113165711416977e-05,
+      "loss": 0.3257,
+      "step": 10600
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 12.179635047912598,
+      "learning_rate": 1.1130820399113083e-05,
+      "loss": 2.1171,
+      "step": 10601
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.754598617553711,
+      "learning_rate": 1.1129983684056395e-05,
+      "loss": 0.7985,
+      "step": 10602
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.699295997619629,
+      "learning_rate": 1.1129146968999709e-05,
+      "loss": 0.3123,
+      "step": 10603
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 14.108746528625488,
+      "learning_rate": 1.112831025394302e-05,
+      "loss": 2.128,
+      "step": 10604
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.265678405761719,
+      "learning_rate": 1.1127473538886333e-05,
+      "loss": 2.2785,
+      "step": 10605
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 37.177345275878906,
+      "learning_rate": 1.1126636823829647e-05,
+      "loss": 1.7067,
+      "step": 10606
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.379622459411621,
+      "learning_rate": 1.1125800108772957e-05,
+      "loss": 0.82,
+      "step": 10607
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 32.17717742919922,
+      "learning_rate": 1.112496339371627e-05,
+      "loss": 1.7148,
+      "step": 10608
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 17.302162170410156,
+      "learning_rate": 1.1124126678659584e-05,
+      "loss": 1.7778,
+      "step": 10609
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.162320137023926,
+      "learning_rate": 1.1123289963602896e-05,
+      "loss": 2.5884,
+      "step": 10610
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 31.90337371826172,
+      "learning_rate": 1.1122453248546208e-05,
+      "loss": 1.3121,
+      "step": 10611
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 20.651403427124023,
+      "learning_rate": 1.1121616533489522e-05,
+      "loss": 1.8125,
+      "step": 10612
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 14.351293563842773,
+      "learning_rate": 1.1120779818432834e-05,
+      "loss": 0.8593,
+      "step": 10613
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.308723449707031,
+      "learning_rate": 1.1119943103376146e-05,
+      "loss": 0.3539,
+      "step": 10614
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.801800727844238,
+      "learning_rate": 1.1119106388319457e-05,
+      "loss": 0.2271,
+      "step": 10615
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 12.822261810302734,
+      "learning_rate": 1.1118269673262771e-05,
+      "loss": 1.3261,
+      "step": 10616
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 30.570140838623047,
+      "learning_rate": 1.1117432958206085e-05,
+      "loss": 2.3794,
+      "step": 10617
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 12.661914825439453,
+      "learning_rate": 1.1116596243149395e-05,
+      "loss": 0.9678,
+      "step": 10618
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 27.10087013244629,
+      "learning_rate": 1.1115759528092709e-05,
+      "loss": 0.6331,
+      "step": 10619
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 24.226333618164062,
+      "learning_rate": 1.1114922813036022e-05,
+      "loss": 0.1215,
+      "step": 10620
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.87596321105957,
+      "learning_rate": 1.1114086097979333e-05,
+      "loss": 1.5549,
+      "step": 10621
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 14.273589134216309,
+      "learning_rate": 1.1113249382922646e-05,
+      "loss": 1.8746,
+      "step": 10622
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.367445945739746,
+      "learning_rate": 1.111241266786596e-05,
+      "loss": 0.1504,
+      "step": 10623
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 46.793949127197266,
+      "learning_rate": 1.1111575952809272e-05,
+      "loss": 0.7179,
+      "step": 10624
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 86.86392974853516,
+      "learning_rate": 1.1110739237752584e-05,
+      "loss": 1.8453,
+      "step": 10625
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 18.575767517089844,
+      "learning_rate": 1.1109902522695897e-05,
+      "loss": 2.5457,
+      "step": 10626
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.418265342712402,
+      "learning_rate": 1.110906580763921e-05,
+      "loss": 0.9466,
+      "step": 10627
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.965066432952881,
+      "learning_rate": 1.1108229092582521e-05,
+      "loss": 1.1486,
+      "step": 10628
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 25.836055755615234,
+      "learning_rate": 1.1107392377525833e-05,
+      "loss": 1.188,
+      "step": 10629
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 32.03029251098633,
+      "learning_rate": 1.1106555662469147e-05,
+      "loss": 3.0007,
+      "step": 10630
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.605908393859863,
+      "learning_rate": 1.110571894741246e-05,
+      "loss": 0.6728,
+      "step": 10631
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.322670936584473,
+      "learning_rate": 1.1104882232355771e-05,
+      "loss": 1.0013,
+      "step": 10632
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 12.611401557922363,
+      "learning_rate": 1.1104045517299085e-05,
+      "loss": 0.9169,
+      "step": 10633
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.119584083557129,
+      "learning_rate": 1.1103208802242398e-05,
+      "loss": 1.8297,
+      "step": 10634
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.377337455749512,
+      "learning_rate": 1.1102372087185708e-05,
+      "loss": 0.898,
+      "step": 10635
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 17.205080032348633,
+      "learning_rate": 1.1101535372129022e-05,
+      "loss": 0.8162,
+      "step": 10636
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 10.054287910461426,
+      "learning_rate": 1.1100698657072336e-05,
+      "loss": 0.9296,
+      "step": 10637
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.22707462310791,
+      "learning_rate": 1.1099861942015648e-05,
+      "loss": 0.8789,
+      "step": 10638
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 5.999391555786133,
+      "learning_rate": 1.109902522695896e-05,
+      "loss": 0.5489,
+      "step": 10639
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 19.555545806884766,
+      "learning_rate": 1.1098188511902273e-05,
+      "loss": 1.0702,
+      "step": 10640
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 29.52223014831543,
+      "learning_rate": 1.1097351796845585e-05,
+      "loss": 2.0596,
+      "step": 10641
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 6.383723258972168,
+      "learning_rate": 1.1096515081788897e-05,
+      "loss": 0.7636,
+      "step": 10642
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 30.461763381958008,
+      "learning_rate": 1.109567836673221e-05,
+      "loss": 2.3186,
+      "step": 10643
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 15.468742370605469,
+      "learning_rate": 1.1094841651675523e-05,
+      "loss": 1.3401,
+      "step": 10644
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 24.03126335144043,
+      "learning_rate": 1.1094004936618836e-05,
+      "loss": 0.7212,
+      "step": 10645
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 64.02912902832031,
+      "learning_rate": 1.1093168221562147e-05,
+      "loss": 1.2535,
+      "step": 10646
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.855770111083984,
+      "learning_rate": 1.109233150650546e-05,
+      "loss": 0.7123,
+      "step": 10647
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 10.665716171264648,
+      "learning_rate": 1.1091494791448774e-05,
+      "loss": 1.0164,
+      "step": 10648
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.271306991577148,
+      "learning_rate": 1.1090658076392084e-05,
+      "loss": 1.5193,
+      "step": 10649
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 5.731505870819092,
+      "learning_rate": 1.1089821361335398e-05,
+      "loss": 0.5393,
+      "step": 10650
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 97.72625732421875,
+      "learning_rate": 1.1088984646278712e-05,
+      "loss": 0.8654,
+      "step": 10651
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 24.18966293334961,
+      "learning_rate": 1.1088147931222024e-05,
+      "loss": 1.3493,
+      "step": 10652
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 52.74606704711914,
+      "learning_rate": 1.1087311216165335e-05,
+      "loss": 1.535,
+      "step": 10653
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 18.01695442199707,
+      "learning_rate": 1.1086474501108649e-05,
+      "loss": 0.5694,
+      "step": 10654
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 5.7204365730285645,
+      "learning_rate": 1.1085637786051961e-05,
+      "loss": 0.2329,
+      "step": 10655
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 5.812489032745361,
+      "learning_rate": 1.1084801070995273e-05,
+      "loss": 0.6321,
+      "step": 10656
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.29145622253418,
+      "learning_rate": 1.1083964355938585e-05,
+      "loss": 1.7102,
+      "step": 10657
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 8.088118553161621,
+      "learning_rate": 1.1083127640881899e-05,
+      "loss": 0.3344,
+      "step": 10658
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 15.177617073059082,
+      "learning_rate": 1.1082290925825212e-05,
+      "loss": 1.1892,
+      "step": 10659
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 17.269393920898438,
+      "learning_rate": 1.1081454210768523e-05,
+      "loss": 2.7234,
+      "step": 10660
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.55111026763916,
+      "learning_rate": 1.1080617495711836e-05,
+      "loss": 0.8772,
+      "step": 10661
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 10.325575828552246,
+      "learning_rate": 1.107978078065515e-05,
+      "loss": 0.6516,
+      "step": 10662
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 11.700462341308594,
+      "learning_rate": 1.107894406559846e-05,
+      "loss": 0.605,
+      "step": 10663
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 163.7694091796875,
+      "learning_rate": 1.1078107350541774e-05,
+      "loss": 2.3894,
+      "step": 10664
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 6.185863018035889,
+      "learning_rate": 1.1077270635485087e-05,
+      "loss": 0.4748,
+      "step": 10665
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.343429565429688,
+      "learning_rate": 1.10764339204284e-05,
+      "loss": 0.5629,
+      "step": 10666
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 6.737290859222412,
+      "learning_rate": 1.1075597205371711e-05,
+      "loss": 0.5619,
+      "step": 10667
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 21.991905212402344,
+      "learning_rate": 1.1074760490315023e-05,
+      "loss": 1.938,
+      "step": 10668
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 8.844432830810547,
+      "learning_rate": 1.1073923775258337e-05,
+      "loss": 1.7115,
+      "step": 10669
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 19.51593780517578,
+      "learning_rate": 1.1073087060201649e-05,
+      "loss": 2.117,
+      "step": 10670
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 25.96725082397461,
+      "learning_rate": 1.107225034514496e-05,
+      "loss": 1.8371,
+      "step": 10671
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 8.808110237121582,
+      "learning_rate": 1.1071413630088274e-05,
+      "loss": 0.4944,
+      "step": 10672
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 43.18079376220703,
+      "learning_rate": 1.1070576915031588e-05,
+      "loss": 0.7812,
+      "step": 10673
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 14.915264129638672,
+      "learning_rate": 1.1069740199974898e-05,
+      "loss": 1.4578,
+      "step": 10674
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.375272750854492,
+      "learning_rate": 1.1068903484918212e-05,
+      "loss": 1.5969,
+      "step": 10675
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 21.682424545288086,
+      "learning_rate": 1.1068066769861526e-05,
+      "loss": 1.688,
+      "step": 10676
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 21.142986297607422,
+      "learning_rate": 1.1067230054804836e-05,
+      "loss": 2.6396,
+      "step": 10677
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 19.430593490600586,
+      "learning_rate": 1.106639333974815e-05,
+      "loss": 0.8929,
+      "step": 10678
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.548630714416504,
+      "learning_rate": 1.1065556624691463e-05,
+      "loss": 0.509,
+      "step": 10679
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 15.138720512390137,
+      "learning_rate": 1.1064719909634775e-05,
+      "loss": 0.878,
+      "step": 10680
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 25.690645217895508,
+      "learning_rate": 1.1063883194578087e-05,
+      "loss": 3.1012,
+      "step": 10681
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 66.32939910888672,
+      "learning_rate": 1.1063046479521399e-05,
+      "loss": 0.5862,
+      "step": 10682
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.385822296142578,
+      "learning_rate": 1.1062209764464713e-05,
+      "loss": 0.6105,
+      "step": 10683
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 40.63554763793945,
+      "learning_rate": 1.1061373049408025e-05,
+      "loss": 2.9037,
+      "step": 10684
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 8.389958381652832,
+      "learning_rate": 1.1060536334351337e-05,
+      "loss": 0.7247,
+      "step": 10685
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.9861979484558105,
+      "learning_rate": 1.105969961929465e-05,
+      "loss": 0.5741,
+      "step": 10686
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 20.282155990600586,
+      "learning_rate": 1.1058862904237964e-05,
+      "loss": 2.1818,
+      "step": 10687
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.458776473999023,
+      "learning_rate": 1.1058026189181274e-05,
+      "loss": 0.3538,
+      "step": 10688
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 10.17504596710205,
+      "learning_rate": 1.1057189474124588e-05,
+      "loss": 0.849,
+      "step": 10689
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 31.185483932495117,
+      "learning_rate": 1.1056352759067902e-05,
+      "loss": 2.2377,
+      "step": 10690
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 6.85310697555542,
+      "learning_rate": 1.1055516044011212e-05,
+      "loss": 0.2518,
+      "step": 10691
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 19.96877098083496,
+      "learning_rate": 1.1054679328954525e-05,
+      "loss": 1.3782,
+      "step": 10692
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.055338859558105,
+      "learning_rate": 1.1053842613897839e-05,
+      "loss": 0.6675,
+      "step": 10693
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 10.827799797058105,
+      "learning_rate": 1.1053005898841151e-05,
+      "loss": 2.3363,
+      "step": 10694
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.236748218536377,
+      "learning_rate": 1.1052169183784463e-05,
+      "loss": 0.9744,
+      "step": 10695
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 17.888032913208008,
+      "learning_rate": 1.1051332468727775e-05,
+      "loss": 0.9261,
+      "step": 10696
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 41.39914321899414,
+      "learning_rate": 1.1050495753671089e-05,
+      "loss": 2.7761,
+      "step": 10697
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 17.63840675354004,
+      "learning_rate": 1.10496590386144e-05,
+      "loss": 0.8235,
+      "step": 10698
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 22.87006378173828,
+      "learning_rate": 1.1048822323557713e-05,
+      "loss": 1.8333,
+      "step": 10699
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 23.068431854248047,
+      "learning_rate": 1.1047985608501026e-05,
+      "loss": 0.9031,
+      "step": 10700
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 10.015175819396973,
+      "learning_rate": 1.104714889344434e-05,
+      "loss": 0.6289,
+      "step": 10701
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 4.441029071807861,
+      "learning_rate": 1.104631217838765e-05,
+      "loss": 0.2909,
+      "step": 10702
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 17.1404972076416,
+      "learning_rate": 1.1045475463330964e-05,
+      "loss": 1.3492,
+      "step": 10703
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 8.014181137084961,
+      "learning_rate": 1.1044638748274277e-05,
+      "loss": 1.2163,
+      "step": 10704
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 5.611763000488281,
+      "learning_rate": 1.1043802033217588e-05,
+      "loss": 0.6437,
+      "step": 10705
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.184558868408203,
+      "learning_rate": 1.1042965318160901e-05,
+      "loss": 2.2818,
+      "step": 10706
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 12.149728775024414,
+      "learning_rate": 1.1042128603104215e-05,
+      "loss": 0.8291,
+      "step": 10707
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 16.432268142700195,
+      "learning_rate": 1.1041291888047527e-05,
+      "loss": 0.4215,
+      "step": 10708
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 15.283656120300293,
+      "learning_rate": 1.1040455172990839e-05,
+      "loss": 1.2255,
+      "step": 10709
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 18.21234130859375,
+      "learning_rate": 1.103961845793415e-05,
+      "loss": 0.9063,
+      "step": 10710
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 17.743425369262695,
+      "learning_rate": 1.1038781742877464e-05,
+      "loss": 1.6894,
+      "step": 10711
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.465819358825684,
+      "learning_rate": 1.1037945027820776e-05,
+      "loss": 0.2748,
+      "step": 10712
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 8.443482398986816,
+      "learning_rate": 1.1037108312764088e-05,
+      "loss": 0.4027,
+      "step": 10713
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 51.657470703125,
+      "learning_rate": 1.1036271597707402e-05,
+      "loss": 1.978,
+      "step": 10714
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 13.823935508728027,
+      "learning_rate": 1.1035434882650716e-05,
+      "loss": 0.5682,
+      "step": 10715
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 5.75874137878418,
+      "learning_rate": 1.1034598167594026e-05,
+      "loss": 0.3957,
+      "step": 10716
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 15.59653091430664,
+      "learning_rate": 1.103376145253734e-05,
+      "loss": 0.864,
+      "step": 10717
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 17.857391357421875,
+      "learning_rate": 1.1032924737480653e-05,
+      "loss": 1.1767,
+      "step": 10718
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 14.209550857543945,
+      "learning_rate": 1.1032088022423963e-05,
+      "loss": 0.9556,
+      "step": 10719
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.322964668273926,
+      "learning_rate": 1.1031251307367277e-05,
+      "loss": 1.8327,
+      "step": 10720
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 13.394015312194824,
+      "learning_rate": 1.1030414592310589e-05,
+      "loss": 0.9464,
+      "step": 10721
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 23.246774673461914,
+      "learning_rate": 1.1029577877253903e-05,
+      "loss": 0.5882,
+      "step": 10722
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.086636543273926,
+      "learning_rate": 1.1028741162197215e-05,
+      "loss": 1.8144,
+      "step": 10723
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 15.44156551361084,
+      "learning_rate": 1.1027904447140527e-05,
+      "loss": 0.8163,
+      "step": 10724
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 30.777740478515625,
+      "learning_rate": 1.102706773208384e-05,
+      "loss": 1.5816,
+      "step": 10725
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 14.217241287231445,
+      "learning_rate": 1.102623101702715e-05,
+      "loss": 1.169,
+      "step": 10726
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.077201843261719,
+      "learning_rate": 1.1025394301970464e-05,
+      "loss": 0.7978,
+      "step": 10727
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.994258880615234,
+      "learning_rate": 1.1024557586913778e-05,
+      "loss": 1.5531,
+      "step": 10728
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 14.76309871673584,
+      "learning_rate": 1.1023720871857091e-05,
+      "loss": 0.5932,
+      "step": 10729
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.74864673614502,
+      "learning_rate": 1.1022884156800402e-05,
+      "loss": 0.3559,
+      "step": 10730
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.553218841552734,
+      "learning_rate": 1.1022047441743715e-05,
+      "loss": 0.5703,
+      "step": 10731
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 17.947389602661133,
+      "learning_rate": 1.1021210726687029e-05,
+      "loss": 1.5157,
+      "step": 10732
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 20.41041374206543,
+      "learning_rate": 1.102037401163034e-05,
+      "loss": 1.24,
+      "step": 10733
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.316654205322266,
+      "learning_rate": 1.1019537296573653e-05,
+      "loss": 0.5193,
+      "step": 10734
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.348031044006348,
+      "learning_rate": 1.1018700581516965e-05,
+      "loss": 0.3684,
+      "step": 10735
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 15.791666984558105,
+      "learning_rate": 1.1017863866460279e-05,
+      "loss": 1.1495,
+      "step": 10736
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.772533416748047,
+      "learning_rate": 1.101702715140359e-05,
+      "loss": 0.6673,
+      "step": 10737
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 5.980607032775879,
+      "learning_rate": 1.1016190436346902e-05,
+      "loss": 0.5814,
+      "step": 10738
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 7.196502208709717,
+      "learning_rate": 1.1015353721290216e-05,
+      "loss": 0.7421,
+      "step": 10739
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 7.806058406829834,
+      "learning_rate": 1.1014517006233526e-05,
+      "loss": 1.1668,
+      "step": 10740
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 22.027751922607422,
+      "learning_rate": 1.101368029117684e-05,
+      "loss": 1.6726,
+      "step": 10741
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 16.02842903137207,
+      "learning_rate": 1.1012843576120154e-05,
+      "loss": 4.1153,
+      "step": 10742
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 10.5230131149292,
+      "learning_rate": 1.1012006861063467e-05,
+      "loss": 0.0575,
+      "step": 10743
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 29.45198631286621,
+      "learning_rate": 1.1011170146006778e-05,
+      "loss": 2.2434,
+      "step": 10744
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 19.144636154174805,
+      "learning_rate": 1.1010333430950091e-05,
+      "loss": 1.2749,
+      "step": 10745
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 26.141258239746094,
+      "learning_rate": 1.1009496715893405e-05,
+      "loss": 1.6285,
+      "step": 10746
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 13.935678482055664,
+      "learning_rate": 1.1008660000836715e-05,
+      "loss": 1.5021,
+      "step": 10747
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 341.974365234375,
+      "learning_rate": 1.1007823285780029e-05,
+      "loss": 1.0623,
+      "step": 10748
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 14.870683670043945,
+      "learning_rate": 1.100698657072334e-05,
+      "loss": 1.3213,
+      "step": 10749
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.464217185974121,
+      "learning_rate": 1.1006149855666654e-05,
+      "loss": 0.7104,
+      "step": 10750
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 38.540260314941406,
+      "learning_rate": 1.1005313140609966e-05,
+      "loss": 3.1378,
+      "step": 10751
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 15.78632926940918,
+      "learning_rate": 1.1004476425553278e-05,
+      "loss": 3.0877,
+      "step": 10752
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 5.118893623352051,
+      "learning_rate": 1.1003639710496592e-05,
+      "loss": 0.3996,
+      "step": 10753
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 25.54676055908203,
+      "learning_rate": 1.1002802995439902e-05,
+      "loss": 1.4233,
+      "step": 10754
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 13.176262855529785,
+      "learning_rate": 1.1001966280383216e-05,
+      "loss": 1.7879,
+      "step": 10755
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 15.778350830078125,
+      "learning_rate": 1.100112956532653e-05,
+      "loss": 1.4188,
+      "step": 10756
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 19.335603713989258,
+      "learning_rate": 1.1000292850269843e-05,
+      "loss": 1.1172,
+      "step": 10757
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 81.1264419555664,
+      "learning_rate": 1.0999456135213153e-05,
+      "loss": 0.6375,
+      "step": 10758
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 4.9337334632873535,
+      "learning_rate": 1.0998619420156467e-05,
+      "loss": 0.1351,
+      "step": 10759
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 26.468721389770508,
+      "learning_rate": 1.0997782705099779e-05,
+      "loss": 1.7092,
+      "step": 10760
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.010913848876953,
+      "learning_rate": 1.0996945990043091e-05,
+      "loss": 1.2734,
+      "step": 10761
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 6.546537399291992,
+      "learning_rate": 1.0996109274986405e-05,
+      "loss": 0.1734,
+      "step": 10762
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 16.221281051635742,
+      "learning_rate": 1.0995272559929717e-05,
+      "loss": 0.7127,
+      "step": 10763
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 6.823152542114258,
+      "learning_rate": 1.099443584487303e-05,
+      "loss": 0.3035,
+      "step": 10764
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.010765075683594,
+      "learning_rate": 1.0993599129816342e-05,
+      "loss": 0.6597,
+      "step": 10765
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 13.101228713989258,
+      "learning_rate": 1.0992762414759654e-05,
+      "loss": 1.7342,
+      "step": 10766
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 20.14594268798828,
+      "learning_rate": 1.0991925699702968e-05,
+      "loss": 1.9099,
+      "step": 10767
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 53.84599304199219,
+      "learning_rate": 1.0991088984646278e-05,
+      "loss": 1.1974,
+      "step": 10768
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 15.755928039550781,
+      "learning_rate": 1.0990252269589592e-05,
+      "loss": 2.1624,
+      "step": 10769
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 6.549153804779053,
+      "learning_rate": 1.0989415554532905e-05,
+      "loss": 0.3828,
+      "step": 10770
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 29.744705200195312,
+      "learning_rate": 1.0988578839476219e-05,
+      "loss": 1.5385,
+      "step": 10771
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.353290557861328,
+      "learning_rate": 1.098774212441953e-05,
+      "loss": 0.7369,
+      "step": 10772
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 44.18974304199219,
+      "learning_rate": 1.0986905409362843e-05,
+      "loss": 1.6053,
+      "step": 10773
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 4.334277153015137,
+      "learning_rate": 1.0986068694306155e-05,
+      "loss": 0.348,
+      "step": 10774
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.039255142211914,
+      "learning_rate": 1.0985231979249467e-05,
+      "loss": 0.4938,
+      "step": 10775
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 30.044965744018555,
+      "learning_rate": 1.098439526419278e-05,
+      "loss": 1.6979,
+      "step": 10776
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.394641876220703,
+      "learning_rate": 1.0983558549136092e-05,
+      "loss": 0.6797,
+      "step": 10777
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 14.829742431640625,
+      "learning_rate": 1.0982721834079406e-05,
+      "loss": 1.0671,
+      "step": 10778
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.1448755264282227,
+      "learning_rate": 1.0981885119022716e-05,
+      "loss": 0.0608,
+      "step": 10779
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 43.15748596191406,
+      "learning_rate": 1.098104840396603e-05,
+      "loss": 1.4878,
+      "step": 10780
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.831615447998047,
+      "learning_rate": 1.0980211688909344e-05,
+      "loss": 0.49,
+      "step": 10781
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 5.491212368011475,
+      "learning_rate": 1.0979374973852654e-05,
+      "loss": 0.7475,
+      "step": 10782
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 52.36024475097656,
+      "learning_rate": 1.0978538258795968e-05,
+      "loss": 1.7538,
+      "step": 10783
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 19.612943649291992,
+      "learning_rate": 1.0977701543739281e-05,
+      "loss": 1.9167,
+      "step": 10784
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 17.72964096069336,
+      "learning_rate": 1.0976864828682595e-05,
+      "loss": 1.9991,
+      "step": 10785
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.766716957092285,
+      "learning_rate": 1.0976028113625905e-05,
+      "loss": 1.0331,
+      "step": 10786
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 17.612449645996094,
+      "learning_rate": 1.0975191398569219e-05,
+      "loss": 0.8623,
+      "step": 10787
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.3621190786361694,
+      "learning_rate": 1.097435468351253e-05,
+      "loss": 0.0174,
+      "step": 10788
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.920506477355957,
+      "learning_rate": 1.0973517968455843e-05,
+      "loss": 0.463,
+      "step": 10789
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.6283490657806396,
+      "learning_rate": 1.0972681253399156e-05,
+      "loss": 0.0628,
+      "step": 10790
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 8.424253463745117,
+      "learning_rate": 1.0971844538342468e-05,
+      "loss": 1.4464,
+      "step": 10791
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 100.44664001464844,
+      "learning_rate": 1.0971007823285782e-05,
+      "loss": 1.607,
+      "step": 10792
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 48.81386947631836,
+      "learning_rate": 1.0970171108229092e-05,
+      "loss": 1.9988,
+      "step": 10793
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 7.60531759262085,
+      "learning_rate": 1.0969334393172406e-05,
+      "loss": 0.278,
+      "step": 10794
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 30.702272415161133,
+      "learning_rate": 1.096849767811572e-05,
+      "loss": 1.2332,
+      "step": 10795
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 13.653675079345703,
+      "learning_rate": 1.096766096305903e-05,
+      "loss": 1.2394,
+      "step": 10796
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 15.487334251403809,
+      "learning_rate": 1.0966824248002343e-05,
+      "loss": 1.075,
+      "step": 10797
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 23.752225875854492,
+      "learning_rate": 1.0965987532945657e-05,
+      "loss": 1.3737,
+      "step": 10798
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.121785163879395,
+      "learning_rate": 1.096515081788897e-05,
+      "loss": 0.4708,
+      "step": 10799
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 6.617895603179932,
+      "learning_rate": 1.0964314102832281e-05,
+      "loss": 0.1338,
+      "step": 10800
+    },
+    {
+      "epoch": 1.36,
+      "eval_loss": 0.09359579533338547,
+      "eval_runtime": 95.2118,
+      "eval_samples_per_second": 37.201,
+      "eval_steps_per_second": 37.201,
+      "step": 10800
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 14.611085891723633,
+      "learning_rate": 1.0963477387775595e-05,
+      "loss": 2.6641,
+      "step": 10801
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.776219367980957,
+      "learning_rate": 1.0962640672718907e-05,
+      "loss": 0.7735,
+      "step": 10802
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 13.162578582763672,
+      "learning_rate": 1.0961803957662218e-05,
+      "loss": 1.9287,
+      "step": 10803
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 36.731101989746094,
+      "learning_rate": 1.0960967242605532e-05,
+      "loss": 1.6382,
+      "step": 10804
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 14.721063613891602,
+      "learning_rate": 1.0960130527548844e-05,
+      "loss": 1.5529,
+      "step": 10805
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 43.59991455078125,
+      "learning_rate": 1.0959293812492158e-05,
+      "loss": 1.7317,
+      "step": 10806
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.519847869873047,
+      "learning_rate": 1.0958457097435468e-05,
+      "loss": 0.8793,
+      "step": 10807
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 31.089527130126953,
+      "learning_rate": 1.0957620382378782e-05,
+      "loss": 1.1497,
+      "step": 10808
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 5.586785316467285,
+      "learning_rate": 1.0956783667322095e-05,
+      "loss": 0.5224,
+      "step": 10809
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 32.68950653076172,
+      "learning_rate": 1.0955946952265406e-05,
+      "loss": 3.5036,
+      "step": 10810
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.807537078857422,
+      "learning_rate": 1.095511023720872e-05,
+      "loss": 1.3255,
+      "step": 10811
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 27.183666229248047,
+      "learning_rate": 1.0954273522152033e-05,
+      "loss": 1.5964,
+      "step": 10812
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 24.39258575439453,
+      "learning_rate": 1.0953436807095345e-05,
+      "loss": 2.093,
+      "step": 10813
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 52.604713439941406,
+      "learning_rate": 1.0952600092038657e-05,
+      "loss": 0.9319,
+      "step": 10814
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 5.523324012756348,
+      "learning_rate": 1.095176337698197e-05,
+      "loss": 0.9474,
+      "step": 10815
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.748193264007568,
+      "learning_rate": 1.0950926661925282e-05,
+      "loss": 0.7595,
+      "step": 10816
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 19.95014762878418,
+      "learning_rate": 1.0950089946868594e-05,
+      "loss": 0.5307,
+      "step": 10817
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 20.722143173217773,
+      "learning_rate": 1.0949253231811908e-05,
+      "loss": 1.3851,
+      "step": 10818
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 16.28141212463379,
+      "learning_rate": 1.094841651675522e-05,
+      "loss": 1.1545,
+      "step": 10819
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 6.727929592132568,
+      "learning_rate": 1.0947579801698534e-05,
+      "loss": 0.4766,
+      "step": 10820
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 15.714207649230957,
+      "learning_rate": 1.0946743086641844e-05,
+      "loss": 0.8382,
+      "step": 10821
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 15.240822792053223,
+      "learning_rate": 1.0945906371585157e-05,
+      "loss": 1.1112,
+      "step": 10822
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 15.163829803466797,
+      "learning_rate": 1.0945069656528471e-05,
+      "loss": 0.7106,
+      "step": 10823
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 41.483150482177734,
+      "learning_rate": 1.0944232941471781e-05,
+      "loss": 0.9613,
+      "step": 10824
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 44.760169982910156,
+      "learning_rate": 1.0943396226415095e-05,
+      "loss": 1.3909,
+      "step": 10825
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.67187237739563,
+      "learning_rate": 1.0942559511358409e-05,
+      "loss": 0.2376,
+      "step": 10826
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 4.086060047149658,
+      "learning_rate": 1.094172279630172e-05,
+      "loss": 0.148,
+      "step": 10827
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 27.953208923339844,
+      "learning_rate": 1.0940886081245033e-05,
+      "loss": 1.6879,
+      "step": 10828
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 95.8250732421875,
+      "learning_rate": 1.0940049366188346e-05,
+      "loss": 1.8391,
+      "step": 10829
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.172741889953613,
+      "learning_rate": 1.0939212651131658e-05,
+      "loss": 1.8034,
+      "step": 10830
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 28.409162521362305,
+      "learning_rate": 1.093837593607497e-05,
+      "loss": 0.6427,
+      "step": 10831
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 5.7281694412231445,
+      "learning_rate": 1.0937539221018282e-05,
+      "loss": 0.7275,
+      "step": 10832
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 41.91596603393555,
+      "learning_rate": 1.0936702505961596e-05,
+      "loss": 2.1048,
+      "step": 10833
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 14.93394947052002,
+      "learning_rate": 1.093586579090491e-05,
+      "loss": 0.6343,
+      "step": 10834
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 35.88250732421875,
+      "learning_rate": 1.093502907584822e-05,
+      "loss": 0.5998,
+      "step": 10835
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 15.611506462097168,
+      "learning_rate": 1.0934192360791533e-05,
+      "loss": 2.218,
+      "step": 10836
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 5.545705318450928,
+      "learning_rate": 1.0933355645734847e-05,
+      "loss": 0.5066,
+      "step": 10837
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 14.528839111328125,
+      "learning_rate": 1.0932518930678157e-05,
+      "loss": 0.8079,
+      "step": 10838
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 10.407254219055176,
+      "learning_rate": 1.0931682215621471e-05,
+      "loss": 0.7827,
+      "step": 10839
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.218513011932373,
+      "learning_rate": 1.0930845500564785e-05,
+      "loss": 1.8934,
+      "step": 10840
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 14.339374542236328,
+      "learning_rate": 1.0930008785508096e-05,
+      "loss": 0.8121,
+      "step": 10841
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 14.267348289489746,
+      "learning_rate": 1.0929172070451408e-05,
+      "loss": 1.1232,
+      "step": 10842
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 15.108508110046387,
+      "learning_rate": 1.0928335355394722e-05,
+      "loss": 0.8803,
+      "step": 10843
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 22.21465492248535,
+      "learning_rate": 1.0927498640338034e-05,
+      "loss": 1.1716,
+      "step": 10844
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 25.122150421142578,
+      "learning_rate": 1.0926661925281346e-05,
+      "loss": 2.0065,
+      "step": 10845
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.775306701660156,
+      "learning_rate": 1.0925825210224658e-05,
+      "loss": 0.9244,
+      "step": 10846
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 56.42924880981445,
+      "learning_rate": 1.0924988495167972e-05,
+      "loss": 1.7071,
+      "step": 10847
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 17.765642166137695,
+      "learning_rate": 1.0924151780111285e-05,
+      "loss": 0.8892,
+      "step": 10848
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.326750755310059,
+      "learning_rate": 1.0923315065054596e-05,
+      "loss": 1.3537,
+      "step": 10849
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 34.46554946899414,
+      "learning_rate": 1.092247834999791e-05,
+      "loss": 0.8488,
+      "step": 10850
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 114.82874298095703,
+      "learning_rate": 1.0921641634941223e-05,
+      "loss": 1.2715,
+      "step": 10851
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 20.8919620513916,
+      "learning_rate": 1.0920804919884533e-05,
+      "loss": 1.28,
+      "step": 10852
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 22.95903778076172,
+      "learning_rate": 1.0919968204827847e-05,
+      "loss": 0.6853,
+      "step": 10853
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 19.324871063232422,
+      "learning_rate": 1.091913148977116e-05,
+      "loss": 0.8568,
+      "step": 10854
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.778809547424316,
+      "learning_rate": 1.0918294774714472e-05,
+      "loss": 0.7449,
+      "step": 10855
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 20.503408432006836,
+      "learning_rate": 1.0917458059657784e-05,
+      "loss": 1.145,
+      "step": 10856
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 22.312541961669922,
+      "learning_rate": 1.0916621344601098e-05,
+      "loss": 0.7269,
+      "step": 10857
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.290511131286621,
+      "learning_rate": 1.091578462954441e-05,
+      "loss": 1.0832,
+      "step": 10858
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 16.26047706604004,
+      "learning_rate": 1.0914947914487722e-05,
+      "loss": 1.5293,
+      "step": 10859
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 29.824007034301758,
+      "learning_rate": 1.0914111199431034e-05,
+      "loss": 1.0911,
+      "step": 10860
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 20.717750549316406,
+      "learning_rate": 1.0913274484374347e-05,
+      "loss": 1.1627,
+      "step": 10861
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 6.122304439544678,
+      "learning_rate": 1.0912437769317661e-05,
+      "loss": 0.4995,
+      "step": 10862
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 20.13656997680664,
+      "learning_rate": 1.0911601054260971e-05,
+      "loss": 1.2077,
+      "step": 10863
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 19.886560440063477,
+      "learning_rate": 1.0910764339204285e-05,
+      "loss": 2.5778,
+      "step": 10864
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 15.786689758300781,
+      "learning_rate": 1.0909927624147599e-05,
+      "loss": 0.9969,
+      "step": 10865
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 12.388118743896484,
+      "learning_rate": 1.0909090909090909e-05,
+      "loss": 2.4683,
+      "step": 10866
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 9.509102821350098,
+      "learning_rate": 1.0908254194034223e-05,
+      "loss": 1.0659,
+      "step": 10867
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.638148307800293,
+      "learning_rate": 1.0907417478977536e-05,
+      "loss": 0.8309,
+      "step": 10868
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 8.701188087463379,
+      "learning_rate": 1.0906580763920848e-05,
+      "loss": 0.3006,
+      "step": 10869
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 14.166561126708984,
+      "learning_rate": 1.090574404886416e-05,
+      "loss": 1.5359,
+      "step": 10870
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 12.052389144897461,
+      "learning_rate": 1.0904907333807472e-05,
+      "loss": 0.7277,
+      "step": 10871
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 23.465457916259766,
+      "learning_rate": 1.0904070618750786e-05,
+      "loss": 1.8724,
+      "step": 10872
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 19.340892791748047,
+      "learning_rate": 1.0903233903694098e-05,
+      "loss": 1.5015,
+      "step": 10873
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.132235050201416,
+      "learning_rate": 1.090239718863741e-05,
+      "loss": 0.8138,
+      "step": 10874
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 34.80180358886719,
+      "learning_rate": 1.0901560473580723e-05,
+      "loss": 2.4774,
+      "step": 10875
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 13.024115562438965,
+      "learning_rate": 1.0900723758524037e-05,
+      "loss": 1.8248,
+      "step": 10876
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 8.109501838684082,
+      "learning_rate": 1.0899887043467347e-05,
+      "loss": 0.5959,
+      "step": 10877
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 18.224985122680664,
+      "learning_rate": 1.089905032841066e-05,
+      "loss": 2.3261,
+      "step": 10878
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 19.332311630249023,
+      "learning_rate": 1.0898213613353974e-05,
+      "loss": 1.1313,
+      "step": 10879
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 48.47065734863281,
+      "learning_rate": 1.0897376898297285e-05,
+      "loss": 1.2985,
+      "step": 10880
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 25.13520050048828,
+      "learning_rate": 1.0896540183240598e-05,
+      "loss": 1.4921,
+      "step": 10881
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.925703048706055,
+      "learning_rate": 1.0895703468183912e-05,
+      "loss": 1.7304,
+      "step": 10882
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 5.960899829864502,
+      "learning_rate": 1.0894866753127222e-05,
+      "loss": 0.3075,
+      "step": 10883
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 40.891258239746094,
+      "learning_rate": 1.0894030038070536e-05,
+      "loss": 1.0237,
+      "step": 10884
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.174378395080566,
+      "learning_rate": 1.0893193323013848e-05,
+      "loss": 1.8066,
+      "step": 10885
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 18.660842895507812,
+      "learning_rate": 1.0892356607957162e-05,
+      "loss": 2.1987,
+      "step": 10886
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 16.45781135559082,
+      "learning_rate": 1.0891519892900474e-05,
+      "loss": 0.7041,
+      "step": 10887
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 15.377516746520996,
+      "learning_rate": 1.0890683177843785e-05,
+      "loss": 1.4118,
+      "step": 10888
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 199.94894409179688,
+      "learning_rate": 1.0889846462787099e-05,
+      "loss": 2.0811,
+      "step": 10889
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 25.48408317565918,
+      "learning_rate": 1.088900974773041e-05,
+      "loss": 2.1387,
+      "step": 10890
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 24.191665649414062,
+      "learning_rate": 1.0888173032673723e-05,
+      "loss": 1.1457,
+      "step": 10891
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 22.73685646057129,
+      "learning_rate": 1.0887336317617037e-05,
+      "loss": 2.3694,
+      "step": 10892
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 21.452775955200195,
+      "learning_rate": 1.088649960256035e-05,
+      "loss": 1.7771,
+      "step": 10893
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 56.104278564453125,
+      "learning_rate": 1.088566288750366e-05,
+      "loss": 2.2139,
+      "step": 10894
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 27.917959213256836,
+      "learning_rate": 1.0884826172446974e-05,
+      "loss": 1.4838,
+      "step": 10895
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.48881721496582,
+      "learning_rate": 1.0883989457390288e-05,
+      "loss": 2.6583,
+      "step": 10896
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.346531867980957,
+      "learning_rate": 1.0883152742333598e-05,
+      "loss": 0.8103,
+      "step": 10897
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 7.896469593048096,
+      "learning_rate": 1.0882316027276912e-05,
+      "loss": 0.9102,
+      "step": 10898
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 6.133947849273682,
+      "learning_rate": 1.0881479312220224e-05,
+      "loss": 0.4329,
+      "step": 10899
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 15.891400337219238,
+      "learning_rate": 1.0880642597163537e-05,
+      "loss": 0.7828,
+      "step": 10900
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 7.322772026062012,
+      "learning_rate": 1.087980588210685e-05,
+      "loss": 1.7073,
+      "step": 10901
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 26.040386199951172,
+      "learning_rate": 1.0878969167050161e-05,
+      "loss": 0.8719,
+      "step": 10902
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.847599029541016,
+      "learning_rate": 1.0878132451993475e-05,
+      "loss": 0.6563,
+      "step": 10903
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 10.403653144836426,
+      "learning_rate": 1.0877295736936785e-05,
+      "loss": 0.9705,
+      "step": 10904
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 113.82823181152344,
+      "learning_rate": 1.0876459021880099e-05,
+      "loss": 1.1203,
+      "step": 10905
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 6.354864597320557,
+      "learning_rate": 1.0875622306823413e-05,
+      "loss": 2.0016,
+      "step": 10906
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 18.746736526489258,
+      "learning_rate": 1.0874785591766726e-05,
+      "loss": 0.6388,
+      "step": 10907
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 33.26611328125,
+      "learning_rate": 1.0873948876710036e-05,
+      "loss": 2.0198,
+      "step": 10908
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.522151947021484,
+      "learning_rate": 1.087311216165335e-05,
+      "loss": 2.2584,
+      "step": 10909
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 32.2423210144043,
+      "learning_rate": 1.0872275446596664e-05,
+      "loss": 1.6917,
+      "step": 10910
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.617437362670898,
+      "learning_rate": 1.0871438731539974e-05,
+      "loss": 0.489,
+      "step": 10911
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 13.178160667419434,
+      "learning_rate": 1.0870602016483288e-05,
+      "loss": 0.6835,
+      "step": 10912
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 21.637924194335938,
+      "learning_rate": 1.08697653014266e-05,
+      "loss": 1.0005,
+      "step": 10913
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.994182586669922,
+      "learning_rate": 1.0868928586369913e-05,
+      "loss": 1.3531,
+      "step": 10914
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 8.365532875061035,
+      "learning_rate": 1.0868091871313225e-05,
+      "loss": 0.7093,
+      "step": 10915
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 53.14915084838867,
+      "learning_rate": 1.0867255156256537e-05,
+      "loss": 1.7739,
+      "step": 10916
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.870243072509766,
+      "learning_rate": 1.086641844119985e-05,
+      "loss": 1.5167,
+      "step": 10917
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 12.425950050354004,
+      "learning_rate": 1.0865581726143161e-05,
+      "loss": 2.6167,
+      "step": 10918
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 18.783540725708008,
+      "learning_rate": 1.0864745011086475e-05,
+      "loss": 1.4072,
+      "step": 10919
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 17.30661964416504,
+      "learning_rate": 1.0863908296029788e-05,
+      "loss": 0.9908,
+      "step": 10920
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 124.06700897216797,
+      "learning_rate": 1.0863071580973102e-05,
+      "loss": 1.6201,
+      "step": 10921
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 25.860363006591797,
+      "learning_rate": 1.0862234865916412e-05,
+      "loss": 1.925,
+      "step": 10922
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 66.43064880371094,
+      "learning_rate": 1.0861398150859726e-05,
+      "loss": 2.0882,
+      "step": 10923
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 32.965850830078125,
+      "learning_rate": 1.0860561435803038e-05,
+      "loss": 3.9624,
+      "step": 10924
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 7.869022369384766,
+      "learning_rate": 1.085972472074635e-05,
+      "loss": 0.5503,
+      "step": 10925
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 8.238611221313477,
+      "learning_rate": 1.0858888005689663e-05,
+      "loss": 1.504,
+      "step": 10926
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 19.855043411254883,
+      "learning_rate": 1.0858051290632975e-05,
+      "loss": 2.174,
+      "step": 10927
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 13.219988822937012,
+      "learning_rate": 1.0857214575576289e-05,
+      "loss": 0.8025,
+      "step": 10928
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 48.4361686706543,
+      "learning_rate": 1.0856377860519601e-05,
+      "loss": 3.9421,
+      "step": 10929
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 51.29350662231445,
+      "learning_rate": 1.0855541145462913e-05,
+      "loss": 1.864,
+      "step": 10930
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 162.08787536621094,
+      "learning_rate": 1.0854704430406227e-05,
+      "loss": 0.8445,
+      "step": 10931
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 5.514544486999512,
+      "learning_rate": 1.0853867715349537e-05,
+      "loss": 0.2165,
+      "step": 10932
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 8.022409439086914,
+      "learning_rate": 1.085303100029285e-05,
+      "loss": 0.4751,
+      "step": 10933
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 13.04383659362793,
+      "learning_rate": 1.0852194285236164e-05,
+      "loss": 1.0201,
+      "step": 10934
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.314271926879883,
+      "learning_rate": 1.0851357570179478e-05,
+      "loss": 1.0127,
+      "step": 10935
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 13.84196949005127,
+      "learning_rate": 1.0850520855122788e-05,
+      "loss": 0.8237,
+      "step": 10936
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 18.14813804626465,
+      "learning_rate": 1.0849684140066102e-05,
+      "loss": 0.6021,
+      "step": 10937
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 31.143245697021484,
+      "learning_rate": 1.0848847425009414e-05,
+      "loss": 2.0824,
+      "step": 10938
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 14.831005096435547,
+      "learning_rate": 1.0848010709952726e-05,
+      "loss": 0.5582,
+      "step": 10939
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 15.000322341918945,
+      "learning_rate": 1.084717399489604e-05,
+      "loss": 1.863,
+      "step": 10940
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 9.42466926574707,
+      "learning_rate": 1.0846337279839351e-05,
+      "loss": 1.4959,
+      "step": 10941
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 28.115360260009766,
+      "learning_rate": 1.0845500564782665e-05,
+      "loss": 1.155,
+      "step": 10942
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 14.936105728149414,
+      "learning_rate": 1.0844663849725975e-05,
+      "loss": 2.4056,
+      "step": 10943
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 13.039700508117676,
+      "learning_rate": 1.0843827134669289e-05,
+      "loss": 0.8834,
+      "step": 10944
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.926374673843384,
+      "learning_rate": 1.0842990419612602e-05,
+      "loss": 1.1018,
+      "step": 10945
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 16.927934646606445,
+      "learning_rate": 1.0842153704555913e-05,
+      "loss": 2.5262,
+      "step": 10946
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 23.388479232788086,
+      "learning_rate": 1.0841316989499226e-05,
+      "loss": 2.0981,
+      "step": 10947
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 5.1547932624816895,
+      "learning_rate": 1.084048027444254e-05,
+      "loss": 0.466,
+      "step": 10948
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 16.516647338867188,
+      "learning_rate": 1.0839643559385854e-05,
+      "loss": 0.9563,
+      "step": 10949
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 4.277315616607666,
+      "learning_rate": 1.0838806844329164e-05,
+      "loss": 0.377,
+      "step": 10950
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 29.24924659729004,
+      "learning_rate": 1.0837970129272478e-05,
+      "loss": 0.8384,
+      "step": 10951
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 7.615672588348389,
+      "learning_rate": 1.083713341421579e-05,
+      "loss": 1.0544,
+      "step": 10952
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 15.231378555297852,
+      "learning_rate": 1.0836296699159101e-05,
+      "loss": 3.7036,
+      "step": 10953
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.526764154434204,
+      "learning_rate": 1.0835459984102415e-05,
+      "loss": 0.1095,
+      "step": 10954
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 67.74472045898438,
+      "learning_rate": 1.0834623269045727e-05,
+      "loss": 1.8468,
+      "step": 10955
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 14.515176773071289,
+      "learning_rate": 1.083378655398904e-05,
+      "loss": 0.5935,
+      "step": 10956
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 28.489931106567383,
+      "learning_rate": 1.0832949838932351e-05,
+      "loss": 2.053,
+      "step": 10957
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 5.610323905944824,
+      "learning_rate": 1.0832113123875665e-05,
+      "loss": 0.5024,
+      "step": 10958
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 141.2397003173828,
+      "learning_rate": 1.0831276408818978e-05,
+      "loss": 1.8588,
+      "step": 10959
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 23.450443267822266,
+      "learning_rate": 1.0830439693762289e-05,
+      "loss": 1.5275,
+      "step": 10960
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 33.665977478027344,
+      "learning_rate": 1.0829602978705602e-05,
+      "loss": 2.2573,
+      "step": 10961
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 4.286736965179443,
+      "learning_rate": 1.0828766263648916e-05,
+      "loss": 0.3182,
+      "step": 10962
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 30.770402908325195,
+      "learning_rate": 1.082792954859223e-05,
+      "loss": 1.2403,
+      "step": 10963
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.3374247550964355,
+      "learning_rate": 1.082709283353554e-05,
+      "loss": 1.1554,
+      "step": 10964
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 28.713409423828125,
+      "learning_rate": 1.0826256118478853e-05,
+      "loss": 1.6937,
+      "step": 10965
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 6.346406936645508,
+      "learning_rate": 1.0825419403422165e-05,
+      "loss": 2.4099,
+      "step": 10966
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.938896179199219,
+      "learning_rate": 1.0824582688365477e-05,
+      "loss": 1.9321,
+      "step": 10967
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 14.221976280212402,
+      "learning_rate": 1.0823745973308791e-05,
+      "loss": 0.5429,
+      "step": 10968
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.394267082214355,
+      "learning_rate": 1.0822909258252103e-05,
+      "loss": 0.7491,
+      "step": 10969
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 6.654685020446777,
+      "learning_rate": 1.0822072543195417e-05,
+      "loss": 0.8189,
+      "step": 10970
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 38.056480407714844,
+      "learning_rate": 1.0821235828138727e-05,
+      "loss": 1.2402,
+      "step": 10971
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.358619689941406,
+      "learning_rate": 1.082039911308204e-05,
+      "loss": 0.9531,
+      "step": 10972
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 15.162528991699219,
+      "learning_rate": 1.0819562398025354e-05,
+      "loss": 1.2263,
+      "step": 10973
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 26.586856842041016,
+      "learning_rate": 1.0818725682968664e-05,
+      "loss": 2.359,
+      "step": 10974
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.965980529785156,
+      "learning_rate": 1.0817888967911978e-05,
+      "loss": 0.2239,
+      "step": 10975
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 26.117725372314453,
+      "learning_rate": 1.0817052252855292e-05,
+      "loss": 2.2138,
+      "step": 10976
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 14.616991996765137,
+      "learning_rate": 1.0816215537798604e-05,
+      "loss": 0.8846,
+      "step": 10977
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.697427749633789,
+      "learning_rate": 1.0815378822741916e-05,
+      "loss": 0.7569,
+      "step": 10978
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.512846946716309,
+      "learning_rate": 1.081454210768523e-05,
+      "loss": 0.5427,
+      "step": 10979
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.949728012084961,
+      "learning_rate": 1.0813705392628541e-05,
+      "loss": 1.1092,
+      "step": 10980
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 19.04229164123535,
+      "learning_rate": 1.0812868677571853e-05,
+      "loss": 0.796,
+      "step": 10981
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.645513534545898,
+      "learning_rate": 1.0812031962515165e-05,
+      "loss": 1.5419,
+      "step": 10982
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.557134628295898,
+      "learning_rate": 1.0811195247458479e-05,
+      "loss": 1.5873,
+      "step": 10983
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.657785892486572,
+      "learning_rate": 1.0810358532401792e-05,
+      "loss": 0.8016,
+      "step": 10984
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 15.9237642288208,
+      "learning_rate": 1.0809521817345103e-05,
+      "loss": 1.4037,
+      "step": 10985
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 11.617383003234863,
+      "learning_rate": 1.0808685102288416e-05,
+      "loss": 0.4267,
+      "step": 10986
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 28.64150047302246,
+      "learning_rate": 1.080784838723173e-05,
+      "loss": 1.7304,
+      "step": 10987
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.376594543457031,
+      "learning_rate": 1.080701167217504e-05,
+      "loss": 1.46,
+      "step": 10988
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.38775634765625,
+      "learning_rate": 1.0806174957118354e-05,
+      "loss": 0.4572,
+      "step": 10989
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 11.072006225585938,
+      "learning_rate": 1.0805338242061668e-05,
+      "loss": 0.5525,
+      "step": 10990
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 5.45393180847168,
+      "learning_rate": 1.080450152700498e-05,
+      "loss": 1.8282,
+      "step": 10991
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.413850784301758,
+      "learning_rate": 1.0803664811948291e-05,
+      "loss": 0.4873,
+      "step": 10992
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 25.184465408325195,
+      "learning_rate": 1.0802828096891605e-05,
+      "loss": 2.4874,
+      "step": 10993
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 21.327537536621094,
+      "learning_rate": 1.0801991381834917e-05,
+      "loss": 1.5459,
+      "step": 10994
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 16.912900924682617,
+      "learning_rate": 1.0801154666778229e-05,
+      "loss": 1.3192,
+      "step": 10995
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 23.991470336914062,
+      "learning_rate": 1.0800317951721541e-05,
+      "loss": 2.3134,
+      "step": 10996
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 6.955298900604248,
+      "learning_rate": 1.0799481236664855e-05,
+      "loss": 0.5381,
+      "step": 10997
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 15.276537895202637,
+      "learning_rate": 1.0798644521608168e-05,
+      "loss": 0.6888,
+      "step": 10998
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 14.224745750427246,
+      "learning_rate": 1.0797807806551479e-05,
+      "loss": 1.1304,
+      "step": 10999
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 14.31340503692627,
+      "learning_rate": 1.0796971091494792e-05,
+      "loss": 1.3738,
+      "step": 11000
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 33.715232849121094,
+      "learning_rate": 1.0796134376438106e-05,
+      "loss": 1.6522,
+      "step": 11001
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.345632076263428,
+      "learning_rate": 1.0795297661381416e-05,
+      "loss": 0.1618,
+      "step": 11002
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 23.19948387145996,
+      "learning_rate": 1.079446094632473e-05,
+      "loss": 1.6154,
+      "step": 11003
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 18.290569305419922,
+      "learning_rate": 1.0793624231268043e-05,
+      "loss": 2.6678,
+      "step": 11004
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 13.243331909179688,
+      "learning_rate": 1.0792787516211355e-05,
+      "loss": 0.7295,
+      "step": 11005
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 25.255516052246094,
+      "learning_rate": 1.0791950801154667e-05,
+      "loss": 2.2459,
+      "step": 11006
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 30.202260971069336,
+      "learning_rate": 1.0791114086097981e-05,
+      "loss": 1.402,
+      "step": 11007
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.659870147705078,
+      "learning_rate": 1.0790277371041293e-05,
+      "loss": 0.3835,
+      "step": 11008
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.047741889953613,
+      "learning_rate": 1.0789440655984605e-05,
+      "loss": 0.9404,
+      "step": 11009
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.712565422058105,
+      "learning_rate": 1.0788603940927917e-05,
+      "loss": 2.0874,
+      "step": 11010
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 4.661409378051758,
+      "learning_rate": 1.078776722587123e-05,
+      "loss": 0.3309,
+      "step": 11011
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 20.807668685913086,
+      "learning_rate": 1.0786930510814544e-05,
+      "loss": 1.5155,
+      "step": 11012
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.328977584838867,
+      "learning_rate": 1.0786093795757854e-05,
+      "loss": 0.8445,
+      "step": 11013
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 37.37929153442383,
+      "learning_rate": 1.0785257080701168e-05,
+      "loss": 4.7872,
+      "step": 11014
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.015878677368164,
+      "learning_rate": 1.0784420365644482e-05,
+      "loss": 2.3438,
+      "step": 11015
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 26.955360412597656,
+      "learning_rate": 1.0783583650587792e-05,
+      "loss": 1.3095,
+      "step": 11016
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 23.8460693359375,
+      "learning_rate": 1.0782746935531106e-05,
+      "loss": 1.2315,
+      "step": 11017
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 13.200667381286621,
+      "learning_rate": 1.078191022047442e-05,
+      "loss": 2.8354,
+      "step": 11018
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 29.953876495361328,
+      "learning_rate": 1.0781073505417731e-05,
+      "loss": 0.7924,
+      "step": 11019
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 5.776995658874512,
+      "learning_rate": 1.0780236790361043e-05,
+      "loss": 1.5278,
+      "step": 11020
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.799798011779785,
+      "learning_rate": 1.0779400075304357e-05,
+      "loss": 2.6669,
+      "step": 11021
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.158820152282715,
+      "learning_rate": 1.0778563360247669e-05,
+      "loss": 1.147,
+      "step": 11022
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 5.1400532722473145,
+      "learning_rate": 1.077772664519098e-05,
+      "loss": 1.771,
+      "step": 11023
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 11.69130802154541,
+      "learning_rate": 1.0776889930134293e-05,
+      "loss": 0.685,
+      "step": 11024
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 25.547908782958984,
+      "learning_rate": 1.0776053215077606e-05,
+      "loss": 1.9566,
+      "step": 11025
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.598379135131836,
+      "learning_rate": 1.077521650002092e-05,
+      "loss": 0.4463,
+      "step": 11026
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 25.001876831054688,
+      "learning_rate": 1.077437978496423e-05,
+      "loss": 0.8894,
+      "step": 11027
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 12.50566577911377,
+      "learning_rate": 1.0773543069907544e-05,
+      "loss": 0.563,
+      "step": 11028
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 11.333840370178223,
+      "learning_rate": 1.0772706354850857e-05,
+      "loss": 0.8401,
+      "step": 11029
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 49.985897064208984,
+      "learning_rate": 1.0771869639794168e-05,
+      "loss": 1.3542,
+      "step": 11030
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 20.68662452697754,
+      "learning_rate": 1.0771032924737481e-05,
+      "loss": 1.6074,
+      "step": 11031
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 5.764721393585205,
+      "learning_rate": 1.0770196209680795e-05,
+      "loss": 1.8161,
+      "step": 11032
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 21.266937255859375,
+      "learning_rate": 1.0769359494624107e-05,
+      "loss": 1.4947,
+      "step": 11033
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 22.41715431213379,
+      "learning_rate": 1.0768522779567419e-05,
+      "loss": 2.498,
+      "step": 11034
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 45.06110382080078,
+      "learning_rate": 1.0767686064510731e-05,
+      "loss": 3.1562,
+      "step": 11035
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.207090377807617,
+      "learning_rate": 1.0766849349454045e-05,
+      "loss": 0.7283,
+      "step": 11036
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 32.75028610229492,
+      "learning_rate": 1.0766012634397357e-05,
+      "loss": 0.9633,
+      "step": 11037
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 11.14129638671875,
+      "learning_rate": 1.0765175919340668e-05,
+      "loss": 1.9259,
+      "step": 11038
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.889047622680664,
+      "learning_rate": 1.0764339204283982e-05,
+      "loss": 0.4887,
+      "step": 11039
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 16.445697784423828,
+      "learning_rate": 1.0763502489227296e-05,
+      "loss": 0.8201,
+      "step": 11040
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 21.291526794433594,
+      "learning_rate": 1.0762665774170606e-05,
+      "loss": 1.5882,
+      "step": 11041
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 6.2344465255737305,
+      "learning_rate": 1.076182905911392e-05,
+      "loss": 0.3961,
+      "step": 11042
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 17.333131790161133,
+      "learning_rate": 1.0760992344057233e-05,
+      "loss": 1.5523,
+      "step": 11043
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 5.335277557373047,
+      "learning_rate": 1.0760155629000544e-05,
+      "loss": 1.0519,
+      "step": 11044
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 7.969385147094727,
+      "learning_rate": 1.0759318913943857e-05,
+      "loss": 0.338,
+      "step": 11045
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.689923286437988,
+      "learning_rate": 1.0758482198887171e-05,
+      "loss": 0.7622,
+      "step": 11046
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 17.758501052856445,
+      "learning_rate": 1.0757645483830483e-05,
+      "loss": 1.8925,
+      "step": 11047
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.584065437316895,
+      "learning_rate": 1.0756808768773795e-05,
+      "loss": 0.677,
+      "step": 11048
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 11.227311134338379,
+      "learning_rate": 1.0755972053717107e-05,
+      "loss": 2.2452,
+      "step": 11049
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 31.221769332885742,
+      "learning_rate": 1.075513533866042e-05,
+      "loss": 0.7633,
+      "step": 11050
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 20.232498168945312,
+      "learning_rate": 1.0754298623603732e-05,
+      "loss": 2.8231,
+      "step": 11051
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 11.236428260803223,
+      "learning_rate": 1.0753461908547044e-05,
+      "loss": 1.6659,
+      "step": 11052
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 5.119511604309082,
+      "learning_rate": 1.0752625193490358e-05,
+      "loss": 2.6232,
+      "step": 11053
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 27.271800994873047,
+      "learning_rate": 1.0751788478433672e-05,
+      "loss": 1.7496,
+      "step": 11054
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 33.42025375366211,
+      "learning_rate": 1.0750951763376982e-05,
+      "loss": 0.9294,
+      "step": 11055
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 41.25328063964844,
+      "learning_rate": 1.0750115048320296e-05,
+      "loss": 2.0432,
+      "step": 11056
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.878971099853516,
+      "learning_rate": 1.074927833326361e-05,
+      "loss": 0.5169,
+      "step": 11057
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 14.663554191589355,
+      "learning_rate": 1.074844161820692e-05,
+      "loss": 1.2426,
+      "step": 11058
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 13.726330757141113,
+      "learning_rate": 1.0747604903150233e-05,
+      "loss": 1.5593,
+      "step": 11059
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 17.34848976135254,
+      "learning_rate": 1.0746768188093547e-05,
+      "loss": 0.9445,
+      "step": 11060
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 13.294726371765137,
+      "learning_rate": 1.0745931473036859e-05,
+      "loss": 2.4288,
+      "step": 11061
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.532093048095703,
+      "learning_rate": 1.074509475798017e-05,
+      "loss": 2.0992,
+      "step": 11062
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 18.581283569335938,
+      "learning_rate": 1.0744258042923483e-05,
+      "loss": 3.5613,
+      "step": 11063
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 13.983625411987305,
+      "learning_rate": 1.0743421327866796e-05,
+      "loss": 1.7882,
+      "step": 11064
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 15.763188362121582,
+      "learning_rate": 1.0742584612810108e-05,
+      "loss": 1.6749,
+      "step": 11065
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 13.039581298828125,
+      "learning_rate": 1.074174789775342e-05,
+      "loss": 2.1293,
+      "step": 11066
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.097599983215332,
+      "learning_rate": 1.0740911182696734e-05,
+      "loss": 0.3257,
+      "step": 11067
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.246635437011719,
+      "learning_rate": 1.0740074467640047e-05,
+      "loss": 1.0761,
+      "step": 11068
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 18.120683670043945,
+      "learning_rate": 1.0739237752583358e-05,
+      "loss": 2.1921,
+      "step": 11069
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.108755111694336,
+      "learning_rate": 1.0738401037526671e-05,
+      "loss": 1.1033,
+      "step": 11070
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 20.601940155029297,
+      "learning_rate": 1.0737564322469985e-05,
+      "loss": 0.8256,
+      "step": 11071
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 13.79460334777832,
+      "learning_rate": 1.0736727607413295e-05,
+      "loss": 1.4347,
+      "step": 11072
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 25.632383346557617,
+      "learning_rate": 1.0735890892356609e-05,
+      "loss": 1.1222,
+      "step": 11073
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.9022481441497803,
+      "learning_rate": 1.0735054177299923e-05,
+      "loss": 0.0795,
+      "step": 11074
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.55939769744873,
+      "learning_rate": 1.0734217462243235e-05,
+      "loss": 2.9203,
+      "step": 11075
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 41.440025329589844,
+      "learning_rate": 1.0733380747186546e-05,
+      "loss": 2.0544,
+      "step": 11076
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.9340813159942627,
+      "learning_rate": 1.0732544032129858e-05,
+      "loss": 0.0388,
+      "step": 11077
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 19.389799118041992,
+      "learning_rate": 1.0731707317073172e-05,
+      "loss": 0.9938,
+      "step": 11078
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.55417251586914,
+      "learning_rate": 1.0730870602016484e-05,
+      "loss": 1.8281,
+      "step": 11079
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.094369888305664,
+      "learning_rate": 1.0730033886959796e-05,
+      "loss": 0.5223,
+      "step": 11080
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 12.736441612243652,
+      "learning_rate": 1.072919717190311e-05,
+      "loss": 0.9671,
+      "step": 11081
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 37.989707946777344,
+      "learning_rate": 1.0728360456846423e-05,
+      "loss": 1.6481,
+      "step": 11082
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.87106704711914,
+      "learning_rate": 1.0727523741789734e-05,
+      "loss": 1.3616,
+      "step": 11083
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 14.872847557067871,
+      "learning_rate": 1.0726687026733047e-05,
+      "loss": 0.4853,
+      "step": 11084
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 36.0228385925293,
+      "learning_rate": 1.072585031167636e-05,
+      "loss": 1.675,
+      "step": 11085
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 14.692157745361328,
+      "learning_rate": 1.0725013596619671e-05,
+      "loss": 1.6181,
+      "step": 11086
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 7.602644920349121,
+      "learning_rate": 1.0724176881562985e-05,
+      "loss": 0.9902,
+      "step": 11087
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 51.961585998535156,
+      "learning_rate": 1.0723340166506297e-05,
+      "loss": 1.2465,
+      "step": 11088
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 20.320905685424805,
+      "learning_rate": 1.072250345144961e-05,
+      "loss": 2.1883,
+      "step": 11089
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.149676322937012,
+      "learning_rate": 1.0721666736392922e-05,
+      "loss": 0.6142,
+      "step": 11090
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 14.76880931854248,
+      "learning_rate": 1.0720830021336234e-05,
+      "loss": 1.689,
+      "step": 11091
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 13.005758285522461,
+      "learning_rate": 1.0719993306279548e-05,
+      "loss": 1.1633,
+      "step": 11092
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 34.54526901245117,
+      "learning_rate": 1.071915659122286e-05,
+      "loss": 1.3603,
+      "step": 11093
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 54.09306335449219,
+      "learning_rate": 1.0718319876166172e-05,
+      "loss": 1.089,
+      "step": 11094
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 13.492703437805176,
+      "learning_rate": 1.0717483161109485e-05,
+      "loss": 0.705,
+      "step": 11095
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.318517684936523,
+      "learning_rate": 1.0716646446052799e-05,
+      "loss": 0.9415,
+      "step": 11096
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 8.012618064880371,
+      "learning_rate": 1.071580973099611e-05,
+      "loss": 0.3978,
+      "step": 11097
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.157938003540039,
+      "learning_rate": 1.0714973015939423e-05,
+      "loss": 0.9549,
+      "step": 11098
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 15.60629940032959,
+      "learning_rate": 1.0714136300882737e-05,
+      "loss": 1.1989,
+      "step": 11099
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 11.698266983032227,
+      "learning_rate": 1.0713299585826047e-05,
+      "loss": 0.4457,
+      "step": 11100
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.133544921875,
+      "learning_rate": 1.071246287076936e-05,
+      "loss": 0.5908,
+      "step": 11101
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 19.813976287841797,
+      "learning_rate": 1.0711626155712673e-05,
+      "loss": 1.8152,
+      "step": 11102
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 22.051944732666016,
+      "learning_rate": 1.0710789440655986e-05,
+      "loss": 1.7517,
+      "step": 11103
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 163.89105224609375,
+      "learning_rate": 1.0709952725599298e-05,
+      "loss": 1.6949,
+      "step": 11104
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 6.643505096435547,
+      "learning_rate": 1.070911601054261e-05,
+      "loss": 0.0331,
+      "step": 11105
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 7.673321723937988,
+      "learning_rate": 1.0708279295485924e-05,
+      "loss": 1.1346,
+      "step": 11106
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 25.48453140258789,
+      "learning_rate": 1.0707442580429234e-05,
+      "loss": 1.3993,
+      "step": 11107
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 54.76433563232422,
+      "learning_rate": 1.0706605865372548e-05,
+      "loss": 1.6626,
+      "step": 11108
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.953798294067383,
+      "learning_rate": 1.0705769150315861e-05,
+      "loss": 2.0811,
+      "step": 11109
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 43.57255554199219,
+      "learning_rate": 1.0704932435259175e-05,
+      "loss": 2.039,
+      "step": 11110
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 6.343830108642578,
+      "learning_rate": 1.0704095720202485e-05,
+      "loss": 0.3893,
+      "step": 11111
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 12.723817825317383,
+      "learning_rate": 1.0703259005145799e-05,
+      "loss": 1.3809,
+      "step": 11112
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 10.008481979370117,
+      "learning_rate": 1.0702422290089113e-05,
+      "loss": 0.3705,
+      "step": 11113
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 16.619836807250977,
+      "learning_rate": 1.0701585575032423e-05,
+      "loss": 0.363,
+      "step": 11114
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 29.2028865814209,
+      "learning_rate": 1.0700748859975736e-05,
+      "loss": 1.9747,
+      "step": 11115
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 12.458796501159668,
+      "learning_rate": 1.0699912144919048e-05,
+      "loss": 1.2696,
+      "step": 11116
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.526585102081299,
+      "learning_rate": 1.0699075429862362e-05,
+      "loss": 0.1859,
+      "step": 11117
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.33280086517334,
+      "learning_rate": 1.0698238714805674e-05,
+      "loss": 1.7716,
+      "step": 11118
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.18589973449707,
+      "learning_rate": 1.0697401999748986e-05,
+      "loss": 1.2223,
+      "step": 11119
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.9433746337890625,
+      "learning_rate": 1.06965652846923e-05,
+      "loss": 1.883,
+      "step": 11120
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 6.184037208557129,
+      "learning_rate": 1.069572856963561e-05,
+      "loss": 0.7084,
+      "step": 11121
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 5.708268165588379,
+      "learning_rate": 1.0694891854578923e-05,
+      "loss": 0.48,
+      "step": 11122
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 21.270519256591797,
+      "learning_rate": 1.0694055139522237e-05,
+      "loss": 1.8448,
+      "step": 11123
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 26.80547332763672,
+      "learning_rate": 1.069321842446555e-05,
+      "loss": 2.0975,
+      "step": 11124
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.265152931213379,
+      "learning_rate": 1.0692381709408861e-05,
+      "loss": 2.047,
+      "step": 11125
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.684669494628906,
+      "learning_rate": 1.0691544994352175e-05,
+      "loss": 1.9874,
+      "step": 11126
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 6.357775688171387,
+      "learning_rate": 1.0690708279295487e-05,
+      "loss": 0.3144,
+      "step": 11127
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 6.835781097412109,
+      "learning_rate": 1.0689871564238799e-05,
+      "loss": 0.5789,
+      "step": 11128
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 29.526168823242188,
+      "learning_rate": 1.0689034849182112e-05,
+      "loss": 1.5356,
+      "step": 11129
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 14.135659217834473,
+      "learning_rate": 1.0688198134125424e-05,
+      "loss": 0.3501,
+      "step": 11130
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 41.2221565246582,
+      "learning_rate": 1.0687361419068738e-05,
+      "loss": 2.5301,
+      "step": 11131
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 25.79400634765625,
+      "learning_rate": 1.068652470401205e-05,
+      "loss": 3.3588,
+      "step": 11132
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.162888526916504,
+      "learning_rate": 1.0685687988955362e-05,
+      "loss": 0.6071,
+      "step": 11133
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 42.55033493041992,
+      "learning_rate": 1.0684851273898675e-05,
+      "loss": 2.1959,
+      "step": 11134
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 4.78655481338501,
+      "learning_rate": 1.0684014558841986e-05,
+      "loss": 0.3336,
+      "step": 11135
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 28.380346298217773,
+      "learning_rate": 1.06831778437853e-05,
+      "loss": 1.1981,
+      "step": 11136
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 25.562061309814453,
+      "learning_rate": 1.0682341128728613e-05,
+      "loss": 0.5148,
+      "step": 11137
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 12.578266143798828,
+      "learning_rate": 1.0681504413671927e-05,
+      "loss": 2.066,
+      "step": 11138
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 9.793967247009277,
+      "learning_rate": 1.0680667698615237e-05,
+      "loss": 1.0688,
+      "step": 11139
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.32377815246582,
+      "learning_rate": 1.067983098355855e-05,
+      "loss": 1.2779,
+      "step": 11140
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 9.000517845153809,
+      "learning_rate": 1.0678994268501862e-05,
+      "loss": 1.5666,
+      "step": 11141
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 8.126484870910645,
+      "learning_rate": 1.0678157553445174e-05,
+      "loss": 1.2407,
+      "step": 11142
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 16.945219039916992,
+      "learning_rate": 1.0677320838388488e-05,
+      "loss": 2.471,
+      "step": 11143
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.179367542266846,
+      "learning_rate": 1.06764841233318e-05,
+      "loss": 0.2999,
+      "step": 11144
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 18.401262283325195,
+      "learning_rate": 1.0675647408275114e-05,
+      "loss": 1.4733,
+      "step": 11145
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 13.894248008728027,
+      "learning_rate": 1.0674810693218424e-05,
+      "loss": 1.1163,
+      "step": 11146
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 11.486808776855469,
+      "learning_rate": 1.0673973978161738e-05,
+      "loss": 2.0351,
+      "step": 11147
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 12.675158500671387,
+      "learning_rate": 1.0673137263105051e-05,
+      "loss": 1.801,
+      "step": 11148
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 18.01869773864746,
+      "learning_rate": 1.0672300548048362e-05,
+      "loss": 1.8798,
+      "step": 11149
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 26.31698989868164,
+      "learning_rate": 1.0671463832991675e-05,
+      "loss": 1.3533,
+      "step": 11150
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 21.525146484375,
+      "learning_rate": 1.0670627117934989e-05,
+      "loss": 1.5077,
+      "step": 11151
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 13.209208488464355,
+      "learning_rate": 1.0669790402878302e-05,
+      "loss": 1.3362,
+      "step": 11152
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 18.235010147094727,
+      "learning_rate": 1.0668953687821613e-05,
+      "loss": 2.3339,
+      "step": 11153
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.60266375541687,
+      "learning_rate": 1.0668116972764926e-05,
+      "loss": 0.0966,
+      "step": 11154
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 5.538740158081055,
+      "learning_rate": 1.0667280257708238e-05,
+      "loss": 0.6292,
+      "step": 11155
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.977618217468262,
+      "learning_rate": 1.066644354265155e-05,
+      "loss": 0.8987,
+      "step": 11156
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.504702091217041,
+      "learning_rate": 1.0665606827594864e-05,
+      "loss": 1.1534,
+      "step": 11157
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 14.682522773742676,
+      "learning_rate": 1.0664770112538176e-05,
+      "loss": 0.4943,
+      "step": 11158
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 28.776384353637695,
+      "learning_rate": 1.066393339748149e-05,
+      "loss": 3.0167,
+      "step": 11159
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 19.993257522583008,
+      "learning_rate": 1.06630966824248e-05,
+      "loss": 2.2105,
+      "step": 11160
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.3054468631744385,
+      "learning_rate": 1.0662259967368113e-05,
+      "loss": 0.3531,
+      "step": 11161
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 18.72955322265625,
+      "learning_rate": 1.0661423252311427e-05,
+      "loss": 2.1389,
+      "step": 11162
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 40.49760818481445,
+      "learning_rate": 1.0660586537254737e-05,
+      "loss": 2.056,
+      "step": 11163
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 22.261398315429688,
+      "learning_rate": 1.0659749822198051e-05,
+      "loss": 2.8086,
+      "step": 11164
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.372966766357422,
+      "learning_rate": 1.0658913107141365e-05,
+      "loss": 0.9135,
+      "step": 11165
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 51.850547790527344,
+      "learning_rate": 1.0658076392084678e-05,
+      "loss": 1.1674,
+      "step": 11166
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 14.914164543151855,
+      "learning_rate": 1.0657239677027989e-05,
+      "loss": 0.8428,
+      "step": 11167
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 11.955036163330078,
+      "learning_rate": 1.0656402961971302e-05,
+      "loss": 1.8275,
+      "step": 11168
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.5064961910247803,
+      "learning_rate": 1.0655566246914614e-05,
+      "loss": 0.1438,
+      "step": 11169
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 26.521604537963867,
+      "learning_rate": 1.0654729531857926e-05,
+      "loss": 0.3152,
+      "step": 11170
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 46.38266372680664,
+      "learning_rate": 1.065389281680124e-05,
+      "loss": 1.2401,
+      "step": 11171
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 13.363889694213867,
+      "learning_rate": 1.0653056101744552e-05,
+      "loss": 1.3346,
+      "step": 11172
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 52.94842529296875,
+      "learning_rate": 1.0652219386687865e-05,
+      "loss": 1.834,
+      "step": 11173
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 101.2232666015625,
+      "learning_rate": 1.0651382671631176e-05,
+      "loss": 1.8024,
+      "step": 11174
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.6217756271362305,
+      "learning_rate": 1.065054595657449e-05,
+      "loss": 0.08,
+      "step": 11175
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 9.53453254699707,
+      "learning_rate": 1.0649709241517803e-05,
+      "loss": 0.6623,
+      "step": 11176
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 38.03897476196289,
+      "learning_rate": 1.0648872526461113e-05,
+      "loss": 1.3396,
+      "step": 11177
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 18.410459518432617,
+      "learning_rate": 1.0648035811404427e-05,
+      "loss": 1.9926,
+      "step": 11178
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 17.08340072631836,
+      "learning_rate": 1.064719909634774e-05,
+      "loss": 1.7154,
+      "step": 11179
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 9.814212799072266,
+      "learning_rate": 1.0646362381291052e-05,
+      "loss": 1.2624,
+      "step": 11180
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 20.458986282348633,
+      "learning_rate": 1.0645525666234364e-05,
+      "loss": 1.3171,
+      "step": 11181
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 22.185209274291992,
+      "learning_rate": 1.0644688951177678e-05,
+      "loss": 2.7426,
+      "step": 11182
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.910929679870605,
+      "learning_rate": 1.064385223612099e-05,
+      "loss": 2.198,
+      "step": 11183
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 22.389389038085938,
+      "learning_rate": 1.0643015521064302e-05,
+      "loss": 1.3108,
+      "step": 11184
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 11.566834449768066,
+      "learning_rate": 1.0642178806007616e-05,
+      "loss": 0.5679,
+      "step": 11185
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 6.975268363952637,
+      "learning_rate": 1.0641342090950928e-05,
+      "loss": 1.1047,
+      "step": 11186
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 143.4335174560547,
+      "learning_rate": 1.0640505375894241e-05,
+      "loss": 1.254,
+      "step": 11187
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 14.49937915802002,
+      "learning_rate": 1.0639668660837551e-05,
+      "loss": 1.3799,
+      "step": 11188
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 11.432032585144043,
+      "learning_rate": 1.0638831945780865e-05,
+      "loss": 0.7361,
+      "step": 11189
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 4.690389156341553,
+      "learning_rate": 1.0637995230724179e-05,
+      "loss": 1.2973,
+      "step": 11190
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 6.342746257781982,
+      "learning_rate": 1.0637158515667489e-05,
+      "loss": 0.4781,
+      "step": 11191
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 24.245380401611328,
+      "learning_rate": 1.0636321800610803e-05,
+      "loss": 2.3932,
+      "step": 11192
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 19.726337432861328,
+      "learning_rate": 1.0635485085554116e-05,
+      "loss": 1.4064,
+      "step": 11193
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 9.738777160644531,
+      "learning_rate": 1.0634648370497428e-05,
+      "loss": 0.4974,
+      "step": 11194
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 16.016401290893555,
+      "learning_rate": 1.063381165544074e-05,
+      "loss": 1.3654,
+      "step": 11195
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 29.034860610961914,
+      "learning_rate": 1.0632974940384054e-05,
+      "loss": 1.8016,
+      "step": 11196
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.642669677734375,
+      "learning_rate": 1.0632138225327366e-05,
+      "loss": 1.2493,
+      "step": 11197
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 18.040664672851562,
+      "learning_rate": 1.0631301510270678e-05,
+      "loss": 1.14,
+      "step": 11198
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 12.934800148010254,
+      "learning_rate": 1.063046479521399e-05,
+      "loss": 0.777,
+      "step": 11199
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 39.06425094604492,
+      "learning_rate": 1.0629628080157303e-05,
+      "loss": 2.6454,
+      "step": 11200
+    },
+    {
+      "epoch": 1.41,
+      "eval_loss": 0.0833502933382988,
+      "eval_runtime": 94.902,
+      "eval_samples_per_second": 37.323,
+      "eval_steps_per_second": 37.323,
+      "step": 11200
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 24.45269012451172,
+      "learning_rate": 1.0628791365100617e-05,
+      "loss": 2.3683,
+      "step": 11201
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 12.949420928955078,
+      "learning_rate": 1.0627954650043927e-05,
+      "loss": 1.3383,
+      "step": 11202
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 5.550119400024414,
+      "learning_rate": 1.0627117934987241e-05,
+      "loss": 0.2897,
+      "step": 11203
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 11.11974048614502,
+      "learning_rate": 1.0626281219930555e-05,
+      "loss": 1.0486,
+      "step": 11204
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 11.411757469177246,
+      "learning_rate": 1.0625444504873865e-05,
+      "loss": 1.0877,
+      "step": 11205
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8041766285896301,
+      "learning_rate": 1.0624607789817179e-05,
+      "loss": 0.014,
+      "step": 11206
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.608766555786133,
+      "learning_rate": 1.0623771074760492e-05,
+      "loss": 1.3749,
+      "step": 11207
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 29.028156280517578,
+      "learning_rate": 1.0622934359703804e-05,
+      "loss": 1.3058,
+      "step": 11208
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 14.396626472473145,
+      "learning_rate": 1.0622097644647116e-05,
+      "loss": 0.8126,
+      "step": 11209
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 11.106209754943848,
+      "learning_rate": 1.062126092959043e-05,
+      "loss": 0.8738,
+      "step": 11210
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 6.783996105194092,
+      "learning_rate": 1.0620424214533742e-05,
+      "loss": 0.5723,
+      "step": 11211
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 126.74786376953125,
+      "learning_rate": 1.0619587499477054e-05,
+      "loss": 0.8913,
+      "step": 11212
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 14.129014015197754,
+      "learning_rate": 1.0618750784420366e-05,
+      "loss": 1.6854,
+      "step": 11213
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 56.18551254272461,
+      "learning_rate": 1.061791406936368e-05,
+      "loss": 3.4491,
+      "step": 11214
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.337648391723633,
+      "learning_rate": 1.0617077354306993e-05,
+      "loss": 0.7548,
+      "step": 11215
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 24.816200256347656,
+      "learning_rate": 1.0616240639250303e-05,
+      "loss": 1.7799,
+      "step": 11216
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 15.044414520263672,
+      "learning_rate": 1.0615403924193617e-05,
+      "loss": 0.9788,
+      "step": 11217
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 30.718883514404297,
+      "learning_rate": 1.061456720913693e-05,
+      "loss": 1.8779,
+      "step": 11218
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 32.816646575927734,
+      "learning_rate": 1.061373049408024e-05,
+      "loss": 1.4192,
+      "step": 11219
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 12.458331108093262,
+      "learning_rate": 1.0612893779023554e-05,
+      "loss": 0.6344,
+      "step": 11220
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 18.22350311279297,
+      "learning_rate": 1.0612057063966868e-05,
+      "loss": 0.9284,
+      "step": 11221
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 16.123369216918945,
+      "learning_rate": 1.061122034891018e-05,
+      "loss": 0.6967,
+      "step": 11222
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 12.499308586120605,
+      "learning_rate": 1.0610383633853492e-05,
+      "loss": 1.1679,
+      "step": 11223
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 4.924980640411377,
+      "learning_rate": 1.0609546918796806e-05,
+      "loss": 0.9552,
+      "step": 11224
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 18.533519744873047,
+      "learning_rate": 1.0608710203740118e-05,
+      "loss": 1.2606,
+      "step": 11225
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 12.271876335144043,
+      "learning_rate": 1.060787348868343e-05,
+      "loss": 0.6614,
+      "step": 11226
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 31.398971557617188,
+      "learning_rate": 1.0607036773626741e-05,
+      "loss": 2.1384,
+      "step": 11227
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 22.711267471313477,
+      "learning_rate": 1.0606200058570055e-05,
+      "loss": 1.2964,
+      "step": 11228
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 6.177417278289795,
+      "learning_rate": 1.0605363343513369e-05,
+      "loss": 0.7655,
+      "step": 11229
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 6.442416667938232,
+      "learning_rate": 1.0604526628456679e-05,
+      "loss": 0.4298,
+      "step": 11230
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.38805866241455,
+      "learning_rate": 1.0603689913399993e-05,
+      "loss": 0.7203,
+      "step": 11231
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 11.661182403564453,
+      "learning_rate": 1.0602853198343306e-05,
+      "loss": 1.6595,
+      "step": 11232
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 16.503828048706055,
+      "learning_rate": 1.0602016483286617e-05,
+      "loss": 1.3584,
+      "step": 11233
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 10.20715618133545,
+      "learning_rate": 1.060117976822993e-05,
+      "loss": 0.533,
+      "step": 11234
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.280577659606934,
+      "learning_rate": 1.0600343053173244e-05,
+      "loss": 0.7534,
+      "step": 11235
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 35.52324676513672,
+      "learning_rate": 1.0599506338116556e-05,
+      "loss": 3.1192,
+      "step": 11236
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 48.30099868774414,
+      "learning_rate": 1.0598669623059868e-05,
+      "loss": 2.1817,
+      "step": 11237
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 10.228086471557617,
+      "learning_rate": 1.0597832908003181e-05,
+      "loss": 0.8891,
+      "step": 11238
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 17.013261795043945,
+      "learning_rate": 1.0596996192946493e-05,
+      "loss": 2.3246,
+      "step": 11239
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 75.33102416992188,
+      "learning_rate": 1.0596159477889805e-05,
+      "loss": 2.7242,
+      "step": 11240
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 20.065509796142578,
+      "learning_rate": 1.0595322762833117e-05,
+      "loss": 1.4334,
+      "step": 11241
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 28.460887908935547,
+      "learning_rate": 1.0594486047776431e-05,
+      "loss": 0.3438,
+      "step": 11242
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 15.307353019714355,
+      "learning_rate": 1.0593649332719743e-05,
+      "loss": 0.8457,
+      "step": 11243
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.002567291259766,
+      "learning_rate": 1.0592812617663055e-05,
+      "loss": 0.7748,
+      "step": 11244
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 5.05307674407959,
+      "learning_rate": 1.0591975902606368e-05,
+      "loss": 0.5225,
+      "step": 11245
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 30.22115135192871,
+      "learning_rate": 1.0591139187549682e-05,
+      "loss": 2.7615,
+      "step": 11246
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 18.640111923217773,
+      "learning_rate": 1.0590302472492992e-05,
+      "loss": 0.8867,
+      "step": 11247
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 12.942728996276855,
+      "learning_rate": 1.0589465757436306e-05,
+      "loss": 0.8576,
+      "step": 11248
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 7.473307132720947,
+      "learning_rate": 1.058862904237962e-05,
+      "loss": 0.7675,
+      "step": 11249
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 6.772233486175537,
+      "learning_rate": 1.058779232732293e-05,
+      "loss": 0.3126,
+      "step": 11250
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 28.95512580871582,
+      "learning_rate": 1.0586955612266244e-05,
+      "loss": 1.0426,
+      "step": 11251
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 7.875672817230225,
+      "learning_rate": 1.0586118897209556e-05,
+      "loss": 0.6801,
+      "step": 11252
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 7.985882759094238,
+      "learning_rate": 1.058528218215287e-05,
+      "loss": 1.1323,
+      "step": 11253
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.809475898742676,
+      "learning_rate": 1.0584445467096181e-05,
+      "loss": 1.0171,
+      "step": 11254
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 9.97216510772705,
+      "learning_rate": 1.0583608752039493e-05,
+      "loss": 0.4769,
+      "step": 11255
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 20.22585105895996,
+      "learning_rate": 1.0582772036982807e-05,
+      "loss": 0.7561,
+      "step": 11256
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 13.346476554870605,
+      "learning_rate": 1.0581935321926117e-05,
+      "loss": 1.3258,
+      "step": 11257
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 17.68963050842285,
+      "learning_rate": 1.058109860686943e-05,
+      "loss": 2.4481,
+      "step": 11258
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 15.243430137634277,
+      "learning_rate": 1.0580261891812744e-05,
+      "loss": 0.9186,
+      "step": 11259
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 10.015549659729004,
+      "learning_rate": 1.0579425176756058e-05,
+      "loss": 0.648,
+      "step": 11260
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 6.309828758239746,
+      "learning_rate": 1.0578588461699368e-05,
+      "loss": 0.7387,
+      "step": 11261
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 7.163145542144775,
+      "learning_rate": 1.0577751746642682e-05,
+      "loss": 0.3292,
+      "step": 11262
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 40.61732864379883,
+      "learning_rate": 1.0576915031585996e-05,
+      "loss": 2.8839,
+      "step": 11263
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 33.57455062866211,
+      "learning_rate": 1.0576078316529306e-05,
+      "loss": 1.0886,
+      "step": 11264
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 7.607451438903809,
+      "learning_rate": 1.057524160147262e-05,
+      "loss": 0.6742,
+      "step": 11265
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 7.639183521270752,
+      "learning_rate": 1.0574404886415931e-05,
+      "loss": 2.3329,
+      "step": 11266
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 6.122906684875488,
+      "learning_rate": 1.0573568171359245e-05,
+      "loss": 0.4457,
+      "step": 11267
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 11.732542037963867,
+      "learning_rate": 1.0572731456302557e-05,
+      "loss": 1.0847,
+      "step": 11268
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 18.822357177734375,
+      "learning_rate": 1.0571894741245869e-05,
+      "loss": 1.259,
+      "step": 11269
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 14.801311492919922,
+      "learning_rate": 1.0571058026189183e-05,
+      "loss": 1.6006,
+      "step": 11270
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 12.911343574523926,
+      "learning_rate": 1.0570221311132493e-05,
+      "loss": 1.0157,
+      "step": 11271
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.495565414428711,
+      "learning_rate": 1.0569384596075806e-05,
+      "loss": 0.7486,
+      "step": 11272
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 15.057435989379883,
+      "learning_rate": 1.056854788101912e-05,
+      "loss": 1.025,
+      "step": 11273
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 14.769805908203125,
+      "learning_rate": 1.0567711165962434e-05,
+      "loss": 1.022,
+      "step": 11274
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 11.487122535705566,
+      "learning_rate": 1.0566874450905744e-05,
+      "loss": 0.6012,
+      "step": 11275
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 13.399752616882324,
+      "learning_rate": 1.0566037735849058e-05,
+      "loss": 1.006,
+      "step": 11276
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 17.783620834350586,
+      "learning_rate": 1.0565201020792371e-05,
+      "loss": 0.493,
+      "step": 11277
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 39.72786331176758,
+      "learning_rate": 1.0564364305735682e-05,
+      "loss": 1.5032,
+      "step": 11278
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.015669822692871,
+      "learning_rate": 1.0563527590678995e-05,
+      "loss": 1.5298,
+      "step": 11279
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 16.162492752075195,
+      "learning_rate": 1.0562690875622307e-05,
+      "loss": 1.7034,
+      "step": 11280
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 42.08264923095703,
+      "learning_rate": 1.0561854160565621e-05,
+      "loss": 3.0007,
+      "step": 11281
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 14.679417610168457,
+      "learning_rate": 1.0561017445508933e-05,
+      "loss": 1.5274,
+      "step": 11282
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 22.4543514251709,
+      "learning_rate": 1.0560180730452245e-05,
+      "loss": 1.086,
+      "step": 11283
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 19.550586700439453,
+      "learning_rate": 1.0559344015395558e-05,
+      "loss": 2.1566,
+      "step": 11284
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 14.509361267089844,
+      "learning_rate": 1.0558507300338869e-05,
+      "loss": 1.9627,
+      "step": 11285
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 37.67739486694336,
+      "learning_rate": 1.0557670585282182e-05,
+      "loss": 0.7337,
+      "step": 11286
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 13.477119445800781,
+      "learning_rate": 1.0556833870225496e-05,
+      "loss": 1.9903,
+      "step": 11287
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.830741882324219,
+      "learning_rate": 1.055599715516881e-05,
+      "loss": 0.3425,
+      "step": 11288
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 4.542238235473633,
+      "learning_rate": 1.055516044011212e-05,
+      "loss": 0.7427,
+      "step": 11289
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 4.7642998695373535,
+      "learning_rate": 1.0554323725055434e-05,
+      "loss": 0.3059,
+      "step": 11290
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 26.863927841186523,
+      "learning_rate": 1.0553487009998745e-05,
+      "loss": 2.5606,
+      "step": 11291
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.374013423919678,
+      "learning_rate": 1.0552650294942057e-05,
+      "loss": 0.6605,
+      "step": 11292
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 11.80407428741455,
+      "learning_rate": 1.0551813579885371e-05,
+      "loss": 1.5701,
+      "step": 11293
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 19.342987060546875,
+      "learning_rate": 1.0550976864828683e-05,
+      "loss": 1.057,
+      "step": 11294
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 16.646146774291992,
+      "learning_rate": 1.0550140149771997e-05,
+      "loss": 1.1462,
+      "step": 11295
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 6.888421058654785,
+      "learning_rate": 1.0549303434715309e-05,
+      "loss": 1.4076,
+      "step": 11296
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.04971170425415,
+      "learning_rate": 1.054846671965862e-05,
+      "loss": 1.6847,
+      "step": 11297
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 41.948150634765625,
+      "learning_rate": 1.0547630004601934e-05,
+      "loss": 1.7488,
+      "step": 11298
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 6.25320291519165,
+      "learning_rate": 1.0546793289545245e-05,
+      "loss": 1.3213,
+      "step": 11299
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 28.94004249572754,
+      "learning_rate": 1.0545956574488558e-05,
+      "loss": 1.1052,
+      "step": 11300
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 41.42261505126953,
+      "learning_rate": 1.0545119859431872e-05,
+      "loss": 1.3953,
+      "step": 11301
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 15.843208312988281,
+      "learning_rate": 1.0544283144375185e-05,
+      "loss": 0.9795,
+      "step": 11302
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 11.01613998413086,
+      "learning_rate": 1.0543446429318496e-05,
+      "loss": 2.2015,
+      "step": 11303
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 13.635912895202637,
+      "learning_rate": 1.054260971426181e-05,
+      "loss": 1.0478,
+      "step": 11304
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 22.0612850189209,
+      "learning_rate": 1.0541772999205121e-05,
+      "loss": 1.3575,
+      "step": 11305
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.714583396911621,
+      "learning_rate": 1.0540936284148433e-05,
+      "loss": 0.6664,
+      "step": 11306
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 20.06359100341797,
+      "learning_rate": 1.0540099569091747e-05,
+      "loss": 0.3482,
+      "step": 11307
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 15.595358848571777,
+      "learning_rate": 1.0539262854035059e-05,
+      "loss": 1.8518,
+      "step": 11308
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 60.89354705810547,
+      "learning_rate": 1.0538426138978373e-05,
+      "loss": 1.7748,
+      "step": 11309
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 11.46818733215332,
+      "learning_rate": 1.0537589423921683e-05,
+      "loss": 0.4397,
+      "step": 11310
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 22.049875259399414,
+      "learning_rate": 1.0536752708864996e-05,
+      "loss": 1.8015,
+      "step": 11311
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 11.30109977722168,
+      "learning_rate": 1.053591599380831e-05,
+      "loss": 0.3747,
+      "step": 11312
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 12.697015762329102,
+      "learning_rate": 1.053507927875162e-05,
+      "loss": 0.8116,
+      "step": 11313
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 14.568947792053223,
+      "learning_rate": 1.0534242563694934e-05,
+      "loss": 1.4176,
+      "step": 11314
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 20.654478073120117,
+      "learning_rate": 1.0533405848638248e-05,
+      "loss": 1.3716,
+      "step": 11315
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 119.02867889404297,
+      "learning_rate": 1.0532569133581561e-05,
+      "loss": 1.7289,
+      "step": 11316
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.667135238647461,
+      "learning_rate": 1.0531732418524872e-05,
+      "loss": 1.968,
+      "step": 11317
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 22.299379348754883,
+      "learning_rate": 1.0530895703468185e-05,
+      "loss": 1.5755,
+      "step": 11318
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 14.984415054321289,
+      "learning_rate": 1.0530058988411497e-05,
+      "loss": 0.8569,
+      "step": 11319
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 92.09161376953125,
+      "learning_rate": 1.0529222273354809e-05,
+      "loss": 3.5519,
+      "step": 11320
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 32.287445068359375,
+      "learning_rate": 1.0528385558298123e-05,
+      "loss": 1.8257,
+      "step": 11321
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 432.5163879394531,
+      "learning_rate": 1.0527548843241435e-05,
+      "loss": 1.5465,
+      "step": 11322
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 15.813030242919922,
+      "learning_rate": 1.0526712128184748e-05,
+      "loss": 0.9823,
+      "step": 11323
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 17.160110473632812,
+      "learning_rate": 1.0525875413128059e-05,
+      "loss": 2.3561,
+      "step": 11324
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 5.169161319732666,
+      "learning_rate": 1.0525038698071372e-05,
+      "loss": 0.2158,
+      "step": 11325
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 12.82089614868164,
+      "learning_rate": 1.0524201983014686e-05,
+      "loss": 1.0913,
+      "step": 11326
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 55.144737243652344,
+      "learning_rate": 1.0523365267957996e-05,
+      "loss": 1.4059,
+      "step": 11327
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 16.044841766357422,
+      "learning_rate": 1.052252855290131e-05,
+      "loss": 1.3942,
+      "step": 11328
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 12.018836975097656,
+      "learning_rate": 1.0521691837844623e-05,
+      "loss": 1.4321,
+      "step": 11329
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 15.334610939025879,
+      "learning_rate": 1.0520855122787937e-05,
+      "loss": 1.2405,
+      "step": 11330
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 11.532013893127441,
+      "learning_rate": 1.0520018407731247e-05,
+      "loss": 0.6072,
+      "step": 11331
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 15.10476303100586,
+      "learning_rate": 1.0519181692674561e-05,
+      "loss": 0.8413,
+      "step": 11332
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 19.77190399169922,
+      "learning_rate": 1.0518344977617873e-05,
+      "loss": 0.9303,
+      "step": 11333
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 11.441058158874512,
+      "learning_rate": 1.0517508262561185e-05,
+      "loss": 1.1298,
+      "step": 11334
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 6.324008941650391,
+      "learning_rate": 1.0516671547504499e-05,
+      "loss": 1.5995,
+      "step": 11335
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 5.152180194854736,
+      "learning_rate": 1.051583483244781e-05,
+      "loss": 0.3173,
+      "step": 11336
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.624914646148682,
+      "learning_rate": 1.0514998117391124e-05,
+      "loss": 0.7505,
+      "step": 11337
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 16.759044647216797,
+      "learning_rate": 1.0514161402334434e-05,
+      "loss": 1.0097,
+      "step": 11338
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 18.782703399658203,
+      "learning_rate": 1.0513324687277748e-05,
+      "loss": 1.6375,
+      "step": 11339
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 4.0037994384765625,
+      "learning_rate": 1.0512487972221062e-05,
+      "loss": 0.1343,
+      "step": 11340
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 19.290443420410156,
+      "learning_rate": 1.0511651257164372e-05,
+      "loss": 1.1627,
+      "step": 11341
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.35206413269043,
+      "learning_rate": 1.0510814542107686e-05,
+      "loss": 1.6862,
+      "step": 11342
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 73.9336166381836,
+      "learning_rate": 1.0509977827051e-05,
+      "loss": 1.1326,
+      "step": 11343
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.469733238220215,
+      "learning_rate": 1.0509141111994311e-05,
+      "loss": 0.8905,
+      "step": 11344
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 41.667808532714844,
+      "learning_rate": 1.0508304396937623e-05,
+      "loss": 1.9519,
+      "step": 11345
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 6.640600681304932,
+      "learning_rate": 1.0507467681880937e-05,
+      "loss": 0.2494,
+      "step": 11346
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 13.451955795288086,
+      "learning_rate": 1.0506630966824249e-05,
+      "loss": 1.1055,
+      "step": 11347
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 27.924213409423828,
+      "learning_rate": 1.050579425176756e-05,
+      "loss": 1.7074,
+      "step": 11348
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.30042839050293,
+      "learning_rate": 1.0504957536710874e-05,
+      "loss": 0.8538,
+      "step": 11349
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.299258708953857,
+      "learning_rate": 1.0504120821654186e-05,
+      "loss": 1.7927,
+      "step": 11350
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 21.970739364624023,
+      "learning_rate": 1.05032841065975e-05,
+      "loss": 1.4202,
+      "step": 11351
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 26.46619415283203,
+      "learning_rate": 1.050244739154081e-05,
+      "loss": 2.9561,
+      "step": 11352
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 17.56070327758789,
+      "learning_rate": 1.0501610676484124e-05,
+      "loss": 1.0894,
+      "step": 11353
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 16.233503341674805,
+      "learning_rate": 1.0500773961427438e-05,
+      "loss": 1.1073,
+      "step": 11354
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.83210277557373,
+      "learning_rate": 1.0499937246370748e-05,
+      "loss": 1.1373,
+      "step": 11355
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 28.475740432739258,
+      "learning_rate": 1.0499100531314062e-05,
+      "loss": 1.1082,
+      "step": 11356
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.7533540725708,
+      "learning_rate": 1.0498263816257375e-05,
+      "loss": 1.259,
+      "step": 11357
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 124.90774536132812,
+      "learning_rate": 1.0497427101200687e-05,
+      "loss": 0.9487,
+      "step": 11358
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.858553886413574,
+      "learning_rate": 1.0496590386143999e-05,
+      "loss": 0.9905,
+      "step": 11359
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 22.065006256103516,
+      "learning_rate": 1.0495753671087313e-05,
+      "loss": 1.4661,
+      "step": 11360
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 12.489311218261719,
+      "learning_rate": 1.0494916956030625e-05,
+      "loss": 1.2853,
+      "step": 11361
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 15.446892738342285,
+      "learning_rate": 1.0494080240973937e-05,
+      "loss": 0.8781,
+      "step": 11362
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.910172462463379,
+      "learning_rate": 1.0493243525917249e-05,
+      "loss": 0.9655,
+      "step": 11363
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 6.618316173553467,
+      "learning_rate": 1.0492406810860562e-05,
+      "loss": 0.7508,
+      "step": 11364
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.195474624633789,
+      "learning_rate": 1.0491570095803876e-05,
+      "loss": 1.1058,
+      "step": 11365
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 26.93873405456543,
+      "learning_rate": 1.0490733380747186e-05,
+      "loss": 1.4308,
+      "step": 11366
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.582056045532227,
+      "learning_rate": 1.04898966656905e-05,
+      "loss": 1.334,
+      "step": 11367
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.099984645843506,
+      "learning_rate": 1.0489059950633813e-05,
+      "loss": 0.1451,
+      "step": 11368
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.376474380493164,
+      "learning_rate": 1.0488223235577124e-05,
+      "loss": 0.6118,
+      "step": 11369
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 15.648659706115723,
+      "learning_rate": 1.0487386520520437e-05,
+      "loss": 1.2294,
+      "step": 11370
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 44.50442123413086,
+      "learning_rate": 1.0486549805463751e-05,
+      "loss": 1.3816,
+      "step": 11371
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 28.336875915527344,
+      "learning_rate": 1.0485713090407063e-05,
+      "loss": 0.7956,
+      "step": 11372
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 6.759217262268066,
+      "learning_rate": 1.0484876375350375e-05,
+      "loss": 1.1031,
+      "step": 11373
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 16.726430892944336,
+      "learning_rate": 1.0484039660293689e-05,
+      "loss": 1.4283,
+      "step": 11374
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 12.317497253417969,
+      "learning_rate": 1.0483202945237e-05,
+      "loss": 1.1722,
+      "step": 11375
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 12.657833099365234,
+      "learning_rate": 1.0482366230180312e-05,
+      "loss": 0.4978,
+      "step": 11376
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 15.169766426086426,
+      "learning_rate": 1.0481529515123624e-05,
+      "loss": 1.3068,
+      "step": 11377
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 4.539684295654297,
+      "learning_rate": 1.0480692800066938e-05,
+      "loss": 0.3645,
+      "step": 11378
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 19.042705535888672,
+      "learning_rate": 1.0479856085010252e-05,
+      "loss": 1.455,
+      "step": 11379
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 7.281583786010742,
+      "learning_rate": 1.0479019369953562e-05,
+      "loss": 0.2026,
+      "step": 11380
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 17.259510040283203,
+      "learning_rate": 1.0478182654896876e-05,
+      "loss": 1.4423,
+      "step": 11381
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 15.83186149597168,
+      "learning_rate": 1.047734593984019e-05,
+      "loss": 1.578,
+      "step": 11382
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 7.90648889541626,
+      "learning_rate": 1.04765092247835e-05,
+      "loss": 1.1944,
+      "step": 11383
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 21.07403564453125,
+      "learning_rate": 1.0475672509726813e-05,
+      "loss": 1.6348,
+      "step": 11384
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 12.450182914733887,
+      "learning_rate": 1.0474835794670127e-05,
+      "loss": 2.2169,
+      "step": 11385
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 6.377939224243164,
+      "learning_rate": 1.0473999079613439e-05,
+      "loss": 0.4044,
+      "step": 11386
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.51230525970459,
+      "learning_rate": 1.047316236455675e-05,
+      "loss": 0.7925,
+      "step": 11387
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 56.4589958190918,
+      "learning_rate": 1.0472325649500064e-05,
+      "loss": 0.9871,
+      "step": 11388
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 26.939266204833984,
+      "learning_rate": 1.0471488934443376e-05,
+      "loss": 1.4693,
+      "step": 11389
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 18.43749237060547,
+      "learning_rate": 1.0470652219386688e-05,
+      "loss": 1.4306,
+      "step": 11390
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 28.045595169067383,
+      "learning_rate": 1.046981550433e-05,
+      "loss": 1.5361,
+      "step": 11391
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.949115753173828,
+      "learning_rate": 1.0468978789273314e-05,
+      "loss": 1.523,
+      "step": 11392
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 51.4660758972168,
+      "learning_rate": 1.0468142074216628e-05,
+      "loss": 2.6285,
+      "step": 11393
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.311626434326172,
+      "learning_rate": 1.0467305359159938e-05,
+      "loss": 1.2782,
+      "step": 11394
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 18.409934997558594,
+      "learning_rate": 1.0466468644103251e-05,
+      "loss": 1.2012,
+      "step": 11395
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 14.347053527832031,
+      "learning_rate": 1.0465631929046565e-05,
+      "loss": 1.9738,
+      "step": 11396
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 16.370105743408203,
+      "learning_rate": 1.0464795213989875e-05,
+      "loss": 0.9552,
+      "step": 11397
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 7.3930182456970215,
+      "learning_rate": 1.0463958498933189e-05,
+      "loss": 0.4915,
+      "step": 11398
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 15.962390899658203,
+      "learning_rate": 1.0463121783876503e-05,
+      "loss": 1.4147,
+      "step": 11399
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 42.52497482299805,
+      "learning_rate": 1.0462285068819815e-05,
+      "loss": 2.0916,
+      "step": 11400
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 34.833011627197266,
+      "learning_rate": 1.0461448353763127e-05,
+      "loss": 2.3191,
+      "step": 11401
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 19.962411880493164,
+      "learning_rate": 1.0460611638706439e-05,
+      "loss": 1.586,
+      "step": 11402
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.79418659210205,
+      "learning_rate": 1.0459774923649752e-05,
+      "loss": 0.8382,
+      "step": 11403
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 32.29881286621094,
+      "learning_rate": 1.0458938208593064e-05,
+      "loss": 1.0914,
+      "step": 11404
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 20.271190643310547,
+      "learning_rate": 1.0458101493536376e-05,
+      "loss": 1.0341,
+      "step": 11405
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 43.88380432128906,
+      "learning_rate": 1.045726477847969e-05,
+      "loss": 0.8645,
+      "step": 11406
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 15.907600402832031,
+      "learning_rate": 1.0456428063423003e-05,
+      "loss": 1.7109,
+      "step": 11407
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 12.382284164428711,
+      "learning_rate": 1.0455591348366314e-05,
+      "loss": 1.8094,
+      "step": 11408
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.19265079498291,
+      "learning_rate": 1.0454754633309627e-05,
+      "loss": 0.7449,
+      "step": 11409
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.750264167785645,
+      "learning_rate": 1.0453917918252941e-05,
+      "loss": 0.4312,
+      "step": 11410
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.6973702907562256,
+      "learning_rate": 1.0453081203196251e-05,
+      "loss": 0.1651,
+      "step": 11411
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 4.4212212562561035,
+      "learning_rate": 1.0452244488139565e-05,
+      "loss": 0.5799,
+      "step": 11412
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.936755180358887,
+      "learning_rate": 1.0451407773082879e-05,
+      "loss": 1.1073,
+      "step": 11413
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.346553802490234,
+      "learning_rate": 1.045057105802619e-05,
+      "loss": 1.0439,
+      "step": 11414
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.141072273254395,
+      "learning_rate": 1.0449734342969502e-05,
+      "loss": 0.6386,
+      "step": 11415
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 16.572446823120117,
+      "learning_rate": 1.0448897627912814e-05,
+      "loss": 0.7752,
+      "step": 11416
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.637643814086914,
+      "learning_rate": 1.0448060912856128e-05,
+      "loss": 0.4291,
+      "step": 11417
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 15.866935729980469,
+      "learning_rate": 1.044722419779944e-05,
+      "loss": 0.701,
+      "step": 11418
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 7.4652018547058105,
+      "learning_rate": 1.0446387482742752e-05,
+      "loss": 0.3932,
+      "step": 11419
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.4305419921875,
+      "learning_rate": 1.0445550767686066e-05,
+      "loss": 1.7346,
+      "step": 11420
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 30.32324981689453,
+      "learning_rate": 1.044471405262938e-05,
+      "loss": 1.9951,
+      "step": 11421
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.34090518951416,
+      "learning_rate": 1.044387733757269e-05,
+      "loss": 0.8096,
+      "step": 11422
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 5.801850318908691,
+      "learning_rate": 1.0443040622516003e-05,
+      "loss": 0.6392,
+      "step": 11423
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 15.56621265411377,
+      "learning_rate": 1.0442203907459317e-05,
+      "loss": 0.9069,
+      "step": 11424
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.361141204833984,
+      "learning_rate": 1.0441367192402627e-05,
+      "loss": 0.6857,
+      "step": 11425
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 11.270186424255371,
+      "learning_rate": 1.044053047734594e-05,
+      "loss": 1.0766,
+      "step": 11426
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 23.772802352905273,
+      "learning_rate": 1.0439693762289254e-05,
+      "loss": 1.2279,
+      "step": 11427
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 5.569838047027588,
+      "learning_rate": 1.0438857047232566e-05,
+      "loss": 0.1386,
+      "step": 11428
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 113.14727783203125,
+      "learning_rate": 1.0438020332175878e-05,
+      "loss": 1.6265,
+      "step": 11429
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.468873977661133,
+      "learning_rate": 1.043718361711919e-05,
+      "loss": 0.8927,
+      "step": 11430
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 21.616683959960938,
+      "learning_rate": 1.0436346902062504e-05,
+      "loss": 0.6899,
+      "step": 11431
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 14.02167797088623,
+      "learning_rate": 1.0435510187005816e-05,
+      "loss": 1.9948,
+      "step": 11432
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 8.072164535522461,
+      "learning_rate": 1.0434673471949128e-05,
+      "loss": 2.6065,
+      "step": 11433
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 31.57025718688965,
+      "learning_rate": 1.0433836756892441e-05,
+      "loss": 3.0612,
+      "step": 11434
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 5.659616470336914,
+      "learning_rate": 1.0433000041835755e-05,
+      "loss": 0.4403,
+      "step": 11435
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.711050987243652,
+      "learning_rate": 1.0432163326779065e-05,
+      "loss": 1.7349,
+      "step": 11436
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.188949584960938,
+      "learning_rate": 1.0431326611722379e-05,
+      "loss": 1.0138,
+      "step": 11437
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 13.183706283569336,
+      "learning_rate": 1.0430489896665693e-05,
+      "loss": 0.8548,
+      "step": 11438
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 7.962393283843994,
+      "learning_rate": 1.0429653181609003e-05,
+      "loss": 0.26,
+      "step": 11439
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 9.664677619934082,
+      "learning_rate": 1.0428816466552317e-05,
+      "loss": 0.7845,
+      "step": 11440
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 14.836224555969238,
+      "learning_rate": 1.042797975149563e-05,
+      "loss": 1.2228,
+      "step": 11441
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.407991409301758,
+      "learning_rate": 1.0427143036438942e-05,
+      "loss": 0.6964,
+      "step": 11442
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 32.23660659790039,
+      "learning_rate": 1.0426306321382254e-05,
+      "loss": 2.0007,
+      "step": 11443
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.350091934204102,
+      "learning_rate": 1.0425469606325566e-05,
+      "loss": 1.1657,
+      "step": 11444
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 6.385477542877197,
+      "learning_rate": 1.042463289126888e-05,
+      "loss": 1.4644,
+      "step": 11445
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.992371559143066,
+      "learning_rate": 1.0423796176212192e-05,
+      "loss": 0.805,
+      "step": 11446
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 16.413516998291016,
+      "learning_rate": 1.0422959461155504e-05,
+      "loss": 1.0079,
+      "step": 11447
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.926766395568848,
+      "learning_rate": 1.0422122746098817e-05,
+      "loss": 1.6126,
+      "step": 11448
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.546387672424316,
+      "learning_rate": 1.0421286031042131e-05,
+      "loss": 0.6899,
+      "step": 11449
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 66.18031311035156,
+      "learning_rate": 1.0420449315985441e-05,
+      "loss": 0.4709,
+      "step": 11450
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 24.744321823120117,
+      "learning_rate": 1.0419612600928755e-05,
+      "loss": 2.5164,
+      "step": 11451
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 25.81559181213379,
+      "learning_rate": 1.0418775885872068e-05,
+      "loss": 1.6183,
+      "step": 11452
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 16.952499389648438,
+      "learning_rate": 1.0417939170815379e-05,
+      "loss": 2.4686,
+      "step": 11453
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 4.533487796783447,
+      "learning_rate": 1.0417102455758692e-05,
+      "loss": 0.341,
+      "step": 11454
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 40.2689208984375,
+      "learning_rate": 1.0416265740702004e-05,
+      "loss": 1.9866,
+      "step": 11455
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.823920249938965,
+      "learning_rate": 1.0415429025645318e-05,
+      "loss": 1.1613,
+      "step": 11456
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.001120567321777,
+      "learning_rate": 1.041459231058863e-05,
+      "loss": 0.9691,
+      "step": 11457
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.59779167175293,
+      "learning_rate": 1.0413755595531942e-05,
+      "loss": 2.6485,
+      "step": 11458
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 51.74159240722656,
+      "learning_rate": 1.0412918880475256e-05,
+      "loss": 0.967,
+      "step": 11459
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 18.291950225830078,
+      "learning_rate": 1.0412082165418567e-05,
+      "loss": 1.0599,
+      "step": 11460
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 6.537196636199951,
+      "learning_rate": 1.041124545036188e-05,
+      "loss": 0.7339,
+      "step": 11461
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 62.63052749633789,
+      "learning_rate": 1.0410408735305193e-05,
+      "loss": 2.2704,
+      "step": 11462
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 9.313514709472656,
+      "learning_rate": 1.0409572020248507e-05,
+      "loss": 1.0498,
+      "step": 11463
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 21.436363220214844,
+      "learning_rate": 1.0408735305191817e-05,
+      "loss": 2.0415,
+      "step": 11464
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 24.772859573364258,
+      "learning_rate": 1.040789859013513e-05,
+      "loss": 1.6862,
+      "step": 11465
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 14.833982467651367,
+      "learning_rate": 1.0407061875078444e-05,
+      "loss": 3.0364,
+      "step": 11466
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 28.4583797454834,
+      "learning_rate": 1.0406225160021755e-05,
+      "loss": 1.5662,
+      "step": 11467
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 20.74558448791504,
+      "learning_rate": 1.0405388444965068e-05,
+      "loss": 2.1873,
+      "step": 11468
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 11.81302547454834,
+      "learning_rate": 1.040455172990838e-05,
+      "loss": 1.2035,
+      "step": 11469
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 21.872716903686523,
+      "learning_rate": 1.0403715014851694e-05,
+      "loss": 3.0724,
+      "step": 11470
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 31.094301223754883,
+      "learning_rate": 1.0402878299795006e-05,
+      "loss": 1.7347,
+      "step": 11471
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 9.362553596496582,
+      "learning_rate": 1.0402041584738318e-05,
+      "loss": 0.4224,
+      "step": 11472
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 16.24514389038086,
+      "learning_rate": 1.0401204869681631e-05,
+      "loss": 2.101,
+      "step": 11473
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.963959693908691,
+      "learning_rate": 1.0400368154624942e-05,
+      "loss": 0.8612,
+      "step": 11474
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 14.617683410644531,
+      "learning_rate": 1.0399531439568255e-05,
+      "loss": 1.9835,
+      "step": 11475
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 13.395903587341309,
+      "learning_rate": 1.0398694724511569e-05,
+      "loss": 1.2385,
+      "step": 11476
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.507243156433105,
+      "learning_rate": 1.0397858009454883e-05,
+      "loss": 2.0466,
+      "step": 11477
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.856938362121582,
+      "learning_rate": 1.0397021294398193e-05,
+      "loss": 0.7299,
+      "step": 11478
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.7521162033081055,
+      "learning_rate": 1.0396184579341506e-05,
+      "loss": 0.196,
+      "step": 11479
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 14.700419425964355,
+      "learning_rate": 1.039534786428482e-05,
+      "loss": 0.9978,
+      "step": 11480
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 7.636246681213379,
+      "learning_rate": 1.039451114922813e-05,
+      "loss": 1.7426,
+      "step": 11481
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.989336013793945,
+      "learning_rate": 1.0393674434171444e-05,
+      "loss": 0.488,
+      "step": 11482
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 27.758371353149414,
+      "learning_rate": 1.0392837719114756e-05,
+      "loss": 0.6158,
+      "step": 11483
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 78.2485122680664,
+      "learning_rate": 1.039200100405807e-05,
+      "loss": 2.7756,
+      "step": 11484
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 21.227731704711914,
+      "learning_rate": 1.0391164289001382e-05,
+      "loss": 1.2272,
+      "step": 11485
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 7.953176021575928,
+      "learning_rate": 1.0390327573944694e-05,
+      "loss": 0.5091,
+      "step": 11486
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.84647274017334,
+      "learning_rate": 1.0389490858888007e-05,
+      "loss": 1.1015,
+      "step": 11487
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.707669258117676,
+      "learning_rate": 1.0388654143831317e-05,
+      "loss": 0.7951,
+      "step": 11488
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 13.149401664733887,
+      "learning_rate": 1.0387817428774631e-05,
+      "loss": 1.0076,
+      "step": 11489
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 7.7071709632873535,
+      "learning_rate": 1.0386980713717945e-05,
+      "loss": 0.8714,
+      "step": 11490
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 10.428049087524414,
+      "learning_rate": 1.0386143998661258e-05,
+      "loss": 0.8757,
+      "step": 11491
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 17.69566535949707,
+      "learning_rate": 1.0385307283604569e-05,
+      "loss": 0.6535,
+      "step": 11492
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 33.63081359863281,
+      "learning_rate": 1.0384470568547882e-05,
+      "loss": 1.939,
+      "step": 11493
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 6.924600124359131,
+      "learning_rate": 1.0383633853491196e-05,
+      "loss": 0.4562,
+      "step": 11494
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 26.665849685668945,
+      "learning_rate": 1.0382797138434506e-05,
+      "loss": 2.8667,
+      "step": 11495
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 20.90088653564453,
+      "learning_rate": 1.038196042337782e-05,
+      "loss": 2.0433,
+      "step": 11496
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.360067367553711,
+      "learning_rate": 1.0381123708321132e-05,
+      "loss": 0.1102,
+      "step": 11497
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 63.318931579589844,
+      "learning_rate": 1.0380286993264445e-05,
+      "loss": 1.4949,
+      "step": 11498
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 9.656840324401855,
+      "learning_rate": 1.0379450278207757e-05,
+      "loss": 0.7261,
+      "step": 11499
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 17.443967819213867,
+      "learning_rate": 1.037861356315107e-05,
+      "loss": 1.3823,
+      "step": 11500
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 17.237642288208008,
+      "learning_rate": 1.0377776848094383e-05,
+      "loss": 0.8791,
+      "step": 11501
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 7.85336971282959,
+      "learning_rate": 1.0376940133037693e-05,
+      "loss": 0.6561,
+      "step": 11502
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 5.917194366455078,
+      "learning_rate": 1.0376103417981007e-05,
+      "loss": 0.2258,
+      "step": 11503
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 25.715185165405273,
+      "learning_rate": 1.037526670292432e-05,
+      "loss": 2.5617,
+      "step": 11504
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 4.817445278167725,
+      "learning_rate": 1.0374429987867634e-05,
+      "loss": 0.4108,
+      "step": 11505
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 103.81026458740234,
+      "learning_rate": 1.0373593272810945e-05,
+      "loss": 1.3861,
+      "step": 11506
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 9.29235553741455,
+      "learning_rate": 1.0372756557754258e-05,
+      "loss": 0.8288,
+      "step": 11507
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 30.044078826904297,
+      "learning_rate": 1.037191984269757e-05,
+      "loss": 0.7739,
+      "step": 11508
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 5.862963676452637,
+      "learning_rate": 1.0371083127640882e-05,
+      "loss": 0.4256,
+      "step": 11509
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 20.0323543548584,
+      "learning_rate": 1.0370246412584196e-05,
+      "loss": 1.6705,
+      "step": 11510
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 27.138111114501953,
+      "learning_rate": 1.0369409697527508e-05,
+      "loss": 0.9324,
+      "step": 11511
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 24.705583572387695,
+      "learning_rate": 1.0368572982470821e-05,
+      "loss": 2.9491,
+      "step": 11512
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 7.6751837730407715,
+      "learning_rate": 1.0367736267414132e-05,
+      "loss": 0.2387,
+      "step": 11513
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.83401107788086,
+      "learning_rate": 1.0366899552357445e-05,
+      "loss": 0.5735,
+      "step": 11514
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.1944580078125,
+      "learning_rate": 1.0366062837300759e-05,
+      "loss": 2.7075,
+      "step": 11515
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 31.396709442138672,
+      "learning_rate": 1.0365226122244069e-05,
+      "loss": 0.9407,
+      "step": 11516
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 17.157712936401367,
+      "learning_rate": 1.0364389407187383e-05,
+      "loss": 1.2364,
+      "step": 11517
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.019781112670898,
+      "learning_rate": 1.0363552692130696e-05,
+      "loss": 1.5119,
+      "step": 11518
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 14.416360855102539,
+      "learning_rate": 1.036271597707401e-05,
+      "loss": 0.9504,
+      "step": 11519
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 6.151710033416748,
+      "learning_rate": 1.036187926201732e-05,
+      "loss": 0.5117,
+      "step": 11520
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 13.85291576385498,
+      "learning_rate": 1.0361042546960634e-05,
+      "loss": 1.6905,
+      "step": 11521
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 4.2402143478393555,
+      "learning_rate": 1.0360205831903946e-05,
+      "loss": 0.3068,
+      "step": 11522
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.13919734954834,
+      "learning_rate": 1.0359369116847258e-05,
+      "loss": 0.7341,
+      "step": 11523
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.106743812561035,
+      "learning_rate": 1.0358532401790572e-05,
+      "loss": 1.064,
+      "step": 11524
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.152956008911133,
+      "learning_rate": 1.0357695686733884e-05,
+      "loss": 1.9516,
+      "step": 11525
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 17.164810180664062,
+      "learning_rate": 1.0356858971677197e-05,
+      "loss": 1.3683,
+      "step": 11526
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.938882827758789,
+      "learning_rate": 1.0356022256620507e-05,
+      "loss": 0.4788,
+      "step": 11527
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 26.89095687866211,
+      "learning_rate": 1.0355185541563821e-05,
+      "loss": 3.3631,
+      "step": 11528
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 7.967386722564697,
+      "learning_rate": 1.0354348826507135e-05,
+      "loss": 1.0033,
+      "step": 11529
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 13.68557357788086,
+      "learning_rate": 1.0353512111450445e-05,
+      "loss": 1.1689,
+      "step": 11530
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 15.132197380065918,
+      "learning_rate": 1.0352675396393759e-05,
+      "loss": 1.3014,
+      "step": 11531
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 45.730281829833984,
+      "learning_rate": 1.0351838681337072e-05,
+      "loss": 2.1266,
+      "step": 11532
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 82.78030395507812,
+      "learning_rate": 1.0351001966280386e-05,
+      "loss": 0.9793,
+      "step": 11533
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 7.21171236038208,
+      "learning_rate": 1.0350165251223696e-05,
+      "loss": 1.0636,
+      "step": 11534
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 908.1190185546875,
+      "learning_rate": 1.034932853616701e-05,
+      "loss": 2.1047,
+      "step": 11535
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 20.707496643066406,
+      "learning_rate": 1.0348491821110322e-05,
+      "loss": 2.0304,
+      "step": 11536
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 181.55435180664062,
+      "learning_rate": 1.0347655106053634e-05,
+      "loss": 1.5087,
+      "step": 11537
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 20.19994354248047,
+      "learning_rate": 1.0346818390996947e-05,
+      "loss": 2.1974,
+      "step": 11538
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 12.254837989807129,
+      "learning_rate": 1.034598167594026e-05,
+      "loss": 1.0284,
+      "step": 11539
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.827713966369629,
+      "learning_rate": 1.0345144960883573e-05,
+      "loss": 0.7259,
+      "step": 11540
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 18.657577514648438,
+      "learning_rate": 1.0344308245826883e-05,
+      "loss": 1.1016,
+      "step": 11541
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 26.329896926879883,
+      "learning_rate": 1.0343471530770197e-05,
+      "loss": 1.3192,
+      "step": 11542
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 13.237192153930664,
+      "learning_rate": 1.034263481571351e-05,
+      "loss": 0.7066,
+      "step": 11543
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 12.47546672821045,
+      "learning_rate": 1.034179810065682e-05,
+      "loss": 0.7557,
+      "step": 11544
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 159.52273559570312,
+      "learning_rate": 1.0340961385600134e-05,
+      "loss": 1.6982,
+      "step": 11545
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 6.027006149291992,
+      "learning_rate": 1.0340124670543448e-05,
+      "loss": 0.5325,
+      "step": 11546
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 13.722332954406738,
+      "learning_rate": 1.033928795548676e-05,
+      "loss": 2.528,
+      "step": 11547
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 18.110464096069336,
+      "learning_rate": 1.0338451240430072e-05,
+      "loss": 0.572,
+      "step": 11548
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.07425308227539,
+      "learning_rate": 1.0337614525373386e-05,
+      "loss": 0.8301,
+      "step": 11549
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 33.45741653442383,
+      "learning_rate": 1.0336777810316698e-05,
+      "loss": 1.9811,
+      "step": 11550
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 13.170585632324219,
+      "learning_rate": 1.033594109526001e-05,
+      "loss": 0.5509,
+      "step": 11551
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 6.260337829589844,
+      "learning_rate": 1.0335104380203323e-05,
+      "loss": 0.1017,
+      "step": 11552
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.019852638244629,
+      "learning_rate": 1.0334267665146635e-05,
+      "loss": 0.6926,
+      "step": 11553
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 8.174002647399902,
+      "learning_rate": 1.0333430950089949e-05,
+      "loss": 0.3792,
+      "step": 11554
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 35.70268249511719,
+      "learning_rate": 1.0332594235033259e-05,
+      "loss": 2.7399,
+      "step": 11555
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 7.015445709228516,
+      "learning_rate": 1.0331757519976573e-05,
+      "loss": 0.2739,
+      "step": 11556
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.410852432250977,
+      "learning_rate": 1.0330920804919886e-05,
+      "loss": 1.3884,
+      "step": 11557
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.94278621673584,
+      "learning_rate": 1.0330084089863197e-05,
+      "loss": 1.0147,
+      "step": 11558
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 20.14786148071289,
+      "learning_rate": 1.032924737480651e-05,
+      "loss": 1.0886,
+      "step": 11559
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.37377643585205,
+      "learning_rate": 1.0328410659749824e-05,
+      "loss": 0.7988,
+      "step": 11560
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 7.83779239654541,
+      "learning_rate": 1.0327573944693136e-05,
+      "loss": 1.6752,
+      "step": 11561
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 15.421342849731445,
+      "learning_rate": 1.0326737229636448e-05,
+      "loss": 1.7482,
+      "step": 11562
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 12.408710479736328,
+      "learning_rate": 1.0325900514579762e-05,
+      "loss": 0.3391,
+      "step": 11563
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 20.381467819213867,
+      "learning_rate": 1.0325063799523073e-05,
+      "loss": 2.3143,
+      "step": 11564
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 13.723130226135254,
+      "learning_rate": 1.0324227084466385e-05,
+      "loss": 1.6073,
+      "step": 11565
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 4.7990264892578125,
+      "learning_rate": 1.0323390369409697e-05,
+      "loss": 0.5784,
+      "step": 11566
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 23.338542938232422,
+      "learning_rate": 1.0322553654353011e-05,
+      "loss": 0.5699,
+      "step": 11567
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 25.580055236816406,
+      "learning_rate": 1.0321716939296325e-05,
+      "loss": 0.2767,
+      "step": 11568
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 18.603832244873047,
+      "learning_rate": 1.0320880224239635e-05,
+      "loss": 1.0041,
+      "step": 11569
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 15.5408353805542,
+      "learning_rate": 1.0320043509182949e-05,
+      "loss": 1.151,
+      "step": 11570
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 20.128328323364258,
+      "learning_rate": 1.0319206794126262e-05,
+      "loss": 1.1813,
+      "step": 11571
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 79.19524383544922,
+      "learning_rate": 1.0318370079069572e-05,
+      "loss": 3.1646,
+      "step": 11572
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 18.595197677612305,
+      "learning_rate": 1.0317533364012886e-05,
+      "loss": 2.187,
+      "step": 11573
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 6.652132034301758,
+      "learning_rate": 1.03166966489562e-05,
+      "loss": 0.3347,
+      "step": 11574
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 33.59093475341797,
+      "learning_rate": 1.0315859933899512e-05,
+      "loss": 3.074,
+      "step": 11575
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.482759475708008,
+      "learning_rate": 1.0315023218842824e-05,
+      "loss": 0.7033,
+      "step": 11576
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 89.27581024169922,
+      "learning_rate": 1.0314186503786137e-05,
+      "loss": 4.2024,
+      "step": 11577
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 17.355918884277344,
+      "learning_rate": 1.031334978872945e-05,
+      "loss": 0.8522,
+      "step": 11578
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.63662052154541,
+      "learning_rate": 1.0312513073672761e-05,
+      "loss": 0.4806,
+      "step": 11579
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 4.776221752166748,
+      "learning_rate": 1.0311676358616073e-05,
+      "loss": 0.1719,
+      "step": 11580
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 5.802083969116211,
+      "learning_rate": 1.0310839643559387e-05,
+      "loss": 1.5277,
+      "step": 11581
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 4.599855899810791,
+      "learning_rate": 1.03100029285027e-05,
+      "loss": 0.6048,
+      "step": 11582
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.42330551147461,
+      "learning_rate": 1.030916621344601e-05,
+      "loss": 0.913,
+      "step": 11583
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 19.721864700317383,
+      "learning_rate": 1.0308329498389324e-05,
+      "loss": 1.861,
+      "step": 11584
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 15.7676362991333,
+      "learning_rate": 1.0307492783332638e-05,
+      "loss": 1.5734,
+      "step": 11585
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 16.899751663208008,
+      "learning_rate": 1.0306656068275948e-05,
+      "loss": 0.8004,
+      "step": 11586
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 16.90408706665039,
+      "learning_rate": 1.0305819353219262e-05,
+      "loss": 1.9167,
+      "step": 11587
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 5.868137359619141,
+      "learning_rate": 1.0304982638162576e-05,
+      "loss": 0.3671,
+      "step": 11588
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 7.04013204574585,
+      "learning_rate": 1.0304145923105888e-05,
+      "loss": 1.5718,
+      "step": 11589
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 9.058857917785645,
+      "learning_rate": 1.03033092080492e-05,
+      "loss": 0.279,
+      "step": 11590
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.890242576599121,
+      "learning_rate": 1.0302472492992513e-05,
+      "loss": 1.5068,
+      "step": 11591
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.963504791259766,
+      "learning_rate": 1.0301635777935825e-05,
+      "loss": 0.6266,
+      "step": 11592
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 70.46163177490234,
+      "learning_rate": 1.0300799062879137e-05,
+      "loss": 0.6158,
+      "step": 11593
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 14.41070556640625,
+      "learning_rate": 1.0299962347822449e-05,
+      "loss": 1.2228,
+      "step": 11594
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 16.657543182373047,
+      "learning_rate": 1.0299125632765763e-05,
+      "loss": 2.6929,
+      "step": 11595
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 6.03270149230957,
+      "learning_rate": 1.0298288917709075e-05,
+      "loss": 0.5633,
+      "step": 11596
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 10.365429878234863,
+      "learning_rate": 1.0297452202652387e-05,
+      "loss": 0.3843,
+      "step": 11597
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 4.687175273895264,
+      "learning_rate": 1.02966154875957e-05,
+      "loss": 0.2333,
+      "step": 11598
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 20.45831298828125,
+      "learning_rate": 1.0295778772539014e-05,
+      "loss": 0.893,
+      "step": 11599
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.342596054077148,
+      "learning_rate": 1.0294942057482324e-05,
+      "loss": 0.4243,
+      "step": 11600
+    },
+    {
+      "epoch": 1.46,
+      "eval_loss": 0.0890585333108902,
+      "eval_runtime": 95.654,
+      "eval_samples_per_second": 37.029,
+      "eval_steps_per_second": 37.029,
+      "step": 11600
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 17.461505889892578,
+      "learning_rate": 1.0294105342425638e-05,
+      "loss": 1.1512,
+      "step": 11601
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 20.406530380249023,
+      "learning_rate": 1.0293268627368951e-05,
+      "loss": 1.7207,
+      "step": 11602
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 17.772703170776367,
+      "learning_rate": 1.0292431912312262e-05,
+      "loss": 2.3751,
+      "step": 11603
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 22.063190460205078,
+      "learning_rate": 1.0291595197255575e-05,
+      "loss": 1.8715,
+      "step": 11604
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 15.428668022155762,
+      "learning_rate": 1.0290758482198889e-05,
+      "loss": 1.2716,
+      "step": 11605
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 13.446057319641113,
+      "learning_rate": 1.0289921767142201e-05,
+      "loss": 0.7319,
+      "step": 11606
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 12.71236515045166,
+      "learning_rate": 1.0289085052085513e-05,
+      "loss": 1.4276,
+      "step": 11607
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 25.381593704223633,
+      "learning_rate": 1.0288248337028825e-05,
+      "loss": 0.5396,
+      "step": 11608
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 5.780124187469482,
+      "learning_rate": 1.0287411621972139e-05,
+      "loss": 0.513,
+      "step": 11609
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 4.070555210113525,
+      "learning_rate": 1.028657490691545e-05,
+      "loss": 1.0341,
+      "step": 11610
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 12.839869499206543,
+      "learning_rate": 1.0285738191858762e-05,
+      "loss": 1.2353,
+      "step": 11611
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 30.747859954833984,
+      "learning_rate": 1.0284901476802076e-05,
+      "loss": 1.0197,
+      "step": 11612
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 18.160499572753906,
+      "learning_rate": 1.028406476174539e-05,
+      "loss": 1.733,
+      "step": 11613
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 13.152453422546387,
+      "learning_rate": 1.02832280466887e-05,
+      "loss": 0.8829,
+      "step": 11614
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 17.087106704711914,
+      "learning_rate": 1.0282391331632014e-05,
+      "loss": 0.381,
+      "step": 11615
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 12.67386245727539,
+      "learning_rate": 1.0281554616575327e-05,
+      "loss": 1.6281,
+      "step": 11616
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 13.952394485473633,
+      "learning_rate": 1.0280717901518638e-05,
+      "loss": 2.1248,
+      "step": 11617
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 32.914794921875,
+      "learning_rate": 1.0279881186461951e-05,
+      "loss": 0.4341,
+      "step": 11618
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 91.3016586303711,
+      "learning_rate": 1.0279044471405263e-05,
+      "loss": 0.8319,
+      "step": 11619
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 21.45525550842285,
+      "learning_rate": 1.0278207756348577e-05,
+      "loss": 1.3928,
+      "step": 11620
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 6.4357380867004395,
+      "learning_rate": 1.0277371041291889e-05,
+      "loss": 0.0834,
+      "step": 11621
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 137.15390014648438,
+      "learning_rate": 1.02765343262352e-05,
+      "loss": 1.8562,
+      "step": 11622
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 14.621424674987793,
+      "learning_rate": 1.0275697611178514e-05,
+      "loss": 0.7229,
+      "step": 11623
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 191.5143280029297,
+      "learning_rate": 1.0274860896121825e-05,
+      "loss": 2.8116,
+      "step": 11624
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 4.596951961517334,
+      "learning_rate": 1.0274024181065138e-05,
+      "loss": 0.1174,
+      "step": 11625
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 19.01567268371582,
+      "learning_rate": 1.0273187466008452e-05,
+      "loss": 0.8861,
+      "step": 11626
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 16.008623123168945,
+      "learning_rate": 1.0272350750951766e-05,
+      "loss": 1.0557,
+      "step": 11627
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 12.518770217895508,
+      "learning_rate": 1.0271514035895076e-05,
+      "loss": 0.4727,
+      "step": 11628
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 6.306007385253906,
+      "learning_rate": 1.027067732083839e-05,
+      "loss": 0.7893,
+      "step": 11629
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 20.78609275817871,
+      "learning_rate": 1.0269840605781703e-05,
+      "loss": 1.0765,
+      "step": 11630
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 5.802038669586182,
+      "learning_rate": 1.0269003890725013e-05,
+      "loss": 1.2745,
+      "step": 11631
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 40.01700210571289,
+      "learning_rate": 1.0268167175668327e-05,
+      "loss": 1.4334,
+      "step": 11632
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 16.79182243347168,
+      "learning_rate": 1.0267330460611639e-05,
+      "loss": 1.6118,
+      "step": 11633
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 14.402016639709473,
+      "learning_rate": 1.0266493745554953e-05,
+      "loss": 0.9737,
+      "step": 11634
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 90.98631286621094,
+      "learning_rate": 1.0265657030498265e-05,
+      "loss": 0.965,
+      "step": 11635
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 13.878726959228516,
+      "learning_rate": 1.0264820315441577e-05,
+      "loss": 1.6883,
+      "step": 11636
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 14.278339385986328,
+      "learning_rate": 1.026398360038489e-05,
+      "loss": 1.9722,
+      "step": 11637
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 14.49180793762207,
+      "learning_rate": 1.02631468853282e-05,
+      "loss": 1.575,
+      "step": 11638
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.01636791229248,
+      "learning_rate": 1.0262310170271514e-05,
+      "loss": 1.0054,
+      "step": 11639
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 16.10940933227539,
+      "learning_rate": 1.0261473455214828e-05,
+      "loss": 2.1795,
+      "step": 11640
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 11.715025901794434,
+      "learning_rate": 1.0260636740158141e-05,
+      "loss": 1.2947,
+      "step": 11641
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 8.381911277770996,
+      "learning_rate": 1.0259800025101452e-05,
+      "loss": 0.3407,
+      "step": 11642
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 7.727115154266357,
+      "learning_rate": 1.0258963310044765e-05,
+      "loss": 0.5429,
+      "step": 11643
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.851998329162598,
+      "learning_rate": 1.0258126594988079e-05,
+      "loss": 0.8527,
+      "step": 11644
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 42.08799743652344,
+      "learning_rate": 1.025728987993139e-05,
+      "loss": 1.1845,
+      "step": 11645
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 3.101151704788208,
+      "learning_rate": 1.0256453164874703e-05,
+      "loss": 0.1981,
+      "step": 11646
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.115474700927734,
+      "learning_rate": 1.0255616449818015e-05,
+      "loss": 0.5998,
+      "step": 11647
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 10.739004135131836,
+      "learning_rate": 1.0254779734761328e-05,
+      "loss": 0.6968,
+      "step": 11648
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 23.58644676208496,
+      "learning_rate": 1.025394301970464e-05,
+      "loss": 1.9058,
+      "step": 11649
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 44.03055191040039,
+      "learning_rate": 1.0253106304647952e-05,
+      "loss": 1.0997,
+      "step": 11650
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 17.007707595825195,
+      "learning_rate": 1.0252269589591266e-05,
+      "loss": 1.3267,
+      "step": 11651
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 6.813281536102295,
+      "learning_rate": 1.0251432874534576e-05,
+      "loss": 0.6019,
+      "step": 11652
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 7.804632186889648,
+      "learning_rate": 1.025059615947789e-05,
+      "loss": 0.4428,
+      "step": 11653
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 51.19334030151367,
+      "learning_rate": 1.0249759444421204e-05,
+      "loss": 3.2358,
+      "step": 11654
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 95.2354965209961,
+      "learning_rate": 1.0248922729364517e-05,
+      "loss": 3.8895,
+      "step": 11655
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 31.65764808654785,
+      "learning_rate": 1.0248086014307828e-05,
+      "loss": 1.4485,
+      "step": 11656
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 15.785746574401855,
+      "learning_rate": 1.0247249299251141e-05,
+      "loss": 0.3068,
+      "step": 11657
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 9.289689064025879,
+      "learning_rate": 1.0246412584194453e-05,
+      "loss": 1.1379,
+      "step": 11658
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 15.987249374389648,
+      "learning_rate": 1.0245575869137765e-05,
+      "loss": 1.2005,
+      "step": 11659
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 17.456310272216797,
+      "learning_rate": 1.0244739154081079e-05,
+      "loss": 1.5594,
+      "step": 11660
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 15.669245719909668,
+      "learning_rate": 1.024390243902439e-05,
+      "loss": 1.0324,
+      "step": 11661
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 6.8458757400512695,
+      "learning_rate": 1.0243065723967704e-05,
+      "loss": 0.7236,
+      "step": 11662
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 16.892261505126953,
+      "learning_rate": 1.0242229008911016e-05,
+      "loss": 0.7325,
+      "step": 11663
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 10.99298095703125,
+      "learning_rate": 1.0241392293854328e-05,
+      "loss": 1.5616,
+      "step": 11664
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 18.590248107910156,
+      "learning_rate": 1.0240555578797642e-05,
+      "loss": 0.6841,
+      "step": 11665
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 37.73554229736328,
+      "learning_rate": 1.0239718863740952e-05,
+      "loss": 2.2575,
+      "step": 11666
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 23.0293025970459,
+      "learning_rate": 1.0238882148684266e-05,
+      "loss": 1.0603,
+      "step": 11667
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 6.224422931671143,
+      "learning_rate": 1.023804543362758e-05,
+      "loss": 0.7531,
+      "step": 11668
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 14.43862533569336,
+      "learning_rate": 1.0237208718570893e-05,
+      "loss": 1.8931,
+      "step": 11669
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 34.57533645629883,
+      "learning_rate": 1.0236372003514203e-05,
+      "loss": 0.9463,
+      "step": 11670
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 18.28287124633789,
+      "learning_rate": 1.0235535288457517e-05,
+      "loss": 1.6143,
+      "step": 11671
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 20.80086898803711,
+      "learning_rate": 1.0234698573400829e-05,
+      "loss": 3.2864,
+      "step": 11672
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 13.415273666381836,
+      "learning_rate": 1.0233861858344141e-05,
+      "loss": 0.6393,
+      "step": 11673
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 12.326879501342773,
+      "learning_rate": 1.0233025143287455e-05,
+      "loss": 2.4646,
+      "step": 11674
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 17.38701629638672,
+      "learning_rate": 1.0232188428230767e-05,
+      "loss": 1.5567,
+      "step": 11675
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 8.564024925231934,
+      "learning_rate": 1.023135171317408e-05,
+      "loss": 1.2473,
+      "step": 11676
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.077590942382812,
+      "learning_rate": 1.023051499811739e-05,
+      "loss": 1.4386,
+      "step": 11677
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.411284446716309,
+      "learning_rate": 1.0229678283060704e-05,
+      "loss": 1.2173,
+      "step": 11678
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 14.678159713745117,
+      "learning_rate": 1.0228841568004018e-05,
+      "loss": 0.8474,
+      "step": 11679
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 6.511855125427246,
+      "learning_rate": 1.0228004852947328e-05,
+      "loss": 0.5764,
+      "step": 11680
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 22.524940490722656,
+      "learning_rate": 1.0227168137890642e-05,
+      "loss": 1.9294,
+      "step": 11681
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 8.448037147521973,
+      "learning_rate": 1.0226331422833955e-05,
+      "loss": 1.3012,
+      "step": 11682
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 16.592552185058594,
+      "learning_rate": 1.0225494707777269e-05,
+      "loss": 1.6058,
+      "step": 11683
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 145.56182861328125,
+      "learning_rate": 1.022465799272058e-05,
+      "loss": 3.3452,
+      "step": 11684
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 19.358047485351562,
+      "learning_rate": 1.0223821277663893e-05,
+      "loss": 1.5458,
+      "step": 11685
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.307069301605225,
+      "learning_rate": 1.0222984562607205e-05,
+      "loss": 0.6352,
+      "step": 11686
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 23.20427131652832,
+      "learning_rate": 1.0222147847550517e-05,
+      "loss": 2.1221,
+      "step": 11687
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 13.527782440185547,
+      "learning_rate": 1.022131113249383e-05,
+      "loss": 1.2126,
+      "step": 11688
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 133.5238494873047,
+      "learning_rate": 1.0220474417437142e-05,
+      "loss": 1.5857,
+      "step": 11689
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 21.087724685668945,
+      "learning_rate": 1.0219637702380456e-05,
+      "loss": 1.7069,
+      "step": 11690
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 60.96128463745117,
+      "learning_rate": 1.0218800987323766e-05,
+      "loss": 1.9141,
+      "step": 11691
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.16915225982666,
+      "learning_rate": 1.021796427226708e-05,
+      "loss": 0.2113,
+      "step": 11692
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 14.474244117736816,
+      "learning_rate": 1.0217127557210394e-05,
+      "loss": 0.7099,
+      "step": 11693
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 13.534197807312012,
+      "learning_rate": 1.0216290842153704e-05,
+      "loss": 0.325,
+      "step": 11694
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.091919898986816,
+      "learning_rate": 1.0215454127097017e-05,
+      "loss": 1.1304,
+      "step": 11695
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 8.387577056884766,
+      "learning_rate": 1.0214617412040331e-05,
+      "loss": 0.4027,
+      "step": 11696
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 20.336320877075195,
+      "learning_rate": 1.0213780696983645e-05,
+      "loss": 0.6635,
+      "step": 11697
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 5.192074298858643,
+      "learning_rate": 1.0212943981926955e-05,
+      "loss": 1.2929,
+      "step": 11698
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 27.580411911010742,
+      "learning_rate": 1.0212107266870269e-05,
+      "loss": 2.0222,
+      "step": 11699
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 30.743709564208984,
+      "learning_rate": 1.021127055181358e-05,
+      "loss": 2.0479,
+      "step": 11700
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 67.473388671875,
+      "learning_rate": 1.0210433836756893e-05,
+      "loss": 2.0701,
+      "step": 11701
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 37.5473518371582,
+      "learning_rate": 1.0209597121700206e-05,
+      "loss": 1.7707,
+      "step": 11702
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 29.353015899658203,
+      "learning_rate": 1.0208760406643518e-05,
+      "loss": 2.1933,
+      "step": 11703
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.22294282913208,
+      "learning_rate": 1.0207923691586832e-05,
+      "loss": 0.662,
+      "step": 11704
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 12.743423461914062,
+      "learning_rate": 1.0207086976530142e-05,
+      "loss": 0.8054,
+      "step": 11705
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.4735307693481445,
+      "learning_rate": 1.0206250261473456e-05,
+      "loss": 0.7347,
+      "step": 11706
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.776078224182129,
+      "learning_rate": 1.020541354641677e-05,
+      "loss": 0.4297,
+      "step": 11707
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 13.706832885742188,
+      "learning_rate": 1.020457683136008e-05,
+      "loss": 0.3821,
+      "step": 11708
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 18.440196990966797,
+      "learning_rate": 1.0203740116303393e-05,
+      "loss": 0.7917,
+      "step": 11709
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 38.80813980102539,
+      "learning_rate": 1.0202903401246707e-05,
+      "loss": 1.9899,
+      "step": 11710
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 13.319124221801758,
+      "learning_rate": 1.0202066686190019e-05,
+      "loss": 1.7305,
+      "step": 11711
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.006318092346191,
+      "learning_rate": 1.0201229971133331e-05,
+      "loss": 2.9458,
+      "step": 11712
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 15.509191513061523,
+      "learning_rate": 1.0200393256076645e-05,
+      "loss": 0.6453,
+      "step": 11713
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 12.68134593963623,
+      "learning_rate": 1.0199556541019956e-05,
+      "loss": 0.6937,
+      "step": 11714
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.146038055419922,
+      "learning_rate": 1.0198719825963268e-05,
+      "loss": 0.6571,
+      "step": 11715
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 26.770910263061523,
+      "learning_rate": 1.0197883110906582e-05,
+      "loss": 2.2434,
+      "step": 11716
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 21.979290008544922,
+      "learning_rate": 1.0197046395849894e-05,
+      "loss": 2.1595,
+      "step": 11717
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 13.98854923248291,
+      "learning_rate": 1.0196209680793208e-05,
+      "loss": 1.1727,
+      "step": 11718
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 12.81928825378418,
+      "learning_rate": 1.0195372965736518e-05,
+      "loss": 2.488,
+      "step": 11719
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 6.362330913543701,
+      "learning_rate": 1.0194536250679832e-05,
+      "loss": 0.8804,
+      "step": 11720
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 17.94073486328125,
+      "learning_rate": 1.0193699535623145e-05,
+      "loss": 0.5014,
+      "step": 11721
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 5.776565074920654,
+      "learning_rate": 1.0192862820566455e-05,
+      "loss": 0.8061,
+      "step": 11722
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 12.142158508300781,
+      "learning_rate": 1.0192026105509769e-05,
+      "loss": 0.6426,
+      "step": 11723
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 17.069988250732422,
+      "learning_rate": 1.0191189390453083e-05,
+      "loss": 0.7519,
+      "step": 11724
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.901285171508789,
+      "learning_rate": 1.0190352675396395e-05,
+      "loss": 0.4464,
+      "step": 11725
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 26.143484115600586,
+      "learning_rate": 1.0189515960339707e-05,
+      "loss": 1.8073,
+      "step": 11726
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 19.156314849853516,
+      "learning_rate": 1.018867924528302e-05,
+      "loss": 1.5151,
+      "step": 11727
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 36.94829559326172,
+      "learning_rate": 1.0187842530226332e-05,
+      "loss": 1.6474,
+      "step": 11728
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 6.899235725402832,
+      "learning_rate": 1.0187005815169644e-05,
+      "loss": 1.2121,
+      "step": 11729
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 12.01082992553711,
+      "learning_rate": 1.0186169100112956e-05,
+      "loss": 0.622,
+      "step": 11730
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 4.53358268737793,
+      "learning_rate": 1.018533238505627e-05,
+      "loss": 0.4688,
+      "step": 11731
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 21.91617774963379,
+      "learning_rate": 1.0184495669999584e-05,
+      "loss": 2.0592,
+      "step": 11732
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 41.274417877197266,
+      "learning_rate": 1.0183658954942894e-05,
+      "loss": 0.9033,
+      "step": 11733
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 8.857417106628418,
+      "learning_rate": 1.0182822239886207e-05,
+      "loss": 0.8782,
+      "step": 11734
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 5.053219795227051,
+      "learning_rate": 1.0181985524829521e-05,
+      "loss": 0.6986,
+      "step": 11735
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 31.58706283569336,
+      "learning_rate": 1.0181148809772831e-05,
+      "loss": 0.7091,
+      "step": 11736
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 13.136528015136719,
+      "learning_rate": 1.0180312094716145e-05,
+      "loss": 0.911,
+      "step": 11737
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 12.264043807983398,
+      "learning_rate": 1.0179475379659459e-05,
+      "loss": 1.101,
+      "step": 11738
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 13.618952751159668,
+      "learning_rate": 1.017863866460277e-05,
+      "loss": 0.866,
+      "step": 11739
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 24.54608154296875,
+      "learning_rate": 1.0177801949546083e-05,
+      "loss": 1.0775,
+      "step": 11740
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 14.495532989501953,
+      "learning_rate": 1.0176965234489396e-05,
+      "loss": 1.7906,
+      "step": 11741
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 16.60027503967285,
+      "learning_rate": 1.0176128519432708e-05,
+      "loss": 1.4876,
+      "step": 11742
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 28.480043411254883,
+      "learning_rate": 1.017529180437602e-05,
+      "loss": 1.5874,
+      "step": 11743
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 15.660050392150879,
+      "learning_rate": 1.0174455089319332e-05,
+      "loss": 1.6086,
+      "step": 11744
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 9.23560619354248,
+      "learning_rate": 1.0173618374262646e-05,
+      "loss": 0.8364,
+      "step": 11745
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 8.3224458694458,
+      "learning_rate": 1.017278165920596e-05,
+      "loss": 0.8419,
+      "step": 11746
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 22.351551055908203,
+      "learning_rate": 1.017194494414927e-05,
+      "loss": 1.2834,
+      "step": 11747
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.089461326599121,
+      "learning_rate": 1.0171108229092583e-05,
+      "loss": 1.9218,
+      "step": 11748
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 15.574318885803223,
+      "learning_rate": 1.0170271514035897e-05,
+      "loss": 0.879,
+      "step": 11749
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.7962749004364014,
+      "learning_rate": 1.0169434798979207e-05,
+      "loss": 0.4112,
+      "step": 11750
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 6.679157733917236,
+      "learning_rate": 1.016859808392252e-05,
+      "loss": 0.5855,
+      "step": 11751
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 17.797319412231445,
+      "learning_rate": 1.0167761368865834e-05,
+      "loss": 2.4238,
+      "step": 11752
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 19.303560256958008,
+      "learning_rate": 1.0166924653809146e-05,
+      "loss": 1.5157,
+      "step": 11753
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 23.224414825439453,
+      "learning_rate": 1.0166087938752458e-05,
+      "loss": 3.1815,
+      "step": 11754
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 27.317115783691406,
+      "learning_rate": 1.0165251223695772e-05,
+      "loss": 1.9885,
+      "step": 11755
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 16.916921615600586,
+      "learning_rate": 1.0164414508639084e-05,
+      "loss": 0.9612,
+      "step": 11756
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.244033813476562,
+      "learning_rate": 1.0163577793582396e-05,
+      "loss": 0.4901,
+      "step": 11757
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 29.34827423095703,
+      "learning_rate": 1.0162741078525708e-05,
+      "loss": 1.8972,
+      "step": 11758
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.552237510681152,
+      "learning_rate": 1.0161904363469022e-05,
+      "loss": 0.9352,
+      "step": 11759
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 17.028141021728516,
+      "learning_rate": 1.0161067648412335e-05,
+      "loss": 0.9987,
+      "step": 11760
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 28.46116828918457,
+      "learning_rate": 1.0160230933355645e-05,
+      "loss": 1.5082,
+      "step": 11761
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.872819185256958,
+      "learning_rate": 1.0159394218298959e-05,
+      "loss": 0.2531,
+      "step": 11762
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 23.133909225463867,
+      "learning_rate": 1.0158557503242273e-05,
+      "loss": 0.9149,
+      "step": 11763
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 20.6193904876709,
+      "learning_rate": 1.0157720788185583e-05,
+      "loss": 0.6258,
+      "step": 11764
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 14.060830116271973,
+      "learning_rate": 1.0156884073128897e-05,
+      "loss": 0.976,
+      "step": 11765
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.146706581115723,
+      "learning_rate": 1.015604735807221e-05,
+      "loss": 0.2321,
+      "step": 11766
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 36.6379280090332,
+      "learning_rate": 1.0155210643015522e-05,
+      "loss": 2.1008,
+      "step": 11767
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 6.3384222984313965,
+      "learning_rate": 1.0154373927958834e-05,
+      "loss": 0.1667,
+      "step": 11768
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 35.882598876953125,
+      "learning_rate": 1.0153537212902146e-05,
+      "loss": 0.6095,
+      "step": 11769
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 15.172981262207031,
+      "learning_rate": 1.015270049784546e-05,
+      "loss": 0.8346,
+      "step": 11770
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 32.03757858276367,
+      "learning_rate": 1.0151863782788772e-05,
+      "loss": 1.836,
+      "step": 11771
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 12.183686256408691,
+      "learning_rate": 1.0151027067732084e-05,
+      "loss": 0.9534,
+      "step": 11772
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 13.156387329101562,
+      "learning_rate": 1.0150190352675397e-05,
+      "loss": 0.6516,
+      "step": 11773
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 78.90950775146484,
+      "learning_rate": 1.0149353637618711e-05,
+      "loss": 3.9082,
+      "step": 11774
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 67.05281066894531,
+      "learning_rate": 1.0148516922562021e-05,
+      "loss": 3.3256,
+      "step": 11775
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 5.540832042694092,
+      "learning_rate": 1.0147680207505335e-05,
+      "loss": 0.3358,
+      "step": 11776
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.787993431091309,
+      "learning_rate": 1.0146843492448649e-05,
+      "loss": 0.7127,
+      "step": 11777
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 62.75838851928711,
+      "learning_rate": 1.0146006777391959e-05,
+      "loss": 2.0421,
+      "step": 11778
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 24.62049674987793,
+      "learning_rate": 1.0145170062335272e-05,
+      "loss": 0.9993,
+      "step": 11779
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 22.030406951904297,
+      "learning_rate": 1.0144333347278586e-05,
+      "loss": 2.3176,
+      "step": 11780
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.144664764404297,
+      "learning_rate": 1.0143496632221898e-05,
+      "loss": 0.5781,
+      "step": 11781
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.777544021606445,
+      "learning_rate": 1.014265991716521e-05,
+      "loss": 0.6678,
+      "step": 11782
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.740224838256836,
+      "learning_rate": 1.0141823202108522e-05,
+      "loss": 0.853,
+      "step": 11783
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 12.109756469726562,
+      "learning_rate": 1.0140986487051836e-05,
+      "loss": 0.6272,
+      "step": 11784
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 14.739649772644043,
+      "learning_rate": 1.0140149771995148e-05,
+      "loss": 0.5708,
+      "step": 11785
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 7.770986557006836,
+      "learning_rate": 1.013931305693846e-05,
+      "loss": 1.0056,
+      "step": 11786
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 13.864361763000488,
+      "learning_rate": 1.0138476341881773e-05,
+      "loss": 0.5382,
+      "step": 11787
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 15.930030822753906,
+      "learning_rate": 1.0137639626825087e-05,
+      "loss": 2.6395,
+      "step": 11788
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.665167808532715,
+      "learning_rate": 1.0136802911768397e-05,
+      "loss": 1.5669,
+      "step": 11789
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.066813468933105,
+      "learning_rate": 1.013596619671171e-05,
+      "loss": 1.2515,
+      "step": 11790
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 59.150367736816406,
+      "learning_rate": 1.0135129481655024e-05,
+      "loss": 2.82,
+      "step": 11791
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.148847579956055,
+      "learning_rate": 1.0134292766598335e-05,
+      "loss": 0.5652,
+      "step": 11792
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.0548095703125,
+      "learning_rate": 1.0133456051541648e-05,
+      "loss": 1.6382,
+      "step": 11793
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.091031074523926,
+      "learning_rate": 1.0132619336484962e-05,
+      "loss": 0.7803,
+      "step": 11794
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 13.253259658813477,
+      "learning_rate": 1.0131782621428274e-05,
+      "loss": 0.6461,
+      "step": 11795
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 33.466670989990234,
+      "learning_rate": 1.0130945906371586e-05,
+      "loss": 1.5478,
+      "step": 11796
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 44.28044509887695,
+      "learning_rate": 1.0130109191314898e-05,
+      "loss": 1.8743,
+      "step": 11797
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 15.960838317871094,
+      "learning_rate": 1.0129272476258211e-05,
+      "loss": 0.6709,
+      "step": 11798
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 18.216751098632812,
+      "learning_rate": 1.0128435761201523e-05,
+      "loss": 1.6836,
+      "step": 11799
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 23.818519592285156,
+      "learning_rate": 1.0127599046144835e-05,
+      "loss": 0.7374,
+      "step": 11800
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 13.40870475769043,
+      "learning_rate": 1.0126762331088149e-05,
+      "loss": 0.9878,
+      "step": 11801
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 33.63993835449219,
+      "learning_rate": 1.0125925616031463e-05,
+      "loss": 1.8692,
+      "step": 11802
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 24.853561401367188,
+      "learning_rate": 1.0125088900974773e-05,
+      "loss": 0.7594,
+      "step": 11803
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 4.262231826782227,
+      "learning_rate": 1.0124252185918087e-05,
+      "loss": 1.6934,
+      "step": 11804
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 17.921628952026367,
+      "learning_rate": 1.01234154708614e-05,
+      "loss": 2.2214,
+      "step": 11805
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 22.366456985473633,
+      "learning_rate": 1.012257875580471e-05,
+      "loss": 2.1575,
+      "step": 11806
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 17.08475112915039,
+      "learning_rate": 1.0121742040748024e-05,
+      "loss": 1.195,
+      "step": 11807
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 28.62654685974121,
+      "learning_rate": 1.0120905325691338e-05,
+      "loss": 1.8959,
+      "step": 11808
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 17.814733505249023,
+      "learning_rate": 1.012006861063465e-05,
+      "loss": 1.4859,
+      "step": 11809
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 39.33662033081055,
+      "learning_rate": 1.0119231895577962e-05,
+      "loss": 1.2965,
+      "step": 11810
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 7.899166584014893,
+      "learning_rate": 1.0118395180521274e-05,
+      "loss": 0.5111,
+      "step": 11811
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 4.4526448249816895,
+      "learning_rate": 1.0117558465464587e-05,
+      "loss": 0.1027,
+      "step": 11812
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 10.049847602844238,
+      "learning_rate": 1.01167217504079e-05,
+      "loss": 1.8753,
+      "step": 11813
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 12.997302055358887,
+      "learning_rate": 1.0115885035351211e-05,
+      "loss": 0.8018,
+      "step": 11814
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.1620454788208,
+      "learning_rate": 1.0115048320294525e-05,
+      "loss": 0.5171,
+      "step": 11815
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 19.684127807617188,
+      "learning_rate": 1.0114211605237839e-05,
+      "loss": 1.2105,
+      "step": 11816
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 32.90684127807617,
+      "learning_rate": 1.0113374890181149e-05,
+      "loss": 1.4728,
+      "step": 11817
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 4.621243953704834,
+      "learning_rate": 1.0112538175124462e-05,
+      "loss": 0.1001,
+      "step": 11818
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.834315299987793,
+      "learning_rate": 1.0111701460067776e-05,
+      "loss": 1.2464,
+      "step": 11819
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 48.37236404418945,
+      "learning_rate": 1.0110864745011086e-05,
+      "loss": 2.0795,
+      "step": 11820
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 24.91161346435547,
+      "learning_rate": 1.01100280299544e-05,
+      "loss": 1.2457,
+      "step": 11821
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.79609489440918,
+      "learning_rate": 1.0109191314897712e-05,
+      "loss": 0.3236,
+      "step": 11822
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 16.0001163482666,
+      "learning_rate": 1.0108354599841026e-05,
+      "loss": 1.9801,
+      "step": 11823
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 6.595298767089844,
+      "learning_rate": 1.0107517884784338e-05,
+      "loss": 1.3833,
+      "step": 11824
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 20.870195388793945,
+      "learning_rate": 1.010668116972765e-05,
+      "loss": 1.9719,
+      "step": 11825
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 30.148893356323242,
+      "learning_rate": 1.0105844454670963e-05,
+      "loss": 0.4573,
+      "step": 11826
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 14.521225929260254,
+      "learning_rate": 1.0105007739614275e-05,
+      "loss": 0.3858,
+      "step": 11827
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 7.839728355407715,
+      "learning_rate": 1.0104171024557587e-05,
+      "loss": 1.4502,
+      "step": 11828
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 11.859395980834961,
+      "learning_rate": 1.01033343095009e-05,
+      "loss": 1.0715,
+      "step": 11829
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 15.987799644470215,
+      "learning_rate": 1.0102497594444214e-05,
+      "loss": 1.0691,
+      "step": 11830
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 7.629037857055664,
+      "learning_rate": 1.0101660879387525e-05,
+      "loss": 0.5356,
+      "step": 11831
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 3.65165638923645,
+      "learning_rate": 1.0100824164330838e-05,
+      "loss": 0.2964,
+      "step": 11832
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.224053382873535,
+      "learning_rate": 1.0099987449274152e-05,
+      "loss": 1.3201,
+      "step": 11833
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 22.002063751220703,
+      "learning_rate": 1.0099150734217462e-05,
+      "loss": 1.6502,
+      "step": 11834
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 13.647278785705566,
+      "learning_rate": 1.0098314019160776e-05,
+      "loss": 0.9005,
+      "step": 11835
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.864240646362305,
+      "learning_rate": 1.0097477304104088e-05,
+      "loss": 0.5403,
+      "step": 11836
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.673495292663574,
+      "learning_rate": 1.0096640589047401e-05,
+      "loss": 0.8332,
+      "step": 11837
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 15.670794486999512,
+      "learning_rate": 1.0095803873990713e-05,
+      "loss": 1.6456,
+      "step": 11838
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 33.64060592651367,
+      "learning_rate": 1.0094967158934025e-05,
+      "loss": 1.1097,
+      "step": 11839
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 7.889852046966553,
+      "learning_rate": 1.0094130443877339e-05,
+      "loss": 0.3595,
+      "step": 11840
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 14.493781089782715,
+      "learning_rate": 1.009329372882065e-05,
+      "loss": 0.4723,
+      "step": 11841
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 21.568572998046875,
+      "learning_rate": 1.0092457013763963e-05,
+      "loss": 0.2411,
+      "step": 11842
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 16.096431732177734,
+      "learning_rate": 1.0091620298707277e-05,
+      "loss": 1.5888,
+      "step": 11843
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.522690773010254,
+      "learning_rate": 1.009078358365059e-05,
+      "loss": 0.1546,
+      "step": 11844
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 44.40749740600586,
+      "learning_rate": 1.00899468685939e-05,
+      "loss": 0.5052,
+      "step": 11845
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 12.55154037475586,
+      "learning_rate": 1.0089110153537214e-05,
+      "loss": 1.7197,
+      "step": 11846
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 15.280006408691406,
+      "learning_rate": 1.0088273438480528e-05,
+      "loss": 1.7976,
+      "step": 11847
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 25.104217529296875,
+      "learning_rate": 1.0087436723423838e-05,
+      "loss": 1.8859,
+      "step": 11848
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 20.182933807373047,
+      "learning_rate": 1.0086600008367152e-05,
+      "loss": 0.8879,
+      "step": 11849
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 10.312606811523438,
+      "learning_rate": 1.0085763293310464e-05,
+      "loss": 1.0224,
+      "step": 11850
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 17.0831298828125,
+      "learning_rate": 1.0084926578253777e-05,
+      "loss": 1.0411,
+      "step": 11851
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.599349975585938,
+      "learning_rate": 1.008408986319709e-05,
+      "loss": 1.3262,
+      "step": 11852
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 16.35350227355957,
+      "learning_rate": 1.0083253148140401e-05,
+      "loss": 0.7316,
+      "step": 11853
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 213.29212951660156,
+      "learning_rate": 1.0082416433083715e-05,
+      "loss": 1.3068,
+      "step": 11854
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 63.58739471435547,
+      "learning_rate": 1.0081579718027025e-05,
+      "loss": 1.922,
+      "step": 11855
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 41.69790267944336,
+      "learning_rate": 1.0080743002970339e-05,
+      "loss": 1.5676,
+      "step": 11856
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 34.60993194580078,
+      "learning_rate": 1.0079906287913652e-05,
+      "loss": 0.8385,
+      "step": 11857
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 67.94532775878906,
+      "learning_rate": 1.0079069572856966e-05,
+      "loss": 2.7407,
+      "step": 11858
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 12.518784523010254,
+      "learning_rate": 1.0078232857800276e-05,
+      "loss": 2.2123,
+      "step": 11859
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 19.069976806640625,
+      "learning_rate": 1.007739614274359e-05,
+      "loss": 1.5177,
+      "step": 11860
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 92.6587142944336,
+      "learning_rate": 1.0076559427686904e-05,
+      "loss": 2.7086,
+      "step": 11861
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 23.413423538208008,
+      "learning_rate": 1.0075722712630214e-05,
+      "loss": 1.2732,
+      "step": 11862
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 15.536075592041016,
+      "learning_rate": 1.0074885997573528e-05,
+      "loss": 2.6961,
+      "step": 11863
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 22.799394607543945,
+      "learning_rate": 1.007404928251684e-05,
+      "loss": 1.3043,
+      "step": 11864
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.75416088104248,
+      "learning_rate": 1.0073212567460153e-05,
+      "loss": 1.0192,
+      "step": 11865
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 19.864070892333984,
+      "learning_rate": 1.0072375852403465e-05,
+      "loss": 2.7687,
+      "step": 11866
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 5.751663684844971,
+      "learning_rate": 1.0071539137346777e-05,
+      "loss": 0.5699,
+      "step": 11867
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 33.524044036865234,
+      "learning_rate": 1.007070242229009e-05,
+      "loss": 1.9562,
+      "step": 11868
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.924762725830078,
+      "learning_rate": 1.0069865707233401e-05,
+      "loss": 2.3523,
+      "step": 11869
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 17.027732849121094,
+      "learning_rate": 1.0069028992176715e-05,
+      "loss": 0.8819,
+      "step": 11870
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 4.360075950622559,
+      "learning_rate": 1.0068192277120028e-05,
+      "loss": 0.2553,
+      "step": 11871
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 7.308431148529053,
+      "learning_rate": 1.0067355562063342e-05,
+      "loss": 0.9271,
+      "step": 11872
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 6.431609630584717,
+      "learning_rate": 1.0066518847006652e-05,
+      "loss": 0.7778,
+      "step": 11873
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 115.9853286743164,
+      "learning_rate": 1.0065682131949966e-05,
+      "loss": 2.2345,
+      "step": 11874
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 18.295310974121094,
+      "learning_rate": 1.0064845416893278e-05,
+      "loss": 2.7087,
+      "step": 11875
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 12.80589485168457,
+      "learning_rate": 1.006400870183659e-05,
+      "loss": 0.4732,
+      "step": 11876
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 15.119963645935059,
+      "learning_rate": 1.0063171986779903e-05,
+      "loss": 1.844,
+      "step": 11877
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 41.419559478759766,
+      "learning_rate": 1.0062335271723215e-05,
+      "loss": 2.2955,
+      "step": 11878
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 14.625669479370117,
+      "learning_rate": 1.0061498556666529e-05,
+      "loss": 2.0124,
+      "step": 11879
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 4.444066524505615,
+      "learning_rate": 1.006066184160984e-05,
+      "loss": 1.4116,
+      "step": 11880
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 12.593189239501953,
+      "learning_rate": 1.0059825126553153e-05,
+      "loss": 0.9024,
+      "step": 11881
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.293560028076172,
+      "learning_rate": 1.0058988411496467e-05,
+      "loss": 0.8693,
+      "step": 11882
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 14.213584899902344,
+      "learning_rate": 1.0058151696439777e-05,
+      "loss": 1.1086,
+      "step": 11883
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 20.005144119262695,
+      "learning_rate": 1.005731498138309e-05,
+      "loss": 1.6349,
+      "step": 11884
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 6.507580757141113,
+      "learning_rate": 1.0056478266326404e-05,
+      "loss": 1.2457,
+      "step": 11885
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 28.846912384033203,
+      "learning_rate": 1.0055641551269718e-05,
+      "loss": 2.2322,
+      "step": 11886
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 16.472408294677734,
+      "learning_rate": 1.0054804836213028e-05,
+      "loss": 1.6491,
+      "step": 11887
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 6.475840091705322,
+      "learning_rate": 1.0053968121156342e-05,
+      "loss": 1.383,
+      "step": 11888
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.640379905700684,
+      "learning_rate": 1.0053131406099654e-05,
+      "loss": 1.076,
+      "step": 11889
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.46177864074707,
+      "learning_rate": 1.0052294691042966e-05,
+      "loss": 0.9874,
+      "step": 11890
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 23.13226318359375,
+      "learning_rate": 1.005145797598628e-05,
+      "loss": 1.8804,
+      "step": 11891
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 76.4865951538086,
+      "learning_rate": 1.0050621260929591e-05,
+      "loss": 2.7426,
+      "step": 11892
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 33.265350341796875,
+      "learning_rate": 1.0049784545872905e-05,
+      "loss": 0.8085,
+      "step": 11893
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 11.67562484741211,
+      "learning_rate": 1.0048947830816215e-05,
+      "loss": 1.2942,
+      "step": 11894
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 10.46027660369873,
+      "learning_rate": 1.0048111115759529e-05,
+      "loss": 0.5026,
+      "step": 11895
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 17.637737274169922,
+      "learning_rate": 1.0047274400702842e-05,
+      "loss": 0.471,
+      "step": 11896
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 76.73388671875,
+      "learning_rate": 1.0046437685646153e-05,
+      "loss": 1.0768,
+      "step": 11897
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.723772048950195,
+      "learning_rate": 1.0045600970589466e-05,
+      "loss": 1.6841,
+      "step": 11898
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 10.220072746276855,
+      "learning_rate": 1.004476425553278e-05,
+      "loss": 2.0881,
+      "step": 11899
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.628190040588379,
+      "learning_rate": 1.0043927540476094e-05,
+      "loss": 0.1753,
+      "step": 11900
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 10.272156715393066,
+      "learning_rate": 1.0043090825419404e-05,
+      "loss": 0.6721,
+      "step": 11901
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.18677806854248,
+      "learning_rate": 1.0042254110362717e-05,
+      "loss": 1.1368,
+      "step": 11902
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 13.048550605773926,
+      "learning_rate": 1.004141739530603e-05,
+      "loss": 0.9223,
+      "step": 11903
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 16.681678771972656,
+      "learning_rate": 1.0040580680249341e-05,
+      "loss": 2.6372,
+      "step": 11904
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 5.473873615264893,
+      "learning_rate": 1.0039743965192655e-05,
+      "loss": 1.0693,
+      "step": 11905
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 14.33332633972168,
+      "learning_rate": 1.0038907250135967e-05,
+      "loss": 0.9577,
+      "step": 11906
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 14.142997741699219,
+      "learning_rate": 1.003807053507928e-05,
+      "loss": 1.1256,
+      "step": 11907
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 21.48374366760254,
+      "learning_rate": 1.0037233820022591e-05,
+      "loss": 2.2423,
+      "step": 11908
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 28.318239212036133,
+      "learning_rate": 1.0036397104965905e-05,
+      "loss": 3.5075,
+      "step": 11909
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 10.63522720336914,
+      "learning_rate": 1.0035560389909218e-05,
+      "loss": 0.3394,
+      "step": 11910
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 15.294273376464844,
+      "learning_rate": 1.0034723674852528e-05,
+      "loss": 1.1554,
+      "step": 11911
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 14.271414756774902,
+      "learning_rate": 1.0033886959795842e-05,
+      "loss": 0.4573,
+      "step": 11912
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.3673518896102905,
+      "learning_rate": 1.0033050244739156e-05,
+      "loss": 0.037,
+      "step": 11913
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.795802116394043,
+      "learning_rate": 1.0032213529682468e-05,
+      "loss": 0.5738,
+      "step": 11914
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 182.52854919433594,
+      "learning_rate": 1.003137681462578e-05,
+      "loss": 1.6654,
+      "step": 11915
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 16.431949615478516,
+      "learning_rate": 1.0030540099569093e-05,
+      "loss": 0.8176,
+      "step": 11916
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 27.6535701751709,
+      "learning_rate": 1.0029703384512405e-05,
+      "loss": 2.2308,
+      "step": 11917
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 16.68084144592285,
+      "learning_rate": 1.0028866669455717e-05,
+      "loss": 1.2446,
+      "step": 11918
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.015377521514893,
+      "learning_rate": 1.0028029954399031e-05,
+      "loss": 1.03,
+      "step": 11919
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 59.408451080322266,
+      "learning_rate": 1.0027193239342343e-05,
+      "loss": 1.3593,
+      "step": 11920
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 112.39595794677734,
+      "learning_rate": 1.0026356524285656e-05,
+      "loss": 2.6855,
+      "step": 11921
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.35781192779541,
+      "learning_rate": 1.0025519809228967e-05,
+      "loss": 2.5156,
+      "step": 11922
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 3.9637932777404785,
+      "learning_rate": 1.002468309417228e-05,
+      "loss": 0.4721,
+      "step": 11923
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 49.070621490478516,
+      "learning_rate": 1.0023846379115594e-05,
+      "loss": 0.9895,
+      "step": 11924
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 17.80061912536621,
+      "learning_rate": 1.0023009664058904e-05,
+      "loss": 0.7227,
+      "step": 11925
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 12.987116813659668,
+      "learning_rate": 1.0022172949002218e-05,
+      "loss": 0.3634,
+      "step": 11926
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 6.7396697998046875,
+      "learning_rate": 1.0021336233945532e-05,
+      "loss": 0.936,
+      "step": 11927
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.231090068817139,
+      "learning_rate": 1.0020499518888844e-05,
+      "loss": 0.8827,
+      "step": 11928
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 4.010375499725342,
+      "learning_rate": 1.0019662803832155e-05,
+      "loss": 0.4028,
+      "step": 11929
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.310550689697266,
+      "learning_rate": 1.0018826088775469e-05,
+      "loss": 0.2468,
+      "step": 11930
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 12.716805458068848,
+      "learning_rate": 1.0017989373718781e-05,
+      "loss": 0.373,
+      "step": 11931
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 98.79090118408203,
+      "learning_rate": 1.0017152658662093e-05,
+      "loss": 1.7562,
+      "step": 11932
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.764037609100342,
+      "learning_rate": 1.0016315943605405e-05,
+      "loss": 0.3759,
+      "step": 11933
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 6.833138465881348,
+      "learning_rate": 1.0015479228548719e-05,
+      "loss": 0.3039,
+      "step": 11934
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.394643783569336,
+      "learning_rate": 1.0014642513492032e-05,
+      "loss": 0.3746,
+      "step": 11935
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 13.278569221496582,
+      "learning_rate": 1.0013805798435343e-05,
+      "loss": 0.7913,
+      "step": 11936
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 123.66014099121094,
+      "learning_rate": 1.0012969083378656e-05,
+      "loss": 0.8893,
+      "step": 11937
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 4.467344284057617,
+      "learning_rate": 1.001213236832197e-05,
+      "loss": 0.2551,
+      "step": 11938
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 34.95133590698242,
+      "learning_rate": 1.001129565326528e-05,
+      "loss": 2.1504,
+      "step": 11939
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 34.456600189208984,
+      "learning_rate": 1.0010458938208594e-05,
+      "loss": 2.4087,
+      "step": 11940
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 15.087719917297363,
+      "learning_rate": 1.0009622223151907e-05,
+      "loss": 1.2157,
+      "step": 11941
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 33.45515441894531,
+      "learning_rate": 1.000878550809522e-05,
+      "loss": 1.5863,
+      "step": 11942
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.268248558044434,
+      "learning_rate": 1.0007948793038531e-05,
+      "loss": 0.7718,
+      "step": 11943
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 197.91110229492188,
+      "learning_rate": 1.0007112077981845e-05,
+      "loss": 2.3887,
+      "step": 11944
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.988476753234863,
+      "learning_rate": 1.0006275362925157e-05,
+      "loss": 1.7683,
+      "step": 11945
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.742283821105957,
+      "learning_rate": 1.0005438647868469e-05,
+      "loss": 1.1903,
+      "step": 11946
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 17.585189819335938,
+      "learning_rate": 1.000460193281178e-05,
+      "loss": 1.2422,
+      "step": 11947
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 4.5751471519470215,
+      "learning_rate": 1.0003765217755094e-05,
+      "loss": 0.0918,
+      "step": 11948
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 12.938639640808105,
+      "learning_rate": 1.0002928502698408e-05,
+      "loss": 0.7853,
+      "step": 11949
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 25.627052307128906,
+      "learning_rate": 1.0002091787641718e-05,
+      "loss": 2.8248,
+      "step": 11950
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 55.588802337646484,
+      "learning_rate": 1.0001255072585032e-05,
+      "loss": 2.367,
+      "step": 11951
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.790284156799316,
+      "learning_rate": 1.0000418357528346e-05,
+      "loss": 0.5722,
+      "step": 11952
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 141.81297302246094,
+      "learning_rate": 9.999581642471658e-06,
+      "loss": 1.7636,
+      "step": 11953
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.381845474243164,
+      "learning_rate": 9.99874492741497e-06,
+      "loss": 0.326,
+      "step": 11954
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 21.792057037353516,
+      "learning_rate": 9.997908212358283e-06,
+      "loss": 1.5584,
+      "step": 11955
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 6.758552074432373,
+      "learning_rate": 9.997071497301595e-06,
+      "loss": 0.5703,
+      "step": 11956
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 15.068915367126465,
+      "learning_rate": 9.996234782244907e-06,
+      "loss": 2.2138,
+      "step": 11957
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 25.37798309326172,
+      "learning_rate": 9.99539806718822e-06,
+      "loss": 2.4027,
+      "step": 11958
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 16.0709171295166,
+      "learning_rate": 9.994561352131533e-06,
+      "loss": 1.8126,
+      "step": 11959
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 19.48594856262207,
+      "learning_rate": 9.993724637074845e-06,
+      "loss": 1.2439,
+      "step": 11960
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.05398178100586,
+      "learning_rate": 9.992887922018157e-06,
+      "loss": 0.5253,
+      "step": 11961
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 17.040876388549805,
+      "learning_rate": 9.99205120696147e-06,
+      "loss": 0.7245,
+      "step": 11962
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 11.07220458984375,
+      "learning_rate": 9.991214491904782e-06,
+      "loss": 1.1667,
+      "step": 11963
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 25.889963150024414,
+      "learning_rate": 9.990377776848094e-06,
+      "loss": 1.5575,
+      "step": 11964
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 6.3265814781188965,
+      "learning_rate": 9.989541061791408e-06,
+      "loss": 0.2746,
+      "step": 11965
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 22.10900115966797,
+      "learning_rate": 9.98870434673472e-06,
+      "loss": 1.2605,
+      "step": 11966
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 15.18563461303711,
+      "learning_rate": 9.987867631678033e-06,
+      "loss": 1.4546,
+      "step": 11967
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 16.048877716064453,
+      "learning_rate": 9.987030916621345e-06,
+      "loss": 0.7405,
+      "step": 11968
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 12.02132511138916,
+      "learning_rate": 9.986194201564659e-06,
+      "loss": 1.2132,
+      "step": 11969
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.891904830932617,
+      "learning_rate": 9.985357486507971e-06,
+      "loss": 2.4866,
+      "step": 11970
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 5.975489139556885,
+      "learning_rate": 9.984520771451283e-06,
+      "loss": 1.2783,
+      "step": 11971
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 5.509067535400391,
+      "learning_rate": 9.983684056394597e-06,
+      "loss": 0.5775,
+      "step": 11972
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.878676891326904,
+      "learning_rate": 9.982847341337909e-06,
+      "loss": 1.0309,
+      "step": 11973
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 9.921243667602539,
+      "learning_rate": 9.98201062628122e-06,
+      "loss": 0.5512,
+      "step": 11974
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 68.78978729248047,
+      "learning_rate": 9.981173911224533e-06,
+      "loss": 2.0138,
+      "step": 11975
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 12.189335823059082,
+      "learning_rate": 9.980337196167846e-06,
+      "loss": 0.6528,
+      "step": 11976
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.658945083618164,
+      "learning_rate": 9.979500481111158e-06,
+      "loss": 0.5057,
+      "step": 11977
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 51.37008285522461,
+      "learning_rate": 9.97866376605447e-06,
+      "loss": 1.6466,
+      "step": 11978
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 6.714479923248291,
+      "learning_rate": 9.977827050997784e-06,
+      "loss": 0.7203,
+      "step": 11979
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 25.759918212890625,
+      "learning_rate": 9.976990335941096e-06,
+      "loss": 1.7984,
+      "step": 11980
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 29.47275161743164,
+      "learning_rate": 9.97615362088441e-06,
+      "loss": 0.9928,
+      "step": 11981
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.146074295043945,
+      "learning_rate": 9.975316905827721e-06,
+      "loss": 0.5818,
+      "step": 11982
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.408407211303711,
+      "learning_rate": 9.974480190771035e-06,
+      "loss": 0.173,
+      "step": 11983
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.282600402832031,
+      "learning_rate": 9.973643475714347e-06,
+      "loss": 1.6004,
+      "step": 11984
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.530019760131836,
+      "learning_rate": 9.972806760657659e-06,
+      "loss": 0.836,
+      "step": 11985
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.769789695739746,
+      "learning_rate": 9.97197004560097e-06,
+      "loss": 0.6612,
+      "step": 11986
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 48.7542610168457,
+      "learning_rate": 9.971133330544284e-06,
+      "loss": 1.4747,
+      "step": 11987
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 24.739336013793945,
+      "learning_rate": 9.970296615487596e-06,
+      "loss": 0.6906,
+      "step": 11988
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 15.494034767150879,
+      "learning_rate": 9.969459900430908e-06,
+      "loss": 0.6135,
+      "step": 11989
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 13.59770393371582,
+      "learning_rate": 9.968623185374222e-06,
+      "loss": 1.0242,
+      "step": 11990
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 17.987462997436523,
+      "learning_rate": 9.967786470317534e-06,
+      "loss": 1.5997,
+      "step": 11991
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 13.37804126739502,
+      "learning_rate": 9.966949755260846e-06,
+      "loss": 0.8806,
+      "step": 11992
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.933453559875488,
+      "learning_rate": 9.96611304020416e-06,
+      "loss": 0.7462,
+      "step": 11993
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 8.850107192993164,
+      "learning_rate": 9.965276325147472e-06,
+      "loss": 0.9161,
+      "step": 11994
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.267704963684082,
+      "learning_rate": 9.964439610090785e-06,
+      "loss": 1.1713,
+      "step": 11995
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 9.207832336425781,
+      "learning_rate": 9.963602895034097e-06,
+      "loss": 0.4839,
+      "step": 11996
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 16.171329498291016,
+      "learning_rate": 9.96276617997741e-06,
+      "loss": 2.0192,
+      "step": 11997
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 16.221424102783203,
+      "learning_rate": 9.961929464920723e-06,
+      "loss": 0.9424,
+      "step": 11998
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 28.946025848388672,
+      "learning_rate": 9.961092749864035e-06,
+      "loss": 1.7608,
+      "step": 11999
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 6.948938369750977,
+      "learning_rate": 9.960256034807347e-06,
+      "loss": 0.6615,
+      "step": 12000
+    },
+    {
+      "epoch": 1.51,
+      "eval_loss": 0.08853369951248169,
+      "eval_runtime": 95.5204,
+      "eval_samples_per_second": 37.081,
+      "eval_steps_per_second": 37.081,
+      "step": 12000
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 8.19046401977539,
+      "learning_rate": 9.95941931975066e-06,
+      "loss": 2.1622,
+      "step": 12001
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 11.977171897888184,
+      "learning_rate": 9.958582604693972e-06,
+      "loss": 1.024,
+      "step": 12002
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 13.06263542175293,
+      "learning_rate": 9.957745889637284e-06,
+      "loss": 0.3926,
+      "step": 12003
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 22.671180725097656,
+      "learning_rate": 9.956909174580598e-06,
+      "loss": 1.0962,
+      "step": 12004
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 23.219268798828125,
+      "learning_rate": 9.95607245952391e-06,
+      "loss": 1.1272,
+      "step": 12005
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 14.273165702819824,
+      "learning_rate": 9.955235744467222e-06,
+      "loss": 1.0359,
+      "step": 12006
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 12.979179382324219,
+      "learning_rate": 9.954399029410535e-06,
+      "loss": 1.2394,
+      "step": 12007
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 22.269634246826172,
+      "learning_rate": 9.953562314353847e-06,
+      "loss": 1.9392,
+      "step": 12008
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 21.703006744384766,
+      "learning_rate": 9.952725599297161e-06,
+      "loss": 2.0548,
+      "step": 12009
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 16.14089584350586,
+      "learning_rate": 9.951888884240473e-06,
+      "loss": 0.9636,
+      "step": 12010
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 17.844032287597656,
+      "learning_rate": 9.951052169183787e-06,
+      "loss": 1.3561,
+      "step": 12011
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 4.559155464172363,
+      "learning_rate": 9.950215454127099e-06,
+      "loss": 0.1477,
+      "step": 12012
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.72029972076416,
+      "learning_rate": 9.94937873907041e-06,
+      "loss": 0.7245,
+      "step": 12013
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 11.713088035583496,
+      "learning_rate": 9.948542024013722e-06,
+      "loss": 1.2342,
+      "step": 12014
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 15.37183666229248,
+      "learning_rate": 9.947705308957034e-06,
+      "loss": 1.161,
+      "step": 12015
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 11.082987785339355,
+      "learning_rate": 9.946868593900348e-06,
+      "loss": 1.2575,
+      "step": 12016
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 16.346601486206055,
+      "learning_rate": 9.94603187884366e-06,
+      "loss": 1.1751,
+      "step": 12017
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 8.998777389526367,
+      "learning_rate": 9.945195163786974e-06,
+      "loss": 2.0404,
+      "step": 12018
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 8.834567070007324,
+      "learning_rate": 9.944358448730286e-06,
+      "loss": 1.4698,
+      "step": 12019
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 37.00133514404297,
+      "learning_rate": 9.943521733673598e-06,
+      "loss": 2.067,
+      "step": 12020
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 13.402094841003418,
+      "learning_rate": 9.942685018616911e-06,
+      "loss": 0.5756,
+      "step": 12021
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 14.236288070678711,
+      "learning_rate": 9.941848303560223e-06,
+      "loss": 1.0936,
+      "step": 12022
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.448648452758789,
+      "learning_rate": 9.941011588503537e-06,
+      "loss": 1.0993,
+      "step": 12023
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 11.812121391296387,
+      "learning_rate": 9.940174873446849e-06,
+      "loss": 0.9132,
+      "step": 12024
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 59.94866943359375,
+      "learning_rate": 9.939338158390162e-06,
+      "loss": 1.2881,
+      "step": 12025
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 8.625561714172363,
+      "learning_rate": 9.938501443333474e-06,
+      "loss": 0.7345,
+      "step": 12026
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 16.29551124572754,
+      "learning_rate": 9.937664728276786e-06,
+      "loss": 0.7359,
+      "step": 12027
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 7.665041446685791,
+      "learning_rate": 9.936828013220098e-06,
+      "loss": 0.733,
+      "step": 12028
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 11.574849128723145,
+      "learning_rate": 9.93599129816341e-06,
+      "loss": 1.3198,
+      "step": 12029
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 6.037437438964844,
+      "learning_rate": 9.935154583106724e-06,
+      "loss": 1.5453,
+      "step": 12030
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 14.004814147949219,
+      "learning_rate": 9.934317868050036e-06,
+      "loss": 0.5832,
+      "step": 12031
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 15.221983909606934,
+      "learning_rate": 9.93348115299335e-06,
+      "loss": 0.9652,
+      "step": 12032
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 5.101906776428223,
+      "learning_rate": 9.932644437936661e-06,
+      "loss": 0.4963,
+      "step": 12033
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 10.096247673034668,
+      "learning_rate": 9.931807722879973e-06,
+      "loss": 0.887,
+      "step": 12034
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 111.1515884399414,
+      "learning_rate": 9.930971007823287e-06,
+      "loss": 4.7195,
+      "step": 12035
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 21.20841407775879,
+      "learning_rate": 9.930134292766599e-06,
+      "loss": 1.7936,
+      "step": 12036
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 27.25119400024414,
+      "learning_rate": 9.929297577709913e-06,
+      "loss": 1.3645,
+      "step": 12037
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 52.72514343261719,
+      "learning_rate": 9.928460862653225e-06,
+      "loss": 0.5084,
+      "step": 12038
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 6.796664714813232,
+      "learning_rate": 9.927624147596537e-06,
+      "loss": 0.2302,
+      "step": 12039
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 13.538002967834473,
+      "learning_rate": 9.92678743253985e-06,
+      "loss": 0.735,
+      "step": 12040
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 68.86786651611328,
+      "learning_rate": 9.925950717483162e-06,
+      "loss": 3.0572,
+      "step": 12041
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 13.591713905334473,
+      "learning_rate": 9.925114002426474e-06,
+      "loss": 2.0597,
+      "step": 12042
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 4.568581581115723,
+      "learning_rate": 9.924277287369786e-06,
+      "loss": 1.7249,
+      "step": 12043
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 11.677054405212402,
+      "learning_rate": 9.9234405723131e-06,
+      "loss": 1.9962,
+      "step": 12044
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 7.531463623046875,
+      "learning_rate": 9.922603857256412e-06,
+      "loss": 0.4709,
+      "step": 12045
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 17.33578872680664,
+      "learning_rate": 9.921767142199724e-06,
+      "loss": 0.9133,
+      "step": 12046
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 14.018386840820312,
+      "learning_rate": 9.920930427143037e-06,
+      "loss": 1.7442,
+      "step": 12047
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 16.10979652404785,
+      "learning_rate": 9.92009371208635e-06,
+      "loss": 0.998,
+      "step": 12048
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 29.893558502197266,
+      "learning_rate": 9.919256997029663e-06,
+      "loss": 2.026,
+      "step": 12049
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 8.823445320129395,
+      "learning_rate": 9.918420281972975e-06,
+      "loss": 1.3437,
+      "step": 12050
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 13.906524658203125,
+      "learning_rate": 9.917583566916289e-06,
+      "loss": 1.0415,
+      "step": 12051
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 5.505490303039551,
+      "learning_rate": 9.9167468518596e-06,
+      "loss": 1.6757,
+      "step": 12052
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 59.318851470947266,
+      "learning_rate": 9.915910136802912e-06,
+      "loss": 0.7066,
+      "step": 12053
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 12.240217208862305,
+      "learning_rate": 9.915073421746226e-06,
+      "loss": 0.3382,
+      "step": 12054
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 30.698877334594727,
+      "learning_rate": 9.914236706689538e-06,
+      "loss": 1.1438,
+      "step": 12055
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 17.633760452270508,
+      "learning_rate": 9.91339999163285e-06,
+      "loss": 1.299,
+      "step": 12056
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 13.139379501342773,
+      "learning_rate": 9.912563276576162e-06,
+      "loss": 0.7885,
+      "step": 12057
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 18.261993408203125,
+      "learning_rate": 9.911726561519476e-06,
+      "loss": 2.2095,
+      "step": 12058
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 9.138144493103027,
+      "learning_rate": 9.910889846462788e-06,
+      "loss": 1.2991,
+      "step": 12059
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 35.964271545410156,
+      "learning_rate": 9.9100531314061e-06,
+      "loss": 1.6079,
+      "step": 12060
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 16.519386291503906,
+      "learning_rate": 9.909216416349413e-06,
+      "loss": 2.1746,
+      "step": 12061
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 46.806793212890625,
+      "learning_rate": 9.908379701292725e-06,
+      "loss": 2.4845,
+      "step": 12062
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 11.36434555053711,
+      "learning_rate": 9.907542986236039e-06,
+      "loss": 0.4807,
+      "step": 12063
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 15.425649642944336,
+      "learning_rate": 9.90670627117935e-06,
+      "loss": 2.0082,
+      "step": 12064
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 14.873329162597656,
+      "learning_rate": 9.905869556122664e-06,
+      "loss": 0.9925,
+      "step": 12065
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 5.44581413269043,
+      "learning_rate": 9.905032841065976e-06,
+      "loss": 0.3184,
+      "step": 12066
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 24.377872467041016,
+      "learning_rate": 9.904196126009288e-06,
+      "loss": 1.3422,
+      "step": 12067
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 16.013059616088867,
+      "learning_rate": 9.9033594109526e-06,
+      "loss": 1.8209,
+      "step": 12068
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 7.558819770812988,
+      "learning_rate": 9.902522695895914e-06,
+      "loss": 1.0303,
+      "step": 12069
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 38.249698638916016,
+      "learning_rate": 9.901685980839226e-06,
+      "loss": 1.2263,
+      "step": 12070
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 5.480305194854736,
+      "learning_rate": 9.900849265782538e-06,
+      "loss": 0.4937,
+      "step": 12071
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 16.048473358154297,
+      "learning_rate": 9.900012550725851e-06,
+      "loss": 2.5661,
+      "step": 12072
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 16.503463745117188,
+      "learning_rate": 9.899175835669163e-06,
+      "loss": 1.272,
+      "step": 12073
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.952848434448242,
+      "learning_rate": 9.898339120612475e-06,
+      "loss": 0.7935,
+      "step": 12074
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 89.3608627319336,
+      "learning_rate": 9.897502405555789e-06,
+      "loss": 1.0207,
+      "step": 12075
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 17.02195167541504,
+      "learning_rate": 9.896665690499101e-06,
+      "loss": 1.5251,
+      "step": 12076
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 4.599850177764893,
+      "learning_rate": 9.895828975442415e-06,
+      "loss": 0.1726,
+      "step": 12077
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 21.801647186279297,
+      "learning_rate": 9.894992260385727e-06,
+      "loss": 1.5907,
+      "step": 12078
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.036595344543457,
+      "learning_rate": 9.89415554532904e-06,
+      "loss": 0.6715,
+      "step": 12079
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 42.203365325927734,
+      "learning_rate": 9.893318830272352e-06,
+      "loss": 1.5699,
+      "step": 12080
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.643888473510742,
+      "learning_rate": 9.892482115215664e-06,
+      "loss": 1.0253,
+      "step": 12081
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.382494926452637,
+      "learning_rate": 9.891645400158976e-06,
+      "loss": 0.5185,
+      "step": 12082
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.488576889038086,
+      "learning_rate": 9.89080868510229e-06,
+      "loss": 0.8755,
+      "step": 12083
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 20.551118850708008,
+      "learning_rate": 9.889971970045602e-06,
+      "loss": 3.2268,
+      "step": 12084
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 18.992177963256836,
+      "learning_rate": 9.889135254988914e-06,
+      "loss": 1.0521,
+      "step": 12085
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.187843322753906,
+      "learning_rate": 9.888298539932227e-06,
+      "loss": 1.2365,
+      "step": 12086
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.308290481567383,
+      "learning_rate": 9.88746182487554e-06,
+      "loss": 0.4318,
+      "step": 12087
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.222276210784912,
+      "learning_rate": 9.886625109818851e-06,
+      "loss": 0.0171,
+      "step": 12088
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 98.22897338867188,
+      "learning_rate": 9.885788394762165e-06,
+      "loss": 0.9283,
+      "step": 12089
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.085124969482422,
+      "learning_rate": 9.884951679705477e-06,
+      "loss": 0.4415,
+      "step": 12090
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 33.56439208984375,
+      "learning_rate": 9.88411496464879e-06,
+      "loss": 2.3724,
+      "step": 12091
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 3.0026347637176514,
+      "learning_rate": 9.883278249592102e-06,
+      "loss": 0.0828,
+      "step": 12092
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 109.67359161376953,
+      "learning_rate": 9.882441534535416e-06,
+      "loss": 1.0562,
+      "step": 12093
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.376280784606934,
+      "learning_rate": 9.881604819478728e-06,
+      "loss": 1.3205,
+      "step": 12094
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.192483901977539,
+      "learning_rate": 9.88076810442204e-06,
+      "loss": 0.6056,
+      "step": 12095
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 75.01764678955078,
+      "learning_rate": 9.879931389365352e-06,
+      "loss": 1.1883,
+      "step": 12096
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.584369659423828,
+      "learning_rate": 9.879094674308664e-06,
+      "loss": 1.0283,
+      "step": 12097
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 27.986286163330078,
+      "learning_rate": 9.878257959251977e-06,
+      "loss": 2.203,
+      "step": 12098
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 14.805858612060547,
+      "learning_rate": 9.87742124419529e-06,
+      "loss": 0.8525,
+      "step": 12099
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 18.489295959472656,
+      "learning_rate": 9.876584529138603e-06,
+      "loss": 1.4147,
+      "step": 12100
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 6.825009822845459,
+      "learning_rate": 9.875747814081915e-06,
+      "loss": 0.8631,
+      "step": 12101
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.646465301513672,
+      "learning_rate": 9.874911099025227e-06,
+      "loss": 0.8423,
+      "step": 12102
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 3.3560781478881836,
+      "learning_rate": 9.87407438396854e-06,
+      "loss": 0.1448,
+      "step": 12103
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 28.39167022705078,
+      "learning_rate": 9.873237668911853e-06,
+      "loss": 0.6005,
+      "step": 12104
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 6.240129470825195,
+      "learning_rate": 9.872400953855166e-06,
+      "loss": 0.9824,
+      "step": 12105
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 47.242679595947266,
+      "learning_rate": 9.871564238798478e-06,
+      "loss": 0.8535,
+      "step": 12106
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 35.06230545043945,
+      "learning_rate": 9.87072752374179e-06,
+      "loss": 1.4405,
+      "step": 12107
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 12.651790618896484,
+      "learning_rate": 9.869890808685104e-06,
+      "loss": 0.5753,
+      "step": 12108
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.116722106933594,
+      "learning_rate": 9.869054093628416e-06,
+      "loss": 0.5279,
+      "step": 12109
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 14.222530364990234,
+      "learning_rate": 9.868217378571728e-06,
+      "loss": 0.3965,
+      "step": 12110
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 22.473535537719727,
+      "learning_rate": 9.86738066351504e-06,
+      "loss": 0.8926,
+      "step": 12111
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 27.788347244262695,
+      "learning_rate": 9.866543948458353e-06,
+      "loss": 1.889,
+      "step": 12112
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 12.01479721069336,
+      "learning_rate": 9.865707233401665e-06,
+      "loss": 0.5314,
+      "step": 12113
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.29796314239502,
+      "learning_rate": 9.864870518344979e-06,
+      "loss": 1.585,
+      "step": 12114
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 13.794572830200195,
+      "learning_rate": 9.864033803288291e-06,
+      "loss": 2.1457,
+      "step": 12115
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 10.506523132324219,
+      "learning_rate": 9.863197088231603e-06,
+      "loss": 1.245,
+      "step": 12116
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.19401741027832,
+      "learning_rate": 9.862360373174916e-06,
+      "loss": 0.4909,
+      "step": 12117
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 17.951099395751953,
+      "learning_rate": 9.861523658118228e-06,
+      "loss": 1.224,
+      "step": 12118
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 7.209202289581299,
+      "learning_rate": 9.860686943061542e-06,
+      "loss": 2.3347,
+      "step": 12119
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 25.041492462158203,
+      "learning_rate": 9.859850228004854e-06,
+      "loss": 0.6295,
+      "step": 12120
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 17.921358108520508,
+      "learning_rate": 9.859013512948166e-06,
+      "loss": 2.9414,
+      "step": 12121
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 24.798927307128906,
+      "learning_rate": 9.85817679789148e-06,
+      "loss": 1.6251,
+      "step": 12122
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 23.21161460876465,
+      "learning_rate": 9.857340082834792e-06,
+      "loss": 1.481,
+      "step": 12123
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 16.68880844116211,
+      "learning_rate": 9.856503367778104e-06,
+      "loss": 2.7578,
+      "step": 12124
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 21.62554931640625,
+      "learning_rate": 9.855666652721416e-06,
+      "loss": 1.2452,
+      "step": 12125
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 44.3754997253418,
+      "learning_rate": 9.85482993766473e-06,
+      "loss": 1.2022,
+      "step": 12126
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 15.518648147583008,
+      "learning_rate": 9.853993222608041e-06,
+      "loss": 1.9063,
+      "step": 12127
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 12.686768531799316,
+      "learning_rate": 9.853156507551355e-06,
+      "loss": 2.1749,
+      "step": 12128
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 19.951292037963867,
+      "learning_rate": 9.852319792494667e-06,
+      "loss": 1.0317,
+      "step": 12129
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 12.443441390991211,
+      "learning_rate": 9.851483077437979e-06,
+      "loss": 0.7286,
+      "step": 12130
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 20.19542694091797,
+      "learning_rate": 9.850646362381292e-06,
+      "loss": 1.6099,
+      "step": 12131
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.66685962677002,
+      "learning_rate": 9.849809647324604e-06,
+      "loss": 1.0508,
+      "step": 12132
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 39.90351867675781,
+      "learning_rate": 9.848972932267918e-06,
+      "loss": 1.2983,
+      "step": 12133
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 17.983383178710938,
+      "learning_rate": 9.84813621721123e-06,
+      "loss": 0.6442,
+      "step": 12134
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 20.91132926940918,
+      "learning_rate": 9.847299502154542e-06,
+      "loss": 2.1058,
+      "step": 12135
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 34.87514877319336,
+      "learning_rate": 9.846462787097855e-06,
+      "loss": 3.0117,
+      "step": 12136
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 17.203081130981445,
+      "learning_rate": 9.845626072041167e-06,
+      "loss": 0.8342,
+      "step": 12137
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 94.10826110839844,
+      "learning_rate": 9.84478935698448e-06,
+      "loss": 1.5379,
+      "step": 12138
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 61.7381706237793,
+      "learning_rate": 9.843952641927791e-06,
+      "loss": 2.9657,
+      "step": 12139
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 62.37911605834961,
+      "learning_rate": 9.843115926871105e-06,
+      "loss": 1.5554,
+      "step": 12140
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.827415943145752,
+      "learning_rate": 9.842279211814417e-06,
+      "loss": 0.3588,
+      "step": 12141
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 55.246517181396484,
+      "learning_rate": 9.84144249675773e-06,
+      "loss": 1.7747,
+      "step": 12142
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 12.423073768615723,
+      "learning_rate": 9.840605781701043e-06,
+      "loss": 0.5489,
+      "step": 12143
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 72.16214752197266,
+      "learning_rate": 9.839769066644355e-06,
+      "loss": 2.5331,
+      "step": 12144
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.032279968261719,
+      "learning_rate": 9.838932351587668e-06,
+      "loss": 0.7265,
+      "step": 12145
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 24.480714797973633,
+      "learning_rate": 9.83809563653098e-06,
+      "loss": 1.3422,
+      "step": 12146
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 7.498961448669434,
+      "learning_rate": 9.837258921474294e-06,
+      "loss": 0.615,
+      "step": 12147
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.6980987787246704,
+      "learning_rate": 9.836422206417606e-06,
+      "loss": 0.0274,
+      "step": 12148
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 40.88421630859375,
+      "learning_rate": 9.835585491360918e-06,
+      "loss": 1.3161,
+      "step": 12149
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.074402809143066,
+      "learning_rate": 9.83474877630423e-06,
+      "loss": 1.012,
+      "step": 12150
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 11.468755722045898,
+      "learning_rate": 9.833912061247543e-06,
+      "loss": 0.818,
+      "step": 12151
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 22.78920555114746,
+      "learning_rate": 9.833075346190855e-06,
+      "loss": 1.5699,
+      "step": 12152
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 6.2296528816223145,
+      "learning_rate": 9.832238631134167e-06,
+      "loss": 0.2785,
+      "step": 12153
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 33.91434097290039,
+      "learning_rate": 9.83140191607748e-06,
+      "loss": 1.6412,
+      "step": 12154
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 9.941302299499512,
+      "learning_rate": 9.830565201020793e-06,
+      "loss": 1.4619,
+      "step": 12155
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.366573333740234,
+      "learning_rate": 9.829728485964106e-06,
+      "loss": 0.5372,
+      "step": 12156
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.877475261688232,
+      "learning_rate": 9.828891770907418e-06,
+      "loss": 1.6623,
+      "step": 12157
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 6.994818687438965,
+      "learning_rate": 9.82805505585073e-06,
+      "loss": 1.0525,
+      "step": 12158
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 16.33742904663086,
+      "learning_rate": 9.827218340794044e-06,
+      "loss": 1.203,
+      "step": 12159
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 9.728849411010742,
+      "learning_rate": 9.826381625737356e-06,
+      "loss": 1.2409,
+      "step": 12160
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.891305923461914,
+      "learning_rate": 9.82554491068067e-06,
+      "loss": 0.1906,
+      "step": 12161
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 17.807693481445312,
+      "learning_rate": 9.824708195623982e-06,
+      "loss": 1.5453,
+      "step": 12162
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 18.572694778442383,
+      "learning_rate": 9.823871480567294e-06,
+      "loss": 1.3513,
+      "step": 12163
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 12.814281463623047,
+      "learning_rate": 9.823034765510605e-06,
+      "loss": 1.9288,
+      "step": 12164
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.732547760009766,
+      "learning_rate": 9.822198050453919e-06,
+      "loss": 0.8764,
+      "step": 12165
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 23.086669921875,
+      "learning_rate": 9.821361335397231e-06,
+      "loss": 0.7889,
+      "step": 12166
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 6.248536586761475,
+      "learning_rate": 9.820524620340543e-06,
+      "loss": 0.1673,
+      "step": 12167
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 13.271327018737793,
+      "learning_rate": 9.819687905283857e-06,
+      "loss": 0.4454,
+      "step": 12168
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 37.87796401977539,
+      "learning_rate": 9.818851190227169e-06,
+      "loss": 1.4341,
+      "step": 12169
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 17.368616104125977,
+      "learning_rate": 9.818014475170482e-06,
+      "loss": 1.4188,
+      "step": 12170
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 20.290699005126953,
+      "learning_rate": 9.817177760113794e-06,
+      "loss": 0.4338,
+      "step": 12171
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 14.502118110656738,
+      "learning_rate": 9.816341045057106e-06,
+      "loss": 1.9969,
+      "step": 12172
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 14.237765312194824,
+      "learning_rate": 9.81550433000042e-06,
+      "loss": 1.4782,
+      "step": 12173
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 20.91703987121582,
+      "learning_rate": 9.814667614943732e-06,
+      "loss": 0.6778,
+      "step": 12174
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 21.437503814697266,
+      "learning_rate": 9.813830899887045e-06,
+      "loss": 1.3012,
+      "step": 12175
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 12.240067481994629,
+      "learning_rate": 9.812994184830357e-06,
+      "loss": 2.2071,
+      "step": 12176
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 6.719193935394287,
+      "learning_rate": 9.81215746977367e-06,
+      "loss": 0.2437,
+      "step": 12177
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 36.355289459228516,
+      "learning_rate": 9.811320754716981e-06,
+      "loss": 0.8935,
+      "step": 12178
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 17.33690643310547,
+      "learning_rate": 9.810484039660293e-06,
+      "loss": 0.8497,
+      "step": 12179
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.777118682861328,
+      "learning_rate": 9.809647324603607e-06,
+      "loss": 1.6085,
+      "step": 12180
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 41.65799331665039,
+      "learning_rate": 9.808810609546919e-06,
+      "loss": 1.9959,
+      "step": 12181
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 69.19049835205078,
+      "learning_rate": 9.807973894490233e-06,
+      "loss": 1.7897,
+      "step": 12182
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.965083122253418,
+      "learning_rate": 9.807137179433544e-06,
+      "loss": 1.1403,
+      "step": 12183
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 20.19078254699707,
+      "learning_rate": 9.806300464376858e-06,
+      "loss": 1.2492,
+      "step": 12184
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 25.6923885345459,
+      "learning_rate": 9.80546374932017e-06,
+      "loss": 0.946,
+      "step": 12185
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 140.13514709472656,
+      "learning_rate": 9.804627034263482e-06,
+      "loss": 1.6308,
+      "step": 12186
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 44.56431579589844,
+      "learning_rate": 9.803790319206796e-06,
+      "loss": 1.266,
+      "step": 12187
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.407170295715332,
+      "learning_rate": 9.802953604150108e-06,
+      "loss": 0.8196,
+      "step": 12188
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 19.636802673339844,
+      "learning_rate": 9.80211688909342e-06,
+      "loss": 1.4294,
+      "step": 12189
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 203.92636108398438,
+      "learning_rate": 9.801280174036733e-06,
+      "loss": 2.3192,
+      "step": 12190
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.777228832244873,
+      "learning_rate": 9.800443458980045e-06,
+      "loss": 0.907,
+      "step": 12191
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.198338508605957,
+      "learning_rate": 9.799606743923357e-06,
+      "loss": 0.3146,
+      "step": 12192
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 28.228805541992188,
+      "learning_rate": 9.798770028866669e-06,
+      "loss": 2.2984,
+      "step": 12193
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 8.053431510925293,
+      "learning_rate": 9.797933313809983e-06,
+      "loss": 0.3273,
+      "step": 12194
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 14.494138717651367,
+      "learning_rate": 9.797096598753295e-06,
+      "loss": 2.056,
+      "step": 12195
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 15.551849365234375,
+      "learning_rate": 9.796259883696608e-06,
+      "loss": 1.273,
+      "step": 12196
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 9.078676223754883,
+      "learning_rate": 9.79542316863992e-06,
+      "loss": 1.6286,
+      "step": 12197
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 22.557003021240234,
+      "learning_rate": 9.794586453583234e-06,
+      "loss": 0.9189,
+      "step": 12198
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 9.545269012451172,
+      "learning_rate": 9.793749738526546e-06,
+      "loss": 0.9631,
+      "step": 12199
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 12.034403800964355,
+      "learning_rate": 9.792913023469858e-06,
+      "loss": 0.7735,
+      "step": 12200
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 73.09227752685547,
+      "learning_rate": 9.792076308413172e-06,
+      "loss": 1.7054,
+      "step": 12201
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 40.55620193481445,
+      "learning_rate": 9.791239593356483e-06,
+      "loss": 1.1453,
+      "step": 12202
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 15.765918731689453,
+      "learning_rate": 9.790402878299795e-06,
+      "loss": 0.5375,
+      "step": 12203
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 16.58881378173828,
+      "learning_rate": 9.789566163243109e-06,
+      "loss": 0.804,
+      "step": 12204
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 4.995907783508301,
+      "learning_rate": 9.788729448186421e-06,
+      "loss": 0.2483,
+      "step": 12205
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.38837718963623,
+      "learning_rate": 9.787892733129733e-06,
+      "loss": 0.9539,
+      "step": 12206
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 16.520999908447266,
+      "learning_rate": 9.787056018073045e-06,
+      "loss": 0.8792,
+      "step": 12207
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 20.41761589050293,
+      "learning_rate": 9.786219303016359e-06,
+      "loss": 1.134,
+      "step": 12208
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 207.6176300048828,
+      "learning_rate": 9.78538258795967e-06,
+      "loss": 1.5874,
+      "step": 12209
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 17.356752395629883,
+      "learning_rate": 9.784545872902984e-06,
+      "loss": 1.6989,
+      "step": 12210
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 17.773801803588867,
+      "learning_rate": 9.783709157846296e-06,
+      "loss": 0.7458,
+      "step": 12211
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 5.917456150054932,
+      "learning_rate": 9.78287244278961e-06,
+      "loss": 1.4179,
+      "step": 12212
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 55.027442932128906,
+      "learning_rate": 9.782035727732922e-06,
+      "loss": 4.6883,
+      "step": 12213
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 33.28546905517578,
+      "learning_rate": 9.781199012676234e-06,
+      "loss": 2.6266,
+      "step": 12214
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 11.39370059967041,
+      "learning_rate": 9.780362297619547e-06,
+      "loss": 1.6386,
+      "step": 12215
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 71.60021209716797,
+      "learning_rate": 9.77952558256286e-06,
+      "loss": 1.3342,
+      "step": 12216
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.1320292949676514,
+      "learning_rate": 9.778688867506171e-06,
+      "loss": 0.0901,
+      "step": 12217
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 36.4505729675293,
+      "learning_rate": 9.777852152449483e-06,
+      "loss": 1.1238,
+      "step": 12218
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 11.367158889770508,
+      "learning_rate": 9.777015437392797e-06,
+      "loss": 2.6591,
+      "step": 12219
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 16.82743263244629,
+      "learning_rate": 9.776178722336109e-06,
+      "loss": 1.5801,
+      "step": 12220
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 11.159351348876953,
+      "learning_rate": 9.77534200727942e-06,
+      "loss": 1.8518,
+      "step": 12221
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 14.667262077331543,
+      "learning_rate": 9.774505292222734e-06,
+      "loss": 1.4004,
+      "step": 12222
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.793519496917725,
+      "learning_rate": 9.773668577166046e-06,
+      "loss": 0.5112,
+      "step": 12223
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 4.275669574737549,
+      "learning_rate": 9.77283186210936e-06,
+      "loss": 0.5828,
+      "step": 12224
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 38.114418029785156,
+      "learning_rate": 9.771995147052672e-06,
+      "loss": 1.5612,
+      "step": 12225
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 6.480970859527588,
+      "learning_rate": 9.771158431995984e-06,
+      "loss": 0.2714,
+      "step": 12226
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 32.20256805419922,
+      "learning_rate": 9.770321716939298e-06,
+      "loss": 0.5955,
+      "step": 12227
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.788589477539062,
+      "learning_rate": 9.76948500188261e-06,
+      "loss": 2.9237,
+      "step": 12228
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 20.747760772705078,
+      "learning_rate": 9.768648286825923e-06,
+      "loss": 1.2273,
+      "step": 12229
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 12.737183570861816,
+      "learning_rate": 9.767811571769235e-06,
+      "loss": 1.8664,
+      "step": 12230
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.263840675354004,
+      "learning_rate": 9.766974856712547e-06,
+      "loss": 0.4733,
+      "step": 12231
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 22.47993278503418,
+      "learning_rate": 9.766138141655859e-06,
+      "loss": 2.0756,
+      "step": 12232
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 15.143706321716309,
+      "learning_rate": 9.765301426599173e-06,
+      "loss": 0.8412,
+      "step": 12233
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 7.365832805633545,
+      "learning_rate": 9.764464711542485e-06,
+      "loss": 0.2762,
+      "step": 12234
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 8.56747055053711,
+      "learning_rate": 9.763627996485797e-06,
+      "loss": 0.4914,
+      "step": 12235
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 56.73297882080078,
+      "learning_rate": 9.76279128142911e-06,
+      "loss": 0.9943,
+      "step": 12236
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.1040303707122803,
+      "learning_rate": 9.761954566372422e-06,
+      "loss": 0.0166,
+      "step": 12237
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 19.92567253112793,
+      "learning_rate": 9.761117851315736e-06,
+      "loss": 0.2826,
+      "step": 12238
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 40.35111618041992,
+      "learning_rate": 9.760281136259048e-06,
+      "loss": 2.1398,
+      "step": 12239
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 7.14793586730957,
+      "learning_rate": 9.75944442120236e-06,
+      "loss": 1.2138,
+      "step": 12240
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 125.70104217529297,
+      "learning_rate": 9.758607706145673e-06,
+      "loss": 1.2716,
+      "step": 12241
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 27.32240867614746,
+      "learning_rate": 9.757770991088985e-06,
+      "loss": 3.5865,
+      "step": 12242
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 28.979501724243164,
+      "learning_rate": 9.756934276032299e-06,
+      "loss": 1.6377,
+      "step": 12243
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.755829811096191,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 1.2625,
+      "step": 12244
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 15.504048347473145,
+      "learning_rate": 9.755260845918923e-06,
+      "loss": 1.3819,
+      "step": 12245
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 13.636133193969727,
+      "learning_rate": 9.754424130862235e-06,
+      "loss": 1.5915,
+      "step": 12246
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.099166870117188,
+      "learning_rate": 9.753587415805549e-06,
+      "loss": 1.1964,
+      "step": 12247
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 18.510513305664062,
+      "learning_rate": 9.75275070074886e-06,
+      "loss": 1.5004,
+      "step": 12248
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 32.700531005859375,
+      "learning_rate": 9.751913985692172e-06,
+      "loss": 1.1932,
+      "step": 12249
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 18.577503204345703,
+      "learning_rate": 9.751077270635486e-06,
+      "loss": 1.8497,
+      "step": 12250
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 8.545149803161621,
+      "learning_rate": 9.750240555578798e-06,
+      "loss": 1.2184,
+      "step": 12251
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 22.767305374145508,
+      "learning_rate": 9.749403840522112e-06,
+      "loss": 1.8692,
+      "step": 12252
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 59.19416046142578,
+      "learning_rate": 9.748567125465424e-06,
+      "loss": 2.6589,
+      "step": 12253
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 6.677204608917236,
+      "learning_rate": 9.747730410408736e-06,
+      "loss": 0.2296,
+      "step": 12254
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.923628807067871,
+      "learning_rate": 9.74689369535205e-06,
+      "loss": 1.023,
+      "step": 12255
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.695693016052246,
+      "learning_rate": 9.746056980295361e-06,
+      "loss": 1.4392,
+      "step": 12256
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 47.62434768676758,
+      "learning_rate": 9.745220265238675e-06,
+      "loss": 2.246,
+      "step": 12257
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 3.9003381729125977,
+      "learning_rate": 9.744383550181987e-06,
+      "loss": 0.0779,
+      "step": 12258
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 20.364971160888672,
+      "learning_rate": 9.743546835125299e-06,
+      "loss": 0.5831,
+      "step": 12259
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.905691146850586,
+      "learning_rate": 9.74271012006861e-06,
+      "loss": 1.1294,
+      "step": 12260
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 24.679933547973633,
+      "learning_rate": 9.741873405011923e-06,
+      "loss": 2.6039,
+      "step": 12261
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.308293342590332,
+      "learning_rate": 9.741036689955236e-06,
+      "loss": 0.9243,
+      "step": 12262
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.862445831298828,
+      "learning_rate": 9.740199974898548e-06,
+      "loss": 2.648,
+      "step": 12263
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.263579368591309,
+      "learning_rate": 9.739363259841862e-06,
+      "loss": 1.5139,
+      "step": 12264
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 17.342958450317383,
+      "learning_rate": 9.738526544785174e-06,
+      "loss": 2.0516,
+      "step": 12265
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.240121841430664,
+      "learning_rate": 9.737689829728488e-06,
+      "loss": 0.8314,
+      "step": 12266
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 8.022343635559082,
+      "learning_rate": 9.7368531146718e-06,
+      "loss": 0.6965,
+      "step": 12267
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 554.3936767578125,
+      "learning_rate": 9.736016399615111e-06,
+      "loss": 2.0198,
+      "step": 12268
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 36.34038543701172,
+      "learning_rate": 9.735179684558425e-06,
+      "loss": 1.6037,
+      "step": 12269
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 19.262128829956055,
+      "learning_rate": 9.734342969501737e-06,
+      "loss": 0.7036,
+      "step": 12270
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 61.7270622253418,
+      "learning_rate": 9.733506254445049e-06,
+      "loss": 0.816,
+      "step": 12271
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 7.3316450119018555,
+      "learning_rate": 9.732669539388363e-06,
+      "loss": 0.7578,
+      "step": 12272
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 26.127300262451172,
+      "learning_rate": 9.731832824331675e-06,
+      "loss": 0.7447,
+      "step": 12273
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 28.84157371520996,
+      "learning_rate": 9.730996109274987e-06,
+      "loss": 1.1236,
+      "step": 12274
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 16.017126083374023,
+      "learning_rate": 9.730159394218299e-06,
+      "loss": 0.8141,
+      "step": 12275
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 3.271279811859131,
+      "learning_rate": 9.729322679161612e-06,
+      "loss": 0.2226,
+      "step": 12276
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.12415885925293,
+      "learning_rate": 9.728485964104924e-06,
+      "loss": 1.3616,
+      "step": 12277
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.111727714538574,
+      "learning_rate": 9.727649249048238e-06,
+      "loss": 1.0948,
+      "step": 12278
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.25155258178711,
+      "learning_rate": 9.72681253399155e-06,
+      "loss": 0.543,
+      "step": 12279
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 28.79137420654297,
+      "learning_rate": 9.725975818934863e-06,
+      "loss": 0.8573,
+      "step": 12280
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 17.782323837280273,
+      "learning_rate": 9.725139103878175e-06,
+      "loss": 1.3399,
+      "step": 12281
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 13.894097328186035,
+      "learning_rate": 9.724302388821487e-06,
+      "loss": 1.2142,
+      "step": 12282
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.390398979187012,
+      "learning_rate": 9.723465673764801e-06,
+      "loss": 0.8592,
+      "step": 12283
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 24.50271987915039,
+      "learning_rate": 9.722628958708113e-06,
+      "loss": 1.0732,
+      "step": 12284
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 16.299976348876953,
+      "learning_rate": 9.721792243651425e-06,
+      "loss": 2.596,
+      "step": 12285
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 8.97154426574707,
+      "learning_rate": 9.720955528594738e-06,
+      "loss": 1.5355,
+      "step": 12286
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 27.228649139404297,
+      "learning_rate": 9.72011881353805e-06,
+      "loss": 1.1823,
+      "step": 12287
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 17.106494903564453,
+      "learning_rate": 9.719282098481362e-06,
+      "loss": 0.4884,
+      "step": 12288
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 20.892436981201172,
+      "learning_rate": 9.718445383424674e-06,
+      "loss": 1.7475,
+      "step": 12289
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 21.541584014892578,
+      "learning_rate": 9.717608668367988e-06,
+      "loss": 2.45,
+      "step": 12290
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 31.563264846801758,
+      "learning_rate": 9.7167719533113e-06,
+      "loss": 1.8688,
+      "step": 12291
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.500767707824707,
+      "learning_rate": 9.715935238254614e-06,
+      "loss": 1.4335,
+      "step": 12292
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 15.389634132385254,
+      "learning_rate": 9.715098523197926e-06,
+      "loss": 1.8416,
+      "step": 12293
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 34.42812728881836,
+      "learning_rate": 9.71426180814124e-06,
+      "loss": 1.1489,
+      "step": 12294
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 7.696777820587158,
+      "learning_rate": 9.713425093084551e-06,
+      "loss": 0.5016,
+      "step": 12295
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 24.3090877532959,
+      "learning_rate": 9.712588378027863e-06,
+      "loss": 0.8746,
+      "step": 12296
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.006400108337402,
+      "learning_rate": 9.711751662971177e-06,
+      "loss": 1.0452,
+      "step": 12297
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.543442726135254,
+      "learning_rate": 9.710914947914489e-06,
+      "loss": 1.1973,
+      "step": 12298
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.206832885742188,
+      "learning_rate": 9.7100782328578e-06,
+      "loss": 2.1005,
+      "step": 12299
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 20.02605438232422,
+      "learning_rate": 9.709241517801113e-06,
+      "loss": 1.5466,
+      "step": 12300
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 14.613446235656738,
+      "learning_rate": 9.708404802744426e-06,
+      "loss": 0.6017,
+      "step": 12301
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 26.277055740356445,
+      "learning_rate": 9.707568087687738e-06,
+      "loss": 1.4789,
+      "step": 12302
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 24.248592376708984,
+      "learning_rate": 9.70673137263105e-06,
+      "loss": 1.5301,
+      "step": 12303
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 11.80296516418457,
+      "learning_rate": 9.705894657574364e-06,
+      "loss": 0.5371,
+      "step": 12304
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.005497932434082,
+      "learning_rate": 9.705057942517676e-06,
+      "loss": 1.7511,
+      "step": 12305
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 14.882233619689941,
+      "learning_rate": 9.70422122746099e-06,
+      "loss": 1.5295,
+      "step": 12306
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 14.475922584533691,
+      "learning_rate": 9.703384512404301e-06,
+      "loss": 0.8347,
+      "step": 12307
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.40028190612793,
+      "learning_rate": 9.702547797347615e-06,
+      "loss": 0.6573,
+      "step": 12308
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 12.468551635742188,
+      "learning_rate": 9.701711082290927e-06,
+      "loss": 1.6914,
+      "step": 12309
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 45.54451370239258,
+      "learning_rate": 9.700874367234239e-06,
+      "loss": 1.5934,
+      "step": 12310
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.441662311553955,
+      "learning_rate": 9.700037652177553e-06,
+      "loss": 0.1754,
+      "step": 12311
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.537643432617188,
+      "learning_rate": 9.699200937120865e-06,
+      "loss": 0.2869,
+      "step": 12312
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 17.57350730895996,
+      "learning_rate": 9.698364222064177e-06,
+      "loss": 2.5289,
+      "step": 12313
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 61.0333366394043,
+      "learning_rate": 9.697527507007488e-06,
+      "loss": 1.6432,
+      "step": 12314
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.310492515563965,
+      "learning_rate": 9.696690791950802e-06,
+      "loss": 0.2658,
+      "step": 12315
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 28.702024459838867,
+      "learning_rate": 9.695854076894114e-06,
+      "loss": 1.0645,
+      "step": 12316
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 15.340718269348145,
+      "learning_rate": 9.695017361837426e-06,
+      "loss": 1.0408,
+      "step": 12317
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 16.828536987304688,
+      "learning_rate": 9.69418064678074e-06,
+      "loss": 2.4937,
+      "step": 12318
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 20.91560935974121,
+      "learning_rate": 9.693343931724052e-06,
+      "loss": 1.9895,
+      "step": 12319
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.444477081298828,
+      "learning_rate": 9.692507216667365e-06,
+      "loss": 0.3833,
+      "step": 12320
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 12.016524314880371,
+      "learning_rate": 9.691670501610677e-06,
+      "loss": 1.4067,
+      "step": 12321
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.002107620239258,
+      "learning_rate": 9.690833786553991e-06,
+      "loss": 0.2779,
+      "step": 12322
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 21.251346588134766,
+      "learning_rate": 9.689997071497303e-06,
+      "loss": 0.6225,
+      "step": 12323
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 17.89685821533203,
+      "learning_rate": 9.689160356440615e-06,
+      "loss": 2.476,
+      "step": 12324
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 15.73957633972168,
+      "learning_rate": 9.688323641383928e-06,
+      "loss": 1.623,
+      "step": 12325
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 5.568100929260254,
+      "learning_rate": 9.68748692632724e-06,
+      "loss": 0.6575,
+      "step": 12326
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.989686012268066,
+      "learning_rate": 9.686650211270552e-06,
+      "loss": 3.0087,
+      "step": 12327
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 17.007863998413086,
+      "learning_rate": 9.685813496213864e-06,
+      "loss": 1.5266,
+      "step": 12328
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 18.261709213256836,
+      "learning_rate": 9.684976781157178e-06,
+      "loss": 1.0027,
+      "step": 12329
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 12.94047737121582,
+      "learning_rate": 9.68414006610049e-06,
+      "loss": 0.7465,
+      "step": 12330
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 16.564481735229492,
+      "learning_rate": 9.683303351043802e-06,
+      "loss": 1.8334,
+      "step": 12331
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 65.0897216796875,
+      "learning_rate": 9.682466635987116e-06,
+      "loss": 0.7476,
+      "step": 12332
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.423556327819824,
+      "learning_rate": 9.681629920930427e-06,
+      "loss": 0.4993,
+      "step": 12333
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 22.15186309814453,
+      "learning_rate": 9.680793205873741e-06,
+      "loss": 1.0165,
+      "step": 12334
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 28.680299758911133,
+      "learning_rate": 9.679956490817053e-06,
+      "loss": 2.543,
+      "step": 12335
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 12.244583129882812,
+      "learning_rate": 9.679119775760367e-06,
+      "loss": 2.632,
+      "step": 12336
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.169486999511719,
+      "learning_rate": 9.678283060703679e-06,
+      "loss": 0.3672,
+      "step": 12337
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.6350631713867188,
+      "learning_rate": 9.67744634564699e-06,
+      "loss": 0.2428,
+      "step": 12338
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 169.20904541015625,
+      "learning_rate": 9.676609630590304e-06,
+      "loss": 2.6483,
+      "step": 12339
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 12.080873489379883,
+      "learning_rate": 9.675772915533616e-06,
+      "loss": 0.572,
+      "step": 12340
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 44.13920974731445,
+      "learning_rate": 9.674936200476928e-06,
+      "loss": 0.9094,
+      "step": 12341
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 21.638214111328125,
+      "learning_rate": 9.67409948542024e-06,
+      "loss": 1.5159,
+      "step": 12342
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.509679794311523,
+      "learning_rate": 9.673262770363554e-06,
+      "loss": 2.5113,
+      "step": 12343
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 27.861263275146484,
+      "learning_rate": 9.672426055306866e-06,
+      "loss": 1.6951,
+      "step": 12344
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.870193481445312,
+      "learning_rate": 9.671589340250178e-06,
+      "loss": 0.6811,
+      "step": 12345
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.583845138549805,
+      "learning_rate": 9.670752625193491e-06,
+      "loss": 0.6466,
+      "step": 12346
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.40594482421875,
+      "learning_rate": 9.669915910136803e-06,
+      "loss": 0.6483,
+      "step": 12347
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 6.930304527282715,
+      "learning_rate": 9.669079195080117e-06,
+      "loss": 1.7079,
+      "step": 12348
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.957571029663086,
+      "learning_rate": 9.668242480023429e-06,
+      "loss": 0.6104,
+      "step": 12349
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.939764976501465,
+      "learning_rate": 9.667405764966743e-06,
+      "loss": 1.1139,
+      "step": 12350
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 7.519237041473389,
+      "learning_rate": 9.666569049910055e-06,
+      "loss": 0.6042,
+      "step": 12351
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 15.627189636230469,
+      "learning_rate": 9.665732334853366e-06,
+      "loss": 1.5947,
+      "step": 12352
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 7.847267150878906,
+      "learning_rate": 9.664895619796678e-06,
+      "loss": 1.2697,
+      "step": 12353
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 7.785482883453369,
+      "learning_rate": 9.664058904739992e-06,
+      "loss": 0.6465,
+      "step": 12354
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 20.836950302124023,
+      "learning_rate": 9.663222189683304e-06,
+      "loss": 1.9692,
+      "step": 12355
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.960698127746582,
+      "learning_rate": 9.662385474626616e-06,
+      "loss": 1.3315,
+      "step": 12356
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 17.833974838256836,
+      "learning_rate": 9.66154875956993e-06,
+      "loss": 1.0398,
+      "step": 12357
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 24.96609115600586,
+      "learning_rate": 9.660712044513242e-06,
+      "loss": 1.0752,
+      "step": 12358
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.910917282104492,
+      "learning_rate": 9.659875329456554e-06,
+      "loss": 1.4323,
+      "step": 12359
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 19.315662384033203,
+      "learning_rate": 9.659038614399867e-06,
+      "loss": 0.7993,
+      "step": 12360
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.594182968139648,
+      "learning_rate": 9.658201899343179e-06,
+      "loss": 1.0659,
+      "step": 12361
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 15.608436584472656,
+      "learning_rate": 9.657365184286493e-06,
+      "loss": 0.6294,
+      "step": 12362
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 88.83906555175781,
+      "learning_rate": 9.656528469229805e-06,
+      "loss": 1.8261,
+      "step": 12363
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 80.29641723632812,
+      "learning_rate": 9.655691754173118e-06,
+      "loss": 1.2879,
+      "step": 12364
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 14.70491886138916,
+      "learning_rate": 9.65485503911643e-06,
+      "loss": 1.9687,
+      "step": 12365
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 15.118184089660645,
+      "learning_rate": 9.654018324059742e-06,
+      "loss": 0.9066,
+      "step": 12366
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 5.322723388671875,
+      "learning_rate": 9.653181609003054e-06,
+      "loss": 0.3755,
+      "step": 12367
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 26.06890869140625,
+      "learning_rate": 9.652344893946368e-06,
+      "loss": 1.4103,
+      "step": 12368
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 22.888280868530273,
+      "learning_rate": 9.65150817888968e-06,
+      "loss": 2.17,
+      "step": 12369
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 20.88540267944336,
+      "learning_rate": 9.650671463832992e-06,
+      "loss": 1.7313,
+      "step": 12370
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.6563138961792,
+      "learning_rate": 9.649834748776305e-06,
+      "loss": 0.4957,
+      "step": 12371
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 44.91715621948242,
+      "learning_rate": 9.648998033719617e-06,
+      "loss": 1.5765,
+      "step": 12372
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 127.05415344238281,
+      "learning_rate": 9.64816131866293e-06,
+      "loss": 3.4356,
+      "step": 12373
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 13.144569396972656,
+      "learning_rate": 9.647324603606243e-06,
+      "loss": 2.4144,
+      "step": 12374
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 16.216175079345703,
+      "learning_rate": 9.646487888549555e-06,
+      "loss": 1.3727,
+      "step": 12375
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 15.892380714416504,
+      "learning_rate": 9.645651173492869e-06,
+      "loss": 0.7439,
+      "step": 12376
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.031373977661133,
+      "learning_rate": 9.64481445843618e-06,
+      "loss": 2.0348,
+      "step": 12377
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 7.820834159851074,
+      "learning_rate": 9.643977743379494e-06,
+      "loss": 0.2667,
+      "step": 12378
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 7.736286163330078,
+      "learning_rate": 9.643141028322806e-06,
+      "loss": 1.1379,
+      "step": 12379
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 90.63997650146484,
+      "learning_rate": 9.642304313266118e-06,
+      "loss": 1.5963,
+      "step": 12380
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.048603057861328,
+      "learning_rate": 9.64146759820943e-06,
+      "loss": 0.8324,
+      "step": 12381
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 15.25834846496582,
+      "learning_rate": 9.640630883152742e-06,
+      "loss": 1.8656,
+      "step": 12382
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 4.8124003410339355,
+      "learning_rate": 9.639794168096056e-06,
+      "loss": 1.7805,
+      "step": 12383
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 12.141240119934082,
+      "learning_rate": 9.638957453039368e-06,
+      "loss": 1.5664,
+      "step": 12384
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 12.14017391204834,
+      "learning_rate": 9.638120737982681e-06,
+      "loss": 2.1406,
+      "step": 12385
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 7.157753944396973,
+      "learning_rate": 9.637284022925993e-06,
+      "loss": 0.9424,
+      "step": 12386
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 20.844449996948242,
+      "learning_rate": 9.636447307869305e-06,
+      "loss": 2.1457,
+      "step": 12387
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 5.263124465942383,
+      "learning_rate": 9.635610592812619e-06,
+      "loss": 0.2392,
+      "step": 12388
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 15.502119064331055,
+      "learning_rate": 9.63477387775593e-06,
+      "loss": 0.6059,
+      "step": 12389
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 5.6785054206848145,
+      "learning_rate": 9.633937162699244e-06,
+      "loss": 1.3948,
+      "step": 12390
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.73088264465332,
+      "learning_rate": 9.633100447642556e-06,
+      "loss": 1.4068,
+      "step": 12391
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 13.06839656829834,
+      "learning_rate": 9.63226373258587e-06,
+      "loss": 1.001,
+      "step": 12392
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.036245346069336,
+      "learning_rate": 9.631427017529182e-06,
+      "loss": 2.2736,
+      "step": 12393
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 15.826783180236816,
+      "learning_rate": 9.630590302472494e-06,
+      "loss": 2.1359,
+      "step": 12394
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 19.438514709472656,
+      "learning_rate": 9.629753587415806e-06,
+      "loss": 1.0111,
+      "step": 12395
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 14.289900779724121,
+      "learning_rate": 9.628916872359118e-06,
+      "loss": 1.2293,
+      "step": 12396
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 72.85143280029297,
+      "learning_rate": 9.628080157302432e-06,
+      "loss": 1.9941,
+      "step": 12397
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 12.056010246276855,
+      "learning_rate": 9.627243442245743e-06,
+      "loss": 0.7306,
+      "step": 12398
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 12.72350788116455,
+      "learning_rate": 9.626406727189055e-06,
+      "loss": 1.0371,
+      "step": 12399
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.367460250854492,
+      "learning_rate": 9.625570012132369e-06,
+      "loss": 0.6634,
+      "step": 12400
+    },
+    {
+      "epoch": 1.56,
+      "eval_loss": 0.09418872743844986,
+      "eval_runtime": 96.357,
+      "eval_samples_per_second": 36.759,
+      "eval_steps_per_second": 36.759,
+      "step": 12400
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.752225875854492,
+      "learning_rate": 9.624733297075681e-06,
+      "loss": 0.6213,
+      "step": 12401
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 12.529314041137695,
+      "learning_rate": 9.623896582018995e-06,
+      "loss": 0.6571,
+      "step": 12402
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 27.533313751220703,
+      "learning_rate": 9.623059866962307e-06,
+      "loss": 1.3716,
+      "step": 12403
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 6.712658405303955,
+      "learning_rate": 9.62222315190562e-06,
+      "loss": 0.6971,
+      "step": 12404
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.645259857177734,
+      "learning_rate": 9.621386436848932e-06,
+      "loss": 0.7859,
+      "step": 12405
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 17.8518123626709,
+      "learning_rate": 9.620549721792244e-06,
+      "loss": 0.8945,
+      "step": 12406
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.712150573730469,
+      "learning_rate": 9.619713006735558e-06,
+      "loss": 0.5132,
+      "step": 12407
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 25.49022674560547,
+      "learning_rate": 9.61887629167887e-06,
+      "loss": 1.8113,
+      "step": 12408
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 30.441619873046875,
+      "learning_rate": 9.618039576622182e-06,
+      "loss": 1.022,
+      "step": 12409
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 13.423223495483398,
+      "learning_rate": 9.617202861565494e-06,
+      "loss": 1.046,
+      "step": 12410
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 36.31425476074219,
+      "learning_rate": 9.616366146508807e-06,
+      "loss": 2.3593,
+      "step": 12411
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 23.63006591796875,
+      "learning_rate": 9.61552943145212e-06,
+      "loss": 0.8683,
+      "step": 12412
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 20.383546829223633,
+      "learning_rate": 9.614692716395431e-06,
+      "loss": 2.1879,
+      "step": 12413
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.273029327392578,
+      "learning_rate": 9.613856001338745e-06,
+      "loss": 0.3497,
+      "step": 12414
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 15.434338569641113,
+      "learning_rate": 9.613019286282057e-06,
+      "loss": 0.7268,
+      "step": 12415
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 14.430562973022461,
+      "learning_rate": 9.61218257122537e-06,
+      "loss": 1.358,
+      "step": 12416
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.518756866455078,
+      "learning_rate": 9.611345856168682e-06,
+      "loss": 0.7091,
+      "step": 12417
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 58.18947982788086,
+      "learning_rate": 9.610509141111996e-06,
+      "loss": 1.5999,
+      "step": 12418
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 16.584684371948242,
+      "learning_rate": 9.609672426055308e-06,
+      "loss": 0.9131,
+      "step": 12419
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 24.425582885742188,
+      "learning_rate": 9.60883571099862e-06,
+      "loss": 1.1493,
+      "step": 12420
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 112.81609344482422,
+      "learning_rate": 9.607998995941934e-06,
+      "loss": 2.2978,
+      "step": 12421
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 34.88551712036133,
+      "learning_rate": 9.607162280885246e-06,
+      "loss": 1.3919,
+      "step": 12422
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.413063049316406,
+      "learning_rate": 9.606325565828558e-06,
+      "loss": 0.5265,
+      "step": 12423
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.930394172668457,
+      "learning_rate": 9.60548885077187e-06,
+      "loss": 0.9577,
+      "step": 12424
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 22.859756469726562,
+      "learning_rate": 9.604652135715183e-06,
+      "loss": 1.2404,
+      "step": 12425
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.711565017700195,
+      "learning_rate": 9.603815420658495e-06,
+      "loss": 0.7393,
+      "step": 12426
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 27.851823806762695,
+      "learning_rate": 9.602978705601807e-06,
+      "loss": 1.8269,
+      "step": 12427
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 35.024620056152344,
+      "learning_rate": 9.60214199054512e-06,
+      "loss": 2.4503,
+      "step": 12428
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.001527786254883,
+      "learning_rate": 9.601305275488433e-06,
+      "loss": 2.1216,
+      "step": 12429
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 14.854911804199219,
+      "learning_rate": 9.600468560431746e-06,
+      "loss": 0.6204,
+      "step": 12430
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 15.91850471496582,
+      "learning_rate": 9.599631845375058e-06,
+      "loss": 1.0912,
+      "step": 12431
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 5.914279460906982,
+      "learning_rate": 9.598795130318372e-06,
+      "loss": 0.6524,
+      "step": 12432
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 15.950666427612305,
+      "learning_rate": 9.597958415261684e-06,
+      "loss": 1.0621,
+      "step": 12433
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 16.08356285095215,
+      "learning_rate": 9.597121700204996e-06,
+      "loss": 1.4678,
+      "step": 12434
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 4.259883880615234,
+      "learning_rate": 9.596284985148308e-06,
+      "loss": 0.2889,
+      "step": 12435
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 19.73095703125,
+      "learning_rate": 9.595448270091621e-06,
+      "loss": 1.0126,
+      "step": 12436
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 12.76375675201416,
+      "learning_rate": 9.594611555034933e-06,
+      "loss": 2.8601,
+      "step": 12437
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 17.41053009033203,
+      "learning_rate": 9.593774839978245e-06,
+      "loss": 0.9852,
+      "step": 12438
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 26.395841598510742,
+      "learning_rate": 9.592938124921559e-06,
+      "loss": 1.3814,
+      "step": 12439
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 17.374435424804688,
+      "learning_rate": 9.592101409864871e-06,
+      "loss": 2.6592,
+      "step": 12440
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 24.25494384765625,
+      "learning_rate": 9.591264694808183e-06,
+      "loss": 1.4085,
+      "step": 12441
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 14.291749954223633,
+      "learning_rate": 9.590427979751497e-06,
+      "loss": 1.3867,
+      "step": 12442
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 43.80595779418945,
+      "learning_rate": 9.589591264694809e-06,
+      "loss": 1.5136,
+      "step": 12443
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 25.88615608215332,
+      "learning_rate": 9.588754549638122e-06,
+      "loss": 1.3238,
+      "step": 12444
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 9.90751838684082,
+      "learning_rate": 9.587917834581434e-06,
+      "loss": 0.4805,
+      "step": 12445
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 25.095247268676758,
+      "learning_rate": 9.587081119524748e-06,
+      "loss": 1.7575,
+      "step": 12446
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 4.303838729858398,
+      "learning_rate": 9.58624440446806e-06,
+      "loss": 0.3622,
+      "step": 12447
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.002155303955078,
+      "learning_rate": 9.585407689411372e-06,
+      "loss": 0.9787,
+      "step": 12448
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 24.475616455078125,
+      "learning_rate": 9.584570974354684e-06,
+      "loss": 1.2444,
+      "step": 12449
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 27.28084373474121,
+      "learning_rate": 9.583734259297997e-06,
+      "loss": 1.5608,
+      "step": 12450
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 18.78505516052246,
+      "learning_rate": 9.58289754424131e-06,
+      "loss": 1.3009,
+      "step": 12451
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 19.323734283447266,
+      "learning_rate": 9.582060829184621e-06,
+      "loss": 0.9721,
+      "step": 12452
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 4.087686061859131,
+      "learning_rate": 9.581224114127935e-06,
+      "loss": 0.2535,
+      "step": 12453
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 19.48501205444336,
+      "learning_rate": 9.580387399071247e-06,
+      "loss": 0.4955,
+      "step": 12454
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 13.740678787231445,
+      "learning_rate": 9.579550684014559e-06,
+      "loss": 0.8021,
+      "step": 12455
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 4.655831813812256,
+      "learning_rate": 9.578713968957872e-06,
+      "loss": 0.6098,
+      "step": 12456
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 15.411494255065918,
+      "learning_rate": 9.577877253901184e-06,
+      "loss": 1.0082,
+      "step": 12457
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 4.377325534820557,
+      "learning_rate": 9.577040538844498e-06,
+      "loss": 0.2156,
+      "step": 12458
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.858933448791504,
+      "learning_rate": 9.57620382378781e-06,
+      "loss": 2.2354,
+      "step": 12459
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 6.7858195304870605,
+      "learning_rate": 9.575367108731124e-06,
+      "loss": 0.2565,
+      "step": 12460
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 6.283478260040283,
+      "learning_rate": 9.574530393674436e-06,
+      "loss": 0.6546,
+      "step": 12461
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.30803394317627,
+      "learning_rate": 9.573693678617748e-06,
+      "loss": 0.3907,
+      "step": 12462
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 178.96331787109375,
+      "learning_rate": 9.57285696356106e-06,
+      "loss": 2.536,
+      "step": 12463
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 17.98655128479004,
+      "learning_rate": 9.572020248504371e-06,
+      "loss": 2.4661,
+      "step": 12464
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 6.036463737487793,
+      "learning_rate": 9.571183533447685e-06,
+      "loss": 0.405,
+      "step": 12465
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.466330528259277,
+      "learning_rate": 9.570346818390997e-06,
+      "loss": 0.5749,
+      "step": 12466
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 25.326618194580078,
+      "learning_rate": 9.56951010333431e-06,
+      "loss": 1.4268,
+      "step": 12467
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 10.319133758544922,
+      "learning_rate": 9.568673388277623e-06,
+      "loss": 2.2784,
+      "step": 12468
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 11.29590892791748,
+      "learning_rate": 9.567836673220935e-06,
+      "loss": 0.749,
+      "step": 12469
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 53.39611053466797,
+      "learning_rate": 9.566999958164248e-06,
+      "loss": 1.4989,
+      "step": 12470
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 12.615883827209473,
+      "learning_rate": 9.56616324310756e-06,
+      "loss": 1.1332,
+      "step": 12471
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.729395866394043,
+      "learning_rate": 9.565326528050874e-06,
+      "loss": 1.5362,
+      "step": 12472
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.843531608581543,
+      "learning_rate": 9.564489812994186e-06,
+      "loss": 0.8666,
+      "step": 12473
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 15.752050399780273,
+      "learning_rate": 9.5636530979375e-06,
+      "loss": 3.0116,
+      "step": 12474
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.753077507019043,
+      "learning_rate": 9.562816382880811e-06,
+      "loss": 0.9885,
+      "step": 12475
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 12.625199317932129,
+      "learning_rate": 9.561979667824123e-06,
+      "loss": 1.6332,
+      "step": 12476
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 12.897058486938477,
+      "learning_rate": 9.561142952767435e-06,
+      "loss": 0.9332,
+      "step": 12477
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 23.377389907836914,
+      "learning_rate": 9.560306237710747e-06,
+      "loss": 1.4842,
+      "step": 12478
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 18.772571563720703,
+      "learning_rate": 9.559469522654061e-06,
+      "loss": 0.9665,
+      "step": 12479
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 15.308571815490723,
+      "learning_rate": 9.558632807597373e-06,
+      "loss": 1.1477,
+      "step": 12480
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 15.69286060333252,
+      "learning_rate": 9.557796092540687e-06,
+      "loss": 0.5971,
+      "step": 12481
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 4.5303850173950195,
+      "learning_rate": 9.556959377483999e-06,
+      "loss": 0.506,
+      "step": 12482
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 22.065988540649414,
+      "learning_rate": 9.55612266242731e-06,
+      "loss": 1.6268,
+      "step": 12483
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 68.5661392211914,
+      "learning_rate": 9.555285947370624e-06,
+      "loss": 1.652,
+      "step": 12484
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 10.060134887695312,
+      "learning_rate": 9.554449232313936e-06,
+      "loss": 0.5053,
+      "step": 12485
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 27.67550277709961,
+      "learning_rate": 9.55361251725725e-06,
+      "loss": 2.0167,
+      "step": 12486
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 11.384492874145508,
+      "learning_rate": 9.552775802200562e-06,
+      "loss": 1.5019,
+      "step": 12487
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 6.32214879989624,
+      "learning_rate": 9.551939087143874e-06,
+      "loss": 0.5649,
+      "step": 12488
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.0320892333984375,
+      "learning_rate": 9.551102372087187e-06,
+      "loss": 0.3363,
+      "step": 12489
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 4.202093601226807,
+      "learning_rate": 9.5502656570305e-06,
+      "loss": 0.1351,
+      "step": 12490
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 13.17863941192627,
+      "learning_rate": 9.549428941973811e-06,
+      "loss": 1.2579,
+      "step": 12491
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 10.442083358764648,
+      "learning_rate": 9.548592226917123e-06,
+      "loss": 1.7067,
+      "step": 12492
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.665885925292969,
+      "learning_rate": 9.547755511860437e-06,
+      "loss": 1.1546,
+      "step": 12493
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 6.95634651184082,
+      "learning_rate": 9.546918796803749e-06,
+      "loss": 0.2485,
+      "step": 12494
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 26.555082321166992,
+      "learning_rate": 9.546082081747062e-06,
+      "loss": 1.2272,
+      "step": 12495
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 12.900822639465332,
+      "learning_rate": 9.545245366690374e-06,
+      "loss": 1.0477,
+      "step": 12496
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.701725959777832,
+      "learning_rate": 9.544408651633686e-06,
+      "loss": 0.8932,
+      "step": 12497
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 35.35837173461914,
+      "learning_rate": 9.543571936577e-06,
+      "loss": 2.5717,
+      "step": 12498
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 589.602783203125,
+      "learning_rate": 9.542735221520312e-06,
+      "loss": 1.1103,
+      "step": 12499
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.06594467163086,
+      "learning_rate": 9.541898506463626e-06,
+      "loss": 0.4326,
+      "step": 12500
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 117.72836303710938,
+      "learning_rate": 9.541061791406938e-06,
+      "loss": 0.6035,
+      "step": 12501
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 83.0579833984375,
+      "learning_rate": 9.54022507635025e-06,
+      "loss": 0.5596,
+      "step": 12502
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 13.687908172607422,
+      "learning_rate": 9.539388361293563e-06,
+      "loss": 1.0665,
+      "step": 12503
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.826162338256836,
+      "learning_rate": 9.538551646236875e-06,
+      "loss": 1.4933,
+      "step": 12504
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 56.926944732666016,
+      "learning_rate": 9.537714931180187e-06,
+      "loss": 0.7264,
+      "step": 12505
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 19.74498176574707,
+      "learning_rate": 9.536878216123499e-06,
+      "loss": 2.171,
+      "step": 12506
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 75.2907943725586,
+      "learning_rate": 9.536041501066813e-06,
+      "loss": 0.723,
+      "step": 12507
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.58877420425415,
+      "learning_rate": 9.535204786010125e-06,
+      "loss": 1.0957,
+      "step": 12508
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 12.512656211853027,
+      "learning_rate": 9.534368070953438e-06,
+      "loss": 1.226,
+      "step": 12509
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.59494400024414,
+      "learning_rate": 9.53353135589675e-06,
+      "loss": 0.4392,
+      "step": 12510
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 16.1082820892334,
+      "learning_rate": 9.532694640840062e-06,
+      "loss": 2.6572,
+      "step": 12511
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 11.349607467651367,
+      "learning_rate": 9.531857925783376e-06,
+      "loss": 2.1882,
+      "step": 12512
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 112.89188385009766,
+      "learning_rate": 9.531021210726688e-06,
+      "loss": 1.2279,
+      "step": 12513
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 17.9888973236084,
+      "learning_rate": 9.530184495670001e-06,
+      "loss": 2.8782,
+      "step": 12514
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 21.969684600830078,
+      "learning_rate": 9.529347780613313e-06,
+      "loss": 1.5233,
+      "step": 12515
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 33.1344108581543,
+      "learning_rate": 9.528511065556625e-06,
+      "loss": 3.0303,
+      "step": 12516
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 4.86375093460083,
+      "learning_rate": 9.527674350499937e-06,
+      "loss": 0.4456,
+      "step": 12517
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 11.052216529846191,
+      "learning_rate": 9.526837635443251e-06,
+      "loss": 1.0125,
+      "step": 12518
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 9.90180778503418,
+      "learning_rate": 9.526000920386563e-06,
+      "loss": 1.4413,
+      "step": 12519
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 16.92300796508789,
+      "learning_rate": 9.525164205329875e-06,
+      "loss": 2.4217,
+      "step": 12520
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 10.172425270080566,
+      "learning_rate": 9.524327490273188e-06,
+      "loss": 0.5772,
+      "step": 12521
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 39.183902740478516,
+      "learning_rate": 9.5234907752165e-06,
+      "loss": 5.5313,
+      "step": 12522
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 10.03840160369873,
+      "learning_rate": 9.522654060159814e-06,
+      "loss": 1.5756,
+      "step": 12523
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 11.910656929016113,
+      "learning_rate": 9.521817345103126e-06,
+      "loss": 0.5367,
+      "step": 12524
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 60.089820861816406,
+      "learning_rate": 9.520980630046438e-06,
+      "loss": 2.7753,
+      "step": 12525
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 12.336292266845703,
+      "learning_rate": 9.520143914989752e-06,
+      "loss": 1.1547,
+      "step": 12526
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.9278714656829834,
+      "learning_rate": 9.519307199933064e-06,
+      "loss": 0.065,
+      "step": 12527
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 10.427318572998047,
+      "learning_rate": 9.518470484876377e-06,
+      "loss": 1.6682,
+      "step": 12528
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 260.3096008300781,
+      "learning_rate": 9.51763376981969e-06,
+      "loss": 0.5499,
+      "step": 12529
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 6.432458400726318,
+      "learning_rate": 9.516797054763001e-06,
+      "loss": 0.1183,
+      "step": 12530
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 19.33864974975586,
+      "learning_rate": 9.515960339706313e-06,
+      "loss": 1.3014,
+      "step": 12531
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.947526931762695,
+      "learning_rate": 9.515123624649627e-06,
+      "loss": 1.0892,
+      "step": 12532
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 27.10247802734375,
+      "learning_rate": 9.514286909592939e-06,
+      "loss": 1.4986,
+      "step": 12533
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 25.001182556152344,
+      "learning_rate": 9.51345019453625e-06,
+      "loss": 2.5495,
+      "step": 12534
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 10.446853637695312,
+      "learning_rate": 9.512613479479564e-06,
+      "loss": 0.5085,
+      "step": 12535
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 6.622890949249268,
+      "learning_rate": 9.511776764422876e-06,
+      "loss": 0.3592,
+      "step": 12536
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 4.06160831451416,
+      "learning_rate": 9.51094004936619e-06,
+      "loss": 0.5385,
+      "step": 12537
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.546213626861572,
+      "learning_rate": 9.510103334309502e-06,
+      "loss": 0.1624,
+      "step": 12538
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 17.955360412597656,
+      "learning_rate": 9.509266619252814e-06,
+      "loss": 1.0174,
+      "step": 12539
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 43.812557220458984,
+      "learning_rate": 9.508429904196127e-06,
+      "loss": 2.9263,
+      "step": 12540
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 14.057463645935059,
+      "learning_rate": 9.50759318913944e-06,
+      "loss": 1.0632,
+      "step": 12541
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.944059371948242,
+      "learning_rate": 9.506756474082753e-06,
+      "loss": 0.5991,
+      "step": 12542
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 12.784570693969727,
+      "learning_rate": 9.505919759026065e-06,
+      "loss": 1.1579,
+      "step": 12543
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 5.623643398284912,
+      "learning_rate": 9.505083043969377e-06,
+      "loss": 0.363,
+      "step": 12544
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.4926652908325195,
+      "learning_rate": 9.504246328912689e-06,
+      "loss": 0.6149,
+      "step": 12545
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 43.37441635131836,
+      "learning_rate": 9.503409613856001e-06,
+      "loss": 1.8677,
+      "step": 12546
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 7.164695739746094,
+      "learning_rate": 9.502572898799315e-06,
+      "loss": 0.3867,
+      "step": 12547
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.881974220275879,
+      "learning_rate": 9.501736183742626e-06,
+      "loss": 1.7908,
+      "step": 12548
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 11.763453483581543,
+      "learning_rate": 9.50089946868594e-06,
+      "loss": 1.5111,
+      "step": 12549
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 16.42769432067871,
+      "learning_rate": 9.500062753629252e-06,
+      "loss": 0.8511,
+      "step": 12550
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 4.932626724243164,
+      "learning_rate": 9.499226038572566e-06,
+      "loss": 0.4485,
+      "step": 12551
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 12.884479522705078,
+      "learning_rate": 9.498389323515878e-06,
+      "loss": 1.6867,
+      "step": 12552
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 28.30213165283203,
+      "learning_rate": 9.49755260845919e-06,
+      "loss": 2.3976,
+      "step": 12553
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 12.66464900970459,
+      "learning_rate": 9.496715893402503e-06,
+      "loss": 1.8408,
+      "step": 12554
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 14.149888038635254,
+      "learning_rate": 9.495879178345815e-06,
+      "loss": 0.6018,
+      "step": 12555
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.6592276096343994,
+      "learning_rate": 9.495042463289127e-06,
+      "loss": 0.1692,
+      "step": 12556
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 130.72134399414062,
+      "learning_rate": 9.494205748232441e-06,
+      "loss": 2.7858,
+      "step": 12557
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 19.557369232177734,
+      "learning_rate": 9.493369033175753e-06,
+      "loss": 0.714,
+      "step": 12558
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 27.16683006286621,
+      "learning_rate": 9.492532318119065e-06,
+      "loss": 1.6699,
+      "step": 12559
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 8.552977561950684,
+      "learning_rate": 9.491695603062377e-06,
+      "loss": 1.3134,
+      "step": 12560
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 32.981197357177734,
+      "learning_rate": 9.49085888800569e-06,
+      "loss": 2.8792,
+      "step": 12561
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 24.212053298950195,
+      "learning_rate": 9.490022172949002e-06,
+      "loss": 1.1681,
+      "step": 12562
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 5.9965362548828125,
+      "learning_rate": 9.489185457892316e-06,
+      "loss": 0.2369,
+      "step": 12563
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 8.74487590789795,
+      "learning_rate": 9.488348742835628e-06,
+      "loss": 0.6477,
+      "step": 12564
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 7.746337413787842,
+      "learning_rate": 9.487512027778942e-06,
+      "loss": 1.4842,
+      "step": 12565
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 6.999999046325684,
+      "learning_rate": 9.486675312722254e-06,
+      "loss": 2.7009,
+      "step": 12566
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.5180628299713135,
+      "learning_rate": 9.485838597665565e-06,
+      "loss": 0.1262,
+      "step": 12567
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 14.253186225891113,
+      "learning_rate": 9.485001882608879e-06,
+      "loss": 1.9534,
+      "step": 12568
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 14.465633392333984,
+      "learning_rate": 9.484165167552191e-06,
+      "loss": 0.9027,
+      "step": 12569
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.833093643188477,
+      "learning_rate": 9.483328452495503e-06,
+      "loss": 1.2568,
+      "step": 12570
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 14.690328598022461,
+      "learning_rate": 9.482491737438817e-06,
+      "loss": 1.6662,
+      "step": 12571
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 12.066746711730957,
+      "learning_rate": 9.481655022382129e-06,
+      "loss": 1.4585,
+      "step": 12572
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 19.365251541137695,
+      "learning_rate": 9.48081830732544e-06,
+      "loss": 1.6843,
+      "step": 12573
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 18.7397518157959,
+      "learning_rate": 9.479981592268753e-06,
+      "loss": 1.2523,
+      "step": 12574
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 13.922237396240234,
+      "learning_rate": 9.479144877212066e-06,
+      "loss": 1.4934,
+      "step": 12575
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 29.804685592651367,
+      "learning_rate": 9.478308162155378e-06,
+      "loss": 1.5592,
+      "step": 12576
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 19.848363876342773,
+      "learning_rate": 9.477471447098692e-06,
+      "loss": 1.0612,
+      "step": 12577
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 18.877689361572266,
+      "learning_rate": 9.476634732042004e-06,
+      "loss": 2.5891,
+      "step": 12578
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 18.858428955078125,
+      "learning_rate": 9.475798016985316e-06,
+      "loss": 3.1228,
+      "step": 12579
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 24.79368782043457,
+      "learning_rate": 9.47496130192863e-06,
+      "loss": 1.2129,
+      "step": 12580
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 12.999039649963379,
+      "learning_rate": 9.474124586871941e-06,
+      "loss": 1.8608,
+      "step": 12581
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 7.107621669769287,
+      "learning_rate": 9.473287871815255e-06,
+      "loss": 0.7415,
+      "step": 12582
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 5.948925971984863,
+      "learning_rate": 9.472451156758567e-06,
+      "loss": 0.3267,
+      "step": 12583
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 28.19964027404785,
+      "learning_rate": 9.471614441701879e-06,
+      "loss": 3.2042,
+      "step": 12584
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.318622589111328,
+      "learning_rate": 9.470777726645193e-06,
+      "loss": 2.0595,
+      "step": 12585
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 17.369403839111328,
+      "learning_rate": 9.469941011588504e-06,
+      "loss": 1.1902,
+      "step": 12586
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 12.831024169921875,
+      "learning_rate": 9.469104296531816e-06,
+      "loss": 0.8384,
+      "step": 12587
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.816106796264648,
+      "learning_rate": 9.468267581475128e-06,
+      "loss": 1.1297,
+      "step": 12588
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 393.5542297363281,
+      "learning_rate": 9.467430866418442e-06,
+      "loss": 1.3763,
+      "step": 12589
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 13.01297664642334,
+      "learning_rate": 9.466594151361754e-06,
+      "loss": 0.5997,
+      "step": 12590
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 12.69599437713623,
+      "learning_rate": 9.465757436305068e-06,
+      "loss": 0.7092,
+      "step": 12591
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.60859489440918,
+      "learning_rate": 9.46492072124838e-06,
+      "loss": 1.2716,
+      "step": 12592
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 10.400321006774902,
+      "learning_rate": 9.464084006191692e-06,
+      "loss": 0.568,
+      "step": 12593
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 140.9966583251953,
+      "learning_rate": 9.463247291135005e-06,
+      "loss": 1.9688,
+      "step": 12594
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 6.2600016593933105,
+      "learning_rate": 9.462410576078317e-06,
+      "loss": 1.351,
+      "step": 12595
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 20.106847763061523,
+      "learning_rate": 9.46157386102163e-06,
+      "loss": 0.7045,
+      "step": 12596
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 5.930684566497803,
+      "learning_rate": 9.460737145964943e-06,
+      "loss": 0.3359,
+      "step": 12597
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 28.652748107910156,
+      "learning_rate": 9.459900430908255e-06,
+      "loss": 1.2072,
+      "step": 12598
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 5.484364986419678,
+      "learning_rate": 9.459063715851567e-06,
+      "loss": 0.5294,
+      "step": 12599
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 24.903980255126953,
+      "learning_rate": 9.45822700079488e-06,
+      "loss": 0.5383,
+      "step": 12600
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.68040943145752,
+      "learning_rate": 9.457390285738192e-06,
+      "loss": 2.5263,
+      "step": 12601
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 12.230031967163086,
+      "learning_rate": 9.456553570681504e-06,
+      "loss": 0.6999,
+      "step": 12602
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 6.316253662109375,
+      "learning_rate": 9.455716855624818e-06,
+      "loss": 1.0758,
+      "step": 12603
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.931807041168213,
+      "learning_rate": 9.45488014056813e-06,
+      "loss": 0.1845,
+      "step": 12604
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 6.719529151916504,
+      "learning_rate": 9.454043425511443e-06,
+      "loss": 0.3294,
+      "step": 12605
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 23.872106552124023,
+      "learning_rate": 9.453206710454755e-06,
+      "loss": 0.8509,
+      "step": 12606
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 56.260189056396484,
+      "learning_rate": 9.452369995398067e-06,
+      "loss": 2.9153,
+      "step": 12607
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 29.488616943359375,
+      "learning_rate": 9.451533280341381e-06,
+      "loss": 1.9577,
+      "step": 12608
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 8.279169082641602,
+      "learning_rate": 9.450696565284693e-06,
+      "loss": 0.3277,
+      "step": 12609
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 56.83429718017578,
+      "learning_rate": 9.449859850228007e-06,
+      "loss": 5.3126,
+      "step": 12610
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 49.203041076660156,
+      "learning_rate": 9.449023135171319e-06,
+      "loss": 1.9299,
+      "step": 12611
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 38.26557159423828,
+      "learning_rate": 9.44818642011463e-06,
+      "loss": 2.201,
+      "step": 12612
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 6.391701698303223,
+      "learning_rate": 9.447349705057943e-06,
+      "loss": 0.5208,
+      "step": 12613
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 26.8066349029541,
+      "learning_rate": 9.446512990001256e-06,
+      "loss": 1.7179,
+      "step": 12614
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 20.002933502197266,
+      "learning_rate": 9.445676274944568e-06,
+      "loss": 1.5349,
+      "step": 12615
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 20.30712890625,
+      "learning_rate": 9.44483955988788e-06,
+      "loss": 1.9082,
+      "step": 12616
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 14.136492729187012,
+      "learning_rate": 9.444002844831194e-06,
+      "loss": 0.9949,
+      "step": 12617
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 22.940683364868164,
+      "learning_rate": 9.443166129774506e-06,
+      "loss": 1.5033,
+      "step": 12618
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 17.115327835083008,
+      "learning_rate": 9.44232941471782e-06,
+      "loss": 1.4698,
+      "step": 12619
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 33.50325012207031,
+      "learning_rate": 9.441492699661131e-06,
+      "loss": 1.637,
+      "step": 12620
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 38.30734634399414,
+      "learning_rate": 9.440655984604443e-06,
+      "loss": 1.4949,
+      "step": 12621
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 5.381844520568848,
+      "learning_rate": 9.439819269547757e-06,
+      "loss": 0.3825,
+      "step": 12622
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 8.267170906066895,
+      "learning_rate": 9.438982554491069e-06,
+      "loss": 0.4952,
+      "step": 12623
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 25.081682205200195,
+      "learning_rate": 9.438145839434382e-06,
+      "loss": 1.7709,
+      "step": 12624
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.4282147884368896,
+      "learning_rate": 9.437309124377694e-06,
+      "loss": 0.2464,
+      "step": 12625
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.6459097862243652,
+      "learning_rate": 9.436472409321006e-06,
+      "loss": 0.2936,
+      "step": 12626
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 25.131540298461914,
+      "learning_rate": 9.435635694264318e-06,
+      "loss": 0.7742,
+      "step": 12627
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 13.032624244689941,
+      "learning_rate": 9.43479897920763e-06,
+      "loss": 1.2999,
+      "step": 12628
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 13.023077011108398,
+      "learning_rate": 9.433962264150944e-06,
+      "loss": 1.355,
+      "step": 12629
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.388630867004395,
+      "learning_rate": 9.433125549094256e-06,
+      "loss": 1.8461,
+      "step": 12630
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.404267311096191,
+      "learning_rate": 9.43228883403757e-06,
+      "loss": 0.9715,
+      "step": 12631
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 38.04804992675781,
+      "learning_rate": 9.431452118980882e-06,
+      "loss": 1.4528,
+      "step": 12632
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 20.062524795532227,
+      "learning_rate": 9.430615403924195e-06,
+      "loss": 1.169,
+      "step": 12633
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 186.04722595214844,
+      "learning_rate": 9.429778688867507e-06,
+      "loss": 1.8812,
+      "step": 12634
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.48676586151123,
+      "learning_rate": 9.428941973810819e-06,
+      "loss": 1.0044,
+      "step": 12635
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.133785724639893,
+      "learning_rate": 9.428105258754133e-06,
+      "loss": 1.0135,
+      "step": 12636
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 19.21123504638672,
+      "learning_rate": 9.427268543697445e-06,
+      "loss": 0.8992,
+      "step": 12637
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 7.901590824127197,
+      "learning_rate": 9.426431828640757e-06,
+      "loss": 0.7217,
+      "step": 12638
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 21.584949493408203,
+      "learning_rate": 9.42559511358407e-06,
+      "loss": 1.4179,
+      "step": 12639
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.24616813659668,
+      "learning_rate": 9.424758398527382e-06,
+      "loss": 0.419,
+      "step": 12640
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 51.427574157714844,
+      "learning_rate": 9.423921683470694e-06,
+      "loss": 1.1515,
+      "step": 12641
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 26.152320861816406,
+      "learning_rate": 9.423084968414006e-06,
+      "loss": 1.4699,
+      "step": 12642
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 30.71882438659668,
+      "learning_rate": 9.42224825335732e-06,
+      "loss": 1.3176,
+      "step": 12643
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 17.129243850708008,
+      "learning_rate": 9.421411538300632e-06,
+      "loss": 1.9202,
+      "step": 12644
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 5.461491107940674,
+      "learning_rate": 9.420574823243945e-06,
+      "loss": 0.377,
+      "step": 12645
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 90.69756317138672,
+      "learning_rate": 9.419738108187257e-06,
+      "loss": 1.5946,
+      "step": 12646
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 11.628805160522461,
+      "learning_rate": 9.418901393130571e-06,
+      "loss": 1.3091,
+      "step": 12647
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 17.364070892333984,
+      "learning_rate": 9.418064678073883e-06,
+      "loss": 1.3447,
+      "step": 12648
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 14.625235557556152,
+      "learning_rate": 9.417227963017195e-06,
+      "loss": 1.1246,
+      "step": 12649
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 22.807085037231445,
+      "learning_rate": 9.416391247960509e-06,
+      "loss": 0.6853,
+      "step": 12650
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 3.4212002754211426,
+      "learning_rate": 9.41555453290382e-06,
+      "loss": 0.1706,
+      "step": 12651
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 3.9466753005981445,
+      "learning_rate": 9.414717817847132e-06,
+      "loss": 0.3402,
+      "step": 12652
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 7.102753162384033,
+      "learning_rate": 9.413881102790446e-06,
+      "loss": 0.689,
+      "step": 12653
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 12.06006908416748,
+      "learning_rate": 9.413044387733758e-06,
+      "loss": 0.6928,
+      "step": 12654
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.17564582824707,
+      "learning_rate": 9.41220767267707e-06,
+      "loss": 0.8857,
+      "step": 12655
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 5.59531831741333,
+      "learning_rate": 9.411370957620382e-06,
+      "loss": 1.2905,
+      "step": 12656
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.778902053833008,
+      "learning_rate": 9.410534242563696e-06,
+      "loss": 0.2996,
+      "step": 12657
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 43.223758697509766,
+      "learning_rate": 9.409697527507008e-06,
+      "loss": 1.7344,
+      "step": 12658
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 14.367379188537598,
+      "learning_rate": 9.408860812450321e-06,
+      "loss": 0.8148,
+      "step": 12659
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 12.753751754760742,
+      "learning_rate": 9.408024097393633e-06,
+      "loss": 1.7903,
+      "step": 12660
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 6.503912448883057,
+      "learning_rate": 9.407187382336947e-06,
+      "loss": 0.5994,
+      "step": 12661
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.3146650791168213,
+      "learning_rate": 9.406350667280259e-06,
+      "loss": 0.0388,
+      "step": 12662
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 12.779787063598633,
+      "learning_rate": 9.40551395222357e-06,
+      "loss": 1.737,
+      "step": 12663
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 21.04336166381836,
+      "learning_rate": 9.404677237166884e-06,
+      "loss": 1.2667,
+      "step": 12664
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 12.546812057495117,
+      "learning_rate": 9.403840522110196e-06,
+      "loss": 0.9729,
+      "step": 12665
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.291451454162598,
+      "learning_rate": 9.403003807053508e-06,
+      "loss": 0.378,
+      "step": 12666
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.457207441329956,
+      "learning_rate": 9.40216709199682e-06,
+      "loss": 0.0792,
+      "step": 12667
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 13.121223449707031,
+      "learning_rate": 9.401330376940134e-06,
+      "loss": 0.7211,
+      "step": 12668
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 63.11892318725586,
+      "learning_rate": 9.400493661883446e-06,
+      "loss": 1.0022,
+      "step": 12669
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 6.179528713226318,
+      "learning_rate": 9.399656946826758e-06,
+      "loss": 0.3718,
+      "step": 12670
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 83.38644409179688,
+      "learning_rate": 9.398820231770071e-06,
+      "loss": 2.4827,
+      "step": 12671
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 50.963565826416016,
+      "learning_rate": 9.397983516713383e-06,
+      "loss": 3.134,
+      "step": 12672
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 36.63359832763672,
+      "learning_rate": 9.397146801656697e-06,
+      "loss": 1.5886,
+      "step": 12673
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 30.252046585083008,
+      "learning_rate": 9.396310086600009e-06,
+      "loss": 3.0493,
+      "step": 12674
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.03751277923584,
+      "learning_rate": 9.395473371543323e-06,
+      "loss": 1.2414,
+      "step": 12675
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 14.713725090026855,
+      "learning_rate": 9.394636656486635e-06,
+      "loss": 2.2616,
+      "step": 12676
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 8.644331932067871,
+      "learning_rate": 9.393799941429947e-06,
+      "loss": 1.9946,
+      "step": 12677
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 24.73609161376953,
+      "learning_rate": 9.39296322637326e-06,
+      "loss": 1.852,
+      "step": 12678
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 23.22011375427246,
+      "learning_rate": 9.392126511316572e-06,
+      "loss": 1.1773,
+      "step": 12679
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 7.167316913604736,
+      "learning_rate": 9.391289796259884e-06,
+      "loss": 0.0825,
+      "step": 12680
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 20.045015335083008,
+      "learning_rate": 9.390453081203196e-06,
+      "loss": 2.0277,
+      "step": 12681
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 50.51588439941406,
+      "learning_rate": 9.38961636614651e-06,
+      "loss": 1.8516,
+      "step": 12682
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 41.1562614440918,
+      "learning_rate": 9.388779651089822e-06,
+      "loss": 0.3866,
+      "step": 12683
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 13.490531921386719,
+      "learning_rate": 9.387942936033134e-06,
+      "loss": 1.4136,
+      "step": 12684
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 15.883825302124023,
+      "learning_rate": 9.387106220976447e-06,
+      "loss": 1.1284,
+      "step": 12685
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 11.150424003601074,
+      "learning_rate": 9.38626950591976e-06,
+      "loss": 0.5869,
+      "step": 12686
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.384401321411133,
+      "learning_rate": 9.385432790863073e-06,
+      "loss": 2.0767,
+      "step": 12687
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 52.81633377075195,
+      "learning_rate": 9.384596075806385e-06,
+      "loss": 1.0207,
+      "step": 12688
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.21898078918457,
+      "learning_rate": 9.383759360749698e-06,
+      "loss": 1.3715,
+      "step": 12689
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 105.48194122314453,
+      "learning_rate": 9.38292264569301e-06,
+      "loss": 1.9724,
+      "step": 12690
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 19.446487426757812,
+      "learning_rate": 9.382085930636322e-06,
+      "loss": 1.4847,
+      "step": 12691
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 12.214441299438477,
+      "learning_rate": 9.381249215579636e-06,
+      "loss": 0.5879,
+      "step": 12692
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 410.44830322265625,
+      "learning_rate": 9.380412500522948e-06,
+      "loss": 1.2085,
+      "step": 12693
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 32.169864654541016,
+      "learning_rate": 9.37957578546626e-06,
+      "loss": 1.2658,
+      "step": 12694
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 18.428232192993164,
+      "learning_rate": 9.378739070409572e-06,
+      "loss": 1.5905,
+      "step": 12695
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 45.33714294433594,
+      "learning_rate": 9.377902355352886e-06,
+      "loss": 1.3951,
+      "step": 12696
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 17.632648468017578,
+      "learning_rate": 9.377065640296198e-06,
+      "loss": 1.3787,
+      "step": 12697
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 7.390748500823975,
+      "learning_rate": 9.37622892523951e-06,
+      "loss": 0.3028,
+      "step": 12698
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 60.25581741333008,
+      "learning_rate": 9.375392210182823e-06,
+      "loss": 2.4936,
+      "step": 12699
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 43.981231689453125,
+      "learning_rate": 9.374555495126135e-06,
+      "loss": 0.7651,
+      "step": 12700
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 11.389113426208496,
+      "learning_rate": 9.373718780069449e-06,
+      "loss": 0.4666,
+      "step": 12701
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 6.8582377433776855,
+      "learning_rate": 9.37288206501276e-06,
+      "loss": 0.7208,
+      "step": 12702
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 24.401554107666016,
+      "learning_rate": 9.372045349956074e-06,
+      "loss": 1.1998,
+      "step": 12703
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.09210467338562,
+      "learning_rate": 9.371208634899386e-06,
+      "loss": 0.0478,
+      "step": 12704
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 7.204529285430908,
+      "learning_rate": 9.370371919842698e-06,
+      "loss": 1.1222,
+      "step": 12705
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 16.0089111328125,
+      "learning_rate": 9.369535204786012e-06,
+      "loss": 1.8902,
+      "step": 12706
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.174287796020508,
+      "learning_rate": 9.368698489729324e-06,
+      "loss": 0.6599,
+      "step": 12707
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 11.618083953857422,
+      "learning_rate": 9.367861774672636e-06,
+      "loss": 1.0567,
+      "step": 12708
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 13.289433479309082,
+      "learning_rate": 9.367025059615948e-06,
+      "loss": 2.0885,
+      "step": 12709
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.8225820064544678,
+      "learning_rate": 9.366188344559261e-06,
+      "loss": 0.8951,
+      "step": 12710
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.53161096572876,
+      "learning_rate": 9.365351629502573e-06,
+      "loss": 0.2615,
+      "step": 12711
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 14.11289119720459,
+      "learning_rate": 9.364514914445885e-06,
+      "loss": 0.9496,
+      "step": 12712
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 89.39127349853516,
+      "learning_rate": 9.363678199389199e-06,
+      "loss": 0.3615,
+      "step": 12713
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 12.24676513671875,
+      "learning_rate": 9.362841484332511e-06,
+      "loss": 0.607,
+      "step": 12714
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 22.23177719116211,
+      "learning_rate": 9.362004769275825e-06,
+      "loss": 1.5199,
+      "step": 12715
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 105.7651138305664,
+      "learning_rate": 9.361168054219137e-06,
+      "loss": 1.6218,
+      "step": 12716
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 12.95954704284668,
+      "learning_rate": 9.36033133916245e-06,
+      "loss": 1.3081,
+      "step": 12717
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 7.558269500732422,
+      "learning_rate": 9.359494624105762e-06,
+      "loss": 1.1606,
+      "step": 12718
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 28.80192756652832,
+      "learning_rate": 9.358657909049074e-06,
+      "loss": 1.4126,
+      "step": 12719
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 14.604012489318848,
+      "learning_rate": 9.357821193992386e-06,
+      "loss": 0.9703,
+      "step": 12720
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 7.633918762207031,
+      "learning_rate": 9.3569844789357e-06,
+      "loss": 0.6268,
+      "step": 12721
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.201777458190918,
+      "learning_rate": 9.356147763879012e-06,
+      "loss": 0.4496,
+      "step": 12722
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 101.30696105957031,
+      "learning_rate": 9.355311048822324e-06,
+      "loss": 1.3412,
+      "step": 12723
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.9050068855285645,
+      "learning_rate": 9.354474333765637e-06,
+      "loss": 0.7253,
+      "step": 12724
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 14.176651954650879,
+      "learning_rate": 9.35363761870895e-06,
+      "loss": 2.8281,
+      "step": 12725
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 20.2751522064209,
+      "learning_rate": 9.352800903652261e-06,
+      "loss": 0.6233,
+      "step": 12726
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.783174991607666,
+      "learning_rate": 9.351964188595575e-06,
+      "loss": 0.1949,
+      "step": 12727
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 21.548534393310547,
+      "learning_rate": 9.351127473538887e-06,
+      "loss": 1.0327,
+      "step": 12728
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 18.70937728881836,
+      "learning_rate": 9.3502907584822e-06,
+      "loss": 1.1648,
+      "step": 12729
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.633247375488281,
+      "learning_rate": 9.349454043425512e-06,
+      "loss": 2.1667,
+      "step": 12730
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 42.88833236694336,
+      "learning_rate": 9.348617328368826e-06,
+      "loss": 1.4255,
+      "step": 12731
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 15.978346824645996,
+      "learning_rate": 9.347780613312138e-06,
+      "loss": 0.6872,
+      "step": 12732
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.080323219299316,
+      "learning_rate": 9.34694389825545e-06,
+      "loss": 0.5269,
+      "step": 12733
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.514498710632324,
+      "learning_rate": 9.346107183198762e-06,
+      "loss": 1.2947,
+      "step": 12734
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 7.079040050506592,
+      "learning_rate": 9.345270468142076e-06,
+      "loss": 0.8225,
+      "step": 12735
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.065547943115234,
+      "learning_rate": 9.344433753085387e-06,
+      "loss": 1.3093,
+      "step": 12736
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 14.474076271057129,
+      "learning_rate": 9.3435970380287e-06,
+      "loss": 0.6239,
+      "step": 12737
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 25.227272033691406,
+      "learning_rate": 9.342760322972013e-06,
+      "loss": 1.5511,
+      "step": 12738
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 36.528316497802734,
+      "learning_rate": 9.341923607915325e-06,
+      "loss": 0.987,
+      "step": 12739
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 86.75886535644531,
+      "learning_rate": 9.341086892858637e-06,
+      "loss": 0.8112,
+      "step": 12740
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.102453231811523,
+      "learning_rate": 9.34025017780195e-06,
+      "loss": 1.2799,
+      "step": 12741
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 21.57074737548828,
+      "learning_rate": 9.339413462745263e-06,
+      "loss": 1.8643,
+      "step": 12742
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 16.9290828704834,
+      "learning_rate": 9.338576747688576e-06,
+      "loss": 0.3976,
+      "step": 12743
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 25.484235763549805,
+      "learning_rate": 9.337740032631888e-06,
+      "loss": 0.6294,
+      "step": 12744
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 28.827713012695312,
+      "learning_rate": 9.336903317575202e-06,
+      "loss": 1.6037,
+      "step": 12745
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.305099487304688,
+      "learning_rate": 9.336066602518514e-06,
+      "loss": 0.8748,
+      "step": 12746
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 17.736371994018555,
+      "learning_rate": 9.335229887461826e-06,
+      "loss": 1.1183,
+      "step": 12747
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 13.301612854003906,
+      "learning_rate": 9.334393172405138e-06,
+      "loss": 1.3793,
+      "step": 12748
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 14.686193466186523,
+      "learning_rate": 9.33355645734845e-06,
+      "loss": 3.1688,
+      "step": 12749
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 15.956156730651855,
+      "learning_rate": 9.332719742291763e-06,
+      "loss": 0.8889,
+      "step": 12750
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 16.69364356994629,
+      "learning_rate": 9.331883027235075e-06,
+      "loss": 1.9165,
+      "step": 12751
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 12.45902156829834,
+      "learning_rate": 9.331046312178389e-06,
+      "loss": 0.7861,
+      "step": 12752
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 17.342580795288086,
+      "learning_rate": 9.330209597121701e-06,
+      "loss": 1.3061,
+      "step": 12753
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.511950492858887,
+      "learning_rate": 9.329372882065013e-06,
+      "loss": 1.1508,
+      "step": 12754
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.777467727661133,
+      "learning_rate": 9.328536167008326e-06,
+      "loss": 0.3007,
+      "step": 12755
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 17.43426513671875,
+      "learning_rate": 9.327699451951638e-06,
+      "loss": 1.1306,
+      "step": 12756
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.546712875366211,
+      "learning_rate": 9.326862736894952e-06,
+      "loss": 0.7657,
+      "step": 12757
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 15.232168197631836,
+      "learning_rate": 9.326026021838264e-06,
+      "loss": 1.9775,
+      "step": 12758
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.149893760681152,
+      "learning_rate": 9.325189306781576e-06,
+      "loss": 0.7729,
+      "step": 12759
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.922850608825684,
+      "learning_rate": 9.32435259172489e-06,
+      "loss": 0.3599,
+      "step": 12760
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 12.676969528198242,
+      "learning_rate": 9.323515876668202e-06,
+      "loss": 0.6969,
+      "step": 12761
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 7.15059757232666,
+      "learning_rate": 9.322679161611514e-06,
+      "loss": 0.2292,
+      "step": 12762
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.097254753112793,
+      "learning_rate": 9.321842446554826e-06,
+      "loss": 1.7266,
+      "step": 12763
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 38.53457260131836,
+      "learning_rate": 9.321005731498139e-06,
+      "loss": 1.8251,
+      "step": 12764
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 19.701183319091797,
+      "learning_rate": 9.320169016441451e-06,
+      "loss": 1.5037,
+      "step": 12765
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 14.709261894226074,
+      "learning_rate": 9.319332301384763e-06,
+      "loss": 0.9079,
+      "step": 12766
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.67050838470459,
+      "learning_rate": 9.318495586328077e-06,
+      "loss": 0.7052,
+      "step": 12767
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 10.067826271057129,
+      "learning_rate": 9.317658871271389e-06,
+      "loss": 0.3172,
+      "step": 12768
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 13.909753799438477,
+      "learning_rate": 9.316822156214702e-06,
+      "loss": 0.6805,
+      "step": 12769
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 15.233804702758789,
+      "learning_rate": 9.315985441158014e-06,
+      "loss": 1.4905,
+      "step": 12770
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 16.987150192260742,
+      "learning_rate": 9.315148726101328e-06,
+      "loss": 1.4539,
+      "step": 12771
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 4.675383567810059,
+      "learning_rate": 9.31431201104464e-06,
+      "loss": 0.47,
+      "step": 12772
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.47659158706665,
+      "learning_rate": 9.313475295987952e-06,
+      "loss": 0.3059,
+      "step": 12773
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.951998710632324,
+      "learning_rate": 9.312638580931265e-06,
+      "loss": 1.4986,
+      "step": 12774
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 25.695837020874023,
+      "learning_rate": 9.311801865874577e-06,
+      "loss": 1.4772,
+      "step": 12775
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 20.428926467895508,
+      "learning_rate": 9.31096515081789e-06,
+      "loss": 1.0381,
+      "step": 12776
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 12.581475257873535,
+      "learning_rate": 9.310128435761201e-06,
+      "loss": 0.7638,
+      "step": 12777
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 12.591764450073242,
+      "learning_rate": 9.309291720704515e-06,
+      "loss": 1.7015,
+      "step": 12778
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 20.13874626159668,
+      "learning_rate": 9.308455005647827e-06,
+      "loss": 1.1436,
+      "step": 12779
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 27.49405860900879,
+      "learning_rate": 9.307618290591139e-06,
+      "loss": 0.7916,
+      "step": 12780
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 73.70000457763672,
+      "learning_rate": 9.306781575534453e-06,
+      "loss": 2.4081,
+      "step": 12781
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 9.437817573547363,
+      "learning_rate": 9.305944860477765e-06,
+      "loss": 0.3377,
+      "step": 12782
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 23.251123428344727,
+      "learning_rate": 9.305108145421078e-06,
+      "loss": 0.8318,
+      "step": 12783
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.730528831481934,
+      "learning_rate": 9.30427143036439e-06,
+      "loss": 0.7024,
+      "step": 12784
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 22.202592849731445,
+      "learning_rate": 9.303434715307704e-06,
+      "loss": 1.2179,
+      "step": 12785
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 11.280617713928223,
+      "learning_rate": 9.302598000251016e-06,
+      "loss": 1.2216,
+      "step": 12786
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 66.64436340332031,
+      "learning_rate": 9.301761285194328e-06,
+      "loss": 1.0517,
+      "step": 12787
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.43831729888916,
+      "learning_rate": 9.300924570137641e-06,
+      "loss": 1.3776,
+      "step": 12788
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 18.07183074951172,
+      "learning_rate": 9.300087855080953e-06,
+      "loss": 1.8422,
+      "step": 12789
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 24.829612731933594,
+      "learning_rate": 9.299251140024265e-06,
+      "loss": 1.9916,
+      "step": 12790
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 4.854472637176514,
+      "learning_rate": 9.298414424967577e-06,
+      "loss": 0.3375,
+      "step": 12791
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.297489166259766,
+      "learning_rate": 9.29757770991089e-06,
+      "loss": 1.4392,
+      "step": 12792
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 6.604464530944824,
+      "learning_rate": 9.296740994854203e-06,
+      "loss": 0.2846,
+      "step": 12793
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 16.88629150390625,
+      "learning_rate": 9.295904279797515e-06,
+      "loss": 1.2431,
+      "step": 12794
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 35.367759704589844,
+      "learning_rate": 9.295067564740828e-06,
+      "loss": 1.9533,
+      "step": 12795
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 20.0457763671875,
+      "learning_rate": 9.29423084968414e-06,
+      "loss": 0.9273,
+      "step": 12796
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 30.456619262695312,
+      "learning_rate": 9.293394134627454e-06,
+      "loss": 1.9967,
+      "step": 12797
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 31.33993911743164,
+      "learning_rate": 9.292557419570766e-06,
+      "loss": 2.6482,
+      "step": 12798
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 12.192071914672852,
+      "learning_rate": 9.29172070451408e-06,
+      "loss": 0.5799,
+      "step": 12799
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 12.041566848754883,
+      "learning_rate": 9.290883989457392e-06,
+      "loss": 0.5665,
+      "step": 12800
+    },
+    {
+      "epoch": 1.61,
+      "eval_loss": 0.08084025979042053,
+      "eval_runtime": 94.8744,
+      "eval_samples_per_second": 37.334,
+      "eval_steps_per_second": 37.334,
+      "step": 12800
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 13.741267204284668,
+      "learning_rate": 9.290047274400704e-06,
+      "loss": 1.1971,
+      "step": 12801
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 12.394692420959473,
+      "learning_rate": 9.289210559344015e-06,
+      "loss": 0.8088,
+      "step": 12802
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 30.322452545166016,
+      "learning_rate": 9.288373844287329e-06,
+      "loss": 1.6056,
+      "step": 12803
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.515654563903809,
+      "learning_rate": 9.287537129230641e-06,
+      "loss": 0.7914,
+      "step": 12804
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 22.28693962097168,
+      "learning_rate": 9.286700414173953e-06,
+      "loss": 1.9402,
+      "step": 12805
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 5.956439971923828,
+      "learning_rate": 9.285863699117267e-06,
+      "loss": 1.3111,
+      "step": 12806
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 40.35560607910156,
+      "learning_rate": 9.285026984060579e-06,
+      "loss": 2.2955,
+      "step": 12807
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 5.762033939361572,
+      "learning_rate": 9.28419026900389e-06,
+      "loss": 0.6368,
+      "step": 12808
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 13.975659370422363,
+      "learning_rate": 9.283353553947204e-06,
+      "loss": 1.6104,
+      "step": 12809
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 12.66861629486084,
+      "learning_rate": 9.282516838890516e-06,
+      "loss": 0.8433,
+      "step": 12810
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 15.702493667602539,
+      "learning_rate": 9.28168012383383e-06,
+      "loss": 1.4411,
+      "step": 12811
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 27.008512496948242,
+      "learning_rate": 9.280843408777142e-06,
+      "loss": 0.9404,
+      "step": 12812
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 14.046055793762207,
+      "learning_rate": 9.280006693720455e-06,
+      "loss": 0.9495,
+      "step": 12813
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 13.30699348449707,
+      "learning_rate": 9.279169978663767e-06,
+      "loss": 0.6348,
+      "step": 12814
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 7.643478870391846,
+      "learning_rate": 9.27833326360708e-06,
+      "loss": 0.2463,
+      "step": 12815
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 10.765046119689941,
+      "learning_rate": 9.277496548550391e-06,
+      "loss": 0.4739,
+      "step": 12816
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 17.187000274658203,
+      "learning_rate": 9.276659833493705e-06,
+      "loss": 1.0207,
+      "step": 12817
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.7416460514068604,
+      "learning_rate": 9.275823118437017e-06,
+      "loss": 0.0402,
+      "step": 12818
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 23.240400314331055,
+      "learning_rate": 9.274986403380329e-06,
+      "loss": 0.8305,
+      "step": 12819
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 7.226999759674072,
+      "learning_rate": 9.274149688323643e-06,
+      "loss": 0.949,
+      "step": 12820
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.760710716247559,
+      "learning_rate": 9.273312973266954e-06,
+      "loss": 0.635,
+      "step": 12821
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.73095989227295,
+      "learning_rate": 9.272476258210266e-06,
+      "loss": 0.5365,
+      "step": 12822
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 15.786834716796875,
+      "learning_rate": 9.27163954315358e-06,
+      "loss": 0.5541,
+      "step": 12823
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 6.006377220153809,
+      "learning_rate": 9.270802828096892e-06,
+      "loss": 0.4464,
+      "step": 12824
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 48.15402603149414,
+      "learning_rate": 9.269966113040206e-06,
+      "loss": 1.9272,
+      "step": 12825
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 20.68983268737793,
+      "learning_rate": 9.269129397983518e-06,
+      "loss": 1.3508,
+      "step": 12826
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 12.626483917236328,
+      "learning_rate": 9.268292682926831e-06,
+      "loss": 1.1429,
+      "step": 12827
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 14.831421852111816,
+      "learning_rate": 9.267455967870143e-06,
+      "loss": 0.8639,
+      "step": 12828
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 11.292095184326172,
+      "learning_rate": 9.266619252813455e-06,
+      "loss": 0.8774,
+      "step": 12829
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.60373592376709,
+      "learning_rate": 9.265782537756767e-06,
+      "loss": 1.5242,
+      "step": 12830
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 27.918676376342773,
+      "learning_rate": 9.264945822700079e-06,
+      "loss": 2.0699,
+      "step": 12831
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 5.252095699310303,
+      "learning_rate": 9.264109107643393e-06,
+      "loss": 0.2138,
+      "step": 12832
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 23.539323806762695,
+      "learning_rate": 9.263272392586705e-06,
+      "loss": 1.0934,
+      "step": 12833
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 22.99410629272461,
+      "learning_rate": 9.262435677530018e-06,
+      "loss": 2.6611,
+      "step": 12834
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.666620254516602,
+      "learning_rate": 9.26159896247333e-06,
+      "loss": 0.3635,
+      "step": 12835
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 14.20853042602539,
+      "learning_rate": 9.260762247416642e-06,
+      "loss": 0.7676,
+      "step": 12836
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 5.007308483123779,
+      "learning_rate": 9.259925532359956e-06,
+      "loss": 0.3601,
+      "step": 12837
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 13.300216674804688,
+      "learning_rate": 9.259088817303268e-06,
+      "loss": 1.1097,
+      "step": 12838
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 29.382566452026367,
+      "learning_rate": 9.258252102246581e-06,
+      "loss": 1.257,
+      "step": 12839
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 128.78793334960938,
+      "learning_rate": 9.257415387189893e-06,
+      "loss": 2.0364,
+      "step": 12840
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 13.95974349975586,
+      "learning_rate": 9.256578672133207e-06,
+      "loss": 1.3131,
+      "step": 12841
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.428278923034668,
+      "learning_rate": 9.255741957076519e-06,
+      "loss": 0.4572,
+      "step": 12842
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.121671676635742,
+      "learning_rate": 9.254905242019831e-06,
+      "loss": 0.5315,
+      "step": 12843
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 6.823456764221191,
+      "learning_rate": 9.254068526963143e-06,
+      "loss": 0.0958,
+      "step": 12844
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 20.92385482788086,
+      "learning_rate": 9.253231811906455e-06,
+      "loss": 1.9386,
+      "step": 12845
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.319921493530273,
+      "learning_rate": 9.252395096849769e-06,
+      "loss": 0.3208,
+      "step": 12846
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 14.565032958984375,
+      "learning_rate": 9.25155838179308e-06,
+      "loss": 1.3546,
+      "step": 12847
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.67790699005127,
+      "learning_rate": 9.250721666736394e-06,
+      "loss": 1.176,
+      "step": 12848
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 18.494850158691406,
+      "learning_rate": 9.249884951679706e-06,
+      "loss": 2.5711,
+      "step": 12849
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 44.53583908081055,
+      "learning_rate": 9.249048236623018e-06,
+      "loss": 1.2291,
+      "step": 12850
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 21.787281036376953,
+      "learning_rate": 9.248211521566332e-06,
+      "loss": 1.064,
+      "step": 12851
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 25.983638763427734,
+      "learning_rate": 9.247374806509644e-06,
+      "loss": 1.1076,
+      "step": 12852
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 15.283540725708008,
+      "learning_rate": 9.246538091452957e-06,
+      "loss": 0.922,
+      "step": 12853
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 23.032045364379883,
+      "learning_rate": 9.24570137639627e-06,
+      "loss": 1.5744,
+      "step": 12854
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 16.80522918701172,
+      "learning_rate": 9.244864661339581e-06,
+      "loss": 0.8887,
+      "step": 12855
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 14.449721336364746,
+      "learning_rate": 9.244027946282895e-06,
+      "loss": 0.625,
+      "step": 12856
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 12.633468627929688,
+      "learning_rate": 9.243191231226207e-06,
+      "loss": 2.3903,
+      "step": 12857
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 20.23297691345215,
+      "learning_rate": 9.242354516169519e-06,
+      "loss": 1.6446,
+      "step": 12858
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 5.653713226318359,
+      "learning_rate": 9.24151780111283e-06,
+      "loss": 0.1913,
+      "step": 12859
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.748419761657715,
+      "learning_rate": 9.240681086056144e-06,
+      "loss": 0.6558,
+      "step": 12860
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 13.776673316955566,
+      "learning_rate": 9.239844370999456e-06,
+      "loss": 0.4821,
+      "step": 12861
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 29.427501678466797,
+      "learning_rate": 9.23900765594277e-06,
+      "loss": 1.1989,
+      "step": 12862
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 8.08488941192627,
+      "learning_rate": 9.238170940886082e-06,
+      "loss": 0.3813,
+      "step": 12863
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 77.74174499511719,
+      "learning_rate": 9.237334225829394e-06,
+      "loss": 2.1877,
+      "step": 12864
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 38.8492317199707,
+      "learning_rate": 9.236497510772708e-06,
+      "loss": 2.0237,
+      "step": 12865
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 15.239294052124023,
+      "learning_rate": 9.23566079571602e-06,
+      "loss": 0.7539,
+      "step": 12866
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 15.638760566711426,
+      "learning_rate": 9.234824080659333e-06,
+      "loss": 1.3254,
+      "step": 12867
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 4.971253871917725,
+      "learning_rate": 9.233987365602645e-06,
+      "loss": 0.238,
+      "step": 12868
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 44.882686614990234,
+      "learning_rate": 9.233150650545957e-06,
+      "loss": 1.2559,
+      "step": 12869
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 8.837841033935547,
+      "learning_rate": 9.23231393548927e-06,
+      "loss": 1.9611,
+      "step": 12870
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 8.826338768005371,
+      "learning_rate": 9.231477220432583e-06,
+      "loss": 1.5011,
+      "step": 12871
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 12.538323402404785,
+      "learning_rate": 9.230640505375895e-06,
+      "loss": 1.4513,
+      "step": 12872
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.48988151550293,
+      "learning_rate": 9.229803790319207e-06,
+      "loss": 2.127,
+      "step": 12873
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 20.119808197021484,
+      "learning_rate": 9.22896707526252e-06,
+      "loss": 1.4636,
+      "step": 12874
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 40.06005096435547,
+      "learning_rate": 9.228130360205832e-06,
+      "loss": 2.06,
+      "step": 12875
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.540440082550049,
+      "learning_rate": 9.227293645149146e-06,
+      "loss": 1.0775,
+      "step": 12876
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 21.5985050201416,
+      "learning_rate": 9.226456930092458e-06,
+      "loss": 1.5787,
+      "step": 12877
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 23.678401947021484,
+      "learning_rate": 9.22562021503577e-06,
+      "loss": 1.7279,
+      "step": 12878
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 16.571250915527344,
+      "learning_rate": 9.224783499979083e-06,
+      "loss": 0.9984,
+      "step": 12879
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 16.851226806640625,
+      "learning_rate": 9.223946784922395e-06,
+      "loss": 2.244,
+      "step": 12880
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 8.822027206420898,
+      "learning_rate": 9.223110069865709e-06,
+      "loss": 0.6714,
+      "step": 12881
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.036003112792969,
+      "learning_rate": 9.222273354809021e-06,
+      "loss": 1.1554,
+      "step": 12882
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 6.548466205596924,
+      "learning_rate": 9.221436639752333e-06,
+      "loss": 0.1496,
+      "step": 12883
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 14.927818298339844,
+      "learning_rate": 9.220599924695645e-06,
+      "loss": 2.1248,
+      "step": 12884
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 35.90271759033203,
+      "learning_rate": 9.219763209638959e-06,
+      "loss": 1.7514,
+      "step": 12885
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 21.67229652404785,
+      "learning_rate": 9.21892649458227e-06,
+      "loss": 1.4736,
+      "step": 12886
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.010519027709961,
+      "learning_rate": 9.218089779525582e-06,
+      "loss": 1.2694,
+      "step": 12887
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 11.781586647033691,
+      "learning_rate": 9.217253064468896e-06,
+      "loss": 0.6613,
+      "step": 12888
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 19.111835479736328,
+      "learning_rate": 9.216416349412208e-06,
+      "loss": 3.3232,
+      "step": 12889
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.765817642211914,
+      "learning_rate": 9.215579634355522e-06,
+      "loss": 0.7193,
+      "step": 12890
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 23.58574867248535,
+      "learning_rate": 9.214742919298834e-06,
+      "loss": 1.5909,
+      "step": 12891
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.807656288146973,
+      "learning_rate": 9.213906204242146e-06,
+      "loss": 0.471,
+      "step": 12892
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 13.614156723022461,
+      "learning_rate": 9.21306948918546e-06,
+      "loss": 0.6731,
+      "step": 12893
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 12.069944381713867,
+      "learning_rate": 9.212232774128771e-06,
+      "loss": 1.1097,
+      "step": 12894
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.385514259338379,
+      "learning_rate": 9.211396059072085e-06,
+      "loss": 0.5647,
+      "step": 12895
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.53537368774414,
+      "learning_rate": 9.210559344015397e-06,
+      "loss": 0.4047,
+      "step": 12896
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 11.166543006896973,
+      "learning_rate": 9.209722628958709e-06,
+      "loss": 0.3255,
+      "step": 12897
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.068735599517822,
+      "learning_rate": 9.20888591390202e-06,
+      "loss": 0.3223,
+      "step": 12898
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 14.783595085144043,
+      "learning_rate": 9.208049198845334e-06,
+      "loss": 0.877,
+      "step": 12899
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 22.187023162841797,
+      "learning_rate": 9.207212483788646e-06,
+      "loss": 2.1927,
+      "step": 12900
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 8.431422233581543,
+      "learning_rate": 9.206375768731958e-06,
+      "loss": 0.9181,
+      "step": 12901
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 103.34065246582031,
+      "learning_rate": 9.205539053675272e-06,
+      "loss": 2.653,
+      "step": 12902
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 13.142014503479004,
+      "learning_rate": 9.204702338618584e-06,
+      "loss": 1.3479,
+      "step": 12903
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 15.372876167297363,
+      "learning_rate": 9.203865623561898e-06,
+      "loss": 2.6061,
+      "step": 12904
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 12.46223258972168,
+      "learning_rate": 9.20302890850521e-06,
+      "loss": 0.6639,
+      "step": 12905
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 115.90477752685547,
+      "learning_rate": 9.202192193448521e-06,
+      "loss": 1.2725,
+      "step": 12906
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 5.032820224761963,
+      "learning_rate": 9.201355478391835e-06,
+      "loss": 0.2912,
+      "step": 12907
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 39.1680908203125,
+      "learning_rate": 9.200518763335147e-06,
+      "loss": 1.3619,
+      "step": 12908
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 17.965621948242188,
+      "learning_rate": 9.19968204827846e-06,
+      "loss": 0.9468,
+      "step": 12909
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 18.22383689880371,
+      "learning_rate": 9.198845333221773e-06,
+      "loss": 2.1563,
+      "step": 12910
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 31.021142959594727,
+      "learning_rate": 9.198008618165085e-06,
+      "loss": 1.0991,
+      "step": 12911
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 18.455698013305664,
+      "learning_rate": 9.197171903108397e-06,
+      "loss": 2.0503,
+      "step": 12912
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 42.6202507019043,
+      "learning_rate": 9.196335188051709e-06,
+      "loss": 1.5054,
+      "step": 12913
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 15.459112167358398,
+      "learning_rate": 9.195498472995022e-06,
+      "loss": 1.3822,
+      "step": 12914
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 18.592182159423828,
+      "learning_rate": 9.194661757938334e-06,
+      "loss": 0.5478,
+      "step": 12915
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 29.801151275634766,
+      "learning_rate": 9.193825042881648e-06,
+      "loss": 2.387,
+      "step": 12916
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 63.17427444458008,
+      "learning_rate": 9.19298832782496e-06,
+      "loss": 3.607,
+      "step": 12917
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 15.072610855102539,
+      "learning_rate": 9.192151612768273e-06,
+      "loss": 1.1563,
+      "step": 12918
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 33.61115264892578,
+      "learning_rate": 9.191314897711585e-06,
+      "loss": 1.1586,
+      "step": 12919
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 15.370922088623047,
+      "learning_rate": 9.190478182654897e-06,
+      "loss": 1.1077,
+      "step": 12920
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 25.674022674560547,
+      "learning_rate": 9.189641467598211e-06,
+      "loss": 1.09,
+      "step": 12921
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 5.827080726623535,
+      "learning_rate": 9.188804752541523e-06,
+      "loss": 0.3103,
+      "step": 12922
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 21.401575088500977,
+      "learning_rate": 9.187968037484837e-06,
+      "loss": 1.5182,
+      "step": 12923
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.65243148803711,
+      "learning_rate": 9.187131322428148e-06,
+      "loss": 1.3113,
+      "step": 12924
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.55691909790039,
+      "learning_rate": 9.18629460737146e-06,
+      "loss": 2.3306,
+      "step": 12925
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 4.405165672302246,
+      "learning_rate": 9.185457892314772e-06,
+      "loss": 0.148,
+      "step": 12926
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 6.446748733520508,
+      "learning_rate": 9.184621177258084e-06,
+      "loss": 0.5479,
+      "step": 12927
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 11.645052909851074,
+      "learning_rate": 9.183784462201398e-06,
+      "loss": 1.109,
+      "step": 12928
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 11.832785606384277,
+      "learning_rate": 9.18294774714471e-06,
+      "loss": 0.9107,
+      "step": 12929
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 8.79918384552002,
+      "learning_rate": 9.182111032088024e-06,
+      "loss": 1.1871,
+      "step": 12930
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 11.487600326538086,
+      "learning_rate": 9.181274317031336e-06,
+      "loss": 0.3511,
+      "step": 12931
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 11.59465503692627,
+      "learning_rate": 9.18043760197465e-06,
+      "loss": 0.4442,
+      "step": 12932
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.8561928272247314,
+      "learning_rate": 9.179600886917961e-06,
+      "loss": 0.1065,
+      "step": 12933
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 16.970701217651367,
+      "learning_rate": 9.178764171861273e-06,
+      "loss": 1.8206,
+      "step": 12934
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 25.955537796020508,
+      "learning_rate": 9.177927456804587e-06,
+      "loss": 3.0464,
+      "step": 12935
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 190.4348602294922,
+      "learning_rate": 9.177090741747899e-06,
+      "loss": 0.8961,
+      "step": 12936
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 8.63735580444336,
+      "learning_rate": 9.17625402669121e-06,
+      "loss": 0.3384,
+      "step": 12937
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 75.52478790283203,
+      "learning_rate": 9.175417311634524e-06,
+      "loss": 0.972,
+      "step": 12938
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 12.855717658996582,
+      "learning_rate": 9.174580596577836e-06,
+      "loss": 1.4988,
+      "step": 12939
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 61.936988830566406,
+      "learning_rate": 9.173743881521148e-06,
+      "loss": 2.4051,
+      "step": 12940
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 20.956762313842773,
+      "learning_rate": 9.17290716646446e-06,
+      "loss": 0.7292,
+      "step": 12941
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 39.11030960083008,
+      "learning_rate": 9.172070451407774e-06,
+      "loss": 0.7878,
+      "step": 12942
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 12.907938003540039,
+      "learning_rate": 9.171233736351086e-06,
+      "loss": 1.0599,
+      "step": 12943
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.939062118530273,
+      "learning_rate": 9.1703970212944e-06,
+      "loss": 0.399,
+      "step": 12944
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 12.606284141540527,
+      "learning_rate": 9.169560306237711e-06,
+      "loss": 1.4578,
+      "step": 12945
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 28.248008728027344,
+      "learning_rate": 9.168723591181023e-06,
+      "loss": 1.1074,
+      "step": 12946
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 12.51881217956543,
+      "learning_rate": 9.167886876124337e-06,
+      "loss": 1.7508,
+      "step": 12947
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 6.950279235839844,
+      "learning_rate": 9.167050161067649e-06,
+      "loss": 0.6877,
+      "step": 12948
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.21493673324585,
+      "learning_rate": 9.166213446010963e-06,
+      "loss": 0.6966,
+      "step": 12949
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 61.48523712158203,
+      "learning_rate": 9.165376730954275e-06,
+      "loss": 1.9934,
+      "step": 12950
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 28.346878051757812,
+      "learning_rate": 9.164540015897587e-06,
+      "loss": 1.3782,
+      "step": 12951
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 14.001502990722656,
+      "learning_rate": 9.1637033008409e-06,
+      "loss": 1.5156,
+      "step": 12952
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 25.5045108795166,
+      "learning_rate": 9.162866585784212e-06,
+      "loss": 2.0468,
+      "step": 12953
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 16.51694679260254,
+      "learning_rate": 9.162029870727524e-06,
+      "loss": 1.3583,
+      "step": 12954
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 17.545406341552734,
+      "learning_rate": 9.161193155670836e-06,
+      "loss": 0.8695,
+      "step": 12955
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.469407081604004,
+      "learning_rate": 9.16035644061415e-06,
+      "loss": 0.3411,
+      "step": 12956
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 9.810575485229492,
+      "learning_rate": 9.159519725557462e-06,
+      "loss": 0.8654,
+      "step": 12957
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.806991577148438,
+      "learning_rate": 9.158683010500775e-06,
+      "loss": 0.2469,
+      "step": 12958
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 12.568473815917969,
+      "learning_rate": 9.157846295444087e-06,
+      "loss": 2.249,
+      "step": 12959
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.245036602020264,
+      "learning_rate": 9.1570095803874e-06,
+      "loss": 0.8974,
+      "step": 12960
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 21.185012817382812,
+      "learning_rate": 9.156172865330713e-06,
+      "loss": 1.611,
+      "step": 12961
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.256817817687988,
+      "learning_rate": 9.155336150274025e-06,
+      "loss": 0.4137,
+      "step": 12962
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 6.60871696472168,
+      "learning_rate": 9.154499435217338e-06,
+      "loss": 0.8335,
+      "step": 12963
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 16.758840560913086,
+      "learning_rate": 9.15366272016065e-06,
+      "loss": 1.1956,
+      "step": 12964
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.283269882202148,
+      "learning_rate": 9.152826005103962e-06,
+      "loss": 0.757,
+      "step": 12965
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.587263584136963,
+      "learning_rate": 9.151989290047274e-06,
+      "loss": 0.3193,
+      "step": 12966
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 15.995939254760742,
+      "learning_rate": 9.151152574990588e-06,
+      "loss": 0.7247,
+      "step": 12967
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.188906192779541,
+      "learning_rate": 9.1503158599339e-06,
+      "loss": 0.3046,
+      "step": 12968
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 19.06035614013672,
+      "learning_rate": 9.149479144877212e-06,
+      "loss": 0.7589,
+      "step": 12969
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 28.602251052856445,
+      "learning_rate": 9.148642429820526e-06,
+      "loss": 0.3866,
+      "step": 12970
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.895336151123047,
+      "learning_rate": 9.147805714763837e-06,
+      "loss": 0.678,
+      "step": 12971
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 23.167448043823242,
+      "learning_rate": 9.146968999707151e-06,
+      "loss": 1.1801,
+      "step": 12972
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 21.483211517333984,
+      "learning_rate": 9.146132284650463e-06,
+      "loss": 0.7266,
+      "step": 12973
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 19.09107208251953,
+      "learning_rate": 9.145295569593775e-06,
+      "loss": 0.7794,
+      "step": 12974
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 15.3024263381958,
+      "learning_rate": 9.144458854537089e-06,
+      "loss": 1.7888,
+      "step": 12975
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 15.789032936096191,
+      "learning_rate": 9.1436221394804e-06,
+      "loss": 1.2744,
+      "step": 12976
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 6.284669399261475,
+      "learning_rate": 9.142785424423714e-06,
+      "loss": 2.9806,
+      "step": 12977
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 15.586581230163574,
+      "learning_rate": 9.141948709367026e-06,
+      "loss": 0.7945,
+      "step": 12978
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 16.861600875854492,
+      "learning_rate": 9.141111994310338e-06,
+      "loss": 1.2064,
+      "step": 12979
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 14.14163589477539,
+      "learning_rate": 9.14027527925365e-06,
+      "loss": 0.5206,
+      "step": 12980
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.704914093017578,
+      "learning_rate": 9.139438564196964e-06,
+      "loss": 0.773,
+      "step": 12981
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 54.307395935058594,
+      "learning_rate": 9.138601849140276e-06,
+      "loss": 2.7865,
+      "step": 12982
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 16.4180965423584,
+      "learning_rate": 9.137765134083588e-06,
+      "loss": 1.4351,
+      "step": 12983
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.057581901550293,
+      "learning_rate": 9.136928419026901e-06,
+      "loss": 0.3731,
+      "step": 12984
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 29.563465118408203,
+      "learning_rate": 9.136091703970213e-06,
+      "loss": 1.1386,
+      "step": 12985
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.984923362731934,
+      "learning_rate": 9.135254988913527e-06,
+      "loss": 0.3085,
+      "step": 12986
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 4.832681179046631,
+      "learning_rate": 9.134418273856839e-06,
+      "loss": 0.5914,
+      "step": 12987
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.778252601623535,
+      "learning_rate": 9.133581558800151e-06,
+      "loss": 0.347,
+      "step": 12988
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 13.914613723754883,
+      "learning_rate": 9.132744843743464e-06,
+      "loss": 2.0648,
+      "step": 12989
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.425278663635254,
+      "learning_rate": 9.131908128686776e-06,
+      "loss": 1.4011,
+      "step": 12990
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 13.942360877990723,
+      "learning_rate": 9.13107141363009e-06,
+      "loss": 1.6094,
+      "step": 12991
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.301091194152832,
+      "learning_rate": 9.130234698573402e-06,
+      "loss": 1.8374,
+      "step": 12992
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 19.2606201171875,
+      "learning_rate": 9.129397983516714e-06,
+      "loss": 1.2653,
+      "step": 12993
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 13.116103172302246,
+      "learning_rate": 9.128561268460026e-06,
+      "loss": 0.5151,
+      "step": 12994
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.208422660827637,
+      "learning_rate": 9.127724553403338e-06,
+      "loss": 2.3458,
+      "step": 12995
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 55.8681640625,
+      "learning_rate": 9.126887838346652e-06,
+      "loss": 1.2972,
+      "step": 12996
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.729084014892578,
+      "learning_rate": 9.126051123289964e-06,
+      "loss": 1.3225,
+      "step": 12997
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.963275909423828,
+      "learning_rate": 9.125214408233277e-06,
+      "loss": 0.9134,
+      "step": 12998
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 18.510528564453125,
+      "learning_rate": 9.124377693176589e-06,
+      "loss": 1.1661,
+      "step": 12999
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 6.595053672790527,
+      "learning_rate": 9.123540978119903e-06,
+      "loss": 0.5564,
+      "step": 13000
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.768023490905762,
+      "learning_rate": 9.122704263063215e-06,
+      "loss": 0.7828,
+      "step": 13001
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.766725540161133,
+      "learning_rate": 9.121867548006527e-06,
+      "loss": 1.9123,
+      "step": 13002
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 9.593929290771484,
+      "learning_rate": 9.12103083294984e-06,
+      "loss": 0.445,
+      "step": 13003
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 35.650943756103516,
+      "learning_rate": 9.120194117893152e-06,
+      "loss": 2.4764,
+      "step": 13004
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.9402058124542236,
+      "learning_rate": 9.119357402836464e-06,
+      "loss": 0.1609,
+      "step": 13005
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 28.445289611816406,
+      "learning_rate": 9.118520687779778e-06,
+      "loss": 2.7993,
+      "step": 13006
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 9.816569328308105,
+      "learning_rate": 9.11768397272309e-06,
+      "loss": 0.4052,
+      "step": 13007
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 150.9856719970703,
+      "learning_rate": 9.116847257666402e-06,
+      "loss": 1.4806,
+      "step": 13008
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 16.420270919799805,
+      "learning_rate": 9.116010542609714e-06,
+      "loss": 0.8386,
+      "step": 13009
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.546241760253906,
+      "learning_rate": 9.115173827553027e-06,
+      "loss": 0.4028,
+      "step": 13010
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 13.879183769226074,
+      "learning_rate": 9.11433711249634e-06,
+      "loss": 0.8947,
+      "step": 13011
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 54.53948211669922,
+      "learning_rate": 9.113500397439653e-06,
+      "loss": 1.996,
+      "step": 13012
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 18.00624656677246,
+      "learning_rate": 9.112663682382965e-06,
+      "loss": 1.6993,
+      "step": 13013
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 15.442643165588379,
+      "learning_rate": 9.111826967326279e-06,
+      "loss": 1.7874,
+      "step": 13014
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 46.516849517822266,
+      "learning_rate": 9.11099025226959e-06,
+      "loss": 1.399,
+      "step": 13015
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 75.43692779541016,
+      "learning_rate": 9.110153537212903e-06,
+      "loss": 1.3915,
+      "step": 13016
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.721510887145996,
+      "learning_rate": 9.109316822156216e-06,
+      "loss": 0.177,
+      "step": 13017
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.118610382080078,
+      "learning_rate": 9.108480107099528e-06,
+      "loss": 1.1014,
+      "step": 13018
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 89.78044891357422,
+      "learning_rate": 9.10764339204284e-06,
+      "loss": 0.9686,
+      "step": 13019
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 6.995342254638672,
+      "learning_rate": 9.106806676986154e-06,
+      "loss": 2.3016,
+      "step": 13020
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 13.39775562286377,
+      "learning_rate": 9.105969961929466e-06,
+      "loss": 1.4211,
+      "step": 13021
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 24.990713119506836,
+      "learning_rate": 9.105133246872778e-06,
+      "loss": 1.2749,
+      "step": 13022
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.517844200134277,
+      "learning_rate": 9.10429653181609e-06,
+      "loss": 0.2187,
+      "step": 13023
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 10.690154075622559,
+      "learning_rate": 9.103459816759403e-06,
+      "loss": 0.395,
+      "step": 13024
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 18.35173988342285,
+      "learning_rate": 9.102623101702715e-06,
+      "loss": 0.6232,
+      "step": 13025
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 101.38148498535156,
+      "learning_rate": 9.101786386646029e-06,
+      "loss": 0.7097,
+      "step": 13026
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 17.962690353393555,
+      "learning_rate": 9.10094967158934e-06,
+      "loss": 0.834,
+      "step": 13027
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 20.415565490722656,
+      "learning_rate": 9.100112956532654e-06,
+      "loss": 1.1577,
+      "step": 13028
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 14.006248474121094,
+      "learning_rate": 9.099276241475966e-06,
+      "loss": 0.7075,
+      "step": 13029
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 13.301826477050781,
+      "learning_rate": 9.098439526419278e-06,
+      "loss": 0.5778,
+      "step": 13030
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.223112106323242,
+      "learning_rate": 9.097602811362592e-06,
+      "loss": 0.6171,
+      "step": 13031
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 20.63963508605957,
+      "learning_rate": 9.096766096305904e-06,
+      "loss": 1.2354,
+      "step": 13032
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 16.198240280151367,
+      "learning_rate": 9.095929381249216e-06,
+      "loss": 1.3007,
+      "step": 13033
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.846941947937012,
+      "learning_rate": 9.09509266619253e-06,
+      "loss": 0.9787,
+      "step": 13034
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 6.430784225463867,
+      "learning_rate": 9.094255951135842e-06,
+      "loss": 0.2251,
+      "step": 13035
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 12.79133415222168,
+      "learning_rate": 9.093419236079153e-06,
+      "loss": 1.9349,
+      "step": 13036
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 30.484493255615234,
+      "learning_rate": 9.092582521022465e-06,
+      "loss": 1.7943,
+      "step": 13037
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 47.64484405517578,
+      "learning_rate": 9.091745805965779e-06,
+      "loss": 1.2628,
+      "step": 13038
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 15.437004089355469,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 1.1696,
+      "step": 13039
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 4.615384101867676,
+      "learning_rate": 9.090072375852405e-06,
+      "loss": 0.4406,
+      "step": 13040
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 15.526065826416016,
+      "learning_rate": 9.089235660795717e-06,
+      "loss": 0.4678,
+      "step": 13041
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 22.352813720703125,
+      "learning_rate": 9.08839894573903e-06,
+      "loss": 1.5878,
+      "step": 13042
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 29.267131805419922,
+      "learning_rate": 9.087562230682342e-06,
+      "loss": 1.5648,
+      "step": 13043
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.587777137756348,
+      "learning_rate": 9.086725515625654e-06,
+      "loss": 1.0474,
+      "step": 13044
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 14.594001770019531,
+      "learning_rate": 9.085888800568968e-06,
+      "loss": 1.7063,
+      "step": 13045
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 18.16721534729004,
+      "learning_rate": 9.08505208551228e-06,
+      "loss": 1.147,
+      "step": 13046
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 27.573862075805664,
+      "learning_rate": 9.084215370455592e-06,
+      "loss": 0.7921,
+      "step": 13047
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 18.00398826599121,
+      "learning_rate": 9.083378655398904e-06,
+      "loss": 0.6198,
+      "step": 13048
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 6.626566410064697,
+      "learning_rate": 9.082541940342217e-06,
+      "loss": 0.7332,
+      "step": 13049
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 39.86878967285156,
+      "learning_rate": 9.08170522528553e-06,
+      "loss": 0.9836,
+      "step": 13050
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 413.9939880371094,
+      "learning_rate": 9.080868510228841e-06,
+      "loss": 1.1279,
+      "step": 13051
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.7134273052215576,
+      "learning_rate": 9.080031795172155e-06,
+      "loss": 0.0985,
+      "step": 13052
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 12.295166969299316,
+      "learning_rate": 9.079195080115467e-06,
+      "loss": 1.1999,
+      "step": 13053
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 47.10236358642578,
+      "learning_rate": 9.07835836505878e-06,
+      "loss": 2.3204,
+      "step": 13054
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 34.14249801635742,
+      "learning_rate": 9.077521650002092e-06,
+      "loss": 2.6892,
+      "step": 13055
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.125469207763672,
+      "learning_rate": 9.076684934945406e-06,
+      "loss": 2.22,
+      "step": 13056
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.195199966430664,
+      "learning_rate": 9.075848219888718e-06,
+      "loss": 1.3924,
+      "step": 13057
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 20.390213012695312,
+      "learning_rate": 9.07501150483203e-06,
+      "loss": 1.5304,
+      "step": 13058
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.653514862060547,
+      "learning_rate": 9.074174789775344e-06,
+      "loss": 0.4206,
+      "step": 13059
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 4.093517303466797,
+      "learning_rate": 9.073338074718656e-06,
+      "loss": 0.2536,
+      "step": 13060
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 4.747400760650635,
+      "learning_rate": 9.072501359661968e-06,
+      "loss": 1.274,
+      "step": 13061
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 65.5945816040039,
+      "learning_rate": 9.07166464460528e-06,
+      "loss": 1.0873,
+      "step": 13062
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 14.872116088867188,
+      "learning_rate": 9.070827929548593e-06,
+      "loss": 1.0578,
+      "step": 13063
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 12.61624813079834,
+      "learning_rate": 9.069991214491905e-06,
+      "loss": 1.4042,
+      "step": 13064
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 15.354135513305664,
+      "learning_rate": 9.069154499435217e-06,
+      "loss": 0.7141,
+      "step": 13065
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 7.654422283172607,
+      "learning_rate": 9.06831778437853e-06,
+      "loss": 1.499,
+      "step": 13066
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 21.095413208007812,
+      "learning_rate": 9.067481069321843e-06,
+      "loss": 1.2577,
+      "step": 13067
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 50.755958557128906,
+      "learning_rate": 9.066644354265156e-06,
+      "loss": 2.6346,
+      "step": 13068
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 16.771076202392578,
+      "learning_rate": 9.065807639208468e-06,
+      "loss": 0.6723,
+      "step": 13069
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 3.3442487716674805,
+      "learning_rate": 9.064970924151782e-06,
+      "loss": 0.2401,
+      "step": 13070
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.364527702331543,
+      "learning_rate": 9.064134209095094e-06,
+      "loss": 1.9107,
+      "step": 13071
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 67.20685577392578,
+      "learning_rate": 9.063297494038406e-06,
+      "loss": 3.1088,
+      "step": 13072
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 12.825528144836426,
+      "learning_rate": 9.06246077898172e-06,
+      "loss": 0.7203,
+      "step": 13073
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 7.3285698890686035,
+      "learning_rate": 9.061624063925031e-06,
+      "loss": 0.3413,
+      "step": 13074
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 4.640127658843994,
+      "learning_rate": 9.060787348868343e-06,
+      "loss": 0.1723,
+      "step": 13075
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.927471160888672,
+      "learning_rate": 9.059950633811655e-06,
+      "loss": 1.1897,
+      "step": 13076
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 29.194480895996094,
+      "learning_rate": 9.059113918754969e-06,
+      "loss": 2.18,
+      "step": 13077
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.097738265991211,
+      "learning_rate": 9.058277203698281e-06,
+      "loss": 0.4861,
+      "step": 13078
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.935216903686523,
+      "learning_rate": 9.057440488641593e-06,
+      "loss": 0.7497,
+      "step": 13079
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 23.826723098754883,
+      "learning_rate": 9.056603773584907e-06,
+      "loss": 1.7941,
+      "step": 13080
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.864187240600586,
+      "learning_rate": 9.055767058528219e-06,
+      "loss": 0.5174,
+      "step": 13081
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.379133224487305,
+      "learning_rate": 9.054930343471532e-06,
+      "loss": 1.9926,
+      "step": 13082
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 21.72909927368164,
+      "learning_rate": 9.054093628414844e-06,
+      "loss": 0.9551,
+      "step": 13083
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 69.47557830810547,
+      "learning_rate": 9.053256913358158e-06,
+      "loss": 1.2902,
+      "step": 13084
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 7.750579357147217,
+      "learning_rate": 9.05242019830147e-06,
+      "loss": 1.019,
+      "step": 13085
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 5.383609771728516,
+      "learning_rate": 9.051583483244782e-06,
+      "loss": 0.272,
+      "step": 13086
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 18.136207580566406,
+      "learning_rate": 9.050746768188094e-06,
+      "loss": 3.1578,
+      "step": 13087
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 33.227691650390625,
+      "learning_rate": 9.049910053131407e-06,
+      "loss": 0.9955,
+      "step": 13088
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 15.877317428588867,
+      "learning_rate": 9.04907333807472e-06,
+      "loss": 0.7077,
+      "step": 13089
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 17.573396682739258,
+      "learning_rate": 9.048236623018031e-06,
+      "loss": 1.6198,
+      "step": 13090
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 14.696701049804688,
+      "learning_rate": 9.047399907961345e-06,
+      "loss": 0.7756,
+      "step": 13091
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.865352630615234,
+      "learning_rate": 9.046563192904657e-06,
+      "loss": 0.5272,
+      "step": 13092
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 13.953130722045898,
+      "learning_rate": 9.045726477847969e-06,
+      "loss": 0.6747,
+      "step": 13093
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 53.01630401611328,
+      "learning_rate": 9.044889762791282e-06,
+      "loss": 1.9975,
+      "step": 13094
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 3.929391860961914,
+      "learning_rate": 9.044053047734594e-06,
+      "loss": 0.1135,
+      "step": 13095
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 56.032047271728516,
+      "learning_rate": 9.043216332677908e-06,
+      "loss": 2.2502,
+      "step": 13096
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 16.776247024536133,
+      "learning_rate": 9.04237961762122e-06,
+      "loss": 0.5896,
+      "step": 13097
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.387946128845215,
+      "learning_rate": 9.041542902564534e-06,
+      "loss": 0.7921,
+      "step": 13098
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 12.66744327545166,
+      "learning_rate": 9.040706187507846e-06,
+      "loss": 1.4705,
+      "step": 13099
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.530153274536133,
+      "learning_rate": 9.039869472451158e-06,
+      "loss": 0.7724,
+      "step": 13100
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 30.687973022460938,
+      "learning_rate": 9.03903275739447e-06,
+      "loss": 1.1171,
+      "step": 13101
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.460206031799316,
+      "learning_rate": 9.038196042337783e-06,
+      "loss": 2.4668,
+      "step": 13102
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.673661231994629,
+      "learning_rate": 9.037359327281095e-06,
+      "loss": 0.3487,
+      "step": 13103
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.060383796691895,
+      "learning_rate": 9.036522612224407e-06,
+      "loss": 1.2103,
+      "step": 13104
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 13.42879581451416,
+      "learning_rate": 9.03568589716772e-06,
+      "loss": 0.8598,
+      "step": 13105
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 438.6036376953125,
+      "learning_rate": 9.034849182111033e-06,
+      "loss": 2.0738,
+      "step": 13106
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 25.97335433959961,
+      "learning_rate": 9.034012467054345e-06,
+      "loss": 0.9358,
+      "step": 13107
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 14.31015682220459,
+      "learning_rate": 9.033175751997658e-06,
+      "loss": 0.6819,
+      "step": 13108
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 15.247424125671387,
+      "learning_rate": 9.03233903694097e-06,
+      "loss": 0.8224,
+      "step": 13109
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 4.5875420570373535,
+      "learning_rate": 9.031502321884284e-06,
+      "loss": 0.1946,
+      "step": 13110
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 17.96656608581543,
+      "learning_rate": 9.030665606827596e-06,
+      "loss": 0.9879,
+      "step": 13111
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 4.172995567321777,
+      "learning_rate": 9.029828891770908e-06,
+      "loss": 0.6168,
+      "step": 13112
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 6.244558811187744,
+      "learning_rate": 9.028992176714221e-06,
+      "loss": 0.1583,
+      "step": 13113
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.22422981262207,
+      "learning_rate": 9.028155461657533e-06,
+      "loss": 0.6554,
+      "step": 13114
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 31.16236114501953,
+      "learning_rate": 9.027318746600845e-06,
+      "loss": 2.6027,
+      "step": 13115
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 51.754554748535156,
+      "learning_rate": 9.026482031544159e-06,
+      "loss": 2.1243,
+      "step": 13116
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 12.304715156555176,
+      "learning_rate": 9.025645316487471e-06,
+      "loss": 1.1691,
+      "step": 13117
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 34.253997802734375,
+      "learning_rate": 9.024808601430783e-06,
+      "loss": 1.2489,
+      "step": 13118
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 18.965248107910156,
+      "learning_rate": 9.023971886374095e-06,
+      "loss": 1.4573,
+      "step": 13119
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 30.697521209716797,
+      "learning_rate": 9.023135171317409e-06,
+      "loss": 1.608,
+      "step": 13120
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 34.999149322509766,
+      "learning_rate": 9.02229845626072e-06,
+      "loss": 1.2003,
+      "step": 13121
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 28.988059997558594,
+      "learning_rate": 9.021461741204034e-06,
+      "loss": 2.2427,
+      "step": 13122
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.598092079162598,
+      "learning_rate": 9.020625026147346e-06,
+      "loss": 0.4843,
+      "step": 13123
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 12.560540199279785,
+      "learning_rate": 9.01978831109066e-06,
+      "loss": 0.4743,
+      "step": 13124
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 14.121764183044434,
+      "learning_rate": 9.018951596033972e-06,
+      "loss": 2.2912,
+      "step": 13125
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 30.48697853088379,
+      "learning_rate": 9.018114880977284e-06,
+      "loss": 1.9225,
+      "step": 13126
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.957478523254395,
+      "learning_rate": 9.017278165920597e-06,
+      "loss": 0.7876,
+      "step": 13127
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.144898414611816,
+      "learning_rate": 9.01644145086391e-06,
+      "loss": 0.69,
+      "step": 13128
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 22.866575241088867,
+      "learning_rate": 9.015604735807221e-06,
+      "loss": 1.3409,
+      "step": 13129
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 13.678271293640137,
+      "learning_rate": 9.014768020750533e-06,
+      "loss": 0.8963,
+      "step": 13130
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 26.363157272338867,
+      "learning_rate": 9.013931305693847e-06,
+      "loss": 1.6555,
+      "step": 13131
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 147.813720703125,
+      "learning_rate": 9.013094590637159e-06,
+      "loss": 2.2164,
+      "step": 13132
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.695622444152832,
+      "learning_rate": 9.01225787558047e-06,
+      "loss": 1.7061,
+      "step": 13133
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 19.920499801635742,
+      "learning_rate": 9.011421160523784e-06,
+      "loss": 1.1819,
+      "step": 13134
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 29.973834991455078,
+      "learning_rate": 9.010584445467096e-06,
+      "loss": 1.4635,
+      "step": 13135
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.446659088134766,
+      "learning_rate": 9.00974773041041e-06,
+      "loss": 1.5663,
+      "step": 13136
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 13.579355239868164,
+      "learning_rate": 9.008911015353722e-06,
+      "loss": 1.9149,
+      "step": 13137
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 4.746458053588867,
+      "learning_rate": 9.008074300297036e-06,
+      "loss": 0.2992,
+      "step": 13138
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 10.081870079040527,
+      "learning_rate": 9.007237585240347e-06,
+      "loss": 0.6807,
+      "step": 13139
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.301534175872803,
+      "learning_rate": 9.00640087018366e-06,
+      "loss": 0.87,
+      "step": 13140
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.21815299987793,
+      "learning_rate": 9.005564155126973e-06,
+      "loss": 1.4987,
+      "step": 13141
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 4.622696876525879,
+      "learning_rate": 9.004727440070285e-06,
+      "loss": 0.1623,
+      "step": 13142
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 29.417076110839844,
+      "learning_rate": 9.003890725013597e-06,
+      "loss": 1.4325,
+      "step": 13143
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 6.1637115478515625,
+      "learning_rate": 9.003054009956909e-06,
+      "loss": 1.247,
+      "step": 13144
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 22.631696701049805,
+      "learning_rate": 9.002217294900223e-06,
+      "loss": 1.1119,
+      "step": 13145
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 17.008960723876953,
+      "learning_rate": 9.001380579843535e-06,
+      "loss": 1.8602,
+      "step": 13146
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 4.248160362243652,
+      "learning_rate": 9.000543864786847e-06,
+      "loss": 1.6219,
+      "step": 13147
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 63.04439163208008,
+      "learning_rate": 8.99970714973016e-06,
+      "loss": 1.4084,
+      "step": 13148
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.603978157043457,
+      "learning_rate": 8.998870434673472e-06,
+      "loss": 0.8025,
+      "step": 13149
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 508.3733215332031,
+      "learning_rate": 8.998033719616786e-06,
+      "loss": 3.8823,
+      "step": 13150
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 39.03932571411133,
+      "learning_rate": 8.997197004560098e-06,
+      "loss": 1.1376,
+      "step": 13151
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 12.622542381286621,
+      "learning_rate": 8.996360289503411e-06,
+      "loss": 0.6019,
+      "step": 13152
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.678735733032227,
+      "learning_rate": 8.995523574446723e-06,
+      "loss": 1.2947,
+      "step": 13153
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 12.827431678771973,
+      "learning_rate": 8.994686859390035e-06,
+      "loss": 0.7529,
+      "step": 13154
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 14.978957176208496,
+      "learning_rate": 8.993850144333349e-06,
+      "loss": 1.5085,
+      "step": 13155
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 12.834829330444336,
+      "learning_rate": 8.993013429276661e-06,
+      "loss": 0.5835,
+      "step": 13156
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.3710126876831055,
+      "learning_rate": 8.992176714219973e-06,
+      "loss": 0.3713,
+      "step": 13157
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 21.46312141418457,
+      "learning_rate": 8.991339999163285e-06,
+      "loss": 1.1262,
+      "step": 13158
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.169536828994751,
+      "learning_rate": 8.990503284106598e-06,
+      "loss": 0.0426,
+      "step": 13159
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 6.690242767333984,
+      "learning_rate": 8.98966656904991e-06,
+      "loss": 1.1321,
+      "step": 13160
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 5.8285040855407715,
+      "learning_rate": 8.988829853993222e-06,
+      "loss": 1.5329,
+      "step": 13161
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 5.963802337646484,
+      "learning_rate": 8.987993138936536e-06,
+      "loss": 0.2929,
+      "step": 13162
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 14.935331344604492,
+      "learning_rate": 8.987156423879848e-06,
+      "loss": 1.0795,
+      "step": 13163
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 13.327879905700684,
+      "learning_rate": 8.986319708823162e-06,
+      "loss": 0.9714,
+      "step": 13164
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 6.610556125640869,
+      "learning_rate": 8.985482993766474e-06,
+      "loss": 0.3407,
+      "step": 13165
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.004228591918945,
+      "learning_rate": 8.984646278709787e-06,
+      "loss": 0.6166,
+      "step": 13166
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.152169227600098,
+      "learning_rate": 8.9838095636531e-06,
+      "loss": 0.4951,
+      "step": 13167
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 28.9377384185791,
+      "learning_rate": 8.982972848596411e-06,
+      "loss": 1.8287,
+      "step": 13168
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 15.092056274414062,
+      "learning_rate": 8.982136133539723e-06,
+      "loss": 0.914,
+      "step": 13169
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 12.820289611816406,
+      "learning_rate": 8.981299418483037e-06,
+      "loss": 0.5046,
+      "step": 13170
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 24.191577911376953,
+      "learning_rate": 8.980462703426349e-06,
+      "loss": 1.9734,
+      "step": 13171
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 5.4591450691223145,
+      "learning_rate": 8.97962598836966e-06,
+      "loss": 0.4638,
+      "step": 13172
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 12.546009063720703,
+      "learning_rate": 8.978789273312974e-06,
+      "loss": 0.9607,
+      "step": 13173
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.58516263961792,
+      "learning_rate": 8.977952558256286e-06,
+      "loss": 0.5153,
+      "step": 13174
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 23.11334991455078,
+      "learning_rate": 8.977115843199598e-06,
+      "loss": 1.3486,
+      "step": 13175
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 30.083221435546875,
+      "learning_rate": 8.976279128142912e-06,
+      "loss": 1.4435,
+      "step": 13176
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.878336906433105,
+      "learning_rate": 8.975442413086224e-06,
+      "loss": 2.6615,
+      "step": 13177
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 12.88902473449707,
+      "learning_rate": 8.974605698029537e-06,
+      "loss": 1.0042,
+      "step": 13178
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 23.017744064331055,
+      "learning_rate": 8.97376898297285e-06,
+      "loss": 1.5633,
+      "step": 13179
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 25.739858627319336,
+      "learning_rate": 8.972932267916163e-06,
+      "loss": 1.3024,
+      "step": 13180
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.158169746398926,
+      "learning_rate": 8.972095552859475e-06,
+      "loss": 1.7064,
+      "step": 13181
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 7.283500671386719,
+      "learning_rate": 8.971258837802787e-06,
+      "loss": 0.5594,
+      "step": 13182
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 18.39688491821289,
+      "learning_rate": 8.970422122746099e-06,
+      "loss": 1.7752,
+      "step": 13183
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.553594589233398,
+      "learning_rate": 8.969585407689413e-06,
+      "loss": 0.4924,
+      "step": 13184
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 18.251686096191406,
+      "learning_rate": 8.968748692632725e-06,
+      "loss": 1.7104,
+      "step": 13185
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 5.3052825927734375,
+      "learning_rate": 8.967911977576036e-06,
+      "loss": 0.5095,
+      "step": 13186
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 6.065149784088135,
+      "learning_rate": 8.96707526251935e-06,
+      "loss": 0.3137,
+      "step": 13187
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.02402114868164,
+      "learning_rate": 8.966238547462662e-06,
+      "loss": 1.0114,
+      "step": 13188
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 13.939468383789062,
+      "learning_rate": 8.965401832405974e-06,
+      "loss": 0.5848,
+      "step": 13189
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.761425018310547,
+      "learning_rate": 8.964565117349288e-06,
+      "loss": 0.784,
+      "step": 13190
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 25.257661819458008,
+      "learning_rate": 8.9637284022926e-06,
+      "loss": 0.819,
+      "step": 13191
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.442473411560059,
+      "learning_rate": 8.962891687235913e-06,
+      "loss": 0.689,
+      "step": 13192
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 14.158611297607422,
+      "learning_rate": 8.962054972179225e-06,
+      "loss": 0.469,
+      "step": 13193
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.559187889099121,
+      "learning_rate": 8.961218257122539e-06,
+      "loss": 0.7824,
+      "step": 13194
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.47706413269043,
+      "learning_rate": 8.960381542065851e-06,
+      "loss": 0.2856,
+      "step": 13195
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 15.341854095458984,
+      "learning_rate": 8.959544827009163e-06,
+      "loss": 1.8322,
+      "step": 13196
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 31.592058181762695,
+      "learning_rate": 8.958708111952475e-06,
+      "loss": 3.7815,
+      "step": 13197
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 21.079078674316406,
+      "learning_rate": 8.957871396895787e-06,
+      "loss": 2.3007,
+      "step": 13198
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 4.746679782867432,
+      "learning_rate": 8.9570346818391e-06,
+      "loss": 0.9576,
+      "step": 13199
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 20.35892105102539,
+      "learning_rate": 8.956197966782412e-06,
+      "loss": 0.6661,
+      "step": 13200
+    },
+    {
+      "epoch": 1.66,
+      "eval_loss": 0.10212861001491547,
+      "eval_runtime": 94.7891,
+      "eval_samples_per_second": 37.367,
+      "eval_steps_per_second": 37.367,
+      "step": 13200
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.064656257629395,
+      "learning_rate": 8.955361251725726e-06,
+      "loss": 1.6586,
+      "step": 13201
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 15.648777961730957,
+      "learning_rate": 8.954524536669038e-06,
+      "loss": 1.0213,
+      "step": 13202
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 34.27608108520508,
+      "learning_rate": 8.95368782161235e-06,
+      "loss": 2.2202,
+      "step": 13203
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 39.48409652709961,
+      "learning_rate": 8.952851106555664e-06,
+      "loss": 2.8217,
+      "step": 13204
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.432174682617188,
+      "learning_rate": 8.952014391498975e-06,
+      "loss": 1.1946,
+      "step": 13205
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 22.67432403564453,
+      "learning_rate": 8.951177676442289e-06,
+      "loss": 1.4657,
+      "step": 13206
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 20.243619918823242,
+      "learning_rate": 8.950340961385601e-06,
+      "loss": 0.729,
+      "step": 13207
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 143.8177947998047,
+      "learning_rate": 8.949504246328915e-06,
+      "loss": 0.6804,
+      "step": 13208
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 90.39649963378906,
+      "learning_rate": 8.948667531272227e-06,
+      "loss": 2.949,
+      "step": 13209
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.051481246948242,
+      "learning_rate": 8.947830816215539e-06,
+      "loss": 0.8819,
+      "step": 13210
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 12.02458381652832,
+      "learning_rate": 8.94699410115885e-06,
+      "loss": 0.8701,
+      "step": 13211
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.880314826965332,
+      "learning_rate": 8.946157386102163e-06,
+      "loss": 1.0519,
+      "step": 13212
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 3.207190990447998,
+      "learning_rate": 8.945320671045476e-06,
+      "loss": 0.1411,
+      "step": 13213
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 13.814050674438477,
+      "learning_rate": 8.944483955988788e-06,
+      "loss": 0.1641,
+      "step": 13214
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.019259452819824,
+      "learning_rate": 8.943647240932102e-06,
+      "loss": 0.7049,
+      "step": 13215
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.67119026184082,
+      "learning_rate": 8.942810525875414e-06,
+      "loss": 0.9873,
+      "step": 13216
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.612390518188477,
+      "learning_rate": 8.941973810818726e-06,
+      "loss": 0.9955,
+      "step": 13217
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.1115264892578125,
+      "learning_rate": 8.94113709576204e-06,
+      "loss": 0.4268,
+      "step": 13218
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 21.119964599609375,
+      "learning_rate": 8.940300380705351e-06,
+      "loss": 1.0977,
+      "step": 13219
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 14.732322692871094,
+      "learning_rate": 8.939463665648665e-06,
+      "loss": 1.9414,
+      "step": 13220
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 18.428756713867188,
+      "learning_rate": 8.938626950591977e-06,
+      "loss": 2.0643,
+      "step": 13221
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.45617389678955,
+      "learning_rate": 8.937790235535289e-06,
+      "loss": 0.7956,
+      "step": 13222
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.30435562133789,
+      "learning_rate": 8.936953520478603e-06,
+      "loss": 0.4185,
+      "step": 13223
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.009489059448242,
+      "learning_rate": 8.936116805421914e-06,
+      "loss": 1.2203,
+      "step": 13224
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.37564468383789,
+      "learning_rate": 8.935280090365226e-06,
+      "loss": 1.103,
+      "step": 13225
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 27.137699127197266,
+      "learning_rate": 8.934443375308538e-06,
+      "loss": 1.3436,
+      "step": 13226
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.834226608276367,
+      "learning_rate": 8.933606660251852e-06,
+      "loss": 2.5409,
+      "step": 13227
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 197.4990692138672,
+      "learning_rate": 8.932769945195164e-06,
+      "loss": 3.1156,
+      "step": 13228
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 52.09612274169922,
+      "learning_rate": 8.931933230138478e-06,
+      "loss": 2.612,
+      "step": 13229
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.936538696289062,
+      "learning_rate": 8.93109651508179e-06,
+      "loss": 1.1146,
+      "step": 13230
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 160.63803100585938,
+      "learning_rate": 8.930259800025102e-06,
+      "loss": 1.8826,
+      "step": 13231
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.45350456237793,
+      "learning_rate": 8.929423084968415e-06,
+      "loss": 1.1375,
+      "step": 13232
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 6.9955573081970215,
+      "learning_rate": 8.928586369911727e-06,
+      "loss": 0.2738,
+      "step": 13233
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.343851089477539,
+      "learning_rate": 8.92774965485504e-06,
+      "loss": 0.9436,
+      "step": 13234
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 13.81751537322998,
+      "learning_rate": 8.926912939798353e-06,
+      "loss": 1.8904,
+      "step": 13235
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.447096347808838,
+      "learning_rate": 8.926076224741665e-06,
+      "loss": 0.1465,
+      "step": 13236
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.009682655334473,
+      "learning_rate": 8.925239509684978e-06,
+      "loss": 0.9183,
+      "step": 13237
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 24.190349578857422,
+      "learning_rate": 8.92440279462829e-06,
+      "loss": 1.2925,
+      "step": 13238
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 124.643310546875,
+      "learning_rate": 8.923566079571602e-06,
+      "loss": 1.3646,
+      "step": 13239
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.799116134643555,
+      "learning_rate": 8.922729364514914e-06,
+      "loss": 0.4123,
+      "step": 13240
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 7.647983074188232,
+      "learning_rate": 8.921892649458228e-06,
+      "loss": 0.699,
+      "step": 13241
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.45411491394043,
+      "learning_rate": 8.92105593440154e-06,
+      "loss": 1.606,
+      "step": 13242
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 19.23842430114746,
+      "learning_rate": 8.920219219344853e-06,
+      "loss": 0.4207,
+      "step": 13243
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 23.47686195373535,
+      "learning_rate": 8.919382504288165e-06,
+      "loss": 1.8671,
+      "step": 13244
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.251797676086426,
+      "learning_rate": 8.918545789231477e-06,
+      "loss": 0.4017,
+      "step": 13245
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 13.252801895141602,
+      "learning_rate": 8.917709074174791e-06,
+      "loss": 0.4089,
+      "step": 13246
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 7.073338985443115,
+      "learning_rate": 8.916872359118103e-06,
+      "loss": 1.9157,
+      "step": 13247
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 32.249420166015625,
+      "learning_rate": 8.916035644061417e-06,
+      "loss": 3.1766,
+      "step": 13248
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 10.403848648071289,
+      "learning_rate": 8.915198929004729e-06,
+      "loss": 1.0762,
+      "step": 13249
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 11.180082321166992,
+      "learning_rate": 8.91436221394804e-06,
+      "loss": 0.5364,
+      "step": 13250
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 12.972611427307129,
+      "learning_rate": 8.913525498891353e-06,
+      "loss": 0.938,
+      "step": 13251
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 6.637202739715576,
+      "learning_rate": 8.912688783834666e-06,
+      "loss": 0.8613,
+      "step": 13252
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 13.172598838806152,
+      "learning_rate": 8.911852068777978e-06,
+      "loss": 0.9947,
+      "step": 13253
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 12.756169319152832,
+      "learning_rate": 8.91101535372129e-06,
+      "loss": 2.4711,
+      "step": 13254
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 14.738645553588867,
+      "learning_rate": 8.910178638664604e-06,
+      "loss": 1.4537,
+      "step": 13255
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 12.010141372680664,
+      "learning_rate": 8.909341923607916e-06,
+      "loss": 1.0155,
+      "step": 13256
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.001840591430664,
+      "learning_rate": 8.90850520855123e-06,
+      "loss": 0.4968,
+      "step": 13257
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 16.674413681030273,
+      "learning_rate": 8.907668493494541e-06,
+      "loss": 1.6073,
+      "step": 13258
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.00567626953125,
+      "learning_rate": 8.906831778437853e-06,
+      "loss": 0.444,
+      "step": 13259
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 36.17286682128906,
+      "learning_rate": 8.905995063381167e-06,
+      "loss": 0.7597,
+      "step": 13260
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.889520645141602,
+      "learning_rate": 8.905158348324479e-06,
+      "loss": 1.3893,
+      "step": 13261
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 22.4814395904541,
+      "learning_rate": 8.904321633267792e-06,
+      "loss": 1.5976,
+      "step": 13262
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 93.91947174072266,
+      "learning_rate": 8.903484918211104e-06,
+      "loss": 2.2526,
+      "step": 13263
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 4.852908134460449,
+      "learning_rate": 8.902648203154416e-06,
+      "loss": 0.1709,
+      "step": 13264
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 9.342089653015137,
+      "learning_rate": 8.901811488097728e-06,
+      "loss": 1.2207,
+      "step": 13265
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 33.74315643310547,
+      "learning_rate": 8.900974773041042e-06,
+      "loss": 0.7485,
+      "step": 13266
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 14.375835418701172,
+      "learning_rate": 8.900138057984354e-06,
+      "loss": 1.0302,
+      "step": 13267
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 14.565969467163086,
+      "learning_rate": 8.899301342927666e-06,
+      "loss": 1.0541,
+      "step": 13268
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 6.327230453491211,
+      "learning_rate": 8.89846462787098e-06,
+      "loss": 0.6158,
+      "step": 13269
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 6.902919769287109,
+      "learning_rate": 8.897627912814292e-06,
+      "loss": 1.111,
+      "step": 13270
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 16.355833053588867,
+      "learning_rate": 8.896791197757605e-06,
+      "loss": 1.2105,
+      "step": 13271
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 14.873172760009766,
+      "learning_rate": 8.895954482700917e-06,
+      "loss": 1.2151,
+      "step": 13272
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.8628973960876465,
+      "learning_rate": 8.895117767644229e-06,
+      "loss": 0.0907,
+      "step": 13273
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 15.422528266906738,
+      "learning_rate": 8.894281052587543e-06,
+      "loss": 0.5023,
+      "step": 13274
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 13.421253204345703,
+      "learning_rate": 8.893444337530855e-06,
+      "loss": 1.842,
+      "step": 13275
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 21.391128540039062,
+      "learning_rate": 8.892607622474168e-06,
+      "loss": 1.4473,
+      "step": 13276
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.939013481140137,
+      "learning_rate": 8.89177090741748e-06,
+      "loss": 1.1062,
+      "step": 13277
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 17.822078704833984,
+      "learning_rate": 8.890934192360792e-06,
+      "loss": 0.4375,
+      "step": 13278
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 3.838214874267578,
+      "learning_rate": 8.890097477304104e-06,
+      "loss": 0.1922,
+      "step": 13279
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 17.83454704284668,
+      "learning_rate": 8.889260762247416e-06,
+      "loss": 1.0488,
+      "step": 13280
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.977375984191895,
+      "learning_rate": 8.88842404719073e-06,
+      "loss": 1.3126,
+      "step": 13281
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.493778228759766,
+      "learning_rate": 8.887587332134042e-06,
+      "loss": 0.9259,
+      "step": 13282
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 7.79358434677124,
+      "learning_rate": 8.886750617077355e-06,
+      "loss": 0.4632,
+      "step": 13283
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 40.83755874633789,
+      "learning_rate": 8.885913902020667e-06,
+      "loss": 0.8441,
+      "step": 13284
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.328298091888428,
+      "learning_rate": 8.885077186963981e-06,
+      "loss": 0.238,
+      "step": 13285
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 27.0789794921875,
+      "learning_rate": 8.884240471907293e-06,
+      "loss": 1.1623,
+      "step": 13286
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.417692184448242,
+      "learning_rate": 8.883403756850605e-06,
+      "loss": 0.6112,
+      "step": 13287
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.841894149780273,
+      "learning_rate": 8.882567041793919e-06,
+      "loss": 0.9748,
+      "step": 13288
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 25.555082321166992,
+      "learning_rate": 8.88173032673723e-06,
+      "loss": 1.5632,
+      "step": 13289
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 13.408820152282715,
+      "learning_rate": 8.880893611680544e-06,
+      "loss": 0.4948,
+      "step": 13290
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 22.0626163482666,
+      "learning_rate": 8.880056896623856e-06,
+      "loss": 1.7045,
+      "step": 13291
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.598587036132812,
+      "learning_rate": 8.879220181567168e-06,
+      "loss": 0.7306,
+      "step": 13292
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 85.8493423461914,
+      "learning_rate": 8.87838346651048e-06,
+      "loss": 1.3975,
+      "step": 13293
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 13.468649864196777,
+      "learning_rate": 8.877546751453792e-06,
+      "loss": 1.0883,
+      "step": 13294
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 13.701459884643555,
+      "learning_rate": 8.876710036397106e-06,
+      "loss": 1.903,
+      "step": 13295
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.358095169067383,
+      "learning_rate": 8.875873321340418e-06,
+      "loss": 0.7231,
+      "step": 13296
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 6.389739513397217,
+      "learning_rate": 8.875036606283731e-06,
+      "loss": 0.8284,
+      "step": 13297
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.046916961669922,
+      "learning_rate": 8.874199891227043e-06,
+      "loss": 1.4727,
+      "step": 13298
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 25.28152084350586,
+      "learning_rate": 8.873363176170355e-06,
+      "loss": 1.1091,
+      "step": 13299
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 20.24308967590332,
+      "learning_rate": 8.872526461113669e-06,
+      "loss": 1.9117,
+      "step": 13300
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 15.938324928283691,
+      "learning_rate": 8.87168974605698e-06,
+      "loss": 1.1477,
+      "step": 13301
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 4.031920433044434,
+      "learning_rate": 8.870853031000294e-06,
+      "loss": 0.1123,
+      "step": 13302
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 31.852489471435547,
+      "learning_rate": 8.870016315943606e-06,
+      "loss": 0.9396,
+      "step": 13303
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 13.530977249145508,
+      "learning_rate": 8.869179600886918e-06,
+      "loss": 1.7093,
+      "step": 13304
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 16.1658878326416,
+      "learning_rate": 8.868342885830232e-06,
+      "loss": 2.0838,
+      "step": 13305
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.248319625854492,
+      "learning_rate": 8.867506170773544e-06,
+      "loss": 2.0007,
+      "step": 13306
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 7.573624610900879,
+      "learning_rate": 8.866669455716856e-06,
+      "loss": 1.0224,
+      "step": 13307
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 21.678936004638672,
+      "learning_rate": 8.865832740660168e-06,
+      "loss": 1.508,
+      "step": 13308
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 6.0260138511657715,
+      "learning_rate": 8.864996025603481e-06,
+      "loss": 0.3533,
+      "step": 13309
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 25.69033432006836,
+      "learning_rate": 8.864159310546793e-06,
+      "loss": 0.589,
+      "step": 13310
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 266.6009521484375,
+      "learning_rate": 8.863322595490107e-06,
+      "loss": 2.3583,
+      "step": 13311
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 23.762489318847656,
+      "learning_rate": 8.862485880433419e-06,
+      "loss": 1.7588,
+      "step": 13312
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.351102828979492,
+      "learning_rate": 8.861649165376731e-06,
+      "loss": 2.2779,
+      "step": 13313
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 23.822011947631836,
+      "learning_rate": 8.860812450320045e-06,
+      "loss": 1.6248,
+      "step": 13314
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 6.911247730255127,
+      "learning_rate": 8.859975735263357e-06,
+      "loss": 0.7206,
+      "step": 13315
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 15.00839614868164,
+      "learning_rate": 8.85913902020667e-06,
+      "loss": 2.2555,
+      "step": 13316
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 37.120361328125,
+      "learning_rate": 8.858302305149982e-06,
+      "loss": 1.3278,
+      "step": 13317
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 23.88784408569336,
+      "learning_rate": 8.857465590093294e-06,
+      "loss": 0.9319,
+      "step": 13318
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.098158836364746,
+      "learning_rate": 8.856628875036608e-06,
+      "loss": 0.6505,
+      "step": 13319
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 162.61207580566406,
+      "learning_rate": 8.85579215997992e-06,
+      "loss": 1.2954,
+      "step": 13320
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 30.69794464111328,
+      "learning_rate": 8.854955444923232e-06,
+      "loss": 2.927,
+      "step": 13321
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.624781608581543,
+      "learning_rate": 8.854118729866544e-06,
+      "loss": 1.0001,
+      "step": 13322
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.311716079711914,
+      "learning_rate": 8.853282014809857e-06,
+      "loss": 0.4252,
+      "step": 13323
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 39.431396484375,
+      "learning_rate": 8.85244529975317e-06,
+      "loss": 1.9132,
+      "step": 13324
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 6.1541218757629395,
+      "learning_rate": 8.851608584696483e-06,
+      "loss": 0.7265,
+      "step": 13325
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.475818634033203,
+      "learning_rate": 8.850771869639795e-06,
+      "loss": 0.4471,
+      "step": 13326
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.667397499084473,
+      "learning_rate": 8.849935154583107e-06,
+      "loss": 0.5298,
+      "step": 13327
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.55826473236084,
+      "learning_rate": 8.84909843952642e-06,
+      "loss": 0.8277,
+      "step": 13328
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 18.122146606445312,
+      "learning_rate": 8.848261724469732e-06,
+      "loss": 1.7542,
+      "step": 13329
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.492593765258789,
+      "learning_rate": 8.847425009413046e-06,
+      "loss": 0.6207,
+      "step": 13330
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 14.764479637145996,
+      "learning_rate": 8.846588294356358e-06,
+      "loss": 1.0586,
+      "step": 13331
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 52.10407638549805,
+      "learning_rate": 8.84575157929967e-06,
+      "loss": 2.8135,
+      "step": 13332
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 61.20310592651367,
+      "learning_rate": 8.844914864242982e-06,
+      "loss": 2.2693,
+      "step": 13333
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 22.537817001342773,
+      "learning_rate": 8.844078149186296e-06,
+      "loss": 1.4059,
+      "step": 13334
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 17.515825271606445,
+      "learning_rate": 8.843241434129608e-06,
+      "loss": 1.8921,
+      "step": 13335
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 108.60758972167969,
+      "learning_rate": 8.84240471907292e-06,
+      "loss": 1.2075,
+      "step": 13336
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 14.8504638671875,
+      "learning_rate": 8.841568004016233e-06,
+      "loss": 0.6504,
+      "step": 13337
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 15.910173416137695,
+      "learning_rate": 8.840731288959545e-06,
+      "loss": 0.8047,
+      "step": 13338
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.04807186126709,
+      "learning_rate": 8.839894573902859e-06,
+      "loss": 0.9502,
+      "step": 13339
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.169631958007812,
+      "learning_rate": 8.83905785884617e-06,
+      "loss": 1.7287,
+      "step": 13340
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.26987075805664,
+      "learning_rate": 8.838221143789483e-06,
+      "loss": 0.4565,
+      "step": 13341
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 11.048150062561035,
+      "learning_rate": 8.837384428732796e-06,
+      "loss": 1.1632,
+      "step": 13342
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 21.97183609008789,
+      "learning_rate": 8.836547713676108e-06,
+      "loss": 2.2033,
+      "step": 13343
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 39.257266998291016,
+      "learning_rate": 8.835710998619422e-06,
+      "loss": 0.8737,
+      "step": 13344
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 17.208297729492188,
+      "learning_rate": 8.834874283562734e-06,
+      "loss": 1.6556,
+      "step": 13345
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.4875246286392212,
+      "learning_rate": 8.834037568506046e-06,
+      "loss": 0.0885,
+      "step": 13346
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 22.558366775512695,
+      "learning_rate": 8.833200853449358e-06,
+      "loss": 0.3735,
+      "step": 13347
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.42983341217041,
+      "learning_rate": 8.832364138392671e-06,
+      "loss": 1.197,
+      "step": 13348
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 11.471229553222656,
+      "learning_rate": 8.831527423335983e-06,
+      "loss": 0.8928,
+      "step": 13349
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 67.52642822265625,
+      "learning_rate": 8.830690708279295e-06,
+      "loss": 1.4679,
+      "step": 13350
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 3.370436429977417,
+      "learning_rate": 8.829853993222609e-06,
+      "loss": 0.2173,
+      "step": 13351
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 56.54008102416992,
+      "learning_rate": 8.829017278165921e-06,
+      "loss": 2.1923,
+      "step": 13352
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 27.789016723632812,
+      "learning_rate": 8.828180563109235e-06,
+      "loss": 2.1442,
+      "step": 13353
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.23800277709961,
+      "learning_rate": 8.827343848052547e-06,
+      "loss": 0.4276,
+      "step": 13354
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 15.000799179077148,
+      "learning_rate": 8.826507132995858e-06,
+      "loss": 1.1826,
+      "step": 13355
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 6.0219831466674805,
+      "learning_rate": 8.825670417939172e-06,
+      "loss": 0.3008,
+      "step": 13356
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 13.526106834411621,
+      "learning_rate": 8.824833702882484e-06,
+      "loss": 1.5786,
+      "step": 13357
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 53.346534729003906,
+      "learning_rate": 8.823996987825798e-06,
+      "loss": 2.1324,
+      "step": 13358
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 51.98458480834961,
+      "learning_rate": 8.82316027276911e-06,
+      "loss": 1.66,
+      "step": 13359
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 11.333806037902832,
+      "learning_rate": 8.822323557712422e-06,
+      "loss": 0.8996,
+      "step": 13360
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 22.819717407226562,
+      "learning_rate": 8.821486842655734e-06,
+      "loss": 1.5081,
+      "step": 13361
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 4.25081729888916,
+      "learning_rate": 8.820650127599046e-06,
+      "loss": 0.0894,
+      "step": 13362
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 15.563796043395996,
+      "learning_rate": 8.81981341254236e-06,
+      "loss": 1.2254,
+      "step": 13363
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.962475776672363,
+      "learning_rate": 8.818976697485671e-06,
+      "loss": 1.3991,
+      "step": 13364
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 11.515885353088379,
+      "learning_rate": 8.818139982428985e-06,
+      "loss": 2.4045,
+      "step": 13365
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 15.969586372375488,
+      "learning_rate": 8.817303267372297e-06,
+      "loss": 3.9123,
+      "step": 13366
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 11.065643310546875,
+      "learning_rate": 8.81646655231561e-06,
+      "loss": 1.8376,
+      "step": 13367
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 25.345840454101562,
+      "learning_rate": 8.815629837258922e-06,
+      "loss": 2.6124,
+      "step": 13368
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 19.20560073852539,
+      "learning_rate": 8.814793122202234e-06,
+      "loss": 2.4694,
+      "step": 13369
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 17.70154571533203,
+      "learning_rate": 8.813956407145548e-06,
+      "loss": 0.9896,
+      "step": 13370
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.05247688293457,
+      "learning_rate": 8.81311969208886e-06,
+      "loss": 0.5865,
+      "step": 13371
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 18.543642044067383,
+      "learning_rate": 8.812282977032174e-06,
+      "loss": 1.2898,
+      "step": 13372
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 87.89289093017578,
+      "learning_rate": 8.811446261975486e-06,
+      "loss": 2.2731,
+      "step": 13373
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 11.179682731628418,
+      "learning_rate": 8.810609546918797e-06,
+      "loss": 1.1016,
+      "step": 13374
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 14.395187377929688,
+      "learning_rate": 8.80977283186211e-06,
+      "loss": 1.9152,
+      "step": 13375
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.498957633972168,
+      "learning_rate": 8.808936116805421e-06,
+      "loss": 0.8089,
+      "step": 13376
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 19.434778213500977,
+      "learning_rate": 8.808099401748735e-06,
+      "loss": 0.9951,
+      "step": 13377
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 4.149975299835205,
+      "learning_rate": 8.807262686692047e-06,
+      "loss": 0.4149,
+      "step": 13378
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 4.292070388793945,
+      "learning_rate": 8.80642597163536e-06,
+      "loss": 0.3905,
+      "step": 13379
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 16.486425399780273,
+      "learning_rate": 8.805589256578673e-06,
+      "loss": 1.3178,
+      "step": 13380
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 13.734532356262207,
+      "learning_rate": 8.804752541521986e-06,
+      "loss": 0.4845,
+      "step": 13381
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.013821601867676,
+      "learning_rate": 8.803915826465298e-06,
+      "loss": 1.778,
+      "step": 13382
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 100.57466125488281,
+      "learning_rate": 8.80307911140861e-06,
+      "loss": 1.6756,
+      "step": 13383
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.50030517578125,
+      "learning_rate": 8.802242396351924e-06,
+      "loss": 0.4496,
+      "step": 13384
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 7.649212837219238,
+      "learning_rate": 8.801405681295236e-06,
+      "loss": 0.3725,
+      "step": 13385
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 13.671998023986816,
+      "learning_rate": 8.800568966238548e-06,
+      "loss": 2.0125,
+      "step": 13386
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 16.365331649780273,
+      "learning_rate": 8.799732251181861e-06,
+      "loss": 0.6007,
+      "step": 13387
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 55.1285400390625,
+      "learning_rate": 8.798895536125173e-06,
+      "loss": 1.6335,
+      "step": 13388
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 7.412749767303467,
+      "learning_rate": 8.798058821068485e-06,
+      "loss": 0.5369,
+      "step": 13389
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 13.199882507324219,
+      "learning_rate": 8.797222106011797e-06,
+      "loss": 0.4003,
+      "step": 13390
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 20.589523315429688,
+      "learning_rate": 8.796385390955111e-06,
+      "loss": 1.498,
+      "step": 13391
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 19.743736267089844,
+      "learning_rate": 8.795548675898423e-06,
+      "loss": 1.5427,
+      "step": 13392
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 17.00905418395996,
+      "learning_rate": 8.794711960841736e-06,
+      "loss": 0.7849,
+      "step": 13393
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 11.008318901062012,
+      "learning_rate": 8.793875245785048e-06,
+      "loss": 1.1313,
+      "step": 13394
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 18.286884307861328,
+      "learning_rate": 8.793038530728362e-06,
+      "loss": 2.9482,
+      "step": 13395
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 29.307876586914062,
+      "learning_rate": 8.792201815671674e-06,
+      "loss": 3.9423,
+      "step": 13396
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 24.198068618774414,
+      "learning_rate": 8.791365100614986e-06,
+      "loss": 1.1835,
+      "step": 13397
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 20.103090286254883,
+      "learning_rate": 8.7905283855583e-06,
+      "loss": 0.616,
+      "step": 13398
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 11.391027450561523,
+      "learning_rate": 8.789691670501612e-06,
+      "loss": 0.8341,
+      "step": 13399
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.7701416015625,
+      "learning_rate": 8.788854955444924e-06,
+      "loss": 0.9425,
+      "step": 13400
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 6.627573013305664,
+      "learning_rate": 8.788018240388237e-06,
+      "loss": 0.8287,
+      "step": 13401
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.134284973144531,
+      "learning_rate": 8.787181525331549e-06,
+      "loss": 0.7226,
+      "step": 13402
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 3.0420546531677246,
+      "learning_rate": 8.786344810274861e-06,
+      "loss": 0.0561,
+      "step": 13403
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 26.259544372558594,
+      "learning_rate": 8.785508095218173e-06,
+      "loss": 2.305,
+      "step": 13404
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 13.82083511352539,
+      "learning_rate": 8.784671380161487e-06,
+      "loss": 0.935,
+      "step": 13405
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 30.167530059814453,
+      "learning_rate": 8.783834665104799e-06,
+      "loss": 2.9084,
+      "step": 13406
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 27.253812789916992,
+      "learning_rate": 8.782997950048112e-06,
+      "loss": 3.7398,
+      "step": 13407
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.45655632019043,
+      "learning_rate": 8.782161234991424e-06,
+      "loss": 1.2718,
+      "step": 13408
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 40.28466796875,
+      "learning_rate": 8.781324519934738e-06,
+      "loss": 1.8136,
+      "step": 13409
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 51.040706634521484,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 1.6632,
+      "step": 13410
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.13768482208252,
+      "learning_rate": 8.779651089821362e-06,
+      "loss": 0.2764,
+      "step": 13411
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 4.894547462463379,
+      "learning_rate": 8.778814374764675e-06,
+      "loss": 0.321,
+      "step": 13412
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 31.434677124023438,
+      "learning_rate": 8.777977659707987e-06,
+      "loss": 1.4748,
+      "step": 13413
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 6.621557712554932,
+      "learning_rate": 8.7771409446513e-06,
+      "loss": 1.5256,
+      "step": 13414
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 26.227516174316406,
+      "learning_rate": 8.776304229594611e-06,
+      "loss": 1.5478,
+      "step": 13415
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 6.326592922210693,
+      "learning_rate": 8.775467514537925e-06,
+      "loss": 0.4077,
+      "step": 13416
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 82.8076171875,
+      "learning_rate": 8.774630799481237e-06,
+      "loss": 2.9694,
+      "step": 13417
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.175357818603516,
+      "learning_rate": 8.773794084424549e-06,
+      "loss": 0.4495,
+      "step": 13418
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.854291915893555,
+      "learning_rate": 8.772957369367863e-06,
+      "loss": 1.138,
+      "step": 13419
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.000757217407227,
+      "learning_rate": 8.772120654311175e-06,
+      "loss": 0.3788,
+      "step": 13420
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 21.460311889648438,
+      "learning_rate": 8.771283939254488e-06,
+      "loss": 2.3919,
+      "step": 13421
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.737377166748047,
+      "learning_rate": 8.7704472241978e-06,
+      "loss": 0.9744,
+      "step": 13422
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 28.311187744140625,
+      "learning_rate": 8.769610509141114e-06,
+      "loss": 1.5108,
+      "step": 13423
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.060797691345215,
+      "learning_rate": 8.768773794084426e-06,
+      "loss": 1.4748,
+      "step": 13424
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 10.785904884338379,
+      "learning_rate": 8.767937079027738e-06,
+      "loss": 0.7211,
+      "step": 13425
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 13.229130744934082,
+      "learning_rate": 8.767100363971051e-06,
+      "loss": 0.6266,
+      "step": 13426
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 18.435352325439453,
+      "learning_rate": 8.766263648914363e-06,
+      "loss": 0.82,
+      "step": 13427
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 122.35250091552734,
+      "learning_rate": 8.765426933857675e-06,
+      "loss": 2.9313,
+      "step": 13428
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.277535438537598,
+      "learning_rate": 8.764590218800987e-06,
+      "loss": 1.6462,
+      "step": 13429
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 6.448642730712891,
+      "learning_rate": 8.7637535037443e-06,
+      "loss": 0.809,
+      "step": 13430
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 28.299129486083984,
+      "learning_rate": 8.762916788687613e-06,
+      "loss": 2.3807,
+      "step": 13431
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 4.615137577056885,
+      "learning_rate": 8.762080073630925e-06,
+      "loss": 1.7854,
+      "step": 13432
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 14.869714736938477,
+      "learning_rate": 8.761243358574238e-06,
+      "loss": 0.9222,
+      "step": 13433
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 5.068790435791016,
+      "learning_rate": 8.76040664351755e-06,
+      "loss": 0.463,
+      "step": 13434
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 244.9357452392578,
+      "learning_rate": 8.759569928460864e-06,
+      "loss": 2.3795,
+      "step": 13435
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 11.517192840576172,
+      "learning_rate": 8.758733213404176e-06,
+      "loss": 1.8589,
+      "step": 13436
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.460245370864868,
+      "learning_rate": 8.75789649834749e-06,
+      "loss": 0.1194,
+      "step": 13437
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 37.955230712890625,
+      "learning_rate": 8.757059783290802e-06,
+      "loss": 0.4742,
+      "step": 13438
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 14.026997566223145,
+      "learning_rate": 8.756223068234113e-06,
+      "loss": 0.9611,
+      "step": 13439
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 13.362473487854004,
+      "learning_rate": 8.755386353177427e-06,
+      "loss": 0.529,
+      "step": 13440
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 23.569236755371094,
+      "learning_rate": 8.754549638120739e-06,
+      "loss": 1.5198,
+      "step": 13441
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 11.897934913635254,
+      "learning_rate": 8.753712923064051e-06,
+      "loss": 0.881,
+      "step": 13442
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 12.302495002746582,
+      "learning_rate": 8.752876208007363e-06,
+      "loss": 0.9799,
+      "step": 13443
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 33.61111068725586,
+      "learning_rate": 8.752039492950677e-06,
+      "loss": 1.3375,
+      "step": 13444
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 29.497589111328125,
+      "learning_rate": 8.751202777893989e-06,
+      "loss": 2.4669,
+      "step": 13445
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 28.58656883239746,
+      "learning_rate": 8.7503660628373e-06,
+      "loss": 1.0672,
+      "step": 13446
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 4.034222602844238,
+      "learning_rate": 8.749529347780614e-06,
+      "loss": 0.1654,
+      "step": 13447
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.035055160522461,
+      "learning_rate": 8.748692632723926e-06,
+      "loss": 0.1728,
+      "step": 13448
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 12.092005729675293,
+      "learning_rate": 8.74785591766724e-06,
+      "loss": 1.5952,
+      "step": 13449
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 12.995216369628906,
+      "learning_rate": 8.747019202610552e-06,
+      "loss": 0.4874,
+      "step": 13450
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 11.660663604736328,
+      "learning_rate": 8.746182487553865e-06,
+      "loss": 0.5306,
+      "step": 13451
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 11.609655380249023,
+      "learning_rate": 8.745345772497177e-06,
+      "loss": 0.9101,
+      "step": 13452
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 11.562085151672363,
+      "learning_rate": 8.74450905744049e-06,
+      "loss": 2.1206,
+      "step": 13453
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 22.926427841186523,
+      "learning_rate": 8.743672342383801e-06,
+      "loss": 1.2818,
+      "step": 13454
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 19.309673309326172,
+      "learning_rate": 8.742835627327115e-06,
+      "loss": 0.8676,
+      "step": 13455
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 11.054777145385742,
+      "learning_rate": 8.741998912270427e-06,
+      "loss": 1.3197,
+      "step": 13456
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 4.592566967010498,
+      "learning_rate": 8.741162197213739e-06,
+      "loss": 0.1495,
+      "step": 13457
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.187602996826172,
+      "learning_rate": 8.740325482157052e-06,
+      "loss": 3.2393,
+      "step": 13458
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 14.268762588500977,
+      "learning_rate": 8.739488767100364e-06,
+      "loss": 0.5333,
+      "step": 13459
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 13.50385570526123,
+      "learning_rate": 8.738652052043676e-06,
+      "loss": 0.6521,
+      "step": 13460
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.107166290283203,
+      "learning_rate": 8.73781533698699e-06,
+      "loss": 0.3628,
+      "step": 13461
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 188.79522705078125,
+      "learning_rate": 8.736978621930302e-06,
+      "loss": 1.6856,
+      "step": 13462
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.642914772033691,
+      "learning_rate": 8.736141906873616e-06,
+      "loss": 1.4633,
+      "step": 13463
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 19.645877838134766,
+      "learning_rate": 8.735305191816928e-06,
+      "loss": 3.3901,
+      "step": 13464
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 50.0162467956543,
+      "learning_rate": 8.734468476760241e-06,
+      "loss": 1.942,
+      "step": 13465
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.19096851348877,
+      "learning_rate": 8.733631761703553e-06,
+      "loss": 0.5044,
+      "step": 13466
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 12.090299606323242,
+      "learning_rate": 8.732795046646865e-06,
+      "loss": 0.871,
+      "step": 13467
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 22.82764434814453,
+      "learning_rate": 8.731958331590177e-06,
+      "loss": 1.4258,
+      "step": 13468
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.270901679992676,
+      "learning_rate": 8.73112161653349e-06,
+      "loss": 0.4719,
+      "step": 13469
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 27.25560760498047,
+      "learning_rate": 8.730284901476803e-06,
+      "loss": 2.2294,
+      "step": 13470
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 11.683439254760742,
+      "learning_rate": 8.729448186420115e-06,
+      "loss": 1.1828,
+      "step": 13471
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 26.004322052001953,
+      "learning_rate": 8.728611471363427e-06,
+      "loss": 0.7483,
+      "step": 13472
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 24.552906036376953,
+      "learning_rate": 8.72777475630674e-06,
+      "loss": 1.509,
+      "step": 13473
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 4.466454029083252,
+      "learning_rate": 8.726938041250052e-06,
+      "loss": 0.2364,
+      "step": 13474
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 12.081656455993652,
+      "learning_rate": 8.726101326193366e-06,
+      "loss": 0.5443,
+      "step": 13475
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.333365440368652,
+      "learning_rate": 8.725264611136678e-06,
+      "loss": 0.5033,
+      "step": 13476
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.717599868774414,
+      "learning_rate": 8.724427896079991e-06,
+      "loss": 0.3139,
+      "step": 13477
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.948026657104492,
+      "learning_rate": 8.723591181023303e-06,
+      "loss": 0.8541,
+      "step": 13478
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.756032943725586,
+      "learning_rate": 8.722754465966615e-06,
+      "loss": 0.3445,
+      "step": 13479
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 6.668010234832764,
+      "learning_rate": 8.721917750909929e-06,
+      "loss": 1.1067,
+      "step": 13480
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 13.908268928527832,
+      "learning_rate": 8.721081035853241e-06,
+      "loss": 1.2556,
+      "step": 13481
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 12.909756660461426,
+      "learning_rate": 8.720244320796553e-06,
+      "loss": 0.3901,
+      "step": 13482
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.64794635772705,
+      "learning_rate": 8.719407605739867e-06,
+      "loss": 1.385,
+      "step": 13483
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 27.266633987426758,
+      "learning_rate": 8.718570890683179e-06,
+      "loss": 1.5704,
+      "step": 13484
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.373408794403076,
+      "learning_rate": 8.71773417562649e-06,
+      "loss": 0.3352,
+      "step": 13485
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.586067199707031,
+      "learning_rate": 8.716897460569802e-06,
+      "loss": 1.0504,
+      "step": 13486
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 19.755983352661133,
+      "learning_rate": 8.716060745513116e-06,
+      "loss": 0.7271,
+      "step": 13487
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 13.576322555541992,
+      "learning_rate": 8.715224030456428e-06,
+      "loss": 0.3314,
+      "step": 13488
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 14.759033203125,
+      "learning_rate": 8.714387315399742e-06,
+      "loss": 0.9154,
+      "step": 13489
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 46.80823516845703,
+      "learning_rate": 8.713550600343054e-06,
+      "loss": 3.4295,
+      "step": 13490
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 14.86972713470459,
+      "learning_rate": 8.712713885286367e-06,
+      "loss": 1.122,
+      "step": 13491
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 21.130163192749023,
+      "learning_rate": 8.71187717022968e-06,
+      "loss": 1.1324,
+      "step": 13492
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 20.94374656677246,
+      "learning_rate": 8.711040455172991e-06,
+      "loss": 1.0438,
+      "step": 13493
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 22.614845275878906,
+      "learning_rate": 8.710203740116305e-06,
+      "loss": 1.3184,
+      "step": 13494
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 20.864540100097656,
+      "learning_rate": 8.709367025059617e-06,
+      "loss": 0.6603,
+      "step": 13495
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 10.283666610717773,
+      "learning_rate": 8.708530310002929e-06,
+      "loss": 0.8978,
+      "step": 13496
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 121.8965072631836,
+      "learning_rate": 8.70769359494624e-06,
+      "loss": 1.1968,
+      "step": 13497
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 26.533832550048828,
+      "learning_rate": 8.706856879889554e-06,
+      "loss": 1.3313,
+      "step": 13498
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 92.5091552734375,
+      "learning_rate": 8.706020164832866e-06,
+      "loss": 1.9148,
+      "step": 13499
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 16.19998550415039,
+      "learning_rate": 8.705183449776178e-06,
+      "loss": 1.619,
+      "step": 13500
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 15.971726417541504,
+      "learning_rate": 8.704346734719492e-06,
+      "loss": 1.1749,
+      "step": 13501
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 19.345623016357422,
+      "learning_rate": 8.703510019662804e-06,
+      "loss": 1.0511,
+      "step": 13502
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 6.182967662811279,
+      "learning_rate": 8.702673304606118e-06,
+      "loss": 0.93,
+      "step": 13503
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.37590503692627,
+      "learning_rate": 8.70183658954943e-06,
+      "loss": 0.3127,
+      "step": 13504
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.4262542724609375,
+      "learning_rate": 8.700999874492743e-06,
+      "loss": 0.2227,
+      "step": 13505
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.645088195800781,
+      "learning_rate": 8.700163159436055e-06,
+      "loss": 0.5829,
+      "step": 13506
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 42.65239715576172,
+      "learning_rate": 8.699326444379367e-06,
+      "loss": 2.4669,
+      "step": 13507
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.191194534301758,
+      "learning_rate": 8.69848972932268e-06,
+      "loss": 2.1358,
+      "step": 13508
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 24.681190490722656,
+      "learning_rate": 8.697653014265993e-06,
+      "loss": 1.0717,
+      "step": 13509
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.34162712097168,
+      "learning_rate": 8.696816299209305e-06,
+      "loss": 0.6754,
+      "step": 13510
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 31.257905960083008,
+      "learning_rate": 8.695979584152617e-06,
+      "loss": 2.1831,
+      "step": 13511
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 41.80419921875,
+      "learning_rate": 8.69514286909593e-06,
+      "loss": 1.0436,
+      "step": 13512
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 17.46442985534668,
+      "learning_rate": 8.694306154039242e-06,
+      "loss": 1.6393,
+      "step": 13513
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 20.784839630126953,
+      "learning_rate": 8.693469438982554e-06,
+      "loss": 1.292,
+      "step": 13514
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 18.792034149169922,
+      "learning_rate": 8.692632723925868e-06,
+      "loss": 1.3554,
+      "step": 13515
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.121496200561523,
+      "learning_rate": 8.69179600886918e-06,
+      "loss": 0.3784,
+      "step": 13516
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 48.75419616699219,
+      "learning_rate": 8.690959293812493e-06,
+      "loss": 1.5,
+      "step": 13517
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 13.551359176635742,
+      "learning_rate": 8.690122578755805e-06,
+      "loss": 0.667,
+      "step": 13518
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 6.288450717926025,
+      "learning_rate": 8.689285863699119e-06,
+      "loss": 0.9067,
+      "step": 13519
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.752936363220215,
+      "learning_rate": 8.688449148642431e-06,
+      "loss": 0.7502,
+      "step": 13520
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.777066707611084,
+      "learning_rate": 8.687612433585743e-06,
+      "loss": 0.2846,
+      "step": 13521
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 42.891937255859375,
+      "learning_rate": 8.686775718529057e-06,
+      "loss": 3.0157,
+      "step": 13522
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 12.691798210144043,
+      "learning_rate": 8.685939003472369e-06,
+      "loss": 1.5813,
+      "step": 13523
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 27.807815551757812,
+      "learning_rate": 8.68510228841568e-06,
+      "loss": 1.3952,
+      "step": 13524
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 13.172384262084961,
+      "learning_rate": 8.684265573358992e-06,
+      "loss": 0.8702,
+      "step": 13525
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 6.7595343589782715,
+      "learning_rate": 8.683428858302306e-06,
+      "loss": 0.22,
+      "step": 13526
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.891069412231445,
+      "learning_rate": 8.682592143245618e-06,
+      "loss": 1.4104,
+      "step": 13527
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 14.212307929992676,
+      "learning_rate": 8.68175542818893e-06,
+      "loss": 1.5403,
+      "step": 13528
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 16.828903198242188,
+      "learning_rate": 8.680918713132244e-06,
+      "loss": 0.9557,
+      "step": 13529
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 49.076622009277344,
+      "learning_rate": 8.680081998075556e-06,
+      "loss": 0.8234,
+      "step": 13530
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 15.557707786560059,
+      "learning_rate": 8.67924528301887e-06,
+      "loss": 1.0391,
+      "step": 13531
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.3461332321167,
+      "learning_rate": 8.678408567962181e-06,
+      "loss": 0.6084,
+      "step": 13532
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.129873275756836,
+      "learning_rate": 8.677571852905495e-06,
+      "loss": 1.2939,
+      "step": 13533
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 32.63777160644531,
+      "learning_rate": 8.676735137848807e-06,
+      "loss": 1.4533,
+      "step": 13534
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.426362991333008,
+      "learning_rate": 8.675898422792119e-06,
+      "loss": 0.9518,
+      "step": 13535
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 19.661296844482422,
+      "learning_rate": 8.67506170773543e-06,
+      "loss": 2.0506,
+      "step": 13536
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 23.262487411499023,
+      "learning_rate": 8.674224992678744e-06,
+      "loss": 1.1738,
+      "step": 13537
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.202347755432129,
+      "learning_rate": 8.673388277622056e-06,
+      "loss": 1.0119,
+      "step": 13538
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 24.753644943237305,
+      "learning_rate": 8.672551562565368e-06,
+      "loss": 1.5677,
+      "step": 13539
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 5.492506980895996,
+      "learning_rate": 8.671714847508682e-06,
+      "loss": 0.655,
+      "step": 13540
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 21.3233585357666,
+      "learning_rate": 8.670878132451994e-06,
+      "loss": 1.0742,
+      "step": 13541
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.688346862792969,
+      "learning_rate": 8.670041417395306e-06,
+      "loss": 0.7168,
+      "step": 13542
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 5.929947853088379,
+      "learning_rate": 8.66920470233862e-06,
+      "loss": 0.6347,
+      "step": 13543
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.265826225280762,
+      "learning_rate": 8.668367987281931e-06,
+      "loss": 1.0965,
+      "step": 13544
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.524782180786133,
+      "learning_rate": 8.667531272225245e-06,
+      "loss": 1.0063,
+      "step": 13545
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.861703872680664,
+      "learning_rate": 8.666694557168557e-06,
+      "loss": 1.071,
+      "step": 13546
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 21.725366592407227,
+      "learning_rate": 8.66585784211187e-06,
+      "loss": 0.9319,
+      "step": 13547
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 20.31465721130371,
+      "learning_rate": 8.665021127055183e-06,
+      "loss": 0.9347,
+      "step": 13548
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 18.960803985595703,
+      "learning_rate": 8.664184411998495e-06,
+      "loss": 1.7244,
+      "step": 13549
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.230573654174805,
+      "learning_rate": 8.663347696941807e-06,
+      "loss": 0.8174,
+      "step": 13550
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.02849006652832,
+      "learning_rate": 8.66251098188512e-06,
+      "loss": 1.5155,
+      "step": 13551
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 18.773893356323242,
+      "learning_rate": 8.661674266828432e-06,
+      "loss": 1.5971,
+      "step": 13552
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 26.453706741333008,
+      "learning_rate": 8.660837551771744e-06,
+      "loss": 1.0111,
+      "step": 13553
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 15.818222999572754,
+      "learning_rate": 8.660000836715058e-06,
+      "loss": 1.148,
+      "step": 13554
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 12.550549507141113,
+      "learning_rate": 8.65916412165837e-06,
+      "loss": 0.7964,
+      "step": 13555
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.0923314094543457,
+      "learning_rate": 8.658327406601682e-06,
+      "loss": 0.1456,
+      "step": 13556
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.524921417236328,
+      "learning_rate": 8.657490691544995e-06,
+      "loss": 1.7639,
+      "step": 13557
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.678040504455566,
+      "learning_rate": 8.656653976488307e-06,
+      "loss": 1.8976,
+      "step": 13558
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 13.557211875915527,
+      "learning_rate": 8.655817261431621e-06,
+      "loss": 0.3274,
+      "step": 13559
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 14.539973258972168,
+      "learning_rate": 8.654980546374933e-06,
+      "loss": 0.8182,
+      "step": 13560
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 228.64242553710938,
+      "learning_rate": 8.654143831318247e-06,
+      "loss": 2.2122,
+      "step": 13561
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 16.100831985473633,
+      "learning_rate": 8.653307116261558e-06,
+      "loss": 1.5305,
+      "step": 13562
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 13.376947402954102,
+      "learning_rate": 8.65247040120487e-06,
+      "loss": 0.4518,
+      "step": 13563
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.532651901245117,
+      "learning_rate": 8.651633686148182e-06,
+      "loss": 0.4058,
+      "step": 13564
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 10.986957550048828,
+      "learning_rate": 8.650796971091496e-06,
+      "loss": 0.8033,
+      "step": 13565
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 19.906335830688477,
+      "learning_rate": 8.649960256034808e-06,
+      "loss": 1.8828,
+      "step": 13566
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.025758743286133,
+      "learning_rate": 8.64912354097812e-06,
+      "loss": 1.6456,
+      "step": 13567
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 39.448062896728516,
+      "learning_rate": 8.648286825921434e-06,
+      "loss": 1.6568,
+      "step": 13568
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 12.004514694213867,
+      "learning_rate": 8.647450110864746e-06,
+      "loss": 0.5564,
+      "step": 13569
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 39.72369384765625,
+      "learning_rate": 8.646613395808058e-06,
+      "loss": 0.8057,
+      "step": 13570
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 6.89996862411499,
+      "learning_rate": 8.645776680751371e-06,
+      "loss": 0.4684,
+      "step": 13571
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 13.687106132507324,
+      "learning_rate": 8.644939965694683e-06,
+      "loss": 0.7281,
+      "step": 13572
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 17.127885818481445,
+      "learning_rate": 8.644103250637997e-06,
+      "loss": 0.953,
+      "step": 13573
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.036026954650879,
+      "learning_rate": 8.643266535581309e-06,
+      "loss": 1.4963,
+      "step": 13574
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.478490829467773,
+      "learning_rate": 8.642429820524622e-06,
+      "loss": 1.2502,
+      "step": 13575
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 14.2973051071167,
+      "learning_rate": 8.641593105467934e-06,
+      "loss": 1.4743,
+      "step": 13576
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 19.452285766601562,
+      "learning_rate": 8.640756390411246e-06,
+      "loss": 1.4742,
+      "step": 13577
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 43.88484573364258,
+      "learning_rate": 8.639919675354558e-06,
+      "loss": 1.2541,
+      "step": 13578
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 44.801395416259766,
+      "learning_rate": 8.63908296029787e-06,
+      "loss": 1.1219,
+      "step": 13579
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 11.87891674041748,
+      "learning_rate": 8.638246245241184e-06,
+      "loss": 0.8183,
+      "step": 13580
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.8933000564575195,
+      "learning_rate": 8.637409530184496e-06,
+      "loss": 0.329,
+      "step": 13581
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 17.78415870666504,
+      "learning_rate": 8.63657281512781e-06,
+      "loss": 0.8824,
+      "step": 13582
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 24.023479461669922,
+      "learning_rate": 8.635736100071121e-06,
+      "loss": 0.4031,
+      "step": 13583
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 8.85848617553711,
+      "learning_rate": 8.634899385014433e-06,
+      "loss": 0.1393,
+      "step": 13584
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 19.534149169921875,
+      "learning_rate": 8.634062669957747e-06,
+      "loss": 2.2703,
+      "step": 13585
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 53.858760833740234,
+      "learning_rate": 8.633225954901059e-06,
+      "loss": 2.1575,
+      "step": 13586
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 15.325075149536133,
+      "learning_rate": 8.632389239844373e-06,
+      "loss": 2.039,
+      "step": 13587
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 12.996599197387695,
+      "learning_rate": 8.631552524787685e-06,
+      "loss": 0.7443,
+      "step": 13588
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 16.511795043945312,
+      "learning_rate": 8.630715809730996e-06,
+      "loss": 1.399,
+      "step": 13589
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 21.71929168701172,
+      "learning_rate": 8.62987909467431e-06,
+      "loss": 0.854,
+      "step": 13590
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 11.38278865814209,
+      "learning_rate": 8.629042379617622e-06,
+      "loss": 1.0666,
+      "step": 13591
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 29.05997085571289,
+      "learning_rate": 8.628205664560934e-06,
+      "loss": 1.1241,
+      "step": 13592
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 16.693775177001953,
+      "learning_rate": 8.627368949504246e-06,
+      "loss": 0.8409,
+      "step": 13593
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.487668991088867,
+      "learning_rate": 8.62653223444756e-06,
+      "loss": 0.4452,
+      "step": 13594
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 11.088615417480469,
+      "learning_rate": 8.625695519390872e-06,
+      "loss": 1.0518,
+      "step": 13595
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.214578628540039,
+      "learning_rate": 8.624858804334185e-06,
+      "loss": 1.0723,
+      "step": 13596
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 23.00535774230957,
+      "learning_rate": 8.624022089277497e-06,
+      "loss": 2.3293,
+      "step": 13597
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 16.56564712524414,
+      "learning_rate": 8.62318537422081e-06,
+      "loss": 1.2909,
+      "step": 13598
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 23.9890079498291,
+      "learning_rate": 8.622348659164123e-06,
+      "loss": 1.214,
+      "step": 13599
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 17.81267738342285,
+      "learning_rate": 8.621511944107435e-06,
+      "loss": 1.1028,
+      "step": 13600
+    },
+    {
+      "epoch": 1.71,
+      "eval_loss": 0.08200293779373169,
+      "eval_runtime": 94.6835,
+      "eval_samples_per_second": 37.409,
+      "eval_steps_per_second": 37.409,
+      "step": 13600
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 15.171854019165039,
+      "learning_rate": 8.620675229050748e-06,
+      "loss": 1.6704,
+      "step": 13601
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 13.904727935791016,
+      "learning_rate": 8.61983851399406e-06,
+      "loss": 0.8487,
+      "step": 13602
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 22.334592819213867,
+      "learning_rate": 8.619001798937372e-06,
+      "loss": 2.0529,
+      "step": 13603
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 17.451332092285156,
+      "learning_rate": 8.618165083880686e-06,
+      "loss": 1.4637,
+      "step": 13604
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 18.929479598999023,
+      "learning_rate": 8.617328368823998e-06,
+      "loss": 1.593,
+      "step": 13605
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 46.291099548339844,
+      "learning_rate": 8.61649165376731e-06,
+      "loss": 1.761,
+      "step": 13606
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 11.363897323608398,
+      "learning_rate": 8.615654938710622e-06,
+      "loss": 1.5096,
+      "step": 13607
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 230.1785888671875,
+      "learning_rate": 8.614818223653935e-06,
+      "loss": 1.7891,
+      "step": 13608
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 7.91306734085083,
+      "learning_rate": 8.613981508597247e-06,
+      "loss": 0.4705,
+      "step": 13609
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 14.959234237670898,
+      "learning_rate": 8.613144793540561e-06,
+      "loss": 1.323,
+      "step": 13610
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 7.350861072540283,
+      "learning_rate": 8.612308078483873e-06,
+      "loss": 1.33,
+      "step": 13611
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 5.810110569000244,
+      "learning_rate": 8.611471363427185e-06,
+      "loss": 0.614,
+      "step": 13612
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 5.175986289978027,
+      "learning_rate": 8.610634648370499e-06,
+      "loss": 0.7721,
+      "step": 13613
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 15.35120964050293,
+      "learning_rate": 8.60979793331381e-06,
+      "loss": 2.745,
+      "step": 13614
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 32.70784378051758,
+      "learning_rate": 8.608961218257124e-06,
+      "loss": 1.5796,
+      "step": 13615
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.75576114654541,
+      "learning_rate": 8.608124503200436e-06,
+      "loss": 0.3475,
+      "step": 13616
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 111.13134002685547,
+      "learning_rate": 8.607287788143748e-06,
+      "loss": 2.3101,
+      "step": 13617
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.689243316650391,
+      "learning_rate": 8.60645107308706e-06,
+      "loss": 0.427,
+      "step": 13618
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.947023868560791,
+      "learning_rate": 8.605614358030374e-06,
+      "loss": 0.0379,
+      "step": 13619
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 23.508211135864258,
+      "learning_rate": 8.604777642973686e-06,
+      "loss": 0.5596,
+      "step": 13620
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 22.981761932373047,
+      "learning_rate": 8.603940927916998e-06,
+      "loss": 1.0385,
+      "step": 13621
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.272953033447266,
+      "learning_rate": 8.603104212860311e-06,
+      "loss": 0.659,
+      "step": 13622
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 35.652923583984375,
+      "learning_rate": 8.602267497803623e-06,
+      "loss": 1.2104,
+      "step": 13623
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 12.808666229248047,
+      "learning_rate": 8.601430782746937e-06,
+      "loss": 1.5748,
+      "step": 13624
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 24.261981964111328,
+      "learning_rate": 8.600594067690249e-06,
+      "loss": 2.4687,
+      "step": 13625
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 18.71622085571289,
+      "learning_rate": 8.599757352633561e-06,
+      "loss": 1.5981,
+      "step": 13626
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.795702934265137,
+      "learning_rate": 8.598920637576874e-06,
+      "loss": 1.047,
+      "step": 13627
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 7.158591270446777,
+      "learning_rate": 8.598083922520186e-06,
+      "loss": 0.5137,
+      "step": 13628
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 22.881824493408203,
+      "learning_rate": 8.5972472074635e-06,
+      "loss": 0.3531,
+      "step": 13629
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.794554710388184,
+      "learning_rate": 8.596410492406812e-06,
+      "loss": 0.9378,
+      "step": 13630
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 24.48370933532715,
+      "learning_rate": 8.595573777350124e-06,
+      "loss": 1.6802,
+      "step": 13631
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.828770637512207,
+      "learning_rate": 8.594737062293436e-06,
+      "loss": 0.6779,
+      "step": 13632
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 28.461143493652344,
+      "learning_rate": 8.59390034723675e-06,
+      "loss": 1.9469,
+      "step": 13633
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 52.104888916015625,
+      "learning_rate": 8.593063632180062e-06,
+      "loss": 3.3121,
+      "step": 13634
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.891456604003906,
+      "learning_rate": 8.592226917123374e-06,
+      "loss": 0.7873,
+      "step": 13635
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 16.9071044921875,
+      "learning_rate": 8.591390202066687e-06,
+      "loss": 1.2607,
+      "step": 13636
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.1126909255981445,
+      "learning_rate": 8.590553487009999e-06,
+      "loss": 0.8182,
+      "step": 13637
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.156928062438965,
+      "learning_rate": 8.589716771953313e-06,
+      "loss": 0.7253,
+      "step": 13638
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 48.94029998779297,
+      "learning_rate": 8.588880056896625e-06,
+      "loss": 2.9874,
+      "step": 13639
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 12.007301330566406,
+      "learning_rate": 8.588043341839937e-06,
+      "loss": 0.8996,
+      "step": 13640
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 19.267536163330078,
+      "learning_rate": 8.58720662678325e-06,
+      "loss": 0.8171,
+      "step": 13641
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 16.437095642089844,
+      "learning_rate": 8.586369911726562e-06,
+      "loss": 1.6608,
+      "step": 13642
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 7.273100852966309,
+      "learning_rate": 8.585533196669876e-06,
+      "loss": 0.2429,
+      "step": 13643
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 20.524356842041016,
+      "learning_rate": 8.584696481613188e-06,
+      "loss": 1.2124,
+      "step": 13644
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 12.111412048339844,
+      "learning_rate": 8.5838597665565e-06,
+      "loss": 1.4774,
+      "step": 13645
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.179340362548828,
+      "learning_rate": 8.583023051499812e-06,
+      "loss": 0.4397,
+      "step": 13646
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.476637840270996,
+      "learning_rate": 8.582186336443124e-06,
+      "loss": 0.6604,
+      "step": 13647
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 34.13114929199219,
+      "learning_rate": 8.581349621386437e-06,
+      "loss": 2.0151,
+      "step": 13648
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.2700515985488892,
+      "learning_rate": 8.58051290632975e-06,
+      "loss": 0.0291,
+      "step": 13649
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.670015335083008,
+      "learning_rate": 8.579676191273063e-06,
+      "loss": 0.7254,
+      "step": 13650
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 58.138404846191406,
+      "learning_rate": 8.578839476216375e-06,
+      "loss": 1.9637,
+      "step": 13651
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 21.40508460998535,
+      "learning_rate": 8.578002761159687e-06,
+      "loss": 1.3869,
+      "step": 13652
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 11.227001190185547,
+      "learning_rate": 8.577166046103e-06,
+      "loss": 0.9253,
+      "step": 13653
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 14.857300758361816,
+      "learning_rate": 8.576329331046313e-06,
+      "loss": 1.2623,
+      "step": 13654
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 17.581310272216797,
+      "learning_rate": 8.575492615989626e-06,
+      "loss": 0.6921,
+      "step": 13655
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 15.472161293029785,
+      "learning_rate": 8.574655900932938e-06,
+      "loss": 0.8102,
+      "step": 13656
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 101.40353393554688,
+      "learning_rate": 8.573819185876252e-06,
+      "loss": 2.4513,
+      "step": 13657
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.423720359802246,
+      "learning_rate": 8.572982470819564e-06,
+      "loss": 0.9824,
+      "step": 13658
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.732719421386719,
+      "learning_rate": 8.572145755762876e-06,
+      "loss": 0.2346,
+      "step": 13659
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.475010395050049,
+      "learning_rate": 8.571309040706188e-06,
+      "loss": 0.4437,
+      "step": 13660
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 7.847319602966309,
+      "learning_rate": 8.5704723256495e-06,
+      "loss": 0.4661,
+      "step": 13661
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 29.919858932495117,
+      "learning_rate": 8.569635610592813e-06,
+      "loss": 0.7812,
+      "step": 13662
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.448214054107666,
+      "learning_rate": 8.568798895536125e-06,
+      "loss": 0.2187,
+      "step": 13663
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 21.727697372436523,
+      "learning_rate": 8.567962180479439e-06,
+      "loss": 1.323,
+      "step": 13664
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 10.879844665527344,
+      "learning_rate": 8.56712546542275e-06,
+      "loss": 0.3115,
+      "step": 13665
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 17.375152587890625,
+      "learning_rate": 8.566288750366063e-06,
+      "loss": 0.339,
+      "step": 13666
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 37.212764739990234,
+      "learning_rate": 8.565452035309376e-06,
+      "loss": 1.396,
+      "step": 13667
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 73.83079528808594,
+      "learning_rate": 8.564615320252688e-06,
+      "loss": 1.2866,
+      "step": 13668
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.035943031311035,
+      "learning_rate": 8.563778605196002e-06,
+      "loss": 1.3252,
+      "step": 13669
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 11.69262409210205,
+      "learning_rate": 8.562941890139314e-06,
+      "loss": 1.0561,
+      "step": 13670
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 18.235258102416992,
+      "learning_rate": 8.562105175082626e-06,
+      "loss": 0.4454,
+      "step": 13671
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.037936210632324,
+      "learning_rate": 8.56126846002594e-06,
+      "loss": 0.4659,
+      "step": 13672
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 22.74379539489746,
+      "learning_rate": 8.560431744969252e-06,
+      "loss": 2.4574,
+      "step": 13673
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 13.192018508911133,
+      "learning_rate": 8.559595029912563e-06,
+      "loss": 1.5641,
+      "step": 13674
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 13.780868530273438,
+      "learning_rate": 8.558758314855875e-06,
+      "loss": 1.4909,
+      "step": 13675
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 71.01911926269531,
+      "learning_rate": 8.557921599799189e-06,
+      "loss": 1.5158,
+      "step": 13676
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.662192344665527,
+      "learning_rate": 8.557084884742501e-06,
+      "loss": 0.4845,
+      "step": 13677
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 29.933422088623047,
+      "learning_rate": 8.556248169685815e-06,
+      "loss": 1.5913,
+      "step": 13678
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.449567794799805,
+      "learning_rate": 8.555411454629127e-06,
+      "loss": 0.3034,
+      "step": 13679
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 13.019527435302734,
+      "learning_rate": 8.554574739572439e-06,
+      "loss": 0.5575,
+      "step": 13680
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 12.01841926574707,
+      "learning_rate": 8.553738024515752e-06,
+      "loss": 1.6908,
+      "step": 13681
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.98511028289795,
+      "learning_rate": 8.552901309459064e-06,
+      "loss": 0.4788,
+      "step": 13682
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.490072250366211,
+      "learning_rate": 8.552064594402378e-06,
+      "loss": 0.3471,
+      "step": 13683
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 11.30111312866211,
+      "learning_rate": 8.55122787934569e-06,
+      "loss": 0.3661,
+      "step": 13684
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 14.25614070892334,
+      "learning_rate": 8.550391164289002e-06,
+      "loss": 0.8016,
+      "step": 13685
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.77916955947876,
+      "learning_rate": 8.549554449232315e-06,
+      "loss": 1.9925,
+      "step": 13686
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 30.912118911743164,
+      "learning_rate": 8.548717734175627e-06,
+      "loss": 1.7488,
+      "step": 13687
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 15.012194633483887,
+      "learning_rate": 8.54788101911894e-06,
+      "loss": 0.7428,
+      "step": 13688
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 20.931856155395508,
+      "learning_rate": 8.547044304062251e-06,
+      "loss": 1.9809,
+      "step": 13689
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 20.445232391357422,
+      "learning_rate": 8.546207589005565e-06,
+      "loss": 0.9766,
+      "step": 13690
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.759213447570801,
+      "learning_rate": 8.545370873948877e-06,
+      "loss": 1.1829,
+      "step": 13691
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 13.30997371673584,
+      "learning_rate": 8.54453415889219e-06,
+      "loss": 0.5945,
+      "step": 13692
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 14.735377311706543,
+      "learning_rate": 8.543697443835502e-06,
+      "loss": 0.8842,
+      "step": 13693
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 18.377967834472656,
+      "learning_rate": 8.542860728778814e-06,
+      "loss": 1.3858,
+      "step": 13694
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.478085041046143,
+      "learning_rate": 8.542024013722128e-06,
+      "loss": 1.8176,
+      "step": 13695
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.661737442016602,
+      "learning_rate": 8.54118729866544e-06,
+      "loss": 1.0031,
+      "step": 13696
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 31.49393081665039,
+      "learning_rate": 8.540350583608754e-06,
+      "loss": 2.1003,
+      "step": 13697
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.8699774742126465,
+      "learning_rate": 8.539513868552066e-06,
+      "loss": 0.6333,
+      "step": 13698
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 23.345783233642578,
+      "learning_rate": 8.538677153495378e-06,
+      "loss": 1.6522,
+      "step": 13699
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 18.031431198120117,
+      "learning_rate": 8.53784043843869e-06,
+      "loss": 1.928,
+      "step": 13700
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 11.421024322509766,
+      "learning_rate": 8.537003723382003e-06,
+      "loss": 1.0393,
+      "step": 13701
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 12.769109725952148,
+      "learning_rate": 8.536167008325315e-06,
+      "loss": 2.8767,
+      "step": 13702
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 19.535921096801758,
+      "learning_rate": 8.535330293268627e-06,
+      "loss": 1.5462,
+      "step": 13703
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.937314987182617,
+      "learning_rate": 8.53449357821194e-06,
+      "loss": 0.1772,
+      "step": 13704
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 14.477164268493652,
+      "learning_rate": 8.533656863155253e-06,
+      "loss": 1.0191,
+      "step": 13705
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 6.0834479331970215,
+      "learning_rate": 8.532820148098566e-06,
+      "loss": 0.2097,
+      "step": 13706
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 41.72673416137695,
+      "learning_rate": 8.531983433041878e-06,
+      "loss": 2.2051,
+      "step": 13707
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 11.719079971313477,
+      "learning_rate": 8.53114671798519e-06,
+      "loss": 0.4026,
+      "step": 13708
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.154730796813965,
+      "learning_rate": 8.530310002928504e-06,
+      "loss": 1.6962,
+      "step": 13709
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.870038032531738,
+      "learning_rate": 8.529473287871816e-06,
+      "loss": 0.6991,
+      "step": 13710
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 28.99807357788086,
+      "learning_rate": 8.52863657281513e-06,
+      "loss": 0.6092,
+      "step": 13711
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 7.3584370613098145,
+      "learning_rate": 8.527799857758441e-06,
+      "loss": 0.3729,
+      "step": 13712
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.6323156356811523,
+      "learning_rate": 8.526963142701753e-06,
+      "loss": 0.4234,
+      "step": 13713
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 13.42520523071289,
+      "learning_rate": 8.526126427645065e-06,
+      "loss": 2.027,
+      "step": 13714
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 6.929305553436279,
+      "learning_rate": 8.525289712588379e-06,
+      "loss": 1.1329,
+      "step": 13715
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 27.86838150024414,
+      "learning_rate": 8.524452997531691e-06,
+      "loss": 1.128,
+      "step": 13716
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 19.303163528442383,
+      "learning_rate": 8.523616282475003e-06,
+      "loss": 1.2456,
+      "step": 13717
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 17.35099220275879,
+      "learning_rate": 8.522779567418317e-06,
+      "loss": 0.7379,
+      "step": 13718
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 17.20098304748535,
+      "learning_rate": 8.521942852361629e-06,
+      "loss": 2.1604,
+      "step": 13719
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 5.011656761169434,
+      "learning_rate": 8.521106137304942e-06,
+      "loss": 0.6209,
+      "step": 13720
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.144037246704102,
+      "learning_rate": 8.520269422248254e-06,
+      "loss": 0.5365,
+      "step": 13721
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 21.16668701171875,
+      "learning_rate": 8.519432707191566e-06,
+      "loss": 1.2462,
+      "step": 13722
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.0023086071014404,
+      "learning_rate": 8.51859599213488e-06,
+      "loss": 1.3551,
+      "step": 13723
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.263258934020996,
+      "learning_rate": 8.517759277078192e-06,
+      "loss": 1.2265,
+      "step": 13724
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 23.832761764526367,
+      "learning_rate": 8.516922562021505e-06,
+      "loss": 0.8474,
+      "step": 13725
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 14.02380084991455,
+      "learning_rate": 8.516085846964817e-06,
+      "loss": 0.7024,
+      "step": 13726
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.211564064025879,
+      "learning_rate": 8.51524913190813e-06,
+      "loss": 1.5636,
+      "step": 13727
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 4.276966094970703,
+      "learning_rate": 8.514412416851441e-06,
+      "loss": 0.2978,
+      "step": 13728
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 155.47015380859375,
+      "learning_rate": 8.513575701794753e-06,
+      "loss": 2.1612,
+      "step": 13729
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 22.023038864135742,
+      "learning_rate": 8.512738986738067e-06,
+      "loss": 1.2634,
+      "step": 13730
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 18.61208724975586,
+      "learning_rate": 8.511902271681379e-06,
+      "loss": 0.8921,
+      "step": 13731
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.686084747314453,
+      "learning_rate": 8.511065556624692e-06,
+      "loss": 1.5567,
+      "step": 13732
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 16.485912322998047,
+      "learning_rate": 8.510228841568004e-06,
+      "loss": 0.8855,
+      "step": 13733
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 17.467750549316406,
+      "learning_rate": 8.509392126511318e-06,
+      "loss": 1.5973,
+      "step": 13734
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 25.251373291015625,
+      "learning_rate": 8.50855541145463e-06,
+      "loss": 0.9983,
+      "step": 13735
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 48.691627502441406,
+      "learning_rate": 8.507718696397942e-06,
+      "loss": 2.9335,
+      "step": 13736
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 35.23447036743164,
+      "learning_rate": 8.506881981341256e-06,
+      "loss": 1.7289,
+      "step": 13737
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.047329902648926,
+      "learning_rate": 8.506045266284568e-06,
+      "loss": 0.5925,
+      "step": 13738
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 23.861122131347656,
+      "learning_rate": 8.505208551227881e-06,
+      "loss": 1.1134,
+      "step": 13739
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 20.802949905395508,
+      "learning_rate": 8.504371836171193e-06,
+      "loss": 0.4285,
+      "step": 13740
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 16.312307357788086,
+      "learning_rate": 8.503535121114505e-06,
+      "loss": 1.0645,
+      "step": 13741
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.014398574829102,
+      "learning_rate": 8.502698406057817e-06,
+      "loss": 1.7962,
+      "step": 13742
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 11.285490989685059,
+      "learning_rate": 8.501861691001129e-06,
+      "loss": 0.7414,
+      "step": 13743
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 9.441317558288574,
+      "learning_rate": 8.501024975944443e-06,
+      "loss": 0.3939,
+      "step": 13744
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 19.6748104095459,
+      "learning_rate": 8.500188260887755e-06,
+      "loss": 1.512,
+      "step": 13745
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 95.89828491210938,
+      "learning_rate": 8.499351545831068e-06,
+      "loss": 1.1321,
+      "step": 13746
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.618378162384033,
+      "learning_rate": 8.49851483077438e-06,
+      "loss": 0.7681,
+      "step": 13747
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.990718841552734,
+      "learning_rate": 8.497678115717694e-06,
+      "loss": 0.6788,
+      "step": 13748
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 12.311456680297852,
+      "learning_rate": 8.496841400661006e-06,
+      "loss": 0.4918,
+      "step": 13749
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 4.860684394836426,
+      "learning_rate": 8.496004685604318e-06,
+      "loss": 1.4039,
+      "step": 13750
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 19.4648380279541,
+      "learning_rate": 8.495167970547631e-06,
+      "loss": 1.4667,
+      "step": 13751
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 4.4415974617004395,
+      "learning_rate": 8.494331255490943e-06,
+      "loss": 0.1931,
+      "step": 13752
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 16.85715675354004,
+      "learning_rate": 8.493494540434255e-06,
+      "loss": 0.9058,
+      "step": 13753
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.172959327697754,
+      "learning_rate": 8.492657825377569e-06,
+      "loss": 1.8469,
+      "step": 13754
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 27.272842407226562,
+      "learning_rate": 8.491821110320881e-06,
+      "loss": 1.3817,
+      "step": 13755
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.073509216308594,
+      "learning_rate": 8.490984395264193e-06,
+      "loss": 1.3283,
+      "step": 13756
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 4.937424182891846,
+      "learning_rate": 8.490147680207505e-06,
+      "loss": 0.0815,
+      "step": 13757
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 46.79994201660156,
+      "learning_rate": 8.489310965150818e-06,
+      "loss": 1.4095,
+      "step": 13758
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.656526565551758,
+      "learning_rate": 8.48847425009413e-06,
+      "loss": 1.0071,
+      "step": 13759
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 27.997211456298828,
+      "learning_rate": 8.487637535037444e-06,
+      "loss": 0.6438,
+      "step": 13760
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 41.18295669555664,
+      "learning_rate": 8.486800819980756e-06,
+      "loss": 0.7046,
+      "step": 13761
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.984874248504639,
+      "learning_rate": 8.48596410492407e-06,
+      "loss": 0.4048,
+      "step": 13762
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 27.08432960510254,
+      "learning_rate": 8.485127389867382e-06,
+      "loss": 1.5601,
+      "step": 13763
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 110.85393524169922,
+      "learning_rate": 8.484290674810694e-06,
+      "loss": 2.1277,
+      "step": 13764
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 45.55110168457031,
+      "learning_rate": 8.483453959754007e-06,
+      "loss": 1.2367,
+      "step": 13765
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 18.700878143310547,
+      "learning_rate": 8.48261724469732e-06,
+      "loss": 0.9085,
+      "step": 13766
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 12.493959426879883,
+      "learning_rate": 8.481780529640631e-06,
+      "loss": 0.6807,
+      "step": 13767
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 24.03819465637207,
+      "learning_rate": 8.480943814583945e-06,
+      "loss": 2.3393,
+      "step": 13768
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 14.463841438293457,
+      "learning_rate": 8.480107099527257e-06,
+      "loss": 1.6939,
+      "step": 13769
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 4.762551307678223,
+      "learning_rate": 8.479270384470569e-06,
+      "loss": 0.3719,
+      "step": 13770
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.250888347625732,
+      "learning_rate": 8.47843366941388e-06,
+      "loss": 0.743,
+      "step": 13771
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.955687522888184,
+      "learning_rate": 8.477596954357194e-06,
+      "loss": 1.356,
+      "step": 13772
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 55.618995666503906,
+      "learning_rate": 8.476760239300506e-06,
+      "loss": 1.9591,
+      "step": 13773
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 5.087319374084473,
+      "learning_rate": 8.47592352424382e-06,
+      "loss": 0.0754,
+      "step": 13774
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 15.81760025024414,
+      "learning_rate": 8.475086809187132e-06,
+      "loss": 0.5238,
+      "step": 13775
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 35.106510162353516,
+      "learning_rate": 8.474250094130446e-06,
+      "loss": 2.2119,
+      "step": 13776
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 19.20872688293457,
+      "learning_rate": 8.473413379073757e-06,
+      "loss": 1.3516,
+      "step": 13777
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 24.926410675048828,
+      "learning_rate": 8.47257666401707e-06,
+      "loss": 1.1482,
+      "step": 13778
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 16.512378692626953,
+      "learning_rate": 8.471739948960383e-06,
+      "loss": 2.4315,
+      "step": 13779
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 26.667551040649414,
+      "learning_rate": 8.470903233903695e-06,
+      "loss": 1.3698,
+      "step": 13780
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.896178722381592,
+      "learning_rate": 8.470066518847007e-06,
+      "loss": 0.0695,
+      "step": 13781
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 16.381025314331055,
+      "learning_rate": 8.469229803790319e-06,
+      "loss": 1.255,
+      "step": 13782
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.191415309906006,
+      "learning_rate": 8.468393088733633e-06,
+      "loss": 0.4244,
+      "step": 13783
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 145.7000274658203,
+      "learning_rate": 8.467556373676945e-06,
+      "loss": 2.1742,
+      "step": 13784
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 17.76723861694336,
+      "learning_rate": 8.466719658620257e-06,
+      "loss": 1.2058,
+      "step": 13785
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 23.901243209838867,
+      "learning_rate": 8.46588294356357e-06,
+      "loss": 1.8144,
+      "step": 13786
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 28.461530685424805,
+      "learning_rate": 8.465046228506882e-06,
+      "loss": 1.5681,
+      "step": 13787
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 6.418713092803955,
+      "learning_rate": 8.464209513450196e-06,
+      "loss": 0.4838,
+      "step": 13788
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 29.05074119567871,
+      "learning_rate": 8.463372798393508e-06,
+      "loss": 2.6298,
+      "step": 13789
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 44.116294860839844,
+      "learning_rate": 8.462536083336821e-06,
+      "loss": 1.7151,
+      "step": 13790
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.961648941040039,
+      "learning_rate": 8.461699368280133e-06,
+      "loss": 1.1424,
+      "step": 13791
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 12.723161697387695,
+      "learning_rate": 8.460862653223445e-06,
+      "loss": 0.7386,
+      "step": 13792
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.818270683288574,
+      "learning_rate": 8.460025938166759e-06,
+      "loss": 1.6343,
+      "step": 13793
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 22.30669593811035,
+      "learning_rate": 8.459189223110071e-06,
+      "loss": 1.8495,
+      "step": 13794
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 14.09034538269043,
+      "learning_rate": 8.458352508053383e-06,
+      "loss": 1.0546,
+      "step": 13795
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.118631362915039,
+      "learning_rate": 8.457515792996695e-06,
+      "loss": 1.7067,
+      "step": 13796
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 12.156725883483887,
+      "learning_rate": 8.456679077940008e-06,
+      "loss": 0.7631,
+      "step": 13797
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 6.303142547607422,
+      "learning_rate": 8.45584236288332e-06,
+      "loss": 0.2077,
+      "step": 13798
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 14.71904182434082,
+      "learning_rate": 8.455005647826632e-06,
+      "loss": 1.9614,
+      "step": 13799
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 19.938352584838867,
+      "learning_rate": 8.454168932769946e-06,
+      "loss": 2.0698,
+      "step": 13800
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 19.06098175048828,
+      "learning_rate": 8.453332217713258e-06,
+      "loss": 1.6484,
+      "step": 13801
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 6.216872692108154,
+      "learning_rate": 8.452495502656572e-06,
+      "loss": 0.2628,
+      "step": 13802
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.316340446472168,
+      "learning_rate": 8.451658787599884e-06,
+      "loss": 1.1693,
+      "step": 13803
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.7603607177734375,
+      "learning_rate": 8.450822072543197e-06,
+      "loss": 0.8073,
+      "step": 13804
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.606012344360352,
+      "learning_rate": 8.44998535748651e-06,
+      "loss": 1.0402,
+      "step": 13805
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 13.379460334777832,
+      "learning_rate": 8.449148642429821e-06,
+      "loss": 1.7186,
+      "step": 13806
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 22.462533950805664,
+      "learning_rate": 8.448311927373135e-06,
+      "loss": 1.9008,
+      "step": 13807
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 12.43236255645752,
+      "learning_rate": 8.447475212316447e-06,
+      "loss": 0.4981,
+      "step": 13808
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 6.2734761238098145,
+      "learning_rate": 8.446638497259759e-06,
+      "loss": 1.7295,
+      "step": 13809
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 5.263993263244629,
+      "learning_rate": 8.44580178220307e-06,
+      "loss": 0.2166,
+      "step": 13810
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 24.654264450073242,
+      "learning_rate": 8.444965067146384e-06,
+      "loss": 1.4845,
+      "step": 13811
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 13.289724349975586,
+      "learning_rate": 8.444128352089696e-06,
+      "loss": 1.5027,
+      "step": 13812
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 31.688879013061523,
+      "learning_rate": 8.443291637033008e-06,
+      "loss": 1.661,
+      "step": 13813
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.262681007385254,
+      "learning_rate": 8.442454921976322e-06,
+      "loss": 0.403,
+      "step": 13814
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 59.12746810913086,
+      "learning_rate": 8.441618206919634e-06,
+      "loss": 2.6707,
+      "step": 13815
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 20.5118408203125,
+      "learning_rate": 8.440781491862947e-06,
+      "loss": 0.8513,
+      "step": 13816
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 12.630598068237305,
+      "learning_rate": 8.43994477680626e-06,
+      "loss": 0.7384,
+      "step": 13817
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.17889404296875,
+      "learning_rate": 8.439108061749573e-06,
+      "loss": 0.7034,
+      "step": 13818
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.370962142944336,
+      "learning_rate": 8.438271346692885e-06,
+      "loss": 1.214,
+      "step": 13819
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 4.087014198303223,
+      "learning_rate": 8.437434631636197e-06,
+      "loss": 0.078,
+      "step": 13820
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 12.483458518981934,
+      "learning_rate": 8.43659791657951e-06,
+      "loss": 2.1561,
+      "step": 13821
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.09365177154541,
+      "learning_rate": 8.435761201522823e-06,
+      "loss": 0.5027,
+      "step": 13822
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 41.06399154663086,
+      "learning_rate": 8.434924486466135e-06,
+      "loss": 2.35,
+      "step": 13823
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 25.1468563079834,
+      "learning_rate": 8.434087771409446e-06,
+      "loss": 1.0067,
+      "step": 13824
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 27.10707664489746,
+      "learning_rate": 8.43325105635276e-06,
+      "loss": 1.8279,
+      "step": 13825
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.643894672393799,
+      "learning_rate": 8.432414341296072e-06,
+      "loss": 1.0992,
+      "step": 13826
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 41.79654312133789,
+      "learning_rate": 8.431577626239384e-06,
+      "loss": 1.5539,
+      "step": 13827
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 40.047645568847656,
+      "learning_rate": 8.430740911182698e-06,
+      "loss": 1.2106,
+      "step": 13828
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 13.573161125183105,
+      "learning_rate": 8.42990419612601e-06,
+      "loss": 0.9888,
+      "step": 13829
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 58.52835464477539,
+      "learning_rate": 8.429067481069323e-06,
+      "loss": 1.9994,
+      "step": 13830
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.157312870025635,
+      "learning_rate": 8.428230766012635e-06,
+      "loss": 0.1587,
+      "step": 13831
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 38.501251220703125,
+      "learning_rate": 8.427394050955947e-06,
+      "loss": 1.1178,
+      "step": 13832
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 26.000242233276367,
+      "learning_rate": 8.42655733589926e-06,
+      "loss": 1.9275,
+      "step": 13833
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 3.8338160514831543,
+      "learning_rate": 8.425720620842573e-06,
+      "loss": 0.2231,
+      "step": 13834
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.635165214538574,
+      "learning_rate": 8.424883905785885e-06,
+      "loss": 0.7556,
+      "step": 13835
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 46.228145599365234,
+      "learning_rate": 8.424047190729198e-06,
+      "loss": 1.097,
+      "step": 13836
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 45.72690963745117,
+      "learning_rate": 8.42321047567251e-06,
+      "loss": 2.284,
+      "step": 13837
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.786887168884277,
+      "learning_rate": 8.422373760615822e-06,
+      "loss": 0.9054,
+      "step": 13838
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 6.442713737487793,
+      "learning_rate": 8.421537045559134e-06,
+      "loss": 0.277,
+      "step": 13839
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 18.807449340820312,
+      "learning_rate": 8.420700330502448e-06,
+      "loss": 1.4255,
+      "step": 13840
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 12.640721321105957,
+      "learning_rate": 8.41986361544576e-06,
+      "loss": 0.7746,
+      "step": 13841
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 11.13254165649414,
+      "learning_rate": 8.419026900389074e-06,
+      "loss": 0.8999,
+      "step": 13842
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 20.162174224853516,
+      "learning_rate": 8.418190185332385e-06,
+      "loss": 2.4199,
+      "step": 13843
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 16.63433074951172,
+      "learning_rate": 8.417353470275699e-06,
+      "loss": 1.207,
+      "step": 13844
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 29.841510772705078,
+      "learning_rate": 8.416516755219011e-06,
+      "loss": 1.2163,
+      "step": 13845
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 15.83189868927002,
+      "learning_rate": 8.415680040162323e-06,
+      "loss": 0.5019,
+      "step": 13846
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 6.463802337646484,
+      "learning_rate": 8.414843325105637e-06,
+      "loss": 0.3681,
+      "step": 13847
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.133244514465332,
+      "learning_rate": 8.414006610048949e-06,
+      "loss": 0.632,
+      "step": 13848
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.25268268585205,
+      "learning_rate": 8.41316989499226e-06,
+      "loss": 0.8419,
+      "step": 13849
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.213932037353516,
+      "learning_rate": 8.412333179935574e-06,
+      "loss": 0.5327,
+      "step": 13850
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 16.98855972290039,
+      "learning_rate": 8.411496464878886e-06,
+      "loss": 0.4425,
+      "step": 13851
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 12.402276039123535,
+      "learning_rate": 8.410659749822198e-06,
+      "loss": 0.9086,
+      "step": 13852
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.999429225921631,
+      "learning_rate": 8.40982303476551e-06,
+      "loss": 0.1145,
+      "step": 13853
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 7.634301662445068,
+      "learning_rate": 8.408986319708824e-06,
+      "loss": 0.1823,
+      "step": 13854
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 31.303504943847656,
+      "learning_rate": 8.408149604652136e-06,
+      "loss": 1.3208,
+      "step": 13855
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.7392351627349854,
+      "learning_rate": 8.40731288959545e-06,
+      "loss": 0.1621,
+      "step": 13856
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.222172737121582,
+      "learning_rate": 8.406476174538761e-06,
+      "loss": 0.4276,
+      "step": 13857
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.697710037231445,
+      "learning_rate": 8.405639459482075e-06,
+      "loss": 1.2623,
+      "step": 13858
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 16.46355628967285,
+      "learning_rate": 8.404802744425387e-06,
+      "loss": 1.2663,
+      "step": 13859
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 19.478023529052734,
+      "learning_rate": 8.403966029368699e-06,
+      "loss": 1.607,
+      "step": 13860
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 13.396600723266602,
+      "learning_rate": 8.403129314312013e-06,
+      "loss": 0.8841,
+      "step": 13861
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 39.11217498779297,
+      "learning_rate": 8.402292599255324e-06,
+      "loss": 2.5331,
+      "step": 13862
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 20.548969268798828,
+      "learning_rate": 8.401455884198636e-06,
+      "loss": 0.7516,
+      "step": 13863
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 14.2288179397583,
+      "learning_rate": 8.400619169141948e-06,
+      "loss": 2.0292,
+      "step": 13864
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 21.902082443237305,
+      "learning_rate": 8.399782454085262e-06,
+      "loss": 1.5894,
+      "step": 13865
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 7.677196979522705,
+      "learning_rate": 8.398945739028574e-06,
+      "loss": 0.302,
+      "step": 13866
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 18.952892303466797,
+      "learning_rate": 8.398109023971886e-06,
+      "loss": 1.3935,
+      "step": 13867
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.475656509399414,
+      "learning_rate": 8.3972723089152e-06,
+      "loss": 0.7635,
+      "step": 13868
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 11.825992584228516,
+      "learning_rate": 8.396435593858512e-06,
+      "loss": 0.7922,
+      "step": 13869
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 7.463834762573242,
+      "learning_rate": 8.395598878801825e-06,
+      "loss": 0.9997,
+      "step": 13870
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 32.768089294433594,
+      "learning_rate": 8.394762163745137e-06,
+      "loss": 1.071,
+      "step": 13871
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 19.2761173248291,
+      "learning_rate": 8.39392544868845e-06,
+      "loss": 1.2451,
+      "step": 13872
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 17.122081756591797,
+      "learning_rate": 8.393088733631763e-06,
+      "loss": 0.5563,
+      "step": 13873
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 15.508161544799805,
+      "learning_rate": 8.392252018575075e-06,
+      "loss": 1.5265,
+      "step": 13874
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.615347862243652,
+      "learning_rate": 8.391415303518388e-06,
+      "loss": 0.7297,
+      "step": 13875
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 36.48480224609375,
+      "learning_rate": 8.3905785884617e-06,
+      "loss": 1.2946,
+      "step": 13876
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 16.75336456298828,
+      "learning_rate": 8.389741873405012e-06,
+      "loss": 0.4269,
+      "step": 13877
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.70810604095459,
+      "learning_rate": 8.388905158348324e-06,
+      "loss": 0.6779,
+      "step": 13878
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 52.989200592041016,
+      "learning_rate": 8.388068443291638e-06,
+      "loss": 0.5792,
+      "step": 13879
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 17.84124755859375,
+      "learning_rate": 8.38723172823495e-06,
+      "loss": 2.1604,
+      "step": 13880
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 15.700440406799316,
+      "learning_rate": 8.386395013178262e-06,
+      "loss": 0.8854,
+      "step": 13881
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.550549030303955,
+      "learning_rate": 8.385558298121575e-06,
+      "loss": 1.4364,
+      "step": 13882
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.984732151031494,
+      "learning_rate": 8.384721583064887e-06,
+      "loss": 0.7948,
+      "step": 13883
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 13.10971450805664,
+      "learning_rate": 8.383884868008201e-06,
+      "loss": 0.768,
+      "step": 13884
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.079219818115234,
+      "learning_rate": 8.383048152951513e-06,
+      "loss": 0.5529,
+      "step": 13885
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 11.177825927734375,
+      "learning_rate": 8.382211437894827e-06,
+      "loss": 0.5724,
+      "step": 13886
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 11.534337043762207,
+      "learning_rate": 8.381374722838139e-06,
+      "loss": 0.91,
+      "step": 13887
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 11.55118179321289,
+      "learning_rate": 8.38053800778145e-06,
+      "loss": 0.3911,
+      "step": 13888
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.919164180755615,
+      "learning_rate": 8.379701292724764e-06,
+      "loss": 0.2172,
+      "step": 13889
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 40.55359649658203,
+      "learning_rate": 8.378864577668076e-06,
+      "loss": 1.3834,
+      "step": 13890
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 27.290895462036133,
+      "learning_rate": 8.378027862611388e-06,
+      "loss": 1.6225,
+      "step": 13891
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 42.276206970214844,
+      "learning_rate": 8.3771911475547e-06,
+      "loss": 0.2779,
+      "step": 13892
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 12.790298461914062,
+      "learning_rate": 8.376354432498014e-06,
+      "loss": 1.0371,
+      "step": 13893
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.467720031738281,
+      "learning_rate": 8.375517717441326e-06,
+      "loss": 0.5272,
+      "step": 13894
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 17.65713882446289,
+      "learning_rate": 8.374681002384638e-06,
+      "loss": 0.5529,
+      "step": 13895
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 70.17169952392578,
+      "learning_rate": 8.373844287327951e-06,
+      "loss": 1.4261,
+      "step": 13896
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 11.62064266204834,
+      "learning_rate": 8.373007572271263e-06,
+      "loss": 1.475,
+      "step": 13897
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.527636528015137,
+      "learning_rate": 8.372170857214577e-06,
+      "loss": 1.585,
+      "step": 13898
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 20.809040069580078,
+      "learning_rate": 8.371334142157889e-06,
+      "loss": 0.7872,
+      "step": 13899
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 33.864967346191406,
+      "learning_rate": 8.370497427101202e-06,
+      "loss": 2.1348,
+      "step": 13900
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 36.15922927856445,
+      "learning_rate": 8.369660712044514e-06,
+      "loss": 1.2589,
+      "step": 13901
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 15.41215705871582,
+      "learning_rate": 8.368823996987826e-06,
+      "loss": 1.1921,
+      "step": 13902
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 21.640539169311523,
+      "learning_rate": 8.36798728193114e-06,
+      "loss": 1.2825,
+      "step": 13903
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 31.704418182373047,
+      "learning_rate": 8.367150566874452e-06,
+      "loss": 1.0447,
+      "step": 13904
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 7.182675361633301,
+      "learning_rate": 8.366313851817764e-06,
+      "loss": 0.4019,
+      "step": 13905
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 32.93462371826172,
+      "learning_rate": 8.365477136761076e-06,
+      "loss": 2.2954,
+      "step": 13906
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 11.313870429992676,
+      "learning_rate": 8.36464042170439e-06,
+      "loss": 1.4137,
+      "step": 13907
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.952693939208984,
+      "learning_rate": 8.363803706647701e-06,
+      "loss": 0.7722,
+      "step": 13908
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.723963737487793,
+      "learning_rate": 8.362966991591013e-06,
+      "loss": 2.7077,
+      "step": 13909
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.148795127868652,
+      "learning_rate": 8.362130276534327e-06,
+      "loss": 2.0539,
+      "step": 13910
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.466438293457031,
+      "learning_rate": 8.361293561477639e-06,
+      "loss": 0.5725,
+      "step": 13911
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 11.479924201965332,
+      "learning_rate": 8.360456846420953e-06,
+      "loss": 0.8503,
+      "step": 13912
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.030116081237793,
+      "learning_rate": 8.359620131364265e-06,
+      "loss": 1.1705,
+      "step": 13913
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 7.061598300933838,
+      "learning_rate": 8.358783416307578e-06,
+      "loss": 0.3001,
+      "step": 13914
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 14.592386245727539,
+      "learning_rate": 8.35794670125089e-06,
+      "loss": 0.9844,
+      "step": 13915
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.461309909820557,
+      "learning_rate": 8.357109986194202e-06,
+      "loss": 0.3957,
+      "step": 13916
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.254886627197266,
+      "learning_rate": 8.356273271137514e-06,
+      "loss": 0.1697,
+      "step": 13917
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 14.68002986907959,
+      "learning_rate": 8.355436556080828e-06,
+      "loss": 1.0483,
+      "step": 13918
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 15.130146980285645,
+      "learning_rate": 8.35459984102414e-06,
+      "loss": 1.4724,
+      "step": 13919
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 20.888988494873047,
+      "learning_rate": 8.353763125967452e-06,
+      "loss": 1.3916,
+      "step": 13920
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.216975212097168,
+      "learning_rate": 8.352926410910765e-06,
+      "loss": 0.3461,
+      "step": 13921
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 22.507368087768555,
+      "learning_rate": 8.352089695854077e-06,
+      "loss": 0.6662,
+      "step": 13922
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 33.56687545776367,
+      "learning_rate": 8.35125298079739e-06,
+      "loss": 0.6495,
+      "step": 13923
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 36.28936767578125,
+      "learning_rate": 8.350416265740703e-06,
+      "loss": 1.6898,
+      "step": 13924
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 11.19839096069336,
+      "learning_rate": 8.349579550684015e-06,
+      "loss": 1.4511,
+      "step": 13925
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.867090225219727,
+      "learning_rate": 8.348742835627329e-06,
+      "loss": 1.0426,
+      "step": 13926
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 8.462395668029785,
+      "learning_rate": 8.34790612057064e-06,
+      "loss": 0.7899,
+      "step": 13927
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.737425804138184,
+      "learning_rate": 8.347069405513954e-06,
+      "loss": 1.1286,
+      "step": 13928
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 15.307694435119629,
+      "learning_rate": 8.346232690457266e-06,
+      "loss": 0.1565,
+      "step": 13929
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.2129364013671875,
+      "learning_rate": 8.345395975400578e-06,
+      "loss": 0.1186,
+      "step": 13930
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 26.382848739624023,
+      "learning_rate": 8.34455926034389e-06,
+      "loss": 1.5627,
+      "step": 13931
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 7.875923156738281,
+      "learning_rate": 8.343722545287204e-06,
+      "loss": 1.8287,
+      "step": 13932
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 20.020475387573242,
+      "learning_rate": 8.342885830230516e-06,
+      "loss": 1.3609,
+      "step": 13933
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.87630033493042,
+      "learning_rate": 8.342049115173828e-06,
+      "loss": 0.3188,
+      "step": 13934
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 32.75434494018555,
+      "learning_rate": 8.341212400117141e-06,
+      "loss": 2.3851,
+      "step": 13935
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 7.332067489624023,
+      "learning_rate": 8.340375685060453e-06,
+      "loss": 0.6301,
+      "step": 13936
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 13.54127025604248,
+      "learning_rate": 8.339538970003765e-06,
+      "loss": 1.4954,
+      "step": 13937
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 73.1351318359375,
+      "learning_rate": 8.338702254947079e-06,
+      "loss": 2.0352,
+      "step": 13938
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 13.21261215209961,
+      "learning_rate": 8.33786553989039e-06,
+      "loss": 1.9045,
+      "step": 13939
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 57.44038772583008,
+      "learning_rate": 8.337028824833704e-06,
+      "loss": 0.7105,
+      "step": 13940
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.058891296386719,
+      "learning_rate": 8.336192109777016e-06,
+      "loss": 0.5623,
+      "step": 13941
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 31.473182678222656,
+      "learning_rate": 8.33535539472033e-06,
+      "loss": 1.6002,
+      "step": 13942
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 13.409120559692383,
+      "learning_rate": 8.334518679663642e-06,
+      "loss": 1.8,
+      "step": 13943
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 7.674248218536377,
+      "learning_rate": 8.333681964606954e-06,
+      "loss": 0.7707,
+      "step": 13944
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 20.194053649902344,
+      "learning_rate": 8.332845249550266e-06,
+      "loss": 0.9419,
+      "step": 13945
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 23.43710708618164,
+      "learning_rate": 8.332008534493578e-06,
+      "loss": 0.7634,
+      "step": 13946
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 28.353042602539062,
+      "learning_rate": 8.331171819436891e-06,
+      "loss": 1.4436,
+      "step": 13947
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.709844589233398,
+      "learning_rate": 8.330335104380203e-06,
+      "loss": 0.6189,
+      "step": 13948
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 7.055968761444092,
+      "learning_rate": 8.329498389323517e-06,
+      "loss": 0.2982,
+      "step": 13949
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.116999626159668,
+      "learning_rate": 8.328661674266829e-06,
+      "loss": 0.6182,
+      "step": 13950
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 17.69551658630371,
+      "learning_rate": 8.327824959210141e-06,
+      "loss": 1.749,
+      "step": 13951
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 23.674076080322266,
+      "learning_rate": 8.326988244153455e-06,
+      "loss": 3.066,
+      "step": 13952
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 22.856245040893555,
+      "learning_rate": 8.326151529096767e-06,
+      "loss": 0.8221,
+      "step": 13953
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 22.558551788330078,
+      "learning_rate": 8.32531481404008e-06,
+      "loss": 0.754,
+      "step": 13954
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 8.679530143737793,
+      "learning_rate": 8.324478098983392e-06,
+      "loss": 0.5943,
+      "step": 13955
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.526307106018066,
+      "learning_rate": 8.323641383926704e-06,
+      "loss": 1.392,
+      "step": 13956
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 18.145200729370117,
+      "learning_rate": 8.322804668870018e-06,
+      "loss": 1.5801,
+      "step": 13957
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.041570663452148,
+      "learning_rate": 8.32196795381333e-06,
+      "loss": 0.6676,
+      "step": 13958
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 34.781551361083984,
+      "learning_rate": 8.321131238756642e-06,
+      "loss": 0.4334,
+      "step": 13959
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 58.627052307128906,
+      "learning_rate": 8.320294523699954e-06,
+      "loss": 0.6873,
+      "step": 13960
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 7.583497047424316,
+      "learning_rate": 8.319457808643267e-06,
+      "loss": 0.4713,
+      "step": 13961
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 15.208681106567383,
+      "learning_rate": 8.31862109358658e-06,
+      "loss": 1.3923,
+      "step": 13962
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.86228084564209,
+      "learning_rate": 8.317784378529893e-06,
+      "loss": 1.069,
+      "step": 13963
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.096314430236816,
+      "learning_rate": 8.316947663473205e-06,
+      "loss": 0.3963,
+      "step": 13964
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 19.38264274597168,
+      "learning_rate": 8.316110948416517e-06,
+      "loss": 0.9396,
+      "step": 13965
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 11.397652626037598,
+      "learning_rate": 8.31527423335983e-06,
+      "loss": 0.5981,
+      "step": 13966
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 3.4790005683898926,
+      "learning_rate": 8.314437518303142e-06,
+      "loss": 0.1858,
+      "step": 13967
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.452469825744629,
+      "learning_rate": 8.313600803246456e-06,
+      "loss": 0.807,
+      "step": 13968
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.185338973999023,
+      "learning_rate": 8.312764088189768e-06,
+      "loss": 0.303,
+      "step": 13969
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.862409591674805,
+      "learning_rate": 8.31192737313308e-06,
+      "loss": 0.4866,
+      "step": 13970
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 25.07809829711914,
+      "learning_rate": 8.311090658076394e-06,
+      "loss": 1.3827,
+      "step": 13971
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 17.37289810180664,
+      "learning_rate": 8.310253943019706e-06,
+      "loss": 0.9669,
+      "step": 13972
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 7.514956951141357,
+      "learning_rate": 8.309417227963018e-06,
+      "loss": 0.3129,
+      "step": 13973
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.670135498046875,
+      "learning_rate": 8.30858051290633e-06,
+      "loss": 0.3078,
+      "step": 13974
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.870745658874512,
+      "learning_rate": 8.307743797849643e-06,
+      "loss": 0.7646,
+      "step": 13975
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 14.9131498336792,
+      "learning_rate": 8.306907082792955e-06,
+      "loss": 0.6213,
+      "step": 13976
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 85.65762329101562,
+      "learning_rate": 8.306070367736269e-06,
+      "loss": 3.9704,
+      "step": 13977
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.706620216369629,
+      "learning_rate": 8.30523365267958e-06,
+      "loss": 1.3252,
+      "step": 13978
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.356450080871582,
+      "learning_rate": 8.304396937622893e-06,
+      "loss": 1.0579,
+      "step": 13979
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 23.153383255004883,
+      "learning_rate": 8.303560222566206e-06,
+      "loss": 2.8506,
+      "step": 13980
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 22.212078094482422,
+      "learning_rate": 8.302723507509518e-06,
+      "loss": 0.7168,
+      "step": 13981
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 7.272219181060791,
+      "learning_rate": 8.301886792452832e-06,
+      "loss": 0.3777,
+      "step": 13982
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 15.351566314697266,
+      "learning_rate": 8.301050077396144e-06,
+      "loss": 2.7114,
+      "step": 13983
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.014077186584473,
+      "learning_rate": 8.300213362339456e-06,
+      "loss": 2.2495,
+      "step": 13984
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.368762969970703,
+      "learning_rate": 8.299376647282768e-06,
+      "loss": 0.302,
+      "step": 13985
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 120.40667724609375,
+      "learning_rate": 8.298539932226081e-06,
+      "loss": 1.5484,
+      "step": 13986
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 15.013331413269043,
+      "learning_rate": 8.297703217169393e-06,
+      "loss": 1.0579,
+      "step": 13987
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.324745178222656,
+      "learning_rate": 8.296866502112705e-06,
+      "loss": 0.6416,
+      "step": 13988
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 6.499182224273682,
+      "learning_rate": 8.296029787056019e-06,
+      "loss": 1.6797,
+      "step": 13989
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.683255195617676,
+      "learning_rate": 8.295193071999331e-06,
+      "loss": 0.8339,
+      "step": 13990
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 3.60404372215271,
+      "learning_rate": 8.294356356942645e-06,
+      "loss": 0.2472,
+      "step": 13991
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 17.4386043548584,
+      "learning_rate": 8.293519641885957e-06,
+      "loss": 1.2046,
+      "step": 13992
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.467900276184082,
+      "learning_rate": 8.292682926829268e-06,
+      "loss": 0.443,
+      "step": 13993
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 21.231168746948242,
+      "learning_rate": 8.291846211772582e-06,
+      "loss": 0.9453,
+      "step": 13994
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 30.515880584716797,
+      "learning_rate": 8.291009496715894e-06,
+      "loss": 1.3454,
+      "step": 13995
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 30.542156219482422,
+      "learning_rate": 8.290172781659208e-06,
+      "loss": 2.3798,
+      "step": 13996
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 43.25813674926758,
+      "learning_rate": 8.28933606660252e-06,
+      "loss": 1.7446,
+      "step": 13997
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 93.33450317382812,
+      "learning_rate": 8.288499351545832e-06,
+      "loss": 0.9311,
+      "step": 13998
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 23.29090118408203,
+      "learning_rate": 8.287662636489144e-06,
+      "loss": 2.2541,
+      "step": 13999
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 13.520791053771973,
+      "learning_rate": 8.286825921432457e-06,
+      "loss": 1.5217,
+      "step": 14000
+    },
+    {
+      "epoch": 1.76,
+      "eval_loss": 0.07692699134349823,
+      "eval_runtime": 97.7806,
+      "eval_samples_per_second": 36.224,
+      "eval_steps_per_second": 36.224,
+      "step": 14000
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 12.751508712768555,
+      "learning_rate": 8.28598920637577e-06,
+      "loss": 0.2391,
+      "step": 14001
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 20.14134979248047,
+      "learning_rate": 8.285152491319081e-06,
+      "loss": 3.0113,
+      "step": 14002
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 63.07255172729492,
+      "learning_rate": 8.284315776262395e-06,
+      "loss": 3.1447,
+      "step": 14003
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 77.31236267089844,
+      "learning_rate": 8.283479061205707e-06,
+      "loss": 1.9684,
+      "step": 14004
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.814595222473145,
+      "learning_rate": 8.282642346149019e-06,
+      "loss": 1.8572,
+      "step": 14005
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.436347007751465,
+      "learning_rate": 8.281805631092332e-06,
+      "loss": 0.6581,
+      "step": 14006
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 57.64645767211914,
+      "learning_rate": 8.280968916035644e-06,
+      "loss": 2.1964,
+      "step": 14007
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.93832778930664,
+      "learning_rate": 8.280132200978958e-06,
+      "loss": 2.741,
+      "step": 14008
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 29.666269302368164,
+      "learning_rate": 8.27929548592227e-06,
+      "loss": 1.2742,
+      "step": 14009
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.4997217655181885,
+      "learning_rate": 8.278458770865584e-06,
+      "loss": 0.0794,
+      "step": 14010
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 14.185319900512695,
+      "learning_rate": 8.277622055808896e-06,
+      "loss": 0.4328,
+      "step": 14011
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.742040634155273,
+      "learning_rate": 8.276785340752207e-06,
+      "loss": 0.6479,
+      "step": 14012
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 14.843554496765137,
+      "learning_rate": 8.27594862569552e-06,
+      "loss": 1.0409,
+      "step": 14013
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.190027236938477,
+      "learning_rate": 8.275111910638833e-06,
+      "loss": 0.1852,
+      "step": 14014
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 154.2579803466797,
+      "learning_rate": 8.274275195582145e-06,
+      "loss": 1.6351,
+      "step": 14015
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.802145957946777,
+      "learning_rate": 8.273438480525457e-06,
+      "loss": 0.9051,
+      "step": 14016
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.853428363800049,
+      "learning_rate": 8.27260176546877e-06,
+      "loss": 0.9526,
+      "step": 14017
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.507288932800293,
+      "learning_rate": 8.271765050412083e-06,
+      "loss": 1.5627,
+      "step": 14018
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.447890281677246,
+      "learning_rate": 8.270928335355395e-06,
+      "loss": 1.6398,
+      "step": 14019
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 13.446758270263672,
+      "learning_rate": 8.270091620298708e-06,
+      "loss": 0.6096,
+      "step": 14020
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 12.474729537963867,
+      "learning_rate": 8.26925490524202e-06,
+      "loss": 1.2577,
+      "step": 14021
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 131.98252868652344,
+      "learning_rate": 8.268418190185334e-06,
+      "loss": 3.8721,
+      "step": 14022
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 19.97817611694336,
+      "learning_rate": 8.267581475128646e-06,
+      "loss": 2.113,
+      "step": 14023
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.92810344696045,
+      "learning_rate": 8.26674476007196e-06,
+      "loss": 0.4914,
+      "step": 14024
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.805947303771973,
+      "learning_rate": 8.265908045015271e-06,
+      "loss": 1.5212,
+      "step": 14025
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 13.755889892578125,
+      "learning_rate": 8.265071329958583e-06,
+      "loss": 0.5217,
+      "step": 14026
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 21.62193489074707,
+      "learning_rate": 8.264234614901895e-06,
+      "loss": 0.7373,
+      "step": 14027
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 63.61070251464844,
+      "learning_rate": 8.263397899845207e-06,
+      "loss": 1.6375,
+      "step": 14028
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.767998218536377,
+      "learning_rate": 8.262561184788521e-06,
+      "loss": 0.022,
+      "step": 14029
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.809958457946777,
+      "learning_rate": 8.261724469731833e-06,
+      "loss": 1.0422,
+      "step": 14030
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 13.731112480163574,
+      "learning_rate": 8.260887754675146e-06,
+      "loss": 1.6344,
+      "step": 14031
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 16.74528694152832,
+      "learning_rate": 8.260051039618458e-06,
+      "loss": 1.2197,
+      "step": 14032
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.615007400512695,
+      "learning_rate": 8.25921432456177e-06,
+      "loss": 0.6503,
+      "step": 14033
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.938024520874023,
+      "learning_rate": 8.258377609505084e-06,
+      "loss": 0.6201,
+      "step": 14034
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.573639869689941,
+      "learning_rate": 8.257540894448396e-06,
+      "loss": 1.1605,
+      "step": 14035
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 4.799045562744141,
+      "learning_rate": 8.25670417939171e-06,
+      "loss": 1.4988,
+      "step": 14036
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.836624145507812,
+      "learning_rate": 8.255867464335022e-06,
+      "loss": 0.992,
+      "step": 14037
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.944649696350098,
+      "learning_rate": 8.255030749278334e-06,
+      "loss": 1.1869,
+      "step": 14038
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.460426330566406,
+      "learning_rate": 8.254194034221647e-06,
+      "loss": 2.0168,
+      "step": 14039
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.50805377960205,
+      "learning_rate": 8.253357319164959e-06,
+      "loss": 2.0397,
+      "step": 14040
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 82.7448959350586,
+      "learning_rate": 8.252520604108271e-06,
+      "loss": 1.589,
+      "step": 14041
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 37.63884735107422,
+      "learning_rate": 8.251683889051583e-06,
+      "loss": 1.0803,
+      "step": 14042
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 41.49068069458008,
+      "learning_rate": 8.250847173994897e-06,
+      "loss": 1.602,
+      "step": 14043
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.409775733947754,
+      "learning_rate": 8.250010458938209e-06,
+      "loss": 0.7266,
+      "step": 14044
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 19.899547576904297,
+      "learning_rate": 8.249173743881522e-06,
+      "loss": 1.347,
+      "step": 14045
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 21.121326446533203,
+      "learning_rate": 8.248337028824834e-06,
+      "loss": 0.9298,
+      "step": 14046
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 13.431509017944336,
+      "learning_rate": 8.247500313768146e-06,
+      "loss": 0.9772,
+      "step": 14047
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 19.138832092285156,
+      "learning_rate": 8.24666359871146e-06,
+      "loss": 0.7483,
+      "step": 14048
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.172293663024902,
+      "learning_rate": 8.245826883654772e-06,
+      "loss": 0.886,
+      "step": 14049
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 15.299717903137207,
+      "learning_rate": 8.244990168598085e-06,
+      "loss": 0.2258,
+      "step": 14050
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 6.301414489746094,
+      "learning_rate": 8.244153453541397e-06,
+      "loss": 0.4489,
+      "step": 14051
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.925097465515137,
+      "learning_rate": 8.24331673848471e-06,
+      "loss": 0.3128,
+      "step": 14052
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 12.037699699401855,
+      "learning_rate": 8.242480023428023e-06,
+      "loss": 1.4974,
+      "step": 14053
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 14.054902076721191,
+      "learning_rate": 8.241643308371335e-06,
+      "loss": 1.1296,
+      "step": 14054
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 75.2170181274414,
+      "learning_rate": 8.240806593314647e-06,
+      "loss": 3.0972,
+      "step": 14055
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 25.608165740966797,
+      "learning_rate": 8.239969878257959e-06,
+      "loss": 0.7163,
+      "step": 14056
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 3.8469531536102295,
+      "learning_rate": 8.239133163201273e-06,
+      "loss": 0.0497,
+      "step": 14057
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 26.958236694335938,
+      "learning_rate": 8.238296448144584e-06,
+      "loss": 2.1268,
+      "step": 14058
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.09694766998291,
+      "learning_rate": 8.237459733087898e-06,
+      "loss": 0.5967,
+      "step": 14059
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 18.695819854736328,
+      "learning_rate": 8.23662301803121e-06,
+      "loss": 1.8521,
+      "step": 14060
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.077091217041016,
+      "learning_rate": 8.235786302974522e-06,
+      "loss": 0.4429,
+      "step": 14061
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.94905948638916,
+      "learning_rate": 8.234949587917836e-06,
+      "loss": 1.2048,
+      "step": 14062
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 62.44631576538086,
+      "learning_rate": 8.234112872861148e-06,
+      "loss": 1.0883,
+      "step": 14063
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 4.850184440612793,
+      "learning_rate": 8.233276157804461e-06,
+      "loss": 0.5964,
+      "step": 14064
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 81.4451675415039,
+      "learning_rate": 8.232439442747773e-06,
+      "loss": 2.5383,
+      "step": 14065
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.646984100341797,
+      "learning_rate": 8.231602727691085e-06,
+      "loss": 1.8475,
+      "step": 14066
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 271.2912902832031,
+      "learning_rate": 8.230766012634397e-06,
+      "loss": 2.6256,
+      "step": 14067
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 33.16635513305664,
+      "learning_rate": 8.22992929757771e-06,
+      "loss": 1.236,
+      "step": 14068
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 15.70523738861084,
+      "learning_rate": 8.229092582521023e-06,
+      "loss": 1.4069,
+      "step": 14069
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 28.73826026916504,
+      "learning_rate": 8.228255867464335e-06,
+      "loss": 1.5104,
+      "step": 14070
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 12.562192916870117,
+      "learning_rate": 8.227419152407648e-06,
+      "loss": 1.5651,
+      "step": 14071
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.837984085083008,
+      "learning_rate": 8.22658243735096e-06,
+      "loss": 1.742,
+      "step": 14072
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.106812000274658,
+      "learning_rate": 8.225745722294274e-06,
+      "loss": 0.6934,
+      "step": 14073
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.599335193634033,
+      "learning_rate": 8.224909007237586e-06,
+      "loss": 0.5154,
+      "step": 14074
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 27.655122756958008,
+      "learning_rate": 8.224072292180898e-06,
+      "loss": 1.2087,
+      "step": 14075
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 18.773101806640625,
+      "learning_rate": 8.223235577124212e-06,
+      "loss": 2.5761,
+      "step": 14076
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.021270275115967,
+      "learning_rate": 8.222398862067523e-06,
+      "loss": 0.5751,
+      "step": 14077
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.0785064697265625,
+      "learning_rate": 8.221562147010837e-06,
+      "loss": 0.3811,
+      "step": 14078
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.324564933776855,
+      "learning_rate": 8.220725431954149e-06,
+      "loss": 0.7991,
+      "step": 14079
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 30.627126693725586,
+      "learning_rate": 8.219888716897461e-06,
+      "loss": 1.1427,
+      "step": 14080
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.819294929504395,
+      "learning_rate": 8.219052001840773e-06,
+      "loss": 0.6441,
+      "step": 14081
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 10.612582206726074,
+      "learning_rate": 8.218215286784087e-06,
+      "loss": 0.8362,
+      "step": 14082
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.291300773620605,
+      "learning_rate": 8.217378571727399e-06,
+      "loss": 0.3432,
+      "step": 14083
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 29.742712020874023,
+      "learning_rate": 8.21654185667071e-06,
+      "loss": 2.0117,
+      "step": 14084
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 19.590259552001953,
+      "learning_rate": 8.215705141614024e-06,
+      "loss": 1.6899,
+      "step": 14085
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 23.724468231201172,
+      "learning_rate": 8.214868426557336e-06,
+      "loss": 2.7822,
+      "step": 14086
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 78.61354064941406,
+      "learning_rate": 8.21403171150065e-06,
+      "loss": 2.2423,
+      "step": 14087
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 30.620384216308594,
+      "learning_rate": 8.213194996443962e-06,
+      "loss": 1.3772,
+      "step": 14088
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 27.140432357788086,
+      "learning_rate": 8.212358281387274e-06,
+      "loss": 2.3649,
+      "step": 14089
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.873764038085938,
+      "learning_rate": 8.211521566330587e-06,
+      "loss": 1.0493,
+      "step": 14090
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 5.370067596435547,
+      "learning_rate": 8.2106848512739e-06,
+      "loss": 0.2943,
+      "step": 14091
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 54.14467239379883,
+      "learning_rate": 8.209848136217213e-06,
+      "loss": 1.8964,
+      "step": 14092
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 4.073047637939453,
+      "learning_rate": 8.209011421160525e-06,
+      "loss": 0.1377,
+      "step": 14093
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 12.148030281066895,
+      "learning_rate": 8.208174706103837e-06,
+      "loss": 0.8864,
+      "step": 14094
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.65673303604126,
+      "learning_rate": 8.207337991047149e-06,
+      "loss": 0.3389,
+      "step": 14095
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.8014116287231445,
+      "learning_rate": 8.20650127599046e-06,
+      "loss": 1.4713,
+      "step": 14096
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 23.743858337402344,
+      "learning_rate": 8.205664560933774e-06,
+      "loss": 0.955,
+      "step": 14097
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 5.458376884460449,
+      "learning_rate": 8.204827845877086e-06,
+      "loss": 0.3562,
+      "step": 14098
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 23.411649703979492,
+      "learning_rate": 8.2039911308204e-06,
+      "loss": 2.3615,
+      "step": 14099
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 16.24994468688965,
+      "learning_rate": 8.203154415763712e-06,
+      "loss": 0.3418,
+      "step": 14100
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 18.841440200805664,
+      "learning_rate": 8.202317700707026e-06,
+      "loss": 2.1813,
+      "step": 14101
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 21.16748046875,
+      "learning_rate": 8.201480985650338e-06,
+      "loss": 0.3784,
+      "step": 14102
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.720735549926758,
+      "learning_rate": 8.20064427059365e-06,
+      "loss": 1.1331,
+      "step": 14103
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.038744926452637,
+      "learning_rate": 8.199807555536963e-06,
+      "loss": 0.2231,
+      "step": 14104
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 26.808597564697266,
+      "learning_rate": 8.198970840480275e-06,
+      "loss": 2.4855,
+      "step": 14105
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 16.5147705078125,
+      "learning_rate": 8.198134125423589e-06,
+      "loss": 1.165,
+      "step": 14106
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 26.96607208251953,
+      "learning_rate": 8.1972974103669e-06,
+      "loss": 1.0483,
+      "step": 14107
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.280531883239746,
+      "learning_rate": 8.196460695310213e-06,
+      "loss": 0.786,
+      "step": 14108
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 41.95281219482422,
+      "learning_rate": 8.195623980253525e-06,
+      "loss": 1.6602,
+      "step": 14109
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.45953369140625,
+      "learning_rate": 8.194787265196837e-06,
+      "loss": 0.4899,
+      "step": 14110
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 4.454230308532715,
+      "learning_rate": 8.19395055014015e-06,
+      "loss": 0.1959,
+      "step": 14111
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.201144695281982,
+      "learning_rate": 8.193113835083462e-06,
+      "loss": 1.1016,
+      "step": 14112
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 34.14670181274414,
+      "learning_rate": 8.192277120026776e-06,
+      "loss": 2.4579,
+      "step": 14113
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.853291511535645,
+      "learning_rate": 8.191440404970088e-06,
+      "loss": 0.8053,
+      "step": 14114
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.906975746154785,
+      "learning_rate": 8.190603689913401e-06,
+      "loss": 1.3813,
+      "step": 14115
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.100425720214844,
+      "learning_rate": 8.189766974856713e-06,
+      "loss": 0.8159,
+      "step": 14116
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 16.280963897705078,
+      "learning_rate": 8.188930259800025e-06,
+      "loss": 0.8809,
+      "step": 14117
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 14.504768371582031,
+      "learning_rate": 8.188093544743339e-06,
+      "loss": 2.6587,
+      "step": 14118
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 34.045616149902344,
+      "learning_rate": 8.187256829686651e-06,
+      "loss": 3.184,
+      "step": 14119
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 20.449115753173828,
+      "learning_rate": 8.186420114629963e-06,
+      "loss": 1.1062,
+      "step": 14120
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.727039337158203,
+      "learning_rate": 8.185583399573277e-06,
+      "loss": 0.9284,
+      "step": 14121
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 10.568921089172363,
+      "learning_rate": 8.184746684516589e-06,
+      "loss": 0.9385,
+      "step": 14122
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 19.89391326904297,
+      "learning_rate": 8.1839099694599e-06,
+      "loss": 0.916,
+      "step": 14123
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 10.833699226379395,
+      "learning_rate": 8.183073254403212e-06,
+      "loss": 0.8083,
+      "step": 14124
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 22.74962615966797,
+      "learning_rate": 8.182236539346526e-06,
+      "loss": 2.1172,
+      "step": 14125
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.44655704498291,
+      "learning_rate": 8.181399824289838e-06,
+      "loss": 1.4098,
+      "step": 14126
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.826528549194336,
+      "learning_rate": 8.180563109233152e-06,
+      "loss": 2.3645,
+      "step": 14127
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 10.453386306762695,
+      "learning_rate": 8.179726394176464e-06,
+      "loss": 1.8492,
+      "step": 14128
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 33.9998664855957,
+      "learning_rate": 8.178889679119777e-06,
+      "loss": 1.2318,
+      "step": 14129
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.9178619384765625,
+      "learning_rate": 8.17805296406309e-06,
+      "loss": 2.0647,
+      "step": 14130
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 267.4306640625,
+      "learning_rate": 8.177216249006401e-06,
+      "loss": 0.9789,
+      "step": 14131
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 15.695507049560547,
+      "learning_rate": 8.176379533949715e-06,
+      "loss": 1.3647,
+      "step": 14132
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 12.526225090026855,
+      "learning_rate": 8.175542818893027e-06,
+      "loss": 1.476,
+      "step": 14133
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 18.775936126708984,
+      "learning_rate": 8.174706103836339e-06,
+      "loss": 1.061,
+      "step": 14134
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 3.695633888244629,
+      "learning_rate": 8.173869388779652e-06,
+      "loss": 0.1435,
+      "step": 14135
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 11.461366653442383,
+      "learning_rate": 8.173032673722964e-06,
+      "loss": 1.7876,
+      "step": 14136
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 16.700559616088867,
+      "learning_rate": 8.172195958666276e-06,
+      "loss": 1.9508,
+      "step": 14137
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 21.389442443847656,
+      "learning_rate": 8.171359243609588e-06,
+      "loss": 3.045,
+      "step": 14138
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 12.076715469360352,
+      "learning_rate": 8.170522528552902e-06,
+      "loss": 1.1381,
+      "step": 14139
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 32.985836029052734,
+      "learning_rate": 8.169685813496214e-06,
+      "loss": 1.6389,
+      "step": 14140
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.656893730163574,
+      "learning_rate": 8.168849098439528e-06,
+      "loss": 0.5599,
+      "step": 14141
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 10.285635948181152,
+      "learning_rate": 8.16801238338284e-06,
+      "loss": 0.6236,
+      "step": 14142
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.626985549926758,
+      "learning_rate": 8.167175668326153e-06,
+      "loss": 2.7633,
+      "step": 14143
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 14.445494651794434,
+      "learning_rate": 8.166338953269465e-06,
+      "loss": 1.8344,
+      "step": 14144
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 12.0984468460083,
+      "learning_rate": 8.165502238212777e-06,
+      "loss": 1.9759,
+      "step": 14145
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 33.258758544921875,
+      "learning_rate": 8.16466552315609e-06,
+      "loss": 2.6521,
+      "step": 14146
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 17.95985984802246,
+      "learning_rate": 8.163828808099403e-06,
+      "loss": 0.6462,
+      "step": 14147
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 11.247654914855957,
+      "learning_rate": 8.162992093042715e-06,
+      "loss": 0.7743,
+      "step": 14148
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 19.912494659423828,
+      "learning_rate": 8.162155377986027e-06,
+      "loss": 0.9983,
+      "step": 14149
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.2008585929870605,
+      "learning_rate": 8.16131866292934e-06,
+      "loss": 0.5633,
+      "step": 14150
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 7.434240341186523,
+      "learning_rate": 8.160481947872652e-06,
+      "loss": 0.2281,
+      "step": 14151
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 47.61603546142578,
+      "learning_rate": 8.159645232815964e-06,
+      "loss": 2.083,
+      "step": 14152
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 11.545709609985352,
+      "learning_rate": 8.158808517759278e-06,
+      "loss": 1.0948,
+      "step": 14153
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 32.54110336303711,
+      "learning_rate": 8.15797180270259e-06,
+      "loss": 2.0602,
+      "step": 14154
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 88.15934753417969,
+      "learning_rate": 8.157135087645903e-06,
+      "loss": 2.3074,
+      "step": 14155
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 3.6791977882385254,
+      "learning_rate": 8.156298372589215e-06,
+      "loss": 0.4666,
+      "step": 14156
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 14.77776050567627,
+      "learning_rate": 8.155461657532529e-06,
+      "loss": 0.6486,
+      "step": 14157
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 12.394466400146484,
+      "learning_rate": 8.154624942475841e-06,
+      "loss": 1.6881,
+      "step": 14158
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 23.44948387145996,
+      "learning_rate": 8.153788227419153e-06,
+      "loss": 1.8498,
+      "step": 14159
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 219.22438049316406,
+      "learning_rate": 8.152951512362467e-06,
+      "loss": 1.3467,
+      "step": 14160
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.641712188720703,
+      "learning_rate": 8.152114797305779e-06,
+      "loss": 0.3743,
+      "step": 14161
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 16.614871978759766,
+      "learning_rate": 8.15127808224909e-06,
+      "loss": 1.1639,
+      "step": 14162
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.148021697998047,
+      "learning_rate": 8.150441367192402e-06,
+      "loss": 0.7238,
+      "step": 14163
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.253996849060059,
+      "learning_rate": 8.149604652135716e-06,
+      "loss": 0.2262,
+      "step": 14164
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 27.66922378540039,
+      "learning_rate": 8.148767937079028e-06,
+      "loss": 1.6873,
+      "step": 14165
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 22.25579071044922,
+      "learning_rate": 8.14793122202234e-06,
+      "loss": 1.5399,
+      "step": 14166
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 13.656011581420898,
+      "learning_rate": 8.147094506965654e-06,
+      "loss": 1.4157,
+      "step": 14167
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 15.185012817382812,
+      "learning_rate": 8.146257791908966e-06,
+      "loss": 0.71,
+      "step": 14168
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.431121826171875,
+      "learning_rate": 8.14542107685228e-06,
+      "loss": 2.5234,
+      "step": 14169
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 7.757891654968262,
+      "learning_rate": 8.144584361795591e-06,
+      "loss": 0.9268,
+      "step": 14170
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 18.855506896972656,
+      "learning_rate": 8.143747646738905e-06,
+      "loss": 0.881,
+      "step": 14171
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 18.44350242614746,
+      "learning_rate": 8.142910931682217e-06,
+      "loss": 1.6093,
+      "step": 14172
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.91901969909668,
+      "learning_rate": 8.142074216625529e-06,
+      "loss": 0.8963,
+      "step": 14173
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 18.291046142578125,
+      "learning_rate": 8.141237501568842e-06,
+      "loss": 1.1103,
+      "step": 14174
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.663735032081604,
+      "learning_rate": 8.140400786512154e-06,
+      "loss": 0.1063,
+      "step": 14175
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.795211791992188,
+      "learning_rate": 8.139564071455466e-06,
+      "loss": 1.5014,
+      "step": 14176
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 26.331180572509766,
+      "learning_rate": 8.138727356398778e-06,
+      "loss": 1.9751,
+      "step": 14177
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 15.649175643920898,
+      "learning_rate": 8.137890641342092e-06,
+      "loss": 0.6444,
+      "step": 14178
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 8.898287773132324,
+      "learning_rate": 8.137053926285404e-06,
+      "loss": 0.8392,
+      "step": 14179
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 25.118547439575195,
+      "learning_rate": 8.136217211228716e-06,
+      "loss": 1.2187,
+      "step": 14180
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 31.6370849609375,
+      "learning_rate": 8.13538049617203e-06,
+      "loss": 1.7968,
+      "step": 14181
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 11.394192695617676,
+      "learning_rate": 8.134543781115341e-06,
+      "loss": 1.9994,
+      "step": 14182
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.50026273727417,
+      "learning_rate": 8.133707066058655e-06,
+      "loss": 0.2194,
+      "step": 14183
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.361149311065674,
+      "learning_rate": 8.132870351001967e-06,
+      "loss": 0.5353,
+      "step": 14184
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 46.48427963256836,
+      "learning_rate": 8.132033635945279e-06,
+      "loss": 0.948,
+      "step": 14185
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.385955810546875,
+      "learning_rate": 8.131196920888593e-06,
+      "loss": 1.5825,
+      "step": 14186
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 90.32432556152344,
+      "learning_rate": 8.130360205831905e-06,
+      "loss": 0.8186,
+      "step": 14187
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 25.5366268157959,
+      "learning_rate": 8.129523490775218e-06,
+      "loss": 2.5684,
+      "step": 14188
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 43.078800201416016,
+      "learning_rate": 8.12868677571853e-06,
+      "loss": 1.8272,
+      "step": 14189
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 7.625771522521973,
+      "learning_rate": 8.127850060661842e-06,
+      "loss": 3.1224,
+      "step": 14190
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 21.902969360351562,
+      "learning_rate": 8.127013345605154e-06,
+      "loss": 0.4124,
+      "step": 14191
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 13.544557571411133,
+      "learning_rate": 8.126176630548466e-06,
+      "loss": 1.4524,
+      "step": 14192
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 62.5804557800293,
+      "learning_rate": 8.12533991549178e-06,
+      "loss": 2.0876,
+      "step": 14193
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.881214618682861,
+      "learning_rate": 8.124503200435092e-06,
+      "loss": 0.6644,
+      "step": 14194
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 13.090020179748535,
+      "learning_rate": 8.123666485378405e-06,
+      "loss": 0.4593,
+      "step": 14195
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 12.378315925598145,
+      "learning_rate": 8.122829770321717e-06,
+      "loss": 1.5352,
+      "step": 14196
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 15.287260055541992,
+      "learning_rate": 8.121993055265031e-06,
+      "loss": 1.8672,
+      "step": 14197
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 17.9168701171875,
+      "learning_rate": 8.121156340208343e-06,
+      "loss": 2.1657,
+      "step": 14198
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 179.5956573486328,
+      "learning_rate": 8.120319625151655e-06,
+      "loss": 0.6685,
+      "step": 14199
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 7.373868465423584,
+      "learning_rate": 8.119482910094968e-06,
+      "loss": 0.5316,
+      "step": 14200
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 18.164833068847656,
+      "learning_rate": 8.11864619503828e-06,
+      "loss": 1.935,
+      "step": 14201
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.35062026977539,
+      "learning_rate": 8.117809479981592e-06,
+      "loss": 0.9188,
+      "step": 14202
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.372532844543457,
+      "learning_rate": 8.116972764924906e-06,
+      "loss": 0.3663,
+      "step": 14203
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 15.599231719970703,
+      "learning_rate": 8.116136049868218e-06,
+      "loss": 0.9538,
+      "step": 14204
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 12.862131118774414,
+      "learning_rate": 8.11529933481153e-06,
+      "loss": 1.077,
+      "step": 14205
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.15827751159668,
+      "learning_rate": 8.114462619754842e-06,
+      "loss": 1.4036,
+      "step": 14206
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 36.890323638916016,
+      "learning_rate": 8.113625904698156e-06,
+      "loss": 0.778,
+      "step": 14207
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 14.244501113891602,
+      "learning_rate": 8.112789189641467e-06,
+      "loss": 0.4864,
+      "step": 14208
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.773188591003418,
+      "learning_rate": 8.111952474584781e-06,
+      "loss": 0.7433,
+      "step": 14209
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 13.432384490966797,
+      "learning_rate": 8.111115759528093e-06,
+      "loss": 1.1146,
+      "step": 14210
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.874054908752441,
+      "learning_rate": 8.110279044471407e-06,
+      "loss": 0.5435,
+      "step": 14211
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 4.427059650421143,
+      "learning_rate": 8.109442329414719e-06,
+      "loss": 0.1933,
+      "step": 14212
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.961760520935059,
+      "learning_rate": 8.10860561435803e-06,
+      "loss": 2.4765,
+      "step": 14213
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 11.333853721618652,
+      "learning_rate": 8.107768899301344e-06,
+      "loss": 0.7581,
+      "step": 14214
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 15.988113403320312,
+      "learning_rate": 8.106932184244656e-06,
+      "loss": 0.8706,
+      "step": 14215
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 11.896188735961914,
+      "learning_rate": 8.106095469187968e-06,
+      "loss": 0.7623,
+      "step": 14216
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 21.775907516479492,
+      "learning_rate": 8.105258754131282e-06,
+      "loss": 1.1781,
+      "step": 14217
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 20.36774444580078,
+      "learning_rate": 8.104422039074594e-06,
+      "loss": 1.7795,
+      "step": 14218
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.5056734085083,
+      "learning_rate": 8.103585324017906e-06,
+      "loss": 0.9233,
+      "step": 14219
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 130.77703857421875,
+      "learning_rate": 8.102748608961218e-06,
+      "loss": 2.7693,
+      "step": 14220
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 13.03814697265625,
+      "learning_rate": 8.101911893904531e-06,
+      "loss": 0.494,
+      "step": 14221
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 11.015776634216309,
+      "learning_rate": 8.101075178847843e-06,
+      "loss": 0.4717,
+      "step": 14222
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 21.49491310119629,
+      "learning_rate": 8.100238463791157e-06,
+      "loss": 2.0776,
+      "step": 14223
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 5.668808460235596,
+      "learning_rate": 8.099401748734469e-06,
+      "loss": 0.5059,
+      "step": 14224
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 6.428109169006348,
+      "learning_rate": 8.098565033677783e-06,
+      "loss": 1.5735,
+      "step": 14225
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 5.883421897888184,
+      "learning_rate": 8.097728318621095e-06,
+      "loss": 0.8716,
+      "step": 14226
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.014955997467041,
+      "learning_rate": 8.096891603564406e-06,
+      "loss": 0.6923,
+      "step": 14227
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 31.45018196105957,
+      "learning_rate": 8.09605488850772e-06,
+      "loss": 0.9072,
+      "step": 14228
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 51.637786865234375,
+      "learning_rate": 8.095218173451032e-06,
+      "loss": 3.0075,
+      "step": 14229
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 3.055673837661743,
+      "learning_rate": 8.094381458394344e-06,
+      "loss": 0.1078,
+      "step": 14230
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 6.919978141784668,
+      "learning_rate": 8.093544743337656e-06,
+      "loss": 0.4284,
+      "step": 14231
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 26.354515075683594,
+      "learning_rate": 8.09270802828097e-06,
+      "loss": 1.4861,
+      "step": 14232
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 38.69197463989258,
+      "learning_rate": 8.091871313224282e-06,
+      "loss": 4.0157,
+      "step": 14233
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 11.464272499084473,
+      "learning_rate": 8.091034598167594e-06,
+      "loss": 1.4697,
+      "step": 14234
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 4.368750095367432,
+      "learning_rate": 8.090197883110907e-06,
+      "loss": 0.1549,
+      "step": 14235
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 5.186683654785156,
+      "learning_rate": 8.08936116805422e-06,
+      "loss": 0.2221,
+      "step": 14236
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 12.998003005981445,
+      "learning_rate": 8.088524452997533e-06,
+      "loss": 1.4524,
+      "step": 14237
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 108.32467651367188,
+      "learning_rate": 8.087687737940845e-06,
+      "loss": 1.4167,
+      "step": 14238
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 65.82601928710938,
+      "learning_rate": 8.086851022884158e-06,
+      "loss": 1.1568,
+      "step": 14239
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.281325340270996,
+      "learning_rate": 8.08601430782747e-06,
+      "loss": 0.4164,
+      "step": 14240
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.825818061828613,
+      "learning_rate": 8.085177592770782e-06,
+      "loss": 0.3728,
+      "step": 14241
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 11.437962532043457,
+      "learning_rate": 8.084340877714096e-06,
+      "loss": 0.3512,
+      "step": 14242
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 18.751483917236328,
+      "learning_rate": 8.083504162657408e-06,
+      "loss": 2.0616,
+      "step": 14243
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 28.451984405517578,
+      "learning_rate": 8.08266744760072e-06,
+      "loss": 1.5825,
+      "step": 14244
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 4.15352725982666,
+      "learning_rate": 8.081830732544032e-06,
+      "loss": 0.2291,
+      "step": 14245
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 47.6756591796875,
+      "learning_rate": 8.080994017487345e-06,
+      "loss": 1.9295,
+      "step": 14246
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 15.957259178161621,
+      "learning_rate": 8.080157302430657e-06,
+      "loss": 1.4346,
+      "step": 14247
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.788030624389648,
+      "learning_rate": 8.07932058737397e-06,
+      "loss": 0.4594,
+      "step": 14248
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 12.14069938659668,
+      "learning_rate": 8.078483872317283e-06,
+      "loss": 2.1253,
+      "step": 14249
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.096846580505371,
+      "learning_rate": 8.077647157260595e-06,
+      "loss": 0.5818,
+      "step": 14250
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 24.04420280456543,
+      "learning_rate": 8.076810442203909e-06,
+      "loss": 2.1495,
+      "step": 14251
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.906283378601074,
+      "learning_rate": 8.07597372714722e-06,
+      "loss": 1.853,
+      "step": 14252
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 6.8881516456604,
+      "learning_rate": 8.075137012090534e-06,
+      "loss": 0.4135,
+      "step": 14253
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 33.854827880859375,
+      "learning_rate": 8.074300297033846e-06,
+      "loss": 1.7228,
+      "step": 14254
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 12.448065757751465,
+      "learning_rate": 8.073463581977158e-06,
+      "loss": 1.63,
+      "step": 14255
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.010289192199707,
+      "learning_rate": 8.072626866920472e-06,
+      "loss": 0.3541,
+      "step": 14256
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 18.160022735595703,
+      "learning_rate": 8.071790151863784e-06,
+      "loss": 1.7952,
+      "step": 14257
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 27.748252868652344,
+      "learning_rate": 8.070953436807096e-06,
+      "loss": 0.6974,
+      "step": 14258
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 17.5703125,
+      "learning_rate": 8.070116721750408e-06,
+      "loss": 0.9125,
+      "step": 14259
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 18.15052032470703,
+      "learning_rate": 8.069280006693721e-06,
+      "loss": 0.5414,
+      "step": 14260
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.297247886657715,
+      "learning_rate": 8.068443291637033e-06,
+      "loss": 2.8713,
+      "step": 14261
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.959202766418457,
+      "learning_rate": 8.067606576580345e-06,
+      "loss": 0.9516,
+      "step": 14262
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 15.244864463806152,
+      "learning_rate": 8.066769861523659e-06,
+      "loss": 0.725,
+      "step": 14263
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 13.854104042053223,
+      "learning_rate": 8.06593314646697e-06,
+      "loss": 0.5473,
+      "step": 14264
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 113.05213165283203,
+      "learning_rate": 8.065096431410284e-06,
+      "loss": 1.9104,
+      "step": 14265
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.714113235473633,
+      "learning_rate": 8.064259716353596e-06,
+      "loss": 1.2544,
+      "step": 14266
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.027963638305664,
+      "learning_rate": 8.06342300129691e-06,
+      "loss": 1.4344,
+      "step": 14267
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 13.1775541305542,
+      "learning_rate": 8.062586286240222e-06,
+      "loss": 1.0498,
+      "step": 14268
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.681795120239258,
+      "learning_rate": 8.061749571183534e-06,
+      "loss": 0.1476,
+      "step": 14269
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 37.30206298828125,
+      "learning_rate": 8.060912856126848e-06,
+      "loss": 2.6026,
+      "step": 14270
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 18.547700881958008,
+      "learning_rate": 8.06007614107016e-06,
+      "loss": 1.3664,
+      "step": 14271
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 28.999473571777344,
+      "learning_rate": 8.059239426013472e-06,
+      "loss": 3.6555,
+      "step": 14272
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 14.13399600982666,
+      "learning_rate": 8.058402710956784e-06,
+      "loss": 0.6258,
+      "step": 14273
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 19.898513793945312,
+      "learning_rate": 8.057565995900097e-06,
+      "loss": 2.7374,
+      "step": 14274
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 22.894289016723633,
+      "learning_rate": 8.056729280843409e-06,
+      "loss": 1.3485,
+      "step": 14275
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.624289512634277,
+      "learning_rate": 8.055892565786721e-06,
+      "loss": 0.9094,
+      "step": 14276
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 14.443469047546387,
+      "learning_rate": 8.055055850730035e-06,
+      "loss": 0.8546,
+      "step": 14277
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 18.679956436157227,
+      "learning_rate": 8.054219135673347e-06,
+      "loss": 1.6593,
+      "step": 14278
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 13.102360725402832,
+      "learning_rate": 8.05338242061666e-06,
+      "loss": 0.5293,
+      "step": 14279
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 9.42329216003418,
+      "learning_rate": 8.052545705559972e-06,
+      "loss": 1.0723,
+      "step": 14280
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 6.173603057861328,
+      "learning_rate": 8.051708990503286e-06,
+      "loss": 0.0611,
+      "step": 14281
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 13.688335418701172,
+      "learning_rate": 8.050872275446598e-06,
+      "loss": 0.5864,
+      "step": 14282
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 11.294967651367188,
+      "learning_rate": 8.05003556038991e-06,
+      "loss": 1.3249,
+      "step": 14283
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 11.166751861572266,
+      "learning_rate": 8.049198845333222e-06,
+      "loss": 1.4536,
+      "step": 14284
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 19.163232803344727,
+      "learning_rate": 8.048362130276535e-06,
+      "loss": 1.6504,
+      "step": 14285
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 13.887179374694824,
+      "learning_rate": 8.047525415219847e-06,
+      "loss": 1.0023,
+      "step": 14286
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 13.513861656188965,
+      "learning_rate": 8.04668870016316e-06,
+      "loss": 2.4072,
+      "step": 14287
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 38.558311462402344,
+      "learning_rate": 8.045851985106473e-06,
+      "loss": 1.6424,
+      "step": 14288
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.030594825744629,
+      "learning_rate": 8.045015270049785e-06,
+      "loss": 0.9199,
+      "step": 14289
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 18.584436416625977,
+      "learning_rate": 8.044178554993097e-06,
+      "loss": 1.5709,
+      "step": 14290
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.418522357940674,
+      "learning_rate": 8.04334183993641e-06,
+      "loss": 0.5635,
+      "step": 14291
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 19.099185943603516,
+      "learning_rate": 8.042505124879723e-06,
+      "loss": 0.1956,
+      "step": 14292
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.259555339813232,
+      "learning_rate": 8.041668409823036e-06,
+      "loss": 0.8623,
+      "step": 14293
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 7.537904739379883,
+      "learning_rate": 8.040831694766348e-06,
+      "loss": 0.4155,
+      "step": 14294
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 5.026714324951172,
+      "learning_rate": 8.039994979709662e-06,
+      "loss": 0.2345,
+      "step": 14295
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 15.554062843322754,
+      "learning_rate": 8.039158264652974e-06,
+      "loss": 1.1563,
+      "step": 14296
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 17.680042266845703,
+      "learning_rate": 8.038321549596286e-06,
+      "loss": 1.4585,
+      "step": 14297
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 32.36421203613281,
+      "learning_rate": 8.037484834539598e-06,
+      "loss": 1.1143,
+      "step": 14298
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 87.64347839355469,
+      "learning_rate": 8.036648119482911e-06,
+      "loss": 2.7058,
+      "step": 14299
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 11.312725067138672,
+      "learning_rate": 8.035811404426223e-06,
+      "loss": 0.7953,
+      "step": 14300
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.867121696472168,
+      "learning_rate": 8.034974689369535e-06,
+      "loss": 0.476,
+      "step": 14301
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 10.188661575317383,
+      "learning_rate": 8.034137974312849e-06,
+      "loss": 0.5231,
+      "step": 14302
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 25.78326988220215,
+      "learning_rate": 8.03330125925616e-06,
+      "loss": 2.0026,
+      "step": 14303
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 3.5009584426879883,
+      "learning_rate": 8.032464544199473e-06,
+      "loss": 0.2596,
+      "step": 14304
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 11.701454162597656,
+      "learning_rate": 8.031627829142786e-06,
+      "loss": 1.0258,
+      "step": 14305
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 33.025184631347656,
+      "learning_rate": 8.030791114086098e-06,
+      "loss": 1.0877,
+      "step": 14306
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 6.063713550567627,
+      "learning_rate": 8.029954399029412e-06,
+      "loss": 0.3886,
+      "step": 14307
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.470044136047363,
+      "learning_rate": 8.029117683972724e-06,
+      "loss": 0.4611,
+      "step": 14308
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 20.49445343017578,
+      "learning_rate": 8.028280968916038e-06,
+      "loss": 1.836,
+      "step": 14309
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 13.73381233215332,
+      "learning_rate": 8.02744425385935e-06,
+      "loss": 0.8764,
+      "step": 14310
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 20.117389678955078,
+      "learning_rate": 8.026607538802662e-06,
+      "loss": 1.4659,
+      "step": 14311
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 12.779775619506836,
+      "learning_rate": 8.025770823745973e-06,
+      "loss": 0.4024,
+      "step": 14312
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.237881660461426,
+      "learning_rate": 8.024934108689285e-06,
+      "loss": 2.0096,
+      "step": 14313
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.170612335205078,
+      "learning_rate": 8.024097393632599e-06,
+      "loss": 1.0357,
+      "step": 14314
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 15.791288375854492,
+      "learning_rate": 8.023260678575911e-06,
+      "loss": 0.7219,
+      "step": 14315
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 12.148241996765137,
+      "learning_rate": 8.022423963519225e-06,
+      "loss": 1.5481,
+      "step": 14316
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 6.0537261962890625,
+      "learning_rate": 8.021587248462537e-06,
+      "loss": 1.1468,
+      "step": 14317
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.948807716369629,
+      "learning_rate": 8.020750533405849e-06,
+      "loss": 0.6462,
+      "step": 14318
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.998846054077148,
+      "learning_rate": 8.019913818349162e-06,
+      "loss": 1.608,
+      "step": 14319
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 7.889427661895752,
+      "learning_rate": 8.019077103292474e-06,
+      "loss": 0.7336,
+      "step": 14320
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 20.464832305908203,
+      "learning_rate": 8.018240388235788e-06,
+      "loss": 1.2012,
+      "step": 14321
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.512769937515259,
+      "learning_rate": 8.0174036731791e-06,
+      "loss": 0.1633,
+      "step": 14322
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.887861251831055,
+      "learning_rate": 8.016566958122412e-06,
+      "loss": 1.2358,
+      "step": 14323
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 13.0720796585083,
+      "learning_rate": 8.015730243065725e-06,
+      "loss": 0.688,
+      "step": 14324
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 14.510822296142578,
+      "learning_rate": 8.014893528009037e-06,
+      "loss": 2.0057,
+      "step": 14325
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.325578689575195,
+      "learning_rate": 8.01405681295235e-06,
+      "loss": 0.8386,
+      "step": 14326
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.9541542530059814,
+      "learning_rate": 8.013220097895661e-06,
+      "loss": 0.1813,
+      "step": 14327
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 13.947324752807617,
+      "learning_rate": 8.012383382838975e-06,
+      "loss": 0.8859,
+      "step": 14328
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 7.45787239074707,
+      "learning_rate": 8.011546667782287e-06,
+      "loss": 2.1223,
+      "step": 14329
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 20.63152503967285,
+      "learning_rate": 8.0107099527256e-06,
+      "loss": 0.9686,
+      "step": 14330
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 14.866303443908691,
+      "learning_rate": 8.009873237668912e-06,
+      "loss": 2.4096,
+      "step": 14331
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.556191444396973,
+      "learning_rate": 8.009036522612224e-06,
+      "loss": 1.2394,
+      "step": 14332
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 11.858455657958984,
+      "learning_rate": 8.008199807555538e-06,
+      "loss": 1.675,
+      "step": 14333
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 7.185389041900635,
+      "learning_rate": 8.00736309249885e-06,
+      "loss": 2.8998,
+      "step": 14334
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.578290939331055,
+      "learning_rate": 8.006526377442164e-06,
+      "loss": 1.0241,
+      "step": 14335
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 13.962796211242676,
+      "learning_rate": 8.005689662385476e-06,
+      "loss": 1.2446,
+      "step": 14336
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 11.544517517089844,
+      "learning_rate": 8.004852947328788e-06,
+      "loss": 2.1161,
+      "step": 14337
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 7.659705638885498,
+      "learning_rate": 8.004016232272101e-06,
+      "loss": 1.0362,
+      "step": 14338
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.644447326660156,
+      "learning_rate": 8.003179517215413e-06,
+      "loss": 0.4681,
+      "step": 14339
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 28.88718032836914,
+      "learning_rate": 8.002342802158725e-06,
+      "loss": 2.6071,
+      "step": 14340
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 13.294441223144531,
+      "learning_rate": 8.001506087102037e-06,
+      "loss": 1.8365,
+      "step": 14341
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 17.310266494750977,
+      "learning_rate": 8.00066937204535e-06,
+      "loss": 0.9588,
+      "step": 14342
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 4.964139938354492,
+      "learning_rate": 7.999832656988663e-06,
+      "loss": 1.5954,
+      "step": 14343
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.747537612915039,
+      "learning_rate": 7.998995941931976e-06,
+      "loss": 2.5247,
+      "step": 14344
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.901177406311035,
+      "learning_rate": 7.998159226875288e-06,
+      "loss": 0.1717,
+      "step": 14345
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 6.909663200378418,
+      "learning_rate": 7.9973225118186e-06,
+      "loss": 0.453,
+      "step": 14346
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 14.984963417053223,
+      "learning_rate": 7.996485796761914e-06,
+      "loss": 0.9857,
+      "step": 14347
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.209060192108154,
+      "learning_rate": 7.995649081705226e-06,
+      "loss": 1.797,
+      "step": 14348
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 61.35810852050781,
+      "learning_rate": 7.99481236664854e-06,
+      "loss": 2.0787,
+      "step": 14349
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.355850219726562,
+      "learning_rate": 7.993975651591851e-06,
+      "loss": 0.2956,
+      "step": 14350
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.559131622314453,
+      "learning_rate": 7.993138936535163e-06,
+      "loss": 0.6137,
+      "step": 14351
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 25.988943099975586,
+      "learning_rate": 7.992302221478477e-06,
+      "loss": 1.1929,
+      "step": 14352
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 29.633085250854492,
+      "learning_rate": 7.991465506421789e-06,
+      "loss": 1.7604,
+      "step": 14353
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 13.114699363708496,
+      "learning_rate": 7.990628791365101e-06,
+      "loss": 1.2965,
+      "step": 14354
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.263970375061035,
+      "learning_rate": 7.989792076308413e-06,
+      "loss": 0.4648,
+      "step": 14355
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 13.104530334472656,
+      "learning_rate": 7.988955361251727e-06,
+      "loss": 1.2807,
+      "step": 14356
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.3938703536987305,
+      "learning_rate": 7.988118646195039e-06,
+      "loss": 1.6159,
+      "step": 14357
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 40.70235824584961,
+      "learning_rate": 7.987281931138352e-06,
+      "loss": 1.9411,
+      "step": 14358
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 23.1292667388916,
+      "learning_rate": 7.986445216081664e-06,
+      "loss": 1.7476,
+      "step": 14359
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 27.007932662963867,
+      "learning_rate": 7.985608501024976e-06,
+      "loss": 1.8868,
+      "step": 14360
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 14.89322566986084,
+      "learning_rate": 7.98477178596829e-06,
+      "loss": 0.9967,
+      "step": 14361
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 28.900524139404297,
+      "learning_rate": 7.983935070911602e-06,
+      "loss": 2.4159,
+      "step": 14362
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 6.286021709442139,
+      "learning_rate": 7.983098355854915e-06,
+      "loss": 0.9095,
+      "step": 14363
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 11.280611991882324,
+      "learning_rate": 7.982261640798227e-06,
+      "loss": 0.294,
+      "step": 14364
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 14.967876434326172,
+      "learning_rate": 7.98142492574154e-06,
+      "loss": 1.035,
+      "step": 14365
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 84.54759979248047,
+      "learning_rate": 7.980588210684851e-06,
+      "loss": 3.8704,
+      "step": 14366
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 79.22369384765625,
+      "learning_rate": 7.979751495628165e-06,
+      "loss": 1.0864,
+      "step": 14367
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.267379760742188,
+      "learning_rate": 7.978914780571477e-06,
+      "loss": 1.2849,
+      "step": 14368
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 18.904296875,
+      "learning_rate": 7.978078065514789e-06,
+      "loss": 0.6259,
+      "step": 14369
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 16.8665771484375,
+      "learning_rate": 7.977241350458102e-06,
+      "loss": 1.8917,
+      "step": 14370
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 31.569623947143555,
+      "learning_rate": 7.976404635401414e-06,
+      "loss": 1.4306,
+      "step": 14371
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 58.07877731323242,
+      "learning_rate": 7.975567920344726e-06,
+      "loss": 3.0194,
+      "step": 14372
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 34.71226501464844,
+      "learning_rate": 7.97473120528804e-06,
+      "loss": 1.6203,
+      "step": 14373
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 10.487889289855957,
+      "learning_rate": 7.973894490231352e-06,
+      "loss": 0.83,
+      "step": 14374
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 13.767682075500488,
+      "learning_rate": 7.973057775174666e-06,
+      "loss": 0.6044,
+      "step": 14375
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 14.913378715515137,
+      "learning_rate": 7.972221060117978e-06,
+      "loss": 1.2755,
+      "step": 14376
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 9.47493839263916,
+      "learning_rate": 7.971384345061291e-06,
+      "loss": 1.0999,
+      "step": 14377
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 13.59195613861084,
+      "learning_rate": 7.970547630004603e-06,
+      "loss": 2.0894,
+      "step": 14378
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 14.368037223815918,
+      "learning_rate": 7.969710914947915e-06,
+      "loss": 1.3127,
+      "step": 14379
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 36.372535705566406,
+      "learning_rate": 7.968874199891227e-06,
+      "loss": 2.1036,
+      "step": 14380
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 17.263235092163086,
+      "learning_rate": 7.96803748483454e-06,
+      "loss": 1.8845,
+      "step": 14381
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 32.459693908691406,
+      "learning_rate": 7.967200769777853e-06,
+      "loss": 1.5034,
+      "step": 14382
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 4.841275215148926,
+      "learning_rate": 7.966364054721165e-06,
+      "loss": 1.1678,
+      "step": 14383
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 5.018244743347168,
+      "learning_rate": 7.965527339664478e-06,
+      "loss": 0.2264,
+      "step": 14384
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.65173053741455,
+      "learning_rate": 7.96469062460779e-06,
+      "loss": 0.8554,
+      "step": 14385
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.671903610229492,
+      "learning_rate": 7.963853909551102e-06,
+      "loss": 1.9722,
+      "step": 14386
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 15.357810020446777,
+      "learning_rate": 7.963017194494416e-06,
+      "loss": 0.979,
+      "step": 14387
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.76091480255127,
+      "learning_rate": 7.962180479437728e-06,
+      "loss": 1.3369,
+      "step": 14388
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.63212776184082,
+      "learning_rate": 7.961343764381041e-06,
+      "loss": 0.4692,
+      "step": 14389
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 21.49897575378418,
+      "learning_rate": 7.960507049324353e-06,
+      "loss": 0.9434,
+      "step": 14390
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 6.799454689025879,
+      "learning_rate": 7.959670334267667e-06,
+      "loss": 0.3093,
+      "step": 14391
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 30.406591415405273,
+      "learning_rate": 7.958833619210979e-06,
+      "loss": 1.7786,
+      "step": 14392
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 14.645602226257324,
+      "learning_rate": 7.957996904154291e-06,
+      "loss": 1.3068,
+      "step": 14393
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 5.332977771759033,
+      "learning_rate": 7.957160189097603e-06,
+      "loss": 1.8079,
+      "step": 14394
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 13.62096881866455,
+      "learning_rate": 7.956323474040915e-06,
+      "loss": 1.9015,
+      "step": 14395
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 15.009598731994629,
+      "learning_rate": 7.955486758984228e-06,
+      "loss": 1.5518,
+      "step": 14396
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 21.20696449279785,
+      "learning_rate": 7.95465004392754e-06,
+      "loss": 1.5061,
+      "step": 14397
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 21.663663864135742,
+      "learning_rate": 7.953813328870854e-06,
+      "loss": 1.268,
+      "step": 14398
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 16.903648376464844,
+      "learning_rate": 7.952976613814166e-06,
+      "loss": 2.0583,
+      "step": 14399
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 60.051414489746094,
+      "learning_rate": 7.952139898757478e-06,
+      "loss": 0.7644,
+      "step": 14400
+    },
+    {
+      "epoch": 1.81,
+      "eval_loss": 0.07709412276744843,
+      "eval_runtime": 98.0422,
+      "eval_samples_per_second": 36.127,
+      "eval_steps_per_second": 36.127,
+      "step": 14400
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 10.575669288635254,
+      "learning_rate": 7.951303183700792e-06,
+      "loss": 3.3803,
+      "step": 14401
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 15.85393238067627,
+      "learning_rate": 7.950466468644104e-06,
+      "loss": 0.9647,
+      "step": 14402
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 6.105818748474121,
+      "learning_rate": 7.949629753587417e-06,
+      "loss": 1.2717,
+      "step": 14403
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 21.075420379638672,
+      "learning_rate": 7.94879303853073e-06,
+      "loss": 2.2108,
+      "step": 14404
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 7.1770453453063965,
+      "learning_rate": 7.947956323474041e-06,
+      "loss": 1.3509,
+      "step": 14405
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 12.135576248168945,
+      "learning_rate": 7.947119608417355e-06,
+      "loss": 2.5509,
+      "step": 14406
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.19250774383545,
+      "learning_rate": 7.946282893360667e-06,
+      "loss": 1.0113,
+      "step": 14407
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.2168388366699219,
+      "learning_rate": 7.945446178303979e-06,
+      "loss": 0.0406,
+      "step": 14408
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 34.41596984863281,
+      "learning_rate": 7.94460946324729e-06,
+      "loss": 3.8037,
+      "step": 14409
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 14.644426345825195,
+      "learning_rate": 7.943772748190604e-06,
+      "loss": 1.299,
+      "step": 14410
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 25.0706844329834,
+      "learning_rate": 7.942936033133916e-06,
+      "loss": 1.041,
+      "step": 14411
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.17798900604248,
+      "learning_rate": 7.94209931807723e-06,
+      "loss": 0.5255,
+      "step": 14412
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 20.191364288330078,
+      "learning_rate": 7.941262603020542e-06,
+      "loss": 0.7884,
+      "step": 14413
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.380793571472168,
+      "learning_rate": 7.940425887963854e-06,
+      "loss": 0.6298,
+      "step": 14414
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 5.755029678344727,
+      "learning_rate": 7.939589172907167e-06,
+      "loss": 0.4435,
+      "step": 14415
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.83551025390625,
+      "learning_rate": 7.93875245785048e-06,
+      "loss": 0.4982,
+      "step": 14416
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.85175609588623,
+      "learning_rate": 7.937915742793793e-06,
+      "loss": 0.5442,
+      "step": 14417
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 13.807879447937012,
+      "learning_rate": 7.937079027737105e-06,
+      "loss": 1.4413,
+      "step": 14418
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 12.331297874450684,
+      "learning_rate": 7.936242312680417e-06,
+      "loss": 0.4959,
+      "step": 14419
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 208.26344299316406,
+      "learning_rate": 7.93540559762373e-06,
+      "loss": 0.5404,
+      "step": 14420
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 23.06793212890625,
+      "learning_rate": 7.934568882567043e-06,
+      "loss": 1.0801,
+      "step": 14421
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.011923789978027,
+      "learning_rate": 7.933732167510355e-06,
+      "loss": 1.3406,
+      "step": 14422
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.207767486572266,
+      "learning_rate": 7.932895452453667e-06,
+      "loss": 0.8914,
+      "step": 14423
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.7527880668640137,
+      "learning_rate": 7.93205873739698e-06,
+      "loss": 0.029,
+      "step": 14424
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.621235847473145,
+      "learning_rate": 7.931222022340292e-06,
+      "loss": 0.2108,
+      "step": 14425
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.2488431930542,
+      "learning_rate": 7.930385307283606e-06,
+      "loss": 0.9488,
+      "step": 14426
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 54.50273132324219,
+      "learning_rate": 7.929548592226918e-06,
+      "loss": 3.4363,
+      "step": 14427
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 527.145263671875,
+      "learning_rate": 7.92871187717023e-06,
+      "loss": 1.9903,
+      "step": 14428
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 17.940338134765625,
+      "learning_rate": 7.927875162113543e-06,
+      "loss": 2.1123,
+      "step": 14429
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 10.652819633483887,
+      "learning_rate": 7.927038447056855e-06,
+      "loss": 0.8332,
+      "step": 14430
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.21216106414795,
+      "learning_rate": 7.926201732000169e-06,
+      "loss": 0.2088,
+      "step": 14431
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 71.3212661743164,
+      "learning_rate": 7.925365016943481e-06,
+      "loss": 2.5172,
+      "step": 14432
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.186037063598633,
+      "learning_rate": 7.924528301886793e-06,
+      "loss": 0.3602,
+      "step": 14433
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 29.011112213134766,
+      "learning_rate": 7.923691586830105e-06,
+      "loss": 1.2023,
+      "step": 14434
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 18.202524185180664,
+      "learning_rate": 7.922854871773418e-06,
+      "loss": 1.8014,
+      "step": 14435
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 20.9152774810791,
+      "learning_rate": 7.92201815671673e-06,
+      "loss": 1.1115,
+      "step": 14436
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.133270263671875,
+      "learning_rate": 7.921181441660042e-06,
+      "loss": 0.6653,
+      "step": 14437
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.92380428314209,
+      "learning_rate": 7.920344726603356e-06,
+      "loss": 0.3128,
+      "step": 14438
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 13.854774475097656,
+      "learning_rate": 7.919508011546668e-06,
+      "loss": 1.9934,
+      "step": 14439
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 18.137493133544922,
+      "learning_rate": 7.918671296489982e-06,
+      "loss": 0.723,
+      "step": 14440
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 13.559948921203613,
+      "learning_rate": 7.917834581433294e-06,
+      "loss": 1.102,
+      "step": 14441
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 6.407212257385254,
+      "learning_rate": 7.916997866376606e-06,
+      "loss": 0.3254,
+      "step": 14442
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 10.264758110046387,
+      "learning_rate": 7.91616115131992e-06,
+      "loss": 1.9795,
+      "step": 14443
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 71.19867706298828,
+      "learning_rate": 7.915324436263231e-06,
+      "loss": 0.6712,
+      "step": 14444
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.976835250854492,
+      "learning_rate": 7.914487721206545e-06,
+      "loss": 0.8841,
+      "step": 14445
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 19.097244262695312,
+      "learning_rate": 7.913651006149857e-06,
+      "loss": 1.1689,
+      "step": 14446
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 5.357875823974609,
+      "learning_rate": 7.912814291093169e-06,
+      "loss": 0.1382,
+      "step": 14447
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 42.744598388671875,
+      "learning_rate": 7.91197757603648e-06,
+      "loss": 2.3061,
+      "step": 14448
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 16.11562728881836,
+      "learning_rate": 7.911140860979794e-06,
+      "loss": 0.6837,
+      "step": 14449
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 75.79932403564453,
+      "learning_rate": 7.910304145923106e-06,
+      "loss": 0.8448,
+      "step": 14450
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 11.990781784057617,
+      "learning_rate": 7.909467430866418e-06,
+      "loss": 0.762,
+      "step": 14451
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 75.82522583007812,
+      "learning_rate": 7.908630715809732e-06,
+      "loss": 1.7107,
+      "step": 14452
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 42.937198638916016,
+      "learning_rate": 7.907794000753044e-06,
+      "loss": 1.0458,
+      "step": 14453
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 5.65687370300293,
+      "learning_rate": 7.906957285696357e-06,
+      "loss": 0.1425,
+      "step": 14454
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 16.94680404663086,
+      "learning_rate": 7.90612057063967e-06,
+      "loss": 0.5582,
+      "step": 14455
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 21.313919067382812,
+      "learning_rate": 7.905283855582981e-06,
+      "loss": 1.1394,
+      "step": 14456
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 31.395153045654297,
+      "learning_rate": 7.904447140526295e-06,
+      "loss": 1.1894,
+      "step": 14457
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 14.642378807067871,
+      "learning_rate": 7.903610425469607e-06,
+      "loss": 0.9938,
+      "step": 14458
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 6.3422017097473145,
+      "learning_rate": 7.90277371041292e-06,
+      "loss": 0.7284,
+      "step": 14459
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.563097953796387,
+      "learning_rate": 7.901936995356233e-06,
+      "loss": 0.4289,
+      "step": 14460
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 31.980749130249023,
+      "learning_rate": 7.901100280299545e-06,
+      "loss": 2.886,
+      "step": 14461
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 16.99601936340332,
+      "learning_rate": 7.900263565242856e-06,
+      "loss": 0.4885,
+      "step": 14462
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 21.302080154418945,
+      "learning_rate": 7.89942685018617e-06,
+      "loss": 0.6471,
+      "step": 14463
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 15.44736099243164,
+      "learning_rate": 7.898590135129482e-06,
+      "loss": 1.4289,
+      "step": 14464
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 30.569433212280273,
+      "learning_rate": 7.897753420072794e-06,
+      "loss": 0.611,
+      "step": 14465
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 22.321868896484375,
+      "learning_rate": 7.896916705016108e-06,
+      "loss": 1.15,
+      "step": 14466
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 10.689000129699707,
+      "learning_rate": 7.89607998995942e-06,
+      "loss": 0.9393,
+      "step": 14467
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 112.66310119628906,
+      "learning_rate": 7.895243274902733e-06,
+      "loss": 1.1678,
+      "step": 14468
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 53.48061752319336,
+      "learning_rate": 7.894406559846045e-06,
+      "loss": 2.8504,
+      "step": 14469
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 17.995603561401367,
+      "learning_rate": 7.893569844789357e-06,
+      "loss": 2.6686,
+      "step": 14470
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.140344619750977,
+      "learning_rate": 7.89273312973267e-06,
+      "loss": 0.9577,
+      "step": 14471
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 12.35569953918457,
+      "learning_rate": 7.891896414675983e-06,
+      "loss": 1.2718,
+      "step": 14472
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 18.828767776489258,
+      "learning_rate": 7.891059699619296e-06,
+      "loss": 3.5721,
+      "step": 14473
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.44033432006836,
+      "learning_rate": 7.890222984562608e-06,
+      "loss": 0.3851,
+      "step": 14474
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.504138946533203,
+      "learning_rate": 7.88938626950592e-06,
+      "loss": 0.5076,
+      "step": 14475
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.0276408195495605,
+      "learning_rate": 7.888549554449232e-06,
+      "loss": 0.4118,
+      "step": 14476
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 11.683658599853516,
+      "learning_rate": 7.887712839392544e-06,
+      "loss": 1.0484,
+      "step": 14477
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 89.66896057128906,
+      "learning_rate": 7.886876124335858e-06,
+      "loss": 1.621,
+      "step": 14478
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 12.588370323181152,
+      "learning_rate": 7.88603940927917e-06,
+      "loss": 0.4962,
+      "step": 14479
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 17.480802536010742,
+      "learning_rate": 7.885202694222484e-06,
+      "loss": 0.9598,
+      "step": 14480
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 20.175277709960938,
+      "learning_rate": 7.884365979165795e-06,
+      "loss": 1.4753,
+      "step": 14481
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.6198923587799072,
+      "learning_rate": 7.883529264109109e-06,
+      "loss": 0.0245,
+      "step": 14482
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 10.016743659973145,
+      "learning_rate": 7.882692549052421e-06,
+      "loss": 0.7644,
+      "step": 14483
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.49308443069458,
+      "learning_rate": 7.881855833995733e-06,
+      "loss": 0.7991,
+      "step": 14484
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 13.122968673706055,
+      "learning_rate": 7.881019118939047e-06,
+      "loss": 1.0511,
+      "step": 14485
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.601958274841309,
+      "learning_rate": 7.880182403882359e-06,
+      "loss": 0.6712,
+      "step": 14486
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 13.161881446838379,
+      "learning_rate": 7.87934568882567e-06,
+      "loss": 0.1473,
+      "step": 14487
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 14.766838073730469,
+      "learning_rate": 7.878508973768984e-06,
+      "loss": 0.6535,
+      "step": 14488
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 35.4913330078125,
+      "learning_rate": 7.877672258712296e-06,
+      "loss": 0.6545,
+      "step": 14489
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 35.55754852294922,
+      "learning_rate": 7.876835543655608e-06,
+      "loss": 2.9493,
+      "step": 14490
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 19.41988754272461,
+      "learning_rate": 7.87599882859892e-06,
+      "loss": 1.2309,
+      "step": 14491
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 13.2708740234375,
+      "learning_rate": 7.875162113542234e-06,
+      "loss": 0.794,
+      "step": 14492
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 22.65144157409668,
+      "learning_rate": 7.874325398485546e-06,
+      "loss": 0.7917,
+      "step": 14493
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 47.246463775634766,
+      "learning_rate": 7.87348868342886e-06,
+      "loss": 2.1365,
+      "step": 14494
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 10.207687377929688,
+      "learning_rate": 7.872651968372171e-06,
+      "loss": 1.2046,
+      "step": 14495
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 27.179967880249023,
+      "learning_rate": 7.871815253315485e-06,
+      "loss": 2.4935,
+      "step": 14496
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 15.798068046569824,
+      "learning_rate": 7.870978538258797e-06,
+      "loss": 0.8742,
+      "step": 14497
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 14.868792533874512,
+      "learning_rate": 7.870141823202109e-06,
+      "loss": 1.1622,
+      "step": 14498
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 23.81890296936035,
+      "learning_rate": 7.869305108145423e-06,
+      "loss": 2.0811,
+      "step": 14499
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 10.45630168914795,
+      "learning_rate": 7.868468393088734e-06,
+      "loss": 0.8939,
+      "step": 14500
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.123615264892578,
+      "learning_rate": 7.867631678032046e-06,
+      "loss": 1.2297,
+      "step": 14501
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 13.614934921264648,
+      "learning_rate": 7.86679496297536e-06,
+      "loss": 1.1254,
+      "step": 14502
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 15.482418060302734,
+      "learning_rate": 7.865958247918672e-06,
+      "loss": 1.7316,
+      "step": 14503
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.831905364990234,
+      "learning_rate": 7.865121532861984e-06,
+      "loss": 1.8212,
+      "step": 14504
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 6.973489284515381,
+      "learning_rate": 7.864284817805296e-06,
+      "loss": 0.3571,
+      "step": 14505
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 18.5034236907959,
+      "learning_rate": 7.86344810274861e-06,
+      "loss": 2.5581,
+      "step": 14506
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 5.724293231964111,
+      "learning_rate": 7.862611387691922e-06,
+      "loss": 0.198,
+      "step": 14507
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 16.614166259765625,
+      "learning_rate": 7.861774672635235e-06,
+      "loss": 1.0352,
+      "step": 14508
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 44.372352600097656,
+      "learning_rate": 7.860937957578547e-06,
+      "loss": 2.7569,
+      "step": 14509
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 46.56011962890625,
+      "learning_rate": 7.86010124252186e-06,
+      "loss": 1.5423,
+      "step": 14510
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.329614639282227,
+      "learning_rate": 7.859264527465173e-06,
+      "loss": 1.6975,
+      "step": 14511
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 11.13457202911377,
+      "learning_rate": 7.858427812408485e-06,
+      "loss": 1.9166,
+      "step": 14512
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 24.75688934326172,
+      "learning_rate": 7.857591097351798e-06,
+      "loss": 0.908,
+      "step": 14513
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 9.076860427856445,
+      "learning_rate": 7.85675438229511e-06,
+      "loss": 1.9261,
+      "step": 14514
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 6.307610034942627,
+      "learning_rate": 7.855917667238422e-06,
+      "loss": 0.2568,
+      "step": 14515
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.085229873657227,
+      "learning_rate": 7.855080952181734e-06,
+      "loss": 0.8695,
+      "step": 14516
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 21.385927200317383,
+      "learning_rate": 7.854244237125048e-06,
+      "loss": 1.48,
+      "step": 14517
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 17.968355178833008,
+      "learning_rate": 7.85340752206836e-06,
+      "loss": 0.8138,
+      "step": 14518
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.746437072753906,
+      "learning_rate": 7.852570807011672e-06,
+      "loss": 0.8136,
+      "step": 14519
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 15.064603805541992,
+      "learning_rate": 7.851734091954985e-06,
+      "loss": 0.8686,
+      "step": 14520
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 11.113859176635742,
+      "learning_rate": 7.850897376898297e-06,
+      "loss": 0.6047,
+      "step": 14521
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 78.57706451416016,
+      "learning_rate": 7.850060661841611e-06,
+      "loss": 0.8011,
+      "step": 14522
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 13.151373863220215,
+      "learning_rate": 7.849223946784923e-06,
+      "loss": 0.7002,
+      "step": 14523
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 12.312122344970703,
+      "learning_rate": 7.848387231728237e-06,
+      "loss": 0.9362,
+      "step": 14524
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 17.823299407958984,
+      "learning_rate": 7.847550516671549e-06,
+      "loss": 0.4392,
+      "step": 14525
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 27.160539627075195,
+      "learning_rate": 7.84671380161486e-06,
+      "loss": 1.3058,
+      "step": 14526
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 4.870408535003662,
+      "learning_rate": 7.845877086558174e-06,
+      "loss": 0.2979,
+      "step": 14527
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 42.98374557495117,
+      "learning_rate": 7.845040371501486e-06,
+      "loss": 1.6995,
+      "step": 14528
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 25.020404815673828,
+      "learning_rate": 7.844203656444798e-06,
+      "loss": 2.2122,
+      "step": 14529
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 27.879436492919922,
+      "learning_rate": 7.84336694138811e-06,
+      "loss": 1.0222,
+      "step": 14530
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 3.8494162559509277,
+      "learning_rate": 7.842530226331424e-06,
+      "loss": 0.392,
+      "step": 14531
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 18.847829818725586,
+      "learning_rate": 7.841693511274736e-06,
+      "loss": 3.4313,
+      "step": 14532
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 18.116872787475586,
+      "learning_rate": 7.840856796218048e-06,
+      "loss": 2.1904,
+      "step": 14533
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 18.75259017944336,
+      "learning_rate": 7.840020081161361e-06,
+      "loss": 0.673,
+      "step": 14534
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.266660213470459,
+      "learning_rate": 7.839183366104673e-06,
+      "loss": 1.7115,
+      "step": 14535
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 5.787752628326416,
+      "learning_rate": 7.838346651047987e-06,
+      "loss": 0.3172,
+      "step": 14536
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.596453666687012,
+      "learning_rate": 7.837509935991299e-06,
+      "loss": 0.4714,
+      "step": 14537
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 10.35500717163086,
+      "learning_rate": 7.836673220934612e-06,
+      "loss": 0.7654,
+      "step": 14538
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 22.067216873168945,
+      "learning_rate": 7.835836505877924e-06,
+      "loss": 3.4565,
+      "step": 14539
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 18.70140266418457,
+      "learning_rate": 7.834999790821236e-06,
+      "loss": 1.6778,
+      "step": 14540
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 11.594501495361328,
+      "learning_rate": 7.83416307576455e-06,
+      "loss": 0.8165,
+      "step": 14541
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 39.64685821533203,
+      "learning_rate": 7.833326360707862e-06,
+      "loss": 3.063,
+      "step": 14542
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.162467002868652,
+      "learning_rate": 7.832489645651174e-06,
+      "loss": 0.5666,
+      "step": 14543
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 105.2118148803711,
+      "learning_rate": 7.831652930594486e-06,
+      "loss": 1.9304,
+      "step": 14544
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 3.5929176807403564,
+      "learning_rate": 7.830816215537798e-06,
+      "loss": 0.1045,
+      "step": 14545
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 6.858730316162109,
+      "learning_rate": 7.829979500481111e-06,
+      "loss": 0.2439,
+      "step": 14546
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 5.487667560577393,
+      "learning_rate": 7.829142785424423e-06,
+      "loss": 0.2675,
+      "step": 14547
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 24.481765747070312,
+      "learning_rate": 7.828306070367737e-06,
+      "loss": 1.3063,
+      "step": 14548
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 15.480400085449219,
+      "learning_rate": 7.827469355311049e-06,
+      "loss": 2.4106,
+      "step": 14549
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.6516752243042,
+      "learning_rate": 7.826632640254363e-06,
+      "loss": 1.8167,
+      "step": 14550
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 32.08213424682617,
+      "learning_rate": 7.825795925197675e-06,
+      "loss": 1.9974,
+      "step": 14551
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 16.49408531188965,
+      "learning_rate": 7.824959210140987e-06,
+      "loss": 1.6478,
+      "step": 14552
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 46.75410842895508,
+      "learning_rate": 7.8241224950843e-06,
+      "loss": 0.192,
+      "step": 14553
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 22.896223068237305,
+      "learning_rate": 7.823285780027612e-06,
+      "loss": 1.2882,
+      "step": 14554
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 14.967391967773438,
+      "learning_rate": 7.822449064970926e-06,
+      "loss": 1.7651,
+      "step": 14555
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 20.13538932800293,
+      "learning_rate": 7.821612349914238e-06,
+      "loss": 2.5185,
+      "step": 14556
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 7.627354621887207,
+      "learning_rate": 7.82077563485755e-06,
+      "loss": 1.418,
+      "step": 14557
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 16.376483917236328,
+      "learning_rate": 7.819938919800862e-06,
+      "loss": 0.2539,
+      "step": 14558
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 19.46470832824707,
+      "learning_rate": 7.819102204744174e-06,
+      "loss": 0.8282,
+      "step": 14559
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.967860221862793,
+      "learning_rate": 7.818265489687487e-06,
+      "loss": 3.0796,
+      "step": 14560
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 29.03691864013672,
+      "learning_rate": 7.8174287746308e-06,
+      "loss": 1.3847,
+      "step": 14561
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 24.95924949645996,
+      "learning_rate": 7.816592059574113e-06,
+      "loss": 0.65,
+      "step": 14562
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 20.359928131103516,
+      "learning_rate": 7.815755344517425e-06,
+      "loss": 1.2683,
+      "step": 14563
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 11.161372184753418,
+      "learning_rate": 7.814918629460739e-06,
+      "loss": 0.7836,
+      "step": 14564
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 23.24300193786621,
+      "learning_rate": 7.81408191440405e-06,
+      "loss": 1.2312,
+      "step": 14565
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 29.766010284423828,
+      "learning_rate": 7.813245199347362e-06,
+      "loss": 1.2391,
+      "step": 14566
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 17.175212860107422,
+      "learning_rate": 7.812408484290676e-06,
+      "loss": 0.8,
+      "step": 14567
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 43.0667724609375,
+      "learning_rate": 7.811571769233988e-06,
+      "loss": 3.6364,
+      "step": 14568
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 46.05887985229492,
+      "learning_rate": 7.8107350541773e-06,
+      "loss": 1.5329,
+      "step": 14569
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 29.1179256439209,
+      "learning_rate": 7.809898339120614e-06,
+      "loss": 1.6498,
+      "step": 14570
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 12.148987770080566,
+      "learning_rate": 7.809061624063926e-06,
+      "loss": 1.7415,
+      "step": 14571
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 23.32232666015625,
+      "learning_rate": 7.808224909007238e-06,
+      "loss": 1.8076,
+      "step": 14572
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 43.80778121948242,
+      "learning_rate": 7.80738819395055e-06,
+      "loss": 2.2583,
+      "step": 14573
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 7.639410972595215,
+      "learning_rate": 7.806551478893863e-06,
+      "loss": 1.4383,
+      "step": 14574
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 15.752108573913574,
+      "learning_rate": 7.805714763837175e-06,
+      "loss": 2.1638,
+      "step": 14575
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 24.567739486694336,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 1.9584,
+      "step": 14576
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 6.122366428375244,
+      "learning_rate": 7.8040413337238e-06,
+      "loss": 0.6865,
+      "step": 14577
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.692014694213867,
+      "learning_rate": 7.803204618667114e-06,
+      "loss": 1.5606,
+      "step": 14578
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 16.712642669677734,
+      "learning_rate": 7.802367903610426e-06,
+      "loss": 0.8948,
+      "step": 14579
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.310211181640625,
+      "learning_rate": 7.801531188553738e-06,
+      "loss": 0.6789,
+      "step": 14580
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 15.829083442687988,
+      "learning_rate": 7.800694473497052e-06,
+      "loss": 0.4798,
+      "step": 14581
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 12.211121559143066,
+      "learning_rate": 7.799857758440364e-06,
+      "loss": 1.0991,
+      "step": 14582
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 49.79775619506836,
+      "learning_rate": 7.799021043383676e-06,
+      "loss": 1.4883,
+      "step": 14583
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 17.090513229370117,
+      "learning_rate": 7.79818432832699e-06,
+      "loss": 1.6109,
+      "step": 14584
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 25.796220779418945,
+      "learning_rate": 7.797347613270301e-06,
+      "loss": 0.9772,
+      "step": 14585
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 4.717153072357178,
+      "learning_rate": 7.796510898213613e-06,
+      "loss": 0.1331,
+      "step": 14586
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 5.1751017570495605,
+      "learning_rate": 7.795674183156925e-06,
+      "loss": 0.1403,
+      "step": 14587
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 94.42440032958984,
+      "learning_rate": 7.794837468100239e-06,
+      "loss": 1.7024,
+      "step": 14588
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 11.68253231048584,
+      "learning_rate": 7.794000753043551e-06,
+      "loss": 0.4139,
+      "step": 14589
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.50575065612793,
+      "learning_rate": 7.793164037986865e-06,
+      "loss": 1.5483,
+      "step": 14590
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 66.21617126464844,
+      "learning_rate": 7.792327322930177e-06,
+      "loss": 1.3414,
+      "step": 14591
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 8.727616310119629,
+      "learning_rate": 7.79149060787349e-06,
+      "loss": 0.9599,
+      "step": 14592
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 27.565093994140625,
+      "learning_rate": 7.790653892816802e-06,
+      "loss": 1.7694,
+      "step": 14593
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 6.212733268737793,
+      "learning_rate": 7.789817177760114e-06,
+      "loss": 0.6293,
+      "step": 14594
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 13.79759407043457,
+      "learning_rate": 7.788980462703428e-06,
+      "loss": 0.9985,
+      "step": 14595
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 4.914252281188965,
+      "learning_rate": 7.78814374764674e-06,
+      "loss": 0.1219,
+      "step": 14596
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 27.71527099609375,
+      "learning_rate": 7.787307032590052e-06,
+      "loss": 2.2464,
+      "step": 14597
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 28.01056671142578,
+      "learning_rate": 7.786470317533364e-06,
+      "loss": 0.8147,
+      "step": 14598
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 29.393339157104492,
+      "learning_rate": 7.785633602476677e-06,
+      "loss": 1.8519,
+      "step": 14599
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 13.588300704956055,
+      "learning_rate": 7.78479688741999e-06,
+      "loss": 1.0838,
+      "step": 14600
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 26.342918395996094,
+      "learning_rate": 7.783960172363301e-06,
+      "loss": 1.3437,
+      "step": 14601
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 21.386323928833008,
+      "learning_rate": 7.783123457306615e-06,
+      "loss": 1.8337,
+      "step": 14602
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 18.504051208496094,
+      "learning_rate": 7.782286742249927e-06,
+      "loss": 2.125,
+      "step": 14603
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 17.427810668945312,
+      "learning_rate": 7.78145002719324e-06,
+      "loss": 0.9518,
+      "step": 14604
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 46.07899856567383,
+      "learning_rate": 7.780613312136552e-06,
+      "loss": 1.5458,
+      "step": 14605
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 15.297252655029297,
+      "learning_rate": 7.779776597079866e-06,
+      "loss": 1.9646,
+      "step": 14606
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 27.815570831298828,
+      "learning_rate": 7.778939882023178e-06,
+      "loss": 2.0745,
+      "step": 14607
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 5.807898998260498,
+      "learning_rate": 7.77810316696649e-06,
+      "loss": 0.5374,
+      "step": 14608
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 22.669795989990234,
+      "learning_rate": 7.777266451909804e-06,
+      "loss": 1.9899,
+      "step": 14609
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 11.786237716674805,
+      "learning_rate": 7.776429736853116e-06,
+      "loss": 1.2184,
+      "step": 14610
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 6.201798915863037,
+      "learning_rate": 7.775593021796428e-06,
+      "loss": 0.2181,
+      "step": 14611
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.68503189086914,
+      "learning_rate": 7.77475630673974e-06,
+      "loss": 1.588,
+      "step": 14612
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 13.119375228881836,
+      "learning_rate": 7.773919591683053e-06,
+      "loss": 0.7835,
+      "step": 14613
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 7.963985443115234,
+      "learning_rate": 7.773082876626365e-06,
+      "loss": 0.2345,
+      "step": 14614
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 26.554868698120117,
+      "learning_rate": 7.772246161569677e-06,
+      "loss": 2.4617,
+      "step": 14615
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 19.229042053222656,
+      "learning_rate": 7.77140944651299e-06,
+      "loss": 1.583,
+      "step": 14616
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 24.596824645996094,
+      "learning_rate": 7.770572731456303e-06,
+      "loss": 0.5835,
+      "step": 14617
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 14.019936561584473,
+      "learning_rate": 7.769736016399616e-06,
+      "loss": 1.1474,
+      "step": 14618
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 14.718634605407715,
+      "learning_rate": 7.768899301342928e-06,
+      "loss": 3.4,
+      "step": 14619
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.23465347290039,
+      "learning_rate": 7.768062586286242e-06,
+      "loss": 1.4429,
+      "step": 14620
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 54.38119888305664,
+      "learning_rate": 7.767225871229554e-06,
+      "loss": 3.5724,
+      "step": 14621
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.59969711303711,
+      "learning_rate": 7.766389156172866e-06,
+      "loss": 0.463,
+      "step": 14622
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 16.999919891357422,
+      "learning_rate": 7.76555244111618e-06,
+      "loss": 0.4796,
+      "step": 14623
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.698204040527344,
+      "learning_rate": 7.764715726059491e-06,
+      "loss": 1.8318,
+      "step": 14624
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 46.9920768737793,
+      "learning_rate": 7.763879011002803e-06,
+      "loss": 1.7167,
+      "step": 14625
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 15.677593231201172,
+      "learning_rate": 7.763042295946115e-06,
+      "loss": 0.9368,
+      "step": 14626
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 10.155533790588379,
+      "learning_rate": 7.762205580889429e-06,
+      "loss": 1.7626,
+      "step": 14627
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.302998542785645,
+      "learning_rate": 7.761368865832741e-06,
+      "loss": 2.3986,
+      "step": 14628
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 29.012435913085938,
+      "learning_rate": 7.760532150776053e-06,
+      "loss": 2.3208,
+      "step": 14629
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 50.5739631652832,
+      "learning_rate": 7.759695435719367e-06,
+      "loss": 2.657,
+      "step": 14630
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 22.324176788330078,
+      "learning_rate": 7.758858720662678e-06,
+      "loss": 1.8601,
+      "step": 14631
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 37.7600212097168,
+      "learning_rate": 7.758022005605992e-06,
+      "loss": 2.0259,
+      "step": 14632
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 22.62678337097168,
+      "learning_rate": 7.757185290549304e-06,
+      "loss": 1.1992,
+      "step": 14633
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.476841449737549,
+      "learning_rate": 7.756348575492618e-06,
+      "loss": 0.6818,
+      "step": 14634
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 16.30000114440918,
+      "learning_rate": 7.75551186043593e-06,
+      "loss": 1.5676,
+      "step": 14635
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.904687404632568,
+      "learning_rate": 7.754675145379242e-06,
+      "loss": 1.7429,
+      "step": 14636
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 6.143351078033447,
+      "learning_rate": 7.753838430322555e-06,
+      "loss": 1.8704,
+      "step": 14637
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.69153118133545,
+      "learning_rate": 7.753001715265867e-06,
+      "loss": 0.5042,
+      "step": 14638
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.439741134643555,
+      "learning_rate": 7.75216500020918e-06,
+      "loss": 0.3078,
+      "step": 14639
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.229491233825684,
+      "learning_rate": 7.751328285152491e-06,
+      "loss": 1.3007,
+      "step": 14640
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 19.291664123535156,
+      "learning_rate": 7.750491570095805e-06,
+      "loss": 0.5928,
+      "step": 14641
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.47659683227539,
+      "learning_rate": 7.749654855039117e-06,
+      "loss": 1.8454,
+      "step": 14642
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.25069808959961,
+      "learning_rate": 7.748818139982429e-06,
+      "loss": 0.8562,
+      "step": 14643
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.662071228027344,
+      "learning_rate": 7.747981424925742e-06,
+      "loss": 0.5109,
+      "step": 14644
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 21.549152374267578,
+      "learning_rate": 7.747144709869054e-06,
+      "loss": 0.6217,
+      "step": 14645
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 9.104865074157715,
+      "learning_rate": 7.746307994812368e-06,
+      "loss": 1.246,
+      "step": 14646
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.59805154800415,
+      "learning_rate": 7.74547127975568e-06,
+      "loss": 0.4916,
+      "step": 14647
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.185526847839355,
+      "learning_rate": 7.744634564698994e-06,
+      "loss": 1.2289,
+      "step": 14648
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 13.667840957641602,
+      "learning_rate": 7.743797849642306e-06,
+      "loss": 1.3791,
+      "step": 14649
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 14.437935829162598,
+      "learning_rate": 7.742961134585617e-06,
+      "loss": 1.9665,
+      "step": 14650
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.515650749206543,
+      "learning_rate": 7.74212441952893e-06,
+      "loss": 0.9934,
+      "step": 14651
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 9.91360855102539,
+      "learning_rate": 7.741287704472243e-06,
+      "loss": 1.5728,
+      "step": 14652
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 13.074822425842285,
+      "learning_rate": 7.740450989415555e-06,
+      "loss": 0.7348,
+      "step": 14653
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 26.621301651000977,
+      "learning_rate": 7.739614274358867e-06,
+      "loss": 2.1105,
+      "step": 14654
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 10.325733184814453,
+      "learning_rate": 7.73877755930218e-06,
+      "loss": 0.3978,
+      "step": 14655
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 6.7510576248168945,
+      "learning_rate": 7.737940844245493e-06,
+      "loss": 1.862,
+      "step": 14656
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.192282676696777,
+      "learning_rate": 7.737104129188805e-06,
+      "loss": 0.5877,
+      "step": 14657
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.965519905090332,
+      "learning_rate": 7.736267414132118e-06,
+      "loss": 0.5455,
+      "step": 14658
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 15.416940689086914,
+      "learning_rate": 7.73543069907543e-06,
+      "loss": 3.6079,
+      "step": 14659
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 27.16056251525879,
+      "learning_rate": 7.734593984018744e-06,
+      "loss": 1.0839,
+      "step": 14660
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 16.25208854675293,
+      "learning_rate": 7.733757268962056e-06,
+      "loss": 0.6543,
+      "step": 14661
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 19.351118087768555,
+      "learning_rate": 7.73292055390537e-06,
+      "loss": 3.4736,
+      "step": 14662
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.313667297363281,
+      "learning_rate": 7.732083838848681e-06,
+      "loss": 0.3627,
+      "step": 14663
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 34.64920425415039,
+      "learning_rate": 7.731247123791993e-06,
+      "loss": 2.8451,
+      "step": 14664
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 156.4160614013672,
+      "learning_rate": 7.730410408735305e-06,
+      "loss": 1.3522,
+      "step": 14665
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 28.074020385742188,
+      "learning_rate": 7.729573693678619e-06,
+      "loss": 1.6769,
+      "step": 14666
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 20.513277053833008,
+      "learning_rate": 7.728736978621931e-06,
+      "loss": 0.4861,
+      "step": 14667
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 22.602537155151367,
+      "learning_rate": 7.727900263565243e-06,
+      "loss": 2.1944,
+      "step": 14668
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 15.476371765136719,
+      "learning_rate": 7.727063548508556e-06,
+      "loss": 1.178,
+      "step": 14669
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 13.422551155090332,
+      "learning_rate": 7.726226833451868e-06,
+      "loss": 0.803,
+      "step": 14670
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 16.891130447387695,
+      "learning_rate": 7.72539011839518e-06,
+      "loss": 3.0551,
+      "step": 14671
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 20.250980377197266,
+      "learning_rate": 7.724553403338494e-06,
+      "loss": 0.2396,
+      "step": 14672
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.72658634185791,
+      "learning_rate": 7.723716688281806e-06,
+      "loss": 0.6455,
+      "step": 14673
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 6.41309118270874,
+      "learning_rate": 7.72287997322512e-06,
+      "loss": 0.7244,
+      "step": 14674
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 79.53043365478516,
+      "learning_rate": 7.722043258168432e-06,
+      "loss": 1.1777,
+      "step": 14675
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 70.87897491455078,
+      "learning_rate": 7.721206543111745e-06,
+      "loss": 1.7493,
+      "step": 14676
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 21.449668884277344,
+      "learning_rate": 7.720369828055057e-06,
+      "loss": 1.5234,
+      "step": 14677
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 88.9782485961914,
+      "learning_rate": 7.719533112998369e-06,
+      "loss": 3.8735,
+      "step": 14678
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 9.213892936706543,
+      "learning_rate": 7.718696397941681e-06,
+      "loss": 0.8011,
+      "step": 14679
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 625.9251708984375,
+      "learning_rate": 7.717859682884993e-06,
+      "loss": 0.577,
+      "step": 14680
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 4.340551376342773,
+      "learning_rate": 7.717022967828307e-06,
+      "loss": 0.2504,
+      "step": 14681
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.4213769435882568,
+      "learning_rate": 7.716186252771619e-06,
+      "loss": 0.0401,
+      "step": 14682
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 67.31068420410156,
+      "learning_rate": 7.715349537714932e-06,
+      "loss": 3.0705,
+      "step": 14683
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.602607727050781,
+      "learning_rate": 7.714512822658244e-06,
+      "loss": 0.7183,
+      "step": 14684
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 21.028234481811523,
+      "learning_rate": 7.713676107601556e-06,
+      "loss": 1.0286,
+      "step": 14685
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.835209846496582,
+      "learning_rate": 7.71283939254487e-06,
+      "loss": 1.2344,
+      "step": 14686
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.893700122833252,
+      "learning_rate": 7.712002677488182e-06,
+      "loss": 1.1592,
+      "step": 14687
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 11.222153663635254,
+      "learning_rate": 7.711165962431495e-06,
+      "loss": 0.6997,
+      "step": 14688
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 10.17320442199707,
+      "learning_rate": 7.710329247374807e-06,
+      "loss": 0.9973,
+      "step": 14689
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 19.59346580505371,
+      "learning_rate": 7.70949253231812e-06,
+      "loss": 0.9439,
+      "step": 14690
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 6.132411956787109,
+      "learning_rate": 7.708655817261433e-06,
+      "loss": 0.5027,
+      "step": 14691
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 9.631757736206055,
+      "learning_rate": 7.707819102204745e-06,
+      "loss": 0.4147,
+      "step": 14692
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.173003196716309,
+      "learning_rate": 7.706982387148057e-06,
+      "loss": 2.1582,
+      "step": 14693
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 104.57372283935547,
+      "learning_rate": 7.706145672091369e-06,
+      "loss": 2.0988,
+      "step": 14694
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 42.085662841796875,
+      "learning_rate": 7.705308957034683e-06,
+      "loss": 1.8397,
+      "step": 14695
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.223543167114258,
+      "learning_rate": 7.704472241977994e-06,
+      "loss": 0.3691,
+      "step": 14696
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 39.029788970947266,
+      "learning_rate": 7.703635526921308e-06,
+      "loss": 1.5826,
+      "step": 14697
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 20.499826431274414,
+      "learning_rate": 7.70279881186462e-06,
+      "loss": 1.1586,
+      "step": 14698
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 28.13475799560547,
+      "learning_rate": 7.701962096807932e-06,
+      "loss": 0.3709,
+      "step": 14699
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 16.042461395263672,
+      "learning_rate": 7.701125381751246e-06,
+      "loss": 0.8329,
+      "step": 14700
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 12.51712703704834,
+      "learning_rate": 7.700288666694558e-06,
+      "loss": 0.8813,
+      "step": 14701
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 29.742050170898438,
+      "learning_rate": 7.699451951637871e-06,
+      "loss": 2.4081,
+      "step": 14702
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 44.979732513427734,
+      "learning_rate": 7.698615236581183e-06,
+      "loss": 0.6644,
+      "step": 14703
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 12.372272491455078,
+      "learning_rate": 7.697778521524495e-06,
+      "loss": 1.7375,
+      "step": 14704
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 22.812719345092773,
+      "learning_rate": 7.696941806467809e-06,
+      "loss": 1.7477,
+      "step": 14705
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 36.66621780395508,
+      "learning_rate": 7.69610509141112e-06,
+      "loss": 2.3893,
+      "step": 14706
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 11.402688026428223,
+      "learning_rate": 7.695268376354433e-06,
+      "loss": 1.1545,
+      "step": 14707
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 6.952476978302002,
+      "learning_rate": 7.694431661297745e-06,
+      "loss": 0.3191,
+      "step": 14708
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 68.07209014892578,
+      "learning_rate": 7.693594946241058e-06,
+      "loss": 1.3018,
+      "step": 14709
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 17.782407760620117,
+      "learning_rate": 7.69275823118437e-06,
+      "loss": 1.2765,
+      "step": 14710
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 64.7061538696289,
+      "learning_rate": 7.691921516127684e-06,
+      "loss": 1.2451,
+      "step": 14711
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.56583023071289,
+      "learning_rate": 7.691084801070996e-06,
+      "loss": 0.5217,
+      "step": 14712
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 43.400611877441406,
+      "learning_rate": 7.690248086014308e-06,
+      "loss": 1.5767,
+      "step": 14713
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 12.612919807434082,
+      "learning_rate": 7.689411370957622e-06,
+      "loss": 1.3096,
+      "step": 14714
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 13.8457612991333,
+      "learning_rate": 7.688574655900933e-06,
+      "loss": 1.7108,
+      "step": 14715
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 16.337207794189453,
+      "learning_rate": 7.687737940844247e-06,
+      "loss": 0.9688,
+      "step": 14716
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 68.52069091796875,
+      "learning_rate": 7.686901225787559e-06,
+      "loss": 1.7413,
+      "step": 14717
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 12.928807258605957,
+      "learning_rate": 7.686064510730871e-06,
+      "loss": 0.5263,
+      "step": 14718
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 12.465153694152832,
+      "learning_rate": 7.685227795674185e-06,
+      "loss": 0.6653,
+      "step": 14719
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 10.703567504882812,
+      "learning_rate": 7.684391080617497e-06,
+      "loss": 1.7437,
+      "step": 14720
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 16.249418258666992,
+      "learning_rate": 7.683554365560809e-06,
+      "loss": 1.6232,
+      "step": 14721
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 12.612680435180664,
+      "learning_rate": 7.68271765050412e-06,
+      "loss": 1.3508,
+      "step": 14722
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 17.31722068786621,
+      "learning_rate": 7.681880935447434e-06,
+      "loss": 1.5709,
+      "step": 14723
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 12.659685134887695,
+      "learning_rate": 7.681044220390746e-06,
+      "loss": 1.0214,
+      "step": 14724
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 69.76463317871094,
+      "learning_rate": 7.680207505334058e-06,
+      "loss": 2.1385,
+      "step": 14725
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.5032670497894287,
+      "learning_rate": 7.679370790277372e-06,
+      "loss": 0.0833,
+      "step": 14726
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 17.35916519165039,
+      "learning_rate": 7.678534075220684e-06,
+      "loss": 1.9738,
+      "step": 14727
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 20.854421615600586,
+      "learning_rate": 7.677697360163997e-06,
+      "loss": 1.0489,
+      "step": 14728
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 15.953429222106934,
+      "learning_rate": 7.67686064510731e-06,
+      "loss": 1.6066,
+      "step": 14729
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.436120986938477,
+      "learning_rate": 7.676023930050623e-06,
+      "loss": 1.6007,
+      "step": 14730
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 17.540395736694336,
+      "learning_rate": 7.675187214993935e-06,
+      "loss": 1.9825,
+      "step": 14731
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 36.389583587646484,
+      "learning_rate": 7.674350499937247e-06,
+      "loss": 1.3227,
+      "step": 14732
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 10.731135368347168,
+      "learning_rate": 7.673513784880559e-06,
+      "loss": 0.6461,
+      "step": 14733
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 68.90753936767578,
+      "learning_rate": 7.672677069823872e-06,
+      "loss": 1.2232,
+      "step": 14734
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 22.537322998046875,
+      "learning_rate": 7.671840354767184e-06,
+      "loss": 2.7363,
+      "step": 14735
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 12.679781913757324,
+      "learning_rate": 7.671003639710496e-06,
+      "loss": 0.39,
+      "step": 14736
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 18.542882919311523,
+      "learning_rate": 7.67016692465381e-06,
+      "loss": 0.7766,
+      "step": 14737
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 13.160587310791016,
+      "learning_rate": 7.669330209597122e-06,
+      "loss": 0.9405,
+      "step": 14738
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 21.700645446777344,
+      "learning_rate": 7.668493494540434e-06,
+      "loss": 0.4254,
+      "step": 14739
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 10.037639617919922,
+      "learning_rate": 7.667656779483748e-06,
+      "loss": 1.0496,
+      "step": 14740
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 55.460689544677734,
+      "learning_rate": 7.66682006442706e-06,
+      "loss": 1.3667,
+      "step": 14741
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 11.1635160446167,
+      "learning_rate": 7.665983349370373e-06,
+      "loss": 0.4446,
+      "step": 14742
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 7.231917858123779,
+      "learning_rate": 7.665146634313685e-06,
+      "loss": 0.2691,
+      "step": 14743
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 11.715499877929688,
+      "learning_rate": 7.664309919256999e-06,
+      "loss": 0.9348,
+      "step": 14744
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 18.822839736938477,
+      "learning_rate": 7.66347320420031e-06,
+      "loss": 1.1269,
+      "step": 14745
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 19.825674057006836,
+      "learning_rate": 7.662636489143623e-06,
+      "loss": 1.0912,
+      "step": 14746
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 11.040472030639648,
+      "learning_rate": 7.661799774086935e-06,
+      "loss": 1.2,
+      "step": 14747
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 23.976301193237305,
+      "learning_rate": 7.660963059030248e-06,
+      "loss": 0.807,
+      "step": 14748
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 64.25426483154297,
+      "learning_rate": 7.66012634397356e-06,
+      "loss": 2.6122,
+      "step": 14749
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.116251945495605,
+      "learning_rate": 7.659289628916872e-06,
+      "loss": 0.7184,
+      "step": 14750
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 15.621927261352539,
+      "learning_rate": 7.658452913860186e-06,
+      "loss": 1.1001,
+      "step": 14751
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.6078104972839355,
+      "learning_rate": 7.657616198803498e-06,
+      "loss": 1.6181,
+      "step": 14752
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.419527530670166,
+      "learning_rate": 7.65677948374681e-06,
+      "loss": 0.1343,
+      "step": 14753
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 18.74810028076172,
+      "learning_rate": 7.655942768690123e-06,
+      "loss": 1.0452,
+      "step": 14754
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 11.465399742126465,
+      "learning_rate": 7.655106053633435e-06,
+      "loss": 0.6569,
+      "step": 14755
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 15.211071014404297,
+      "learning_rate": 7.654269338576749e-06,
+      "loss": 0.6556,
+      "step": 14756
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 13.307765007019043,
+      "learning_rate": 7.653432623520061e-06,
+      "loss": 0.6282,
+      "step": 14757
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 6.284524917602539,
+      "learning_rate": 7.652595908463375e-06,
+      "loss": 0.4561,
+      "step": 14758
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.023004531860352,
+      "learning_rate": 7.651759193406687e-06,
+      "loss": 0.5253,
+      "step": 14759
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 20.6444149017334,
+      "learning_rate": 7.650922478349999e-06,
+      "loss": 0.6476,
+      "step": 14760
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 6.272350311279297,
+      "learning_rate": 7.65008576329331e-06,
+      "loss": 0.4519,
+      "step": 14761
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 19.79265785217285,
+      "learning_rate": 7.649249048236622e-06,
+      "loss": 1.0469,
+      "step": 14762
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 12.268784523010254,
+      "learning_rate": 7.648412333179936e-06,
+      "loss": 0.7808,
+      "step": 14763
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 11.571566581726074,
+      "learning_rate": 7.647575618123248e-06,
+      "loss": 0.6196,
+      "step": 14764
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 42.86050033569336,
+      "learning_rate": 7.646738903066562e-06,
+      "loss": 2.3287,
+      "step": 14765
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 36.44502258300781,
+      "learning_rate": 7.645902188009874e-06,
+      "loss": 2.2601,
+      "step": 14766
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 20.920602798461914,
+      "learning_rate": 7.645065472953186e-06,
+      "loss": 1.1987,
+      "step": 14767
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 15.131333351135254,
+      "learning_rate": 7.6442287578965e-06,
+      "loss": 0.5331,
+      "step": 14768
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 60.51508712768555,
+      "learning_rate": 7.643392042839811e-06,
+      "loss": 0.4155,
+      "step": 14769
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 61.69184875488281,
+      "learning_rate": 7.642555327783125e-06,
+      "loss": 1.8619,
+      "step": 14770
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 14.843707084655762,
+      "learning_rate": 7.641718612726437e-06,
+      "loss": 0.518,
+      "step": 14771
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 38.67082214355469,
+      "learning_rate": 7.640881897669749e-06,
+      "loss": 1.871,
+      "step": 14772
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 31.577425003051758,
+      "learning_rate": 7.640045182613062e-06,
+      "loss": 1.8729,
+      "step": 14773
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 34.70793914794922,
+      "learning_rate": 7.639208467556374e-06,
+      "loss": 1.9818,
+      "step": 14774
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 9.932486534118652,
+      "learning_rate": 7.638371752499686e-06,
+      "loss": 1.055,
+      "step": 14775
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 4.367156505584717,
+      "learning_rate": 7.637535037442998e-06,
+      "loss": 0.2814,
+      "step": 14776
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 26.68075180053711,
+      "learning_rate": 7.636698322386312e-06,
+      "loss": 1.1387,
+      "step": 14777
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 11.871581077575684,
+      "learning_rate": 7.635861607329624e-06,
+      "loss": 1.7331,
+      "step": 14778
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 10.713318824768066,
+      "learning_rate": 7.635024892272938e-06,
+      "loss": 1.1925,
+      "step": 14779
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 17.177345275878906,
+      "learning_rate": 7.63418817721625e-06,
+      "loss": 0.8947,
+      "step": 14780
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 20.661283493041992,
+      "learning_rate": 7.633351462159561e-06,
+      "loss": 2.3745,
+      "step": 14781
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 17.319683074951172,
+      "learning_rate": 7.632514747102875e-06,
+      "loss": 0.6588,
+      "step": 14782
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 29.432462692260742,
+      "learning_rate": 7.631678032046187e-06,
+      "loss": 2.2025,
+      "step": 14783
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 28.53501319885254,
+      "learning_rate": 7.6308413169895e-06,
+      "loss": 1.0598,
+      "step": 14784
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 25.438735961914062,
+      "learning_rate": 7.630004601932813e-06,
+      "loss": 0.5378,
+      "step": 14785
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 53.4843864440918,
+      "learning_rate": 7.6291678868761255e-06,
+      "loss": 1.4765,
+      "step": 14786
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 4.252384185791016,
+      "learning_rate": 7.6283311718194374e-06,
+      "loss": 0.461,
+      "step": 14787
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 7.987371444702148,
+      "learning_rate": 7.627494456762749e-06,
+      "loss": 0.4523,
+      "step": 14788
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 21.252737045288086,
+      "learning_rate": 7.626657741706063e-06,
+      "loss": 1.0041,
+      "step": 14789
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.794703483581543,
+      "learning_rate": 7.625821026649375e-06,
+      "loss": 0.7371,
+      "step": 14790
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.99533462524414,
+      "learning_rate": 7.624984311592688e-06,
+      "loss": 1.334,
+      "step": 14791
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.046095848083496,
+      "learning_rate": 7.624147596536e-06,
+      "loss": 0.7383,
+      "step": 14792
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 3.4559438228607178,
+      "learning_rate": 7.623310881479313e-06,
+      "loss": 0.1237,
+      "step": 14793
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.042539596557617,
+      "learning_rate": 7.622474166422625e-06,
+      "loss": 0.5694,
+      "step": 14794
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 94.61004638671875,
+      "learning_rate": 7.621637451365937e-06,
+      "loss": 1.7079,
+      "step": 14795
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.512558937072754,
+      "learning_rate": 7.620800736309251e-06,
+      "loss": 0.2887,
+      "step": 14796
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 73.2804183959961,
+      "learning_rate": 7.619964021252563e-06,
+      "loss": 2.4657,
+      "step": 14797
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 16.300159454345703,
+      "learning_rate": 7.619127306195876e-06,
+      "loss": 0.6152,
+      "step": 14798
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 20.082332611083984,
+      "learning_rate": 7.618290591139188e-06,
+      "loss": 1.8323,
+      "step": 14799
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.368802070617676,
+      "learning_rate": 7.617453876082501e-06,
+      "loss": 1.3725,
+      "step": 14800
+    },
+    {
+      "epoch": 1.86,
+      "eval_loss": 0.07997239381074905,
+      "eval_runtime": 97.6038,
+      "eval_samples_per_second": 36.29,
+      "eval_steps_per_second": 36.29,
+      "step": 14800
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 246.64675903320312,
+      "learning_rate": 7.616617161025813e-06,
+      "loss": 1.6255,
+      "step": 14801
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 6.828473091125488,
+      "learning_rate": 7.615780445969125e-06,
+      "loss": 1.6192,
+      "step": 14802
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.192270278930664,
+      "learning_rate": 7.614943730912439e-06,
+      "loss": 1.0769,
+      "step": 14803
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 15.74098014831543,
+      "learning_rate": 7.614107015855751e-06,
+      "loss": 1.4209,
+      "step": 14804
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.044049263000488,
+      "learning_rate": 7.613270300799064e-06,
+      "loss": 1.6556,
+      "step": 14805
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 14.061810493469238,
+      "learning_rate": 7.612433585742376e-06,
+      "loss": 0.8869,
+      "step": 14806
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 15.723001480102539,
+      "learning_rate": 7.611596870685689e-06,
+      "loss": 0.9136,
+      "step": 14807
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 16.836034774780273,
+      "learning_rate": 7.610760155629001e-06,
+      "loss": 1.0389,
+      "step": 14808
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.526412963867188,
+      "learning_rate": 7.609923440572313e-06,
+      "loss": 0.8741,
+      "step": 14809
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 40.982303619384766,
+      "learning_rate": 7.609086725515627e-06,
+      "loss": 1.6724,
+      "step": 14810
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 48.365840911865234,
+      "learning_rate": 7.608250010458939e-06,
+      "loss": 0.829,
+      "step": 14811
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 13.846466064453125,
+      "learning_rate": 7.6074132954022516e-06,
+      "loss": 1.6373,
+      "step": 14812
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 10.954010963439941,
+      "learning_rate": 7.6065765803455635e-06,
+      "loss": 1.8855,
+      "step": 14813
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 13.028228759765625,
+      "learning_rate": 7.605739865288877e-06,
+      "loss": 1.7035,
+      "step": 14814
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.14072036743164,
+      "learning_rate": 7.604903150232189e-06,
+      "loss": 0.2946,
+      "step": 14815
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 21.410184860229492,
+      "learning_rate": 7.604066435175501e-06,
+      "loss": 0.5266,
+      "step": 14816
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.846100807189941,
+      "learning_rate": 7.603229720118815e-06,
+      "loss": 0.5662,
+      "step": 14817
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.277616500854492,
+      "learning_rate": 7.602393005062127e-06,
+      "loss": 0.3525,
+      "step": 14818
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 13.545755386352539,
+      "learning_rate": 7.6015562900054395e-06,
+      "loss": 1.0353,
+      "step": 14819
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.16395092010498,
+      "learning_rate": 7.6007195749487514e-06,
+      "loss": 0.9289,
+      "step": 14820
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.918290138244629,
+      "learning_rate": 7.599882859892065e-06,
+      "loss": 2.2482,
+      "step": 14821
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.14470100402832,
+      "learning_rate": 7.599046144835377e-06,
+      "loss": 0.9612,
+      "step": 14822
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 35.48596954345703,
+      "learning_rate": 7.598209429778689e-06,
+      "loss": 1.5181,
+      "step": 14823
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 15.159586906433105,
+      "learning_rate": 7.597372714722003e-06,
+      "loss": 2.5833,
+      "step": 14824
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.813267230987549,
+      "learning_rate": 7.596535999665315e-06,
+      "loss": 0.232,
+      "step": 14825
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 18.843538284301758,
+      "learning_rate": 7.595699284608627e-06,
+      "loss": 0.7198,
+      "step": 14826
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 13.109396934509277,
+      "learning_rate": 7.594862569551939e-06,
+      "loss": 0.8969,
+      "step": 14827
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 7.410354137420654,
+      "learning_rate": 7.594025854495253e-06,
+      "loss": 1.4,
+      "step": 14828
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 10.625185012817383,
+      "learning_rate": 7.593189139438565e-06,
+      "loss": 0.6854,
+      "step": 14829
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.607718467712402,
+      "learning_rate": 7.592352424381877e-06,
+      "loss": 0.724,
+      "step": 14830
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 23.821897506713867,
+      "learning_rate": 7.59151570932519e-06,
+      "loss": 1.3364,
+      "step": 14831
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 22.2536563873291,
+      "learning_rate": 7.5906789942685025e-06,
+      "loss": 1.5433,
+      "step": 14832
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 17.018207550048828,
+      "learning_rate": 7.589842279211815e-06,
+      "loss": 1.29,
+      "step": 14833
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 36.413246154785156,
+      "learning_rate": 7.589005564155127e-06,
+      "loss": 2.2479,
+      "step": 14834
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 13.229951858520508,
+      "learning_rate": 7.588168849098441e-06,
+      "loss": 1.2976,
+      "step": 14835
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 34.4698371887207,
+      "learning_rate": 7.587332134041753e-06,
+      "loss": 2.5385,
+      "step": 14836
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.363869667053223,
+      "learning_rate": 7.586495418985065e-06,
+      "loss": 1.5394,
+      "step": 14837
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.833475112915039,
+      "learning_rate": 7.585658703928378e-06,
+      "loss": 2.091,
+      "step": 14838
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.779874801635742,
+      "learning_rate": 7.5848219888716904e-06,
+      "loss": 0.6732,
+      "step": 14839
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 24.601621627807617,
+      "learning_rate": 7.583985273815003e-06,
+      "loss": 0.7512,
+      "step": 14840
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.069605827331543,
+      "learning_rate": 7.583148558758315e-06,
+      "loss": 0.4151,
+      "step": 14841
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 24.539474487304688,
+      "learning_rate": 7.582311843701629e-06,
+      "loss": 0.7814,
+      "step": 14842
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 25.05148696899414,
+      "learning_rate": 7.581475128644941e-06,
+      "loss": 2.6374,
+      "step": 14843
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 20.171566009521484,
+      "learning_rate": 7.580638413588253e-06,
+      "loss": 1.0501,
+      "step": 14844
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 20.1003475189209,
+      "learning_rate": 7.5798016985315655e-06,
+      "loss": 1.4832,
+      "step": 14845
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 6.150067329406738,
+      "learning_rate": 7.578964983474878e-06,
+      "loss": 0.4083,
+      "step": 14846
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 29.662729263305664,
+      "learning_rate": 7.578128268418191e-06,
+      "loss": 3.1539,
+      "step": 14847
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 4.108884811401367,
+      "learning_rate": 7.577291553361503e-06,
+      "loss": 0.3868,
+      "step": 14848
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 11.957849502563477,
+      "learning_rate": 7.576454838304817e-06,
+      "loss": 0.7825,
+      "step": 14849
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 24.528156280517578,
+      "learning_rate": 7.575618123248129e-06,
+      "loss": 2.3116,
+      "step": 14850
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 30.471540451049805,
+      "learning_rate": 7.574781408191441e-06,
+      "loss": 1.2928,
+      "step": 14851
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 25.57601547241211,
+      "learning_rate": 7.5739446931347535e-06,
+      "loss": 2.7023,
+      "step": 14852
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 9.855649948120117,
+      "learning_rate": 7.573107978078066e-06,
+      "loss": 0.5518,
+      "step": 14853
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 22.2115535736084,
+      "learning_rate": 7.572271263021379e-06,
+      "loss": 1.4069,
+      "step": 14854
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 18.855382919311523,
+      "learning_rate": 7.571434547964691e-06,
+      "loss": 0.9062,
+      "step": 14855
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.688556671142578,
+      "learning_rate": 7.570597832908005e-06,
+      "loss": 0.486,
+      "step": 14856
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 18.32562255859375,
+      "learning_rate": 7.569761117851317e-06,
+      "loss": 0.9634,
+      "step": 14857
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 16.361801147460938,
+      "learning_rate": 7.568924402794629e-06,
+      "loss": 1.1458,
+      "step": 14858
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 4.865877628326416,
+      "learning_rate": 7.568087687737941e-06,
+      "loss": 0.247,
+      "step": 14859
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 22.873950958251953,
+      "learning_rate": 7.567250972681253e-06,
+      "loss": 0.5436,
+      "step": 14860
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.281028747558594,
+      "learning_rate": 7.566414257624567e-06,
+      "loss": 1.5352,
+      "step": 14861
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 65.8516845703125,
+      "learning_rate": 7.565577542567879e-06,
+      "loss": 0.5066,
+      "step": 14862
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 21.525156021118164,
+      "learning_rate": 7.564740827511193e-06,
+      "loss": 0.531,
+      "step": 14863
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 14.380889892578125,
+      "learning_rate": 7.5639041124545045e-06,
+      "loss": 1.3095,
+      "step": 14864
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 13.246685981750488,
+      "learning_rate": 7.5630673973978165e-06,
+      "loss": 0.7462,
+      "step": 14865
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 14.796537399291992,
+      "learning_rate": 7.562230682341129e-06,
+      "loss": 0.4701,
+      "step": 14866
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 11.796875953674316,
+      "learning_rate": 7.561393967284441e-06,
+      "loss": 1.0644,
+      "step": 14867
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 6.4520769119262695,
+      "learning_rate": 7.560557252227755e-06,
+      "loss": 0.2443,
+      "step": 14868
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 19.781051635742188,
+      "learning_rate": 7.559720537171067e-06,
+      "loss": 0.8506,
+      "step": 14869
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 65.25953674316406,
+      "learning_rate": 7.5588838221143805e-06,
+      "loss": 1.7756,
+      "step": 14870
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.863962173461914,
+      "learning_rate": 7.5580471070576925e-06,
+      "loss": 0.4296,
+      "step": 14871
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 25.320886611938477,
+      "learning_rate": 7.557210392001004e-06,
+      "loss": 1.2524,
+      "step": 14872
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.225039958953857,
+      "learning_rate": 7.556373676944317e-06,
+      "loss": 0.2312,
+      "step": 14873
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 24.97355842590332,
+      "learning_rate": 7.555536961887629e-06,
+      "loss": 2.759,
+      "step": 14874
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 26.371421813964844,
+      "learning_rate": 7.554700246830943e-06,
+      "loss": 1.3567,
+      "step": 14875
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.538215637207031,
+      "learning_rate": 7.553863531774255e-06,
+      "loss": 0.6993,
+      "step": 14876
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 67.82234954833984,
+      "learning_rate": 7.5530268167175676e-06,
+      "loss": 0.7644,
+      "step": 14877
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 37.28622817993164,
+      "learning_rate": 7.55219010166088e-06,
+      "loss": 1.1606,
+      "step": 14878
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 19.709482192993164,
+      "learning_rate": 7.551353386604192e-06,
+      "loss": 1.4561,
+      "step": 14879
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 11.090790748596191,
+      "learning_rate": 7.550516671547505e-06,
+      "loss": 1.7676,
+      "step": 14880
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 4.8185553550720215,
+      "learning_rate": 7.549679956490817e-06,
+      "loss": 0.1238,
+      "step": 14881
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.536447048187256,
+      "learning_rate": 7.548843241434131e-06,
+      "loss": 0.1998,
+      "step": 14882
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.016610145568848,
+      "learning_rate": 7.548006526377443e-06,
+      "loss": 1.1862,
+      "step": 14883
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.89788293838501,
+      "learning_rate": 7.5471698113207555e-06,
+      "loss": 0.4358,
+      "step": 14884
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 65.45915985107422,
+      "learning_rate": 7.546333096264068e-06,
+      "loss": 2.5424,
+      "step": 14885
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 52.495174407958984,
+      "learning_rate": 7.54549638120738e-06,
+      "loss": 2.0097,
+      "step": 14886
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 20.664783477783203,
+      "learning_rate": 7.544659666150693e-06,
+      "loss": 2.3188,
+      "step": 14887
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.102899551391602,
+      "learning_rate": 7.543822951094005e-06,
+      "loss": 0.3336,
+      "step": 14888
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.261075496673584,
+      "learning_rate": 7.542986236037319e-06,
+      "loss": 0.4018,
+      "step": 14889
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.724687576293945,
+      "learning_rate": 7.542149520980631e-06,
+      "loss": 1.7554,
+      "step": 14890
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 17.16265106201172,
+      "learning_rate": 7.541312805923943e-06,
+      "loss": 1.215,
+      "step": 14891
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 6.207948684692383,
+      "learning_rate": 7.540476090867256e-06,
+      "loss": 0.3008,
+      "step": 14892
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 35.358238220214844,
+      "learning_rate": 7.539639375810568e-06,
+      "loss": 1.6874,
+      "step": 14893
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 70.42288208007812,
+      "learning_rate": 7.538802660753881e-06,
+      "loss": 0.8292,
+      "step": 14894
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 29.4013671875,
+      "learning_rate": 7.537965945697193e-06,
+      "loss": 0.617,
+      "step": 14895
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 11.098967552185059,
+      "learning_rate": 7.5371292306405066e-06,
+      "loss": 1.0755,
+      "step": 14896
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 31.135793685913086,
+      "learning_rate": 7.5362925155838185e-06,
+      "loss": 1.7696,
+      "step": 14897
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 17.028921127319336,
+      "learning_rate": 7.5354558005271305e-06,
+      "loss": 1.865,
+      "step": 14898
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 23.53152847290039,
+      "learning_rate": 7.534619085470444e-06,
+      "loss": 1.4702,
+      "step": 14899
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 6.146441459655762,
+      "learning_rate": 7.533782370413756e-06,
+      "loss": 1.9333,
+      "step": 14900
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.910538673400879,
+      "learning_rate": 7.532945655357069e-06,
+      "loss": 1.1973,
+      "step": 14901
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 45.51909637451172,
+      "learning_rate": 7.532108940300381e-06,
+      "loss": 2.0465,
+      "step": 14902
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 14.263692855834961,
+      "learning_rate": 7.5312722252436945e-06,
+      "loss": 0.4382,
+      "step": 14903
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 15.351764678955078,
+      "learning_rate": 7.5304355101870064e-06,
+      "loss": 2.218,
+      "step": 14904
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 6.3323845863342285,
+      "learning_rate": 7.529598795130318e-06,
+      "loss": 0.2431,
+      "step": 14905
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 17.696409225463867,
+      "learning_rate": 7.528762080073632e-06,
+      "loss": 0.4567,
+      "step": 14906
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 30.03977394104004,
+      "learning_rate": 7.527925365016944e-06,
+      "loss": 1.8367,
+      "step": 14907
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 13.985361099243164,
+      "learning_rate": 7.527088649960257e-06,
+      "loss": 2.163,
+      "step": 14908
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 20.119407653808594,
+      "learning_rate": 7.526251934903569e-06,
+      "loss": 1.9892,
+      "step": 14909
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 14.874587059020996,
+      "learning_rate": 7.525415219846882e-06,
+      "loss": 0.9434,
+      "step": 14910
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 19.314739227294922,
+      "learning_rate": 7.524578504790194e-06,
+      "loss": 1.0257,
+      "step": 14911
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 16.053003311157227,
+      "learning_rate": 7.523741789733506e-06,
+      "loss": 1.4353,
+      "step": 14912
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.82081937789917,
+      "learning_rate": 7.522905074676819e-06,
+      "loss": 0.7296,
+      "step": 14913
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.054506301879883,
+      "learning_rate": 7.522068359620132e-06,
+      "loss": 1.1525,
+      "step": 14914
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 16.969287872314453,
+      "learning_rate": 7.521231644563445e-06,
+      "loss": 3.0622,
+      "step": 14915
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 15.36079216003418,
+      "learning_rate": 7.520394929506757e-06,
+      "loss": 0.9112,
+      "step": 14916
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 38.61628341674805,
+      "learning_rate": 7.51955821445007e-06,
+      "loss": 0.5906,
+      "step": 14917
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.0408730506896973,
+      "learning_rate": 7.518721499393382e-06,
+      "loss": 0.3868,
+      "step": 14918
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.081964015960693,
+      "learning_rate": 7.517884784336694e-06,
+      "loss": 0.1102,
+      "step": 14919
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.888787269592285,
+      "learning_rate": 7.517048069280007e-06,
+      "loss": 0.3224,
+      "step": 14920
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.768947124481201,
+      "learning_rate": 7.51621135422332e-06,
+      "loss": 0.2688,
+      "step": 14921
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 35.43925476074219,
+      "learning_rate": 7.515374639166633e-06,
+      "loss": 0.97,
+      "step": 14922
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 9.788805961608887,
+      "learning_rate": 7.514537924109945e-06,
+      "loss": 0.705,
+      "step": 14923
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 20.753562927246094,
+      "learning_rate": 7.513701209053258e-06,
+      "loss": 3.1368,
+      "step": 14924
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 63.22145080566406,
+      "learning_rate": 7.51286449399657e-06,
+      "loss": 1.0398,
+      "step": 14925
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 19.292924880981445,
+      "learning_rate": 7.512027778939882e-06,
+      "loss": 2.1925,
+      "step": 14926
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.6630377769470215,
+      "learning_rate": 7.511191063883195e-06,
+      "loss": 0.9697,
+      "step": 14927
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 19.1313533782959,
+      "learning_rate": 7.510354348826508e-06,
+      "loss": 0.4675,
+      "step": 14928
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 18.08177947998047,
+      "learning_rate": 7.5095176337698206e-06,
+      "loss": 1.5367,
+      "step": 14929
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 19.45388412475586,
+      "learning_rate": 7.5086809187131325e-06,
+      "loss": 1.5596,
+      "step": 14930
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 18.87152099609375,
+      "learning_rate": 7.507844203656446e-06,
+      "loss": 2.3846,
+      "step": 14931
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 30.16594696044922,
+      "learning_rate": 7.507007488599758e-06,
+      "loss": 2.878,
+      "step": 14932
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 11.567892074584961,
+      "learning_rate": 7.50617077354307e-06,
+      "loss": 0.8096,
+      "step": 14933
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 23.418529510498047,
+      "learning_rate": 7.505334058486383e-06,
+      "loss": 1.8027,
+      "step": 14934
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 24.442947387695312,
+      "learning_rate": 7.504497343429696e-06,
+      "loss": 2.7443,
+      "step": 14935
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 63.38848876953125,
+      "learning_rate": 7.5036606283730085e-06,
+      "loss": 2.4058,
+      "step": 14936
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 19.26790428161621,
+      "learning_rate": 7.5028239133163204e-06,
+      "loss": 1.8722,
+      "step": 14937
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 15.698874473571777,
+      "learning_rate": 7.501987198259634e-06,
+      "loss": 0.7961,
+      "step": 14938
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 4.914449691772461,
+      "learning_rate": 7.501150483202946e-06,
+      "loss": 0.325,
+      "step": 14939
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 6.667304992675781,
+      "learning_rate": 7.500313768146258e-06,
+      "loss": 0.247,
+      "step": 14940
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 34.39765930175781,
+      "learning_rate": 7.499477053089571e-06,
+      "loss": 1.1508,
+      "step": 14941
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 33.020774841308594,
+      "learning_rate": 7.498640338032883e-06,
+      "loss": 1.1243,
+      "step": 14942
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 7.784989356994629,
+      "learning_rate": 7.497803622976196e-06,
+      "loss": 0.6662,
+      "step": 14943
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 24.58091926574707,
+      "learning_rate": 7.496966907919508e-06,
+      "loss": 1.3466,
+      "step": 14944
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 25.048736572265625,
+      "learning_rate": 7.496130192862822e-06,
+      "loss": 0.866,
+      "step": 14945
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 4.25439977645874,
+      "learning_rate": 7.495293477806134e-06,
+      "loss": 0.5579,
+      "step": 14946
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 14.87307071685791,
+      "learning_rate": 7.494456762749446e-06,
+      "loss": 2.0097,
+      "step": 14947
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 12.519543647766113,
+      "learning_rate": 7.493620047692759e-06,
+      "loss": 0.6585,
+      "step": 14948
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 29.968774795532227,
+      "learning_rate": 7.492783332636071e-06,
+      "loss": 1.0066,
+      "step": 14949
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 26.526437759399414,
+      "learning_rate": 7.491946617579384e-06,
+      "loss": 0.8447,
+      "step": 14950
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 7.1481852531433105,
+      "learning_rate": 7.491109902522696e-06,
+      "loss": 0.5562,
+      "step": 14951
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 43.571720123291016,
+      "learning_rate": 7.49027318746601e-06,
+      "loss": 2.229,
+      "step": 14952
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.739995956420898,
+      "learning_rate": 7.489436472409322e-06,
+      "loss": 1.0044,
+      "step": 14953
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 34.558101654052734,
+      "learning_rate": 7.488599757352634e-06,
+      "loss": 0.8945,
+      "step": 14954
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 6.550328254699707,
+      "learning_rate": 7.487763042295947e-06,
+      "loss": 0.2706,
+      "step": 14955
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 15.401753425598145,
+      "learning_rate": 7.486926327239259e-06,
+      "loss": 0.3218,
+      "step": 14956
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.214600563049316,
+      "learning_rate": 7.486089612182572e-06,
+      "loss": 0.6701,
+      "step": 14957
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 7.7430524826049805,
+      "learning_rate": 7.485252897125884e-06,
+      "loss": 0.5822,
+      "step": 14958
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 27.569425582885742,
+      "learning_rate": 7.484416182069197e-06,
+      "loss": 1.1034,
+      "step": 14959
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 14.088428497314453,
+      "learning_rate": 7.48357946701251e-06,
+      "loss": 1.3161,
+      "step": 14960
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 5.511892795562744,
+      "learning_rate": 7.482742751955822e-06,
+      "loss": 0.2117,
+      "step": 14961
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 11.08498764038086,
+      "learning_rate": 7.4819060368991346e-06,
+      "loss": 0.9814,
+      "step": 14962
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.185216903686523,
+      "learning_rate": 7.4810693218424465e-06,
+      "loss": 0.4156,
+      "step": 14963
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 13.83070182800293,
+      "learning_rate": 7.48023260678576e-06,
+      "loss": 0.9413,
+      "step": 14964
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 5.967977523803711,
+      "learning_rate": 7.479395891729072e-06,
+      "loss": 0.3474,
+      "step": 14965
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 64.0623550415039,
+      "learning_rate": 7.478559176672385e-06,
+      "loss": 2.2951,
+      "step": 14966
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 6.03565788269043,
+      "learning_rate": 7.477722461615698e-06,
+      "loss": 0.3642,
+      "step": 14967
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.3975555896759033,
+      "learning_rate": 7.47688574655901e-06,
+      "loss": 0.0542,
+      "step": 14968
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 15.115584373474121,
+      "learning_rate": 7.4760490315023225e-06,
+      "loss": 1.0409,
+      "step": 14969
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 43.474605560302734,
+      "learning_rate": 7.4752123164456344e-06,
+      "loss": 1.5636,
+      "step": 14970
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.10328197479248,
+      "learning_rate": 7.474375601388948e-06,
+      "loss": 0.2745,
+      "step": 14971
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 16.16593360900879,
+      "learning_rate": 7.47353888633226e-06,
+      "loss": 0.4139,
+      "step": 14972
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.691909790039062,
+      "learning_rate": 7.472702171275573e-06,
+      "loss": 1.9511,
+      "step": 14973
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 38.42649841308594,
+      "learning_rate": 7.471865456218886e-06,
+      "loss": 2.803,
+      "step": 14974
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 13.04145622253418,
+      "learning_rate": 7.471028741162198e-06,
+      "loss": 0.2814,
+      "step": 14975
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 12.061103820800781,
+      "learning_rate": 7.47019202610551e-06,
+      "loss": 1.7101,
+      "step": 14976
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 11.907960891723633,
+      "learning_rate": 7.469355311048822e-06,
+      "loss": 0.8512,
+      "step": 14977
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 26.97811508178711,
+      "learning_rate": 7.468518595992136e-06,
+      "loss": 1.2657,
+      "step": 14978
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 11.2202730178833,
+      "learning_rate": 7.467681880935448e-06,
+      "loss": 1.148,
+      "step": 14979
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 7.339235305786133,
+      "learning_rate": 7.466845165878761e-06,
+      "loss": 0.4095,
+      "step": 14980
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 23.894771575927734,
+      "learning_rate": 7.4660084508220736e-06,
+      "loss": 0.4987,
+      "step": 14981
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.5060096383094788,
+      "learning_rate": 7.4651717357653855e-06,
+      "loss": 0.014,
+      "step": 14982
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.960077166557312,
+      "learning_rate": 7.464335020708698e-06,
+      "loss": 0.0285,
+      "step": 14983
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 114.88871765136719,
+      "learning_rate": 7.46349830565201e-06,
+      "loss": 4.6805,
+      "step": 14984
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 11.071863174438477,
+      "learning_rate": 7.462661590595324e-06,
+      "loss": 0.3641,
+      "step": 14985
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 12.264084815979004,
+      "learning_rate": 7.461824875538636e-06,
+      "loss": 0.6126,
+      "step": 14986
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 11.309356689453125,
+      "learning_rate": 7.460988160481949e-06,
+      "loss": 1.2135,
+      "step": 14987
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 22.1566219329834,
+      "learning_rate": 7.460151445425261e-06,
+      "loss": 0.7779,
+      "step": 14988
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 156.8177490234375,
+      "learning_rate": 7.4593147303685734e-06,
+      "loss": 2.935,
+      "step": 14989
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 22.175682067871094,
+      "learning_rate": 7.458478015311886e-06,
+      "loss": 0.5215,
+      "step": 14990
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 28.10517120361328,
+      "learning_rate": 7.457641300255198e-06,
+      "loss": 3.2197,
+      "step": 14991
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 14.341082572937012,
+      "learning_rate": 7.456804585198512e-06,
+      "loss": 0.9033,
+      "step": 14992
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 40.654903411865234,
+      "learning_rate": 7.455967870141824e-06,
+      "loss": 1.9905,
+      "step": 14993
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 6.580481052398682,
+      "learning_rate": 7.455131155085137e-06,
+      "loss": 0.1788,
+      "step": 14994
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 26.085426330566406,
+      "learning_rate": 7.4542944400284485e-06,
+      "loss": 2.4308,
+      "step": 14995
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.547032356262207,
+      "learning_rate": 7.453457724971761e-06,
+      "loss": 1.7124,
+      "step": 14996
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 34.72230529785156,
+      "learning_rate": 7.452621009915074e-06,
+      "loss": 1.6396,
+      "step": 14997
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 7.088932037353516,
+      "learning_rate": 7.451784294858386e-06,
+      "loss": 0.739,
+      "step": 14998
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 18.540529251098633,
+      "learning_rate": 7.4509475798017e-06,
+      "loss": 3.2027,
+      "step": 14999
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 12.922123908996582,
+      "learning_rate": 7.450110864745012e-06,
+      "loss": 0.4767,
+      "step": 15000
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 30.11306381225586,
+      "learning_rate": 7.4492741496883245e-06,
+      "loss": 1.2894,
+      "step": 15001
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 35.27677536010742,
+      "learning_rate": 7.4484374346316365e-06,
+      "loss": 1.9374,
+      "step": 15002
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 14.58852767944336,
+      "learning_rate": 7.447600719574949e-06,
+      "loss": 2.1957,
+      "step": 15003
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 6.9970383644104,
+      "learning_rate": 7.446764004518262e-06,
+      "loss": 0.9583,
+      "step": 15004
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 39.45263671875,
+      "learning_rate": 7.445927289461574e-06,
+      "loss": 0.8669,
+      "step": 15005
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 18.60435676574707,
+      "learning_rate": 7.445090574404888e-06,
+      "loss": 1.7102,
+      "step": 15006
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 16.23671531677246,
+      "learning_rate": 7.4442538593482e-06,
+      "loss": 0.9465,
+      "step": 15007
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 53.88591384887695,
+      "learning_rate": 7.4434171442915124e-06,
+      "loss": 1.9764,
+      "step": 15008
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 26.555509567260742,
+      "learning_rate": 7.442580429234824e-06,
+      "loss": 1.1828,
+      "step": 15009
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8898343443870544,
+      "learning_rate": 7.441743714178137e-06,
+      "loss": 0.0484,
+      "step": 15010
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.331498146057129,
+      "learning_rate": 7.44090699912145e-06,
+      "loss": 0.8278,
+      "step": 15011
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 16.52419662475586,
+      "learning_rate": 7.440070284064762e-06,
+      "loss": 1.8094,
+      "step": 15012
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.045432090759277,
+      "learning_rate": 7.439233569008076e-06,
+      "loss": 0.6667,
+      "step": 15013
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 13.238429069519043,
+      "learning_rate": 7.4383968539513875e-06,
+      "loss": 1.7813,
+      "step": 15014
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 9.393712997436523,
+      "learning_rate": 7.4375601388947e-06,
+      "loss": 0.6944,
+      "step": 15015
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 18.744245529174805,
+      "learning_rate": 7.436723423838012e-06,
+      "loss": 1.1589,
+      "step": 15016
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 11.48732852935791,
+      "learning_rate": 7.435886708781325e-06,
+      "loss": 1.0309,
+      "step": 15017
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 58.54405975341797,
+      "learning_rate": 7.435049993724638e-06,
+      "loss": 0.9708,
+      "step": 15018
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.87790298461914,
+      "learning_rate": 7.43421327866795e-06,
+      "loss": 2.8177,
+      "step": 15019
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.245800971984863,
+      "learning_rate": 7.4333765636112635e-06,
+      "loss": 0.6433,
+      "step": 15020
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 42.300506591796875,
+      "learning_rate": 7.4325398485545755e-06,
+      "loss": 3.357,
+      "step": 15021
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 13.989384651184082,
+      "learning_rate": 7.431703133497888e-06,
+      "loss": 1.7183,
+      "step": 15022
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.733213424682617,
+      "learning_rate": 7.4308664184412e-06,
+      "loss": 0.5374,
+      "step": 15023
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 13.210402488708496,
+      "learning_rate": 7.430029703384512e-06,
+      "loss": 0.8232,
+      "step": 15024
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 31.181640625,
+      "learning_rate": 7.429192988327826e-06,
+      "loss": 1.2312,
+      "step": 15025
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 74.49955749511719,
+      "learning_rate": 7.428356273271138e-06,
+      "loss": 1.7963,
+      "step": 15026
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 56.21243667602539,
+      "learning_rate": 7.427519558214451e-06,
+      "loss": 1.9133,
+      "step": 15027
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 12.333606719970703,
+      "learning_rate": 7.426682843157763e-06,
+      "loss": 1.6646,
+      "step": 15028
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 24.455326080322266,
+      "learning_rate": 7.425846128101076e-06,
+      "loss": 1.2796,
+      "step": 15029
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.309059143066406,
+      "learning_rate": 7.425009413044388e-06,
+      "loss": 0.1407,
+      "step": 15030
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 18.7886962890625,
+      "learning_rate": 7.4241726979877e-06,
+      "loss": 2.037,
+      "step": 15031
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 4.938962459564209,
+      "learning_rate": 7.423335982931014e-06,
+      "loss": 0.4311,
+      "step": 15032
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.151268005371094,
+      "learning_rate": 7.422499267874326e-06,
+      "loss": 0.6303,
+      "step": 15033
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 93.7016830444336,
+      "learning_rate": 7.421662552817639e-06,
+      "loss": 1.1028,
+      "step": 15034
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 18.35399627685547,
+      "learning_rate": 7.420825837760951e-06,
+      "loss": 1.5585,
+      "step": 15035
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.213486671447754,
+      "learning_rate": 7.419989122704264e-06,
+      "loss": 0.8414,
+      "step": 15036
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 23.497493743896484,
+      "learning_rate": 7.419152407647576e-06,
+      "loss": 0.9,
+      "step": 15037
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 25.89967918395996,
+      "learning_rate": 7.418315692590888e-06,
+      "loss": 1.3119,
+      "step": 15038
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 18.524890899658203,
+      "learning_rate": 7.417478977534202e-06,
+      "loss": 0.8211,
+      "step": 15039
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 30.518224716186523,
+      "learning_rate": 7.416642262477514e-06,
+      "loss": 1.0983,
+      "step": 15040
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.779489994049072,
+      "learning_rate": 7.415805547420826e-06,
+      "loss": 1.4563,
+      "step": 15041
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.428210735321045,
+      "learning_rate": 7.414968832364139e-06,
+      "loss": 0.208,
+      "step": 15042
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.045461654663086,
+      "learning_rate": 7.414132117307452e-06,
+      "loss": 0.2831,
+      "step": 15043
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 18.02309226989746,
+      "learning_rate": 7.413295402250764e-06,
+      "loss": 0.7516,
+      "step": 15044
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 94.65620422363281,
+      "learning_rate": 7.412458687194076e-06,
+      "loss": 1.8608,
+      "step": 15045
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 25.21953582763672,
+      "learning_rate": 7.4116219721373896e-06,
+      "loss": 1.3827,
+      "step": 15046
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.448881149291992,
+      "learning_rate": 7.4107852570807015e-06,
+      "loss": 2.3396,
+      "step": 15047
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.742633819580078,
+      "learning_rate": 7.409948542024014e-06,
+      "loss": 2.5715,
+      "step": 15048
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 31.669166564941406,
+      "learning_rate": 7.409111826967327e-06,
+      "loss": 1.3321,
+      "step": 15049
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 3.381666660308838,
+      "learning_rate": 7.40827511191064e-06,
+      "loss": 0.0671,
+      "step": 15050
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 17.08733558654785,
+      "learning_rate": 7.407438396853952e-06,
+      "loss": 0.93,
+      "step": 15051
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 24.145172119140625,
+      "learning_rate": 7.406601681797264e-06,
+      "loss": 1.7006,
+      "step": 15052
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 15.978361129760742,
+      "learning_rate": 7.4057649667405775e-06,
+      "loss": 1.4769,
+      "step": 15053
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.425533294677734,
+      "learning_rate": 7.4049282516838894e-06,
+      "loss": 0.4073,
+      "step": 15054
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 19.693145751953125,
+      "learning_rate": 7.404091536627202e-06,
+      "loss": 1.3016,
+      "step": 15055
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 4.42262601852417,
+      "learning_rate": 7.403254821570515e-06,
+      "loss": 0.1591,
+      "step": 15056
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 30.48455238342285,
+      "learning_rate": 7.402418106513828e-06,
+      "loss": 0.6781,
+      "step": 15057
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 6.9818339347839355,
+      "learning_rate": 7.40158139145714e-06,
+      "loss": 0.719,
+      "step": 15058
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 21.614707946777344,
+      "learning_rate": 7.400744676400452e-06,
+      "loss": 0.8311,
+      "step": 15059
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 19.429691314697266,
+      "learning_rate": 7.399907961343765e-06,
+      "loss": 2.7223,
+      "step": 15060
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 6.083532810211182,
+      "learning_rate": 7.399071246287077e-06,
+      "loss": 0.5901,
+      "step": 15061
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 25.171422958374023,
+      "learning_rate": 7.39823453123039e-06,
+      "loss": 0.969,
+      "step": 15062
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 18.150936126708984,
+      "learning_rate": 7.397397816173703e-06,
+      "loss": 1.1038,
+      "step": 15063
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.097476959228516,
+      "learning_rate": 7.396561101117016e-06,
+      "loss": 0.5034,
+      "step": 15064
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.264492511749268,
+      "learning_rate": 7.395724386060328e-06,
+      "loss": 0.1194,
+      "step": 15065
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.850740909576416,
+      "learning_rate": 7.39488767100364e-06,
+      "loss": 1.5162,
+      "step": 15066
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 12.903263092041016,
+      "learning_rate": 7.394050955946953e-06,
+      "loss": 0.3627,
+      "step": 15067
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 176.21908569335938,
+      "learning_rate": 7.393214240890265e-06,
+      "loss": 0.3115,
+      "step": 15068
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 32.34139633178711,
+      "learning_rate": 7.392377525833578e-06,
+      "loss": 0.6036,
+      "step": 15069
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 6.2513604164123535,
+      "learning_rate": 7.39154081077689e-06,
+      "loss": 0.7679,
+      "step": 15070
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 21.014720916748047,
+      "learning_rate": 7.390704095720204e-06,
+      "loss": 1.3911,
+      "step": 15071
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 6.944918155670166,
+      "learning_rate": 7.389867380663516e-06,
+      "loss": 0.6498,
+      "step": 15072
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 25.82847785949707,
+      "learning_rate": 7.389030665606828e-06,
+      "loss": 1.5971,
+      "step": 15073
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 15.153432846069336,
+      "learning_rate": 7.388193950550141e-06,
+      "loss": 0.7888,
+      "step": 15074
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 20.675094604492188,
+      "learning_rate": 7.387357235493453e-06,
+      "loss": 0.6522,
+      "step": 15075
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.721600532531738,
+      "learning_rate": 7.386520520436766e-06,
+      "loss": 0.4128,
+      "step": 15076
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 18.053680419921875,
+      "learning_rate": 7.385683805380078e-06,
+      "loss": 1.6978,
+      "step": 15077
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.639769554138184,
+      "learning_rate": 7.384847090323391e-06,
+      "loss": 0.7635,
+      "step": 15078
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 12.367140769958496,
+      "learning_rate": 7.3840103752667036e-06,
+      "loss": 0.7246,
+      "step": 15079
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 49.35564422607422,
+      "learning_rate": 7.3831736602100155e-06,
+      "loss": 0.6651,
+      "step": 15080
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 40.579952239990234,
+      "learning_rate": 7.382336945153329e-06,
+      "loss": 1.3458,
+      "step": 15081
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 28.484344482421875,
+      "learning_rate": 7.381500230096641e-06,
+      "loss": 2.1591,
+      "step": 15082
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.652926445007324,
+      "learning_rate": 7.380663515039954e-06,
+      "loss": 0.3406,
+      "step": 15083
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 18.412296295166016,
+      "learning_rate": 7.379826799983266e-06,
+      "loss": 1.7563,
+      "step": 15084
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 16.60628318786621,
+      "learning_rate": 7.378990084926579e-06,
+      "loss": 0.7008,
+      "step": 15085
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 21.525320053100586,
+      "learning_rate": 7.3781533698698915e-06,
+      "loss": 0.7776,
+      "step": 15086
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.889859199523926,
+      "learning_rate": 7.3773166548132034e-06,
+      "loss": 1.0566,
+      "step": 15087
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 12.126493453979492,
+      "learning_rate": 7.376479939756517e-06,
+      "loss": 1.1783,
+      "step": 15088
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.531156539916992,
+      "learning_rate": 7.375643224699829e-06,
+      "loss": 1.9034,
+      "step": 15089
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 18.80508041381836,
+      "learning_rate": 7.374806509643142e-06,
+      "loss": 0.9211,
+      "step": 15090
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 2.0429701805114746,
+      "learning_rate": 7.373969794586454e-06,
+      "loss": 0.1388,
+      "step": 15091
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.433183670043945,
+      "learning_rate": 7.373133079529767e-06,
+      "loss": 0.3499,
+      "step": 15092
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.792083740234375,
+      "learning_rate": 7.372296364473079e-06,
+      "loss": 0.7006,
+      "step": 15093
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.4274983406066895,
+      "learning_rate": 7.371459649416391e-06,
+      "loss": 0.2147,
+      "step": 15094
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 22.65265655517578,
+      "learning_rate": 7.370622934359705e-06,
+      "loss": 0.9307,
+      "step": 15095
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 74.88692474365234,
+      "learning_rate": 7.369786219303017e-06,
+      "loss": 2.2369,
+      "step": 15096
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 43.321231842041016,
+      "learning_rate": 7.36894950424633e-06,
+      "loss": 1.5066,
+      "step": 15097
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.069525718688965,
+      "learning_rate": 7.368112789189642e-06,
+      "loss": 1.406,
+      "step": 15098
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 24.34798240661621,
+      "learning_rate": 7.367276074132954e-06,
+      "loss": 0.8286,
+      "step": 15099
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.9306001663208,
+      "learning_rate": 7.366439359076267e-06,
+      "loss": 0.3151,
+      "step": 15100
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 10.898557662963867,
+      "learning_rate": 7.365602644019579e-06,
+      "loss": 0.6529,
+      "step": 15101
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 14.822643280029297,
+      "learning_rate": 7.364765928962893e-06,
+      "loss": 0.6295,
+      "step": 15102
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.52712631225586,
+      "learning_rate": 7.363929213906205e-06,
+      "loss": 0.4436,
+      "step": 15103
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.008702278137207,
+      "learning_rate": 7.363092498849518e-06,
+      "loss": 1.3962,
+      "step": 15104
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 49.62347412109375,
+      "learning_rate": 7.36225578379283e-06,
+      "loss": 1.4771,
+      "step": 15105
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 21.786096572875977,
+      "learning_rate": 7.361419068736142e-06,
+      "loss": 1.0213,
+      "step": 15106
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 21.74210548400879,
+      "learning_rate": 7.360582353679455e-06,
+      "loss": 1.9559,
+      "step": 15107
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 11.14261245727539,
+      "learning_rate": 7.359745638622767e-06,
+      "loss": 1.6207,
+      "step": 15108
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.294177293777466,
+      "learning_rate": 7.358908923566081e-06,
+      "loss": 0.0516,
+      "step": 15109
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 11.590841293334961,
+      "learning_rate": 7.358072208509393e-06,
+      "loss": 0.5179,
+      "step": 15110
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 18.835065841674805,
+      "learning_rate": 7.357235493452706e-06,
+      "loss": 1.0335,
+      "step": 15111
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 10.532160758972168,
+      "learning_rate": 7.3563987783960176e-06,
+      "loss": 1.2372,
+      "step": 15112
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 14.514026641845703,
+      "learning_rate": 7.3555620633393295e-06,
+      "loss": 1.355,
+      "step": 15113
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 16.543455123901367,
+      "learning_rate": 7.354725348282643e-06,
+      "loss": 0.6673,
+      "step": 15114
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 16.31077766418457,
+      "learning_rate": 7.353888633225955e-06,
+      "loss": 0.9401,
+      "step": 15115
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.899883270263672,
+      "learning_rate": 7.353051918169268e-06,
+      "loss": 0.6144,
+      "step": 15116
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 56.374244689941406,
+      "learning_rate": 7.352215203112581e-06,
+      "loss": 1.9511,
+      "step": 15117
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 17.845420837402344,
+      "learning_rate": 7.3513784880558935e-06,
+      "loss": 0.3596,
+      "step": 15118
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 11.164807319641113,
+      "learning_rate": 7.3505417729992055e-06,
+      "loss": 1.0834,
+      "step": 15119
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 23.736774444580078,
+      "learning_rate": 7.3497050579425174e-06,
+      "loss": 0.6354,
+      "step": 15120
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 69.24571990966797,
+      "learning_rate": 7.348868342885831e-06,
+      "loss": 1.576,
+      "step": 15121
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 17.8673152923584,
+      "learning_rate": 7.348031627829143e-06,
+      "loss": 0.4724,
+      "step": 15122
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 24.305837631225586,
+      "learning_rate": 7.347194912772456e-06,
+      "loss": 0.5262,
+      "step": 15123
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.139628410339355,
+      "learning_rate": 7.346358197715769e-06,
+      "loss": 0.8911,
+      "step": 15124
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.655332565307617,
+      "learning_rate": 7.3455214826590814e-06,
+      "loss": 0.4515,
+      "step": 15125
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 11.019580841064453,
+      "learning_rate": 7.344684767602393e-06,
+      "loss": 0.5824,
+      "step": 15126
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 14.877920150756836,
+      "learning_rate": 7.343848052545705e-06,
+      "loss": 1.5006,
+      "step": 15127
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 40.81086349487305,
+      "learning_rate": 7.343011337489019e-06,
+      "loss": 1.0782,
+      "step": 15128
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.504908561706543,
+      "learning_rate": 7.342174622432331e-06,
+      "loss": 0.4803,
+      "step": 15129
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 13.374335289001465,
+      "learning_rate": 7.341337907375644e-06,
+      "loss": 1.976,
+      "step": 15130
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 20.366533279418945,
+      "learning_rate": 7.3405011923189566e-06,
+      "loss": 0.9099,
+      "step": 15131
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 10.132173538208008,
+      "learning_rate": 7.339664477262269e-06,
+      "loss": 1.0805,
+      "step": 15132
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.006936073303223,
+      "learning_rate": 7.338827762205581e-06,
+      "loss": 0.4156,
+      "step": 15133
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 7.405896186828613,
+      "learning_rate": 7.337991047148893e-06,
+      "loss": 1.38,
+      "step": 15134
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 4.1786627769470215,
+      "learning_rate": 7.337154332092207e-06,
+      "loss": 0.2808,
+      "step": 15135
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.891201496124268,
+      "learning_rate": 7.336317617035519e-06,
+      "loss": 0.2731,
+      "step": 15136
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 11.198047637939453,
+      "learning_rate": 7.335480901978832e-06,
+      "loss": 1.3121,
+      "step": 15137
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.8166863918304443,
+      "learning_rate": 7.3346441869221445e-06,
+      "loss": 0.2421,
+      "step": 15138
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.097291946411133,
+      "learning_rate": 7.333807471865457e-06,
+      "loss": 0.2636,
+      "step": 15139
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 31.219444274902344,
+      "learning_rate": 7.332970756808769e-06,
+      "loss": 1.0549,
+      "step": 15140
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.707075595855713,
+      "learning_rate": 7.332134041752081e-06,
+      "loss": 0.6725,
+      "step": 15141
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 30.624589920043945,
+      "learning_rate": 7.331297326695395e-06,
+      "loss": 2.4987,
+      "step": 15142
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 36.42421340942383,
+      "learning_rate": 7.330460611638707e-06,
+      "loss": 1.2836,
+      "step": 15143
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 21.4605655670166,
+      "learning_rate": 7.32962389658202e-06,
+      "loss": 0.5328,
+      "step": 15144
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 11.258630752563477,
+      "learning_rate": 7.328787181525332e-06,
+      "loss": 0.3049,
+      "step": 15145
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.895224571228027,
+      "learning_rate": 7.327950466468645e-06,
+      "loss": 1.7441,
+      "step": 15146
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 47.341033935546875,
+      "learning_rate": 7.327113751411957e-06,
+      "loss": 0.7324,
+      "step": 15147
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 13.817826271057129,
+      "learning_rate": 7.326277036355269e-06,
+      "loss": 2.5864,
+      "step": 15148
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 72.39340209960938,
+      "learning_rate": 7.325440321298583e-06,
+      "loss": 2.7951,
+      "step": 15149
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 10.49294376373291,
+      "learning_rate": 7.324603606241895e-06,
+      "loss": 0.3464,
+      "step": 15150
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 53.07066345214844,
+      "learning_rate": 7.3237668911852075e-06,
+      "loss": 1.9196,
+      "step": 15151
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 12.413329124450684,
+      "learning_rate": 7.3229301761285195e-06,
+      "loss": 2.0227,
+      "step": 15152
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.606522560119629,
+      "learning_rate": 7.322093461071833e-06,
+      "loss": 0.4902,
+      "step": 15153
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 17.33139419555664,
+      "learning_rate": 7.321256746015145e-06,
+      "loss": 1.5737,
+      "step": 15154
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 24.492815017700195,
+      "learning_rate": 7.320420030958457e-06,
+      "loss": 1.4284,
+      "step": 15155
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 7.170698642730713,
+      "learning_rate": 7.319583315901771e-06,
+      "loss": 0.328,
+      "step": 15156
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 20.43840789794922,
+      "learning_rate": 7.318746600845083e-06,
+      "loss": 1.3311,
+      "step": 15157
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 22.837284088134766,
+      "learning_rate": 7.3179098857883954e-06,
+      "loss": 1.1056,
+      "step": 15158
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 17.995685577392578,
+      "learning_rate": 7.317073170731707e-06,
+      "loss": 1.1954,
+      "step": 15159
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 18.641700744628906,
+      "learning_rate": 7.316236455675021e-06,
+      "loss": 0.83,
+      "step": 15160
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 10.630788803100586,
+      "learning_rate": 7.315399740618333e-06,
+      "loss": 1.4478,
+      "step": 15161
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 16.64458465576172,
+      "learning_rate": 7.314563025561645e-06,
+      "loss": 1.4767,
+      "step": 15162
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 26.294815063476562,
+      "learning_rate": 7.313726310504959e-06,
+      "loss": 1.6531,
+      "step": 15163
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 13.200927734375,
+      "learning_rate": 7.3128895954482705e-06,
+      "loss": 1.0693,
+      "step": 15164
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 13.845024108886719,
+      "learning_rate": 7.312052880391583e-06,
+      "loss": 4.2466,
+      "step": 15165
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.46701717376709,
+      "learning_rate": 7.311216165334895e-06,
+      "loss": 0.6261,
+      "step": 15166
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 31.682180404663086,
+      "learning_rate": 7.310379450278209e-06,
+      "loss": 2.2054,
+      "step": 15167
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.807620048522949,
+      "learning_rate": 7.309542735221521e-06,
+      "loss": 1.3925,
+      "step": 15168
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 25.933950424194336,
+      "learning_rate": 7.308706020164833e-06,
+      "loss": 0.6675,
+      "step": 15169
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 28.577220916748047,
+      "learning_rate": 7.3078693051081465e-06,
+      "loss": 1.0649,
+      "step": 15170
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.963960647583008,
+      "learning_rate": 7.3070325900514585e-06,
+      "loss": 0.9175,
+      "step": 15171
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 7.93988561630249,
+      "learning_rate": 7.306195874994771e-06,
+      "loss": 0.485,
+      "step": 15172
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.721303939819336,
+      "learning_rate": 7.305359159938083e-06,
+      "loss": 0.4843,
+      "step": 15173
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.980069160461426,
+      "learning_rate": 7.304522444881397e-06,
+      "loss": 0.7523,
+      "step": 15174
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.0261625051498413,
+      "learning_rate": 7.303685729824709e-06,
+      "loss": 0.0355,
+      "step": 15175
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 28.56754493713379,
+      "learning_rate": 7.302849014768021e-06,
+      "loss": 1.4144,
+      "step": 15176
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.409519672393799,
+      "learning_rate": 7.302012299711334e-06,
+      "loss": 0.4377,
+      "step": 15177
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 11.686038970947266,
+      "learning_rate": 7.301175584654646e-06,
+      "loss": 2.361,
+      "step": 15178
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 16.717464447021484,
+      "learning_rate": 7.300338869597959e-06,
+      "loss": 4.0278,
+      "step": 15179
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 15.546704292297363,
+      "learning_rate": 7.299502154541271e-06,
+      "loss": 0.6647,
+      "step": 15180
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.070075511932373,
+      "learning_rate": 7.298665439484585e-06,
+      "loss": 0.2537,
+      "step": 15181
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 29.783416748046875,
+      "learning_rate": 7.297828724427897e-06,
+      "loss": 1.9954,
+      "step": 15182
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 18.873929977416992,
+      "learning_rate": 7.296992009371209e-06,
+      "loss": 0.3738,
+      "step": 15183
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.223883152008057,
+      "learning_rate": 7.296155294314522e-06,
+      "loss": 0.7467,
+      "step": 15184
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 15.374015808105469,
+      "learning_rate": 7.295318579257834e-06,
+      "loss": 1.0972,
+      "step": 15185
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.549311637878418,
+      "learning_rate": 7.294481864201147e-06,
+      "loss": 0.3452,
+      "step": 15186
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 16.02145004272461,
+      "learning_rate": 7.293645149144459e-06,
+      "loss": 0.3838,
+      "step": 15187
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 20.3539981842041,
+      "learning_rate": 7.292808434087773e-06,
+      "loss": 2.1246,
+      "step": 15188
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.821852684020996,
+      "learning_rate": 7.291971719031085e-06,
+      "loss": 0.5058,
+      "step": 15189
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 12.831040382385254,
+      "learning_rate": 7.291135003974397e-06,
+      "loss": 0.2566,
+      "step": 15190
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 15.333756446838379,
+      "learning_rate": 7.29029828891771e-06,
+      "loss": 2.0446,
+      "step": 15191
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 44.381813049316406,
+      "learning_rate": 7.289461573861022e-06,
+      "loss": 2.7031,
+      "step": 15192
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 52.53483963012695,
+      "learning_rate": 7.288624858804335e-06,
+      "loss": 1.6683,
+      "step": 15193
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 23.234952926635742,
+      "learning_rate": 7.287788143747647e-06,
+      "loss": 0.887,
+      "step": 15194
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 119.96180725097656,
+      "learning_rate": 7.286951428690961e-06,
+      "loss": 1.6006,
+      "step": 15195
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 18.631879806518555,
+      "learning_rate": 7.2861147136342726e-06,
+      "loss": 1.9097,
+      "step": 15196
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 50.981788635253906,
+      "learning_rate": 7.2852779985775845e-06,
+      "loss": 1.9735,
+      "step": 15197
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.808567523956299,
+      "learning_rate": 7.284441283520897e-06,
+      "loss": 0.3778,
+      "step": 15198
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 17.7376651763916,
+      "learning_rate": 7.28360456846421e-06,
+      "loss": 0.9597,
+      "step": 15199
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 142.1316375732422,
+      "learning_rate": 7.282767853407523e-06,
+      "loss": 0.846,
+      "step": 15200
+    },
+    {
+      "epoch": 1.91,
+      "eval_loss": 0.07876706123352051,
+      "eval_runtime": 96.4561,
+      "eval_samples_per_second": 36.721,
+      "eval_steps_per_second": 36.721,
+      "step": 15200
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 6.942018032073975,
+      "learning_rate": 7.281931138350835e-06,
+      "loss": 0.2998,
+      "step": 15201
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 39.050254821777344,
+      "learning_rate": 7.2810944232941485e-06,
+      "loss": 1.4191,
+      "step": 15202
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 12.845418930053711,
+      "learning_rate": 7.2802577082374605e-06,
+      "loss": 0.9449,
+      "step": 15203
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 237.6184539794922,
+      "learning_rate": 7.2794209931807724e-06,
+      "loss": 3.3873,
+      "step": 15204
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 991.9547119140625,
+      "learning_rate": 7.278584278124085e-06,
+      "loss": 1.8497,
+      "step": 15205
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 22.550867080688477,
+      "learning_rate": 7.277747563067398e-06,
+      "loss": 0.6088,
+      "step": 15206
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 15.08200740814209,
+      "learning_rate": 7.276910848010711e-06,
+      "loss": 2.0656,
+      "step": 15207
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.465595722198486,
+      "learning_rate": 7.276074132954023e-06,
+      "loss": 0.1612,
+      "step": 15208
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 21.401498794555664,
+      "learning_rate": 7.2752374178973365e-06,
+      "loss": 1.6728,
+      "step": 15209
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 25.57209587097168,
+      "learning_rate": 7.274400702840648e-06,
+      "loss": 1.3603,
+      "step": 15210
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.294596195220947,
+      "learning_rate": 7.27356398778396e-06,
+      "loss": 0.1678,
+      "step": 15211
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 56.66082763671875,
+      "learning_rate": 7.272727272727273e-06,
+      "loss": 2.7304,
+      "step": 15212
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 11.751508712768555,
+      "learning_rate": 7.271890557670586e-06,
+      "loss": 0.881,
+      "step": 15213
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 20.95166778564453,
+      "learning_rate": 7.271053842613899e-06,
+      "loss": 1.5158,
+      "step": 15214
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 16.030107498168945,
+      "learning_rate": 7.270217127557211e-06,
+      "loss": 0.8177,
+      "step": 15215
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 10.919445991516113,
+      "learning_rate": 7.269380412500524e-06,
+      "loss": 1.8145,
+      "step": 15216
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 36.17577362060547,
+      "learning_rate": 7.268543697443836e-06,
+      "loss": 0.9956,
+      "step": 15217
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 11.49276351928711,
+      "learning_rate": 7.267706982387148e-06,
+      "loss": 2.2279,
+      "step": 15218
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 11.070647239685059,
+      "learning_rate": 7.266870267330461e-06,
+      "loss": 0.3389,
+      "step": 15219
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 19.18267059326172,
+      "learning_rate": 7.266033552273774e-06,
+      "loss": 1.5976,
+      "step": 15220
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 12.499030113220215,
+      "learning_rate": 7.265196837217087e-06,
+      "loss": 0.986,
+      "step": 15221
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 14.461274147033691,
+      "learning_rate": 7.264360122160399e-06,
+      "loss": 0.7536,
+      "step": 15222
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 22.120290756225586,
+      "learning_rate": 7.263523407103712e-06,
+      "loss": 1.0406,
+      "step": 15223
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 43.58599090576172,
+      "learning_rate": 7.262686692047024e-06,
+      "loss": 3.0559,
+      "step": 15224
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 55.521148681640625,
+      "learning_rate": 7.261849976990336e-06,
+      "loss": 1.18,
+      "step": 15225
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 35.34981918334961,
+      "learning_rate": 7.261013261933649e-06,
+      "loss": 1.4666,
+      "step": 15226
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 6.730340003967285,
+      "learning_rate": 7.260176546876962e-06,
+      "loss": 0.325,
+      "step": 15227
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.402364730834961,
+      "learning_rate": 7.259339831820275e-06,
+      "loss": 0.263,
+      "step": 15228
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 10.066123962402344,
+      "learning_rate": 7.2585031167635866e-06,
+      "loss": 0.4633,
+      "step": 15229
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 40.3651008605957,
+      "learning_rate": 7.2576664017069e-06,
+      "loss": 1.5189,
+      "step": 15230
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 18.799203872680664,
+      "learning_rate": 7.256829686650212e-06,
+      "loss": 1.1197,
+      "step": 15231
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 15.619763374328613,
+      "learning_rate": 7.255992971593524e-06,
+      "loss": 2.1876,
+      "step": 15232
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 12.761515617370605,
+      "learning_rate": 7.255156256536837e-06,
+      "loss": 0.7248,
+      "step": 15233
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 11.928339004516602,
+      "learning_rate": 7.254319541480149e-06,
+      "loss": 0.4087,
+      "step": 15234
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 23.60286521911621,
+      "learning_rate": 7.2534828264234625e-06,
+      "loss": 1.801,
+      "step": 15235
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.796131134033203,
+      "learning_rate": 7.2526461113667745e-06,
+      "loss": 0.2426,
+      "step": 15236
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 54.70196533203125,
+      "learning_rate": 7.251809396310088e-06,
+      "loss": 3.1836,
+      "step": 15237
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 79.63700103759766,
+      "learning_rate": 7.2509726812534e-06,
+      "loss": 1.9657,
+      "step": 15238
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.026930809020996,
+      "learning_rate": 7.250135966196712e-06,
+      "loss": 0.7326,
+      "step": 15239
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 10.02031135559082,
+      "learning_rate": 7.249299251140025e-06,
+      "loss": 1.1231,
+      "step": 15240
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 21.590858459472656,
+      "learning_rate": 7.248462536083337e-06,
+      "loss": 3.2721,
+      "step": 15241
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 12.597800254821777,
+      "learning_rate": 7.2476258210266504e-06,
+      "loss": 1.2757,
+      "step": 15242
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 42.85508728027344,
+      "learning_rate": 7.246789105969962e-06,
+      "loss": 0.5524,
+      "step": 15243
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 12.83395004272461,
+      "learning_rate": 7.245952390913275e-06,
+      "loss": 2.0516,
+      "step": 15244
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 24.339702606201172,
+      "learning_rate": 7.245115675856588e-06,
+      "loss": 2.0179,
+      "step": 15245
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 160.06024169921875,
+      "learning_rate": 7.2442789607999e-06,
+      "loss": 3.1359,
+      "step": 15246
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 16.193593978881836,
+      "learning_rate": 7.243442245743213e-06,
+      "loss": 0.4034,
+      "step": 15247
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 49.94191360473633,
+      "learning_rate": 7.242605530686525e-06,
+      "loss": 2.3597,
+      "step": 15248
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 12.985690116882324,
+      "learning_rate": 7.241768815629838e-06,
+      "loss": 0.4545,
+      "step": 15249
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 19.354537963867188,
+      "learning_rate": 7.24093210057315e-06,
+      "loss": 1.873,
+      "step": 15250
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 13.181794166564941,
+      "learning_rate": 7.240095385516463e-06,
+      "loss": 1.2042,
+      "step": 15251
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 61.11602020263672,
+      "learning_rate": 7.239258670459776e-06,
+      "loss": 0.8935,
+      "step": 15252
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 18.95747184753418,
+      "learning_rate": 7.238421955403088e-06,
+      "loss": 1.0299,
+      "step": 15253
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 9.337152481079102,
+      "learning_rate": 7.237585240346401e-06,
+      "loss": 1.2961,
+      "step": 15254
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.195521354675293,
+      "learning_rate": 7.236748525289713e-06,
+      "loss": 2.0989,
+      "step": 15255
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 12.675674438476562,
+      "learning_rate": 7.235911810233026e-06,
+      "loss": 1.7304,
+      "step": 15256
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 8.78879451751709,
+      "learning_rate": 7.235075095176338e-06,
+      "loss": 0.2985,
+      "step": 15257
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 18.082355499267578,
+      "learning_rate": 7.23423838011965e-06,
+      "loss": 1.065,
+      "step": 15258
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 24.055559158325195,
+      "learning_rate": 7.233401665062964e-06,
+      "loss": 0.9967,
+      "step": 15259
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 11.412266731262207,
+      "learning_rate": 7.232564950006276e-06,
+      "loss": 1.2657,
+      "step": 15260
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 5.17311954498291,
+      "learning_rate": 7.231728234949589e-06,
+      "loss": 0.6104,
+      "step": 15261
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.313230991363525,
+      "learning_rate": 7.2308915198929006e-06,
+      "loss": 0.3799,
+      "step": 15262
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.846939086914062,
+      "learning_rate": 7.230054804836214e-06,
+      "loss": 0.9638,
+      "step": 15263
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 12.357636451721191,
+      "learning_rate": 7.229218089779526e-06,
+      "loss": 1.0083,
+      "step": 15264
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.515074729919434,
+      "learning_rate": 7.228381374722838e-06,
+      "loss": 0.3895,
+      "step": 15265
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 26.114919662475586,
+      "learning_rate": 7.227544659666152e-06,
+      "loss": 1.9338,
+      "step": 15266
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 15.958468437194824,
+      "learning_rate": 7.226707944609464e-06,
+      "loss": 0.7954,
+      "step": 15267
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 7.361175537109375,
+      "learning_rate": 7.2258712295527765e-06,
+      "loss": 0.2709,
+      "step": 15268
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.929686546325684,
+      "learning_rate": 7.2250345144960885e-06,
+      "loss": 0.7763,
+      "step": 15269
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 30.771678924560547,
+      "learning_rate": 7.224197799439402e-06,
+      "loss": 0.9929,
+      "step": 15270
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 16.85222816467285,
+      "learning_rate": 7.223361084382714e-06,
+      "loss": 1.3375,
+      "step": 15271
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 18.63667869567871,
+      "learning_rate": 7.222524369326026e-06,
+      "loss": 1.1272,
+      "step": 15272
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 8.972474098205566,
+      "learning_rate": 7.22168765426934e-06,
+      "loss": 0.5789,
+      "step": 15273
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 17.44340705871582,
+      "learning_rate": 7.220850939212652e-06,
+      "loss": 1.8272,
+      "step": 15274
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 16.283367156982422,
+      "learning_rate": 7.2200142241559644e-06,
+      "loss": 0.3297,
+      "step": 15275
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.77719646692276,
+      "learning_rate": 7.219177509099276e-06,
+      "loss": 0.0125,
+      "step": 15276
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 18.31374740600586,
+      "learning_rate": 7.21834079404259e-06,
+      "loss": 3.0636,
+      "step": 15277
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 50.88170623779297,
+      "learning_rate": 7.217504078985902e-06,
+      "loss": 2.3115,
+      "step": 15278
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 78.62443542480469,
+      "learning_rate": 7.216667363929214e-06,
+      "loss": 0.7992,
+      "step": 15279
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 16.376123428344727,
+      "learning_rate": 7.215830648872527e-06,
+      "loss": 0.6161,
+      "step": 15280
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 27.0351619720459,
+      "learning_rate": 7.2149939338158396e-06,
+      "loss": 1.6892,
+      "step": 15281
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 13.475540161132812,
+      "learning_rate": 7.214157218759152e-06,
+      "loss": 0.7147,
+      "step": 15282
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 18.50657081604004,
+      "learning_rate": 7.213320503702464e-06,
+      "loss": 1.8595,
+      "step": 15283
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 11.895668983459473,
+      "learning_rate": 7.212483788645778e-06,
+      "loss": 0.3789,
+      "step": 15284
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.596177101135254,
+      "learning_rate": 7.21164707358909e-06,
+      "loss": 1.4801,
+      "step": 15285
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 7.016933441162109,
+      "learning_rate": 7.210810358532402e-06,
+      "loss": 1.1085,
+      "step": 15286
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 8.173614501953125,
+      "learning_rate": 7.209973643475715e-06,
+      "loss": 0.6395,
+      "step": 15287
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 16.760009765625,
+      "learning_rate": 7.2091369284190275e-06,
+      "loss": 0.949,
+      "step": 15288
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 47.62588119506836,
+      "learning_rate": 7.20830021336234e-06,
+      "loss": 2.1585,
+      "step": 15289
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.460638999938965,
+      "learning_rate": 7.207463498305652e-06,
+      "loss": 1.4834,
+      "step": 15290
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 16.595172882080078,
+      "learning_rate": 7.206626783248966e-06,
+      "loss": 0.5579,
+      "step": 15291
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 16.40092658996582,
+      "learning_rate": 7.205790068192278e-06,
+      "loss": 1.733,
+      "step": 15292
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 45.9400520324707,
+      "learning_rate": 7.20495335313559e-06,
+      "loss": 1.4993,
+      "step": 15293
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 4.556918144226074,
+      "learning_rate": 7.204116638078903e-06,
+      "loss": 0.093,
+      "step": 15294
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.307995557785034,
+      "learning_rate": 7.203279923022215e-06,
+      "loss": 0.2106,
+      "step": 15295
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 24.323028564453125,
+      "learning_rate": 7.202443207965528e-06,
+      "loss": 0.9309,
+      "step": 15296
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 13.639983177185059,
+      "learning_rate": 7.20160649290884e-06,
+      "loss": 1.3806,
+      "step": 15297
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 19.285390853881836,
+      "learning_rate": 7.200769777852154e-06,
+      "loss": 0.9875,
+      "step": 15298
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.064143180847168,
+      "learning_rate": 7.199933062795466e-06,
+      "loss": 1.4574,
+      "step": 15299
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 16.76412582397461,
+      "learning_rate": 7.199096347738778e-06,
+      "loss": 1.0419,
+      "step": 15300
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 33.953739166259766,
+      "learning_rate": 7.1982596326820905e-06,
+      "loss": 2.0381,
+      "step": 15301
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 12.692020416259766,
+      "learning_rate": 7.197422917625403e-06,
+      "loss": 0.7316,
+      "step": 15302
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 28.04511260986328,
+      "learning_rate": 7.196586202568716e-06,
+      "loss": 2.1434,
+      "step": 15303
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.163270473480225,
+      "learning_rate": 7.195749487512028e-06,
+      "loss": 0.8395,
+      "step": 15304
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.6385579109191895,
+      "learning_rate": 7.194912772455342e-06,
+      "loss": 0.3598,
+      "step": 15305
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 19.119504928588867,
+      "learning_rate": 7.194076057398654e-06,
+      "loss": 2.1298,
+      "step": 15306
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 7.362306594848633,
+      "learning_rate": 7.193239342341966e-06,
+      "loss": 0.4124,
+      "step": 15307
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.695954322814941,
+      "learning_rate": 7.1924026272852784e-06,
+      "loss": 0.2428,
+      "step": 15308
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 17.593093872070312,
+      "learning_rate": 7.19156591222859e-06,
+      "loss": 1.2185,
+      "step": 15309
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 11.084668159484863,
+      "learning_rate": 7.190729197171904e-06,
+      "loss": 1.2277,
+      "step": 15310
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 15.913836479187012,
+      "learning_rate": 7.189892482115216e-06,
+      "loss": 0.9599,
+      "step": 15311
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 8.229795455932617,
+      "learning_rate": 7.18905576705853e-06,
+      "loss": 0.5386,
+      "step": 15312
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 19.191814422607422,
+      "learning_rate": 7.188219052001842e-06,
+      "loss": 1.045,
+      "step": 15313
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 22.096803665161133,
+      "learning_rate": 7.1873823369451535e-06,
+      "loss": 1.4529,
+      "step": 15314
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 18.706634521484375,
+      "learning_rate": 7.186545621888466e-06,
+      "loss": 0.8223,
+      "step": 15315
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.793954849243164,
+      "learning_rate": 7.185708906831778e-06,
+      "loss": 0.1147,
+      "step": 15316
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 13.102605819702148,
+      "learning_rate": 7.184872191775092e-06,
+      "loss": 0.7827,
+      "step": 15317
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.377617835998535,
+      "learning_rate": 7.184035476718404e-06,
+      "loss": 0.4478,
+      "step": 15318
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 22.061677932739258,
+      "learning_rate": 7.1831987616617175e-06,
+      "loss": 0.964,
+      "step": 15319
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 7.3718719482421875,
+      "learning_rate": 7.1823620466050295e-06,
+      "loss": 0.9412,
+      "step": 15320
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 92.04766082763672,
+      "learning_rate": 7.1815253315483415e-06,
+      "loss": 1.7727,
+      "step": 15321
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 13.109322547912598,
+      "learning_rate": 7.180688616491654e-06,
+      "loss": 1.5429,
+      "step": 15322
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 22.95648765563965,
+      "learning_rate": 7.179851901434966e-06,
+      "loss": 1.0019,
+      "step": 15323
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.840642929077148,
+      "learning_rate": 7.17901518637828e-06,
+      "loss": 1.2736,
+      "step": 15324
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.248290061950684,
+      "learning_rate": 7.178178471321592e-06,
+      "loss": 1.3105,
+      "step": 15325
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 15.547542572021484,
+      "learning_rate": 7.177341756264905e-06,
+      "loss": 0.5836,
+      "step": 15326
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.253751277923584,
+      "learning_rate": 7.176505041208217e-06,
+      "loss": 0.0978,
+      "step": 15327
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.300126552581787,
+      "learning_rate": 7.175668326151529e-06,
+      "loss": 1.7964,
+      "step": 15328
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 14.57581615447998,
+      "learning_rate": 7.174831611094842e-06,
+      "loss": 0.7965,
+      "step": 15329
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 35.33696746826172,
+      "learning_rate": 7.173994896038154e-06,
+      "loss": 3.6189,
+      "step": 15330
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 4.974849224090576,
+      "learning_rate": 7.173158180981468e-06,
+      "loss": 0.6593,
+      "step": 15331
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 15.684173583984375,
+      "learning_rate": 7.17232146592478e-06,
+      "loss": 2.0062,
+      "step": 15332
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.59436321258545,
+      "learning_rate": 7.1714847508680925e-06,
+      "loss": 0.7349,
+      "step": 15333
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 24.182329177856445,
+      "learning_rate": 7.170648035811405e-06,
+      "loss": 1.2789,
+      "step": 15334
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 20.643909454345703,
+      "learning_rate": 7.169811320754717e-06,
+      "loss": 0.5253,
+      "step": 15335
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 15.558892250061035,
+      "learning_rate": 7.16897460569803e-06,
+      "loss": 1.6174,
+      "step": 15336
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 36.112823486328125,
+      "learning_rate": 7.168137890641342e-06,
+      "loss": 0.7389,
+      "step": 15337
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 25.463340759277344,
+      "learning_rate": 7.167301175584656e-06,
+      "loss": 1.8058,
+      "step": 15338
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 7.937673091888428,
+      "learning_rate": 7.166464460527968e-06,
+      "loss": 0.1034,
+      "step": 15339
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.738503456115723,
+      "learning_rate": 7.1656277454712805e-06,
+      "loss": 0.548,
+      "step": 15340
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.6526103019714355,
+      "learning_rate": 7.164791030414593e-06,
+      "loss": 0.3518,
+      "step": 15341
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 25.582611083984375,
+      "learning_rate": 7.163954315357905e-06,
+      "loss": 0.8825,
+      "step": 15342
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 15.33735466003418,
+      "learning_rate": 7.163117600301218e-06,
+      "loss": 1.0733,
+      "step": 15343
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 15.540820121765137,
+      "learning_rate": 7.16228088524453e-06,
+      "loss": 1.082,
+      "step": 15344
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 14.650261878967285,
+      "learning_rate": 7.161444170187844e-06,
+      "loss": 1.4717,
+      "step": 15345
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 15.091980934143066,
+      "learning_rate": 7.1606074551311556e-06,
+      "loss": 0.8659,
+      "step": 15346
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.12157154083252,
+      "learning_rate": 7.159770740074468e-06,
+      "loss": 0.437,
+      "step": 15347
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.351181983947754,
+      "learning_rate": 7.158934025017781e-06,
+      "loss": 1.6069,
+      "step": 15348
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 90.13201904296875,
+      "learning_rate": 7.158097309961093e-06,
+      "loss": 3.6957,
+      "step": 15349
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 10.546927452087402,
+      "learning_rate": 7.157260594904406e-06,
+      "loss": 0.5513,
+      "step": 15350
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 19.79357147216797,
+      "learning_rate": 7.156423879847718e-06,
+      "loss": 1.0419,
+      "step": 15351
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.656234741210938,
+      "learning_rate": 7.1555871647910315e-06,
+      "loss": 2.2094,
+      "step": 15352
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 31.641164779663086,
+      "learning_rate": 7.1547504497343435e-06,
+      "loss": 1.1596,
+      "step": 15353
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 17.413711547851562,
+      "learning_rate": 7.153913734677656e-06,
+      "loss": 1.6971,
+      "step": 15354
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 4.659546375274658,
+      "learning_rate": 7.153077019620969e-06,
+      "loss": 1.0103,
+      "step": 15355
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 4.0066752433776855,
+      "learning_rate": 7.152240304564281e-06,
+      "loss": 0.1789,
+      "step": 15356
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.486143589019775,
+      "learning_rate": 7.151403589507594e-06,
+      "loss": 0.3966,
+      "step": 15357
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.760356426239014,
+      "learning_rate": 7.150566874450906e-06,
+      "loss": 1.0819,
+      "step": 15358
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 12.972009658813477,
+      "learning_rate": 7.1497301593942195e-06,
+      "loss": 0.9274,
+      "step": 15359
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 21.619810104370117,
+      "learning_rate": 7.148893444337531e-06,
+      "loss": 1.5404,
+      "step": 15360
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 4.129205226898193,
+      "learning_rate": 7.148056729280844e-06,
+      "loss": 0.3018,
+      "step": 15361
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.901113510131836,
+      "learning_rate": 7.147220014224156e-06,
+      "loss": 0.748,
+      "step": 15362
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 12.46020793914795,
+      "learning_rate": 7.146383299167469e-06,
+      "loss": 0.5956,
+      "step": 15363
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.097681045532227,
+      "learning_rate": 7.145546584110782e-06,
+      "loss": 1.1116,
+      "step": 15364
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.668359756469727,
+      "learning_rate": 7.144709869054094e-06,
+      "loss": 1.9949,
+      "step": 15365
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 7.611221790313721,
+      "learning_rate": 7.143873153997407e-06,
+      "loss": 0.3213,
+      "step": 15366
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 64.34578704833984,
+      "learning_rate": 7.143036438940719e-06,
+      "loss": 2.5571,
+      "step": 15367
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 7.546182632446289,
+      "learning_rate": 7.142199723884032e-06,
+      "loss": 2.1841,
+      "step": 15368
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.2903261184692383,
+      "learning_rate": 7.141363008827344e-06,
+      "loss": 0.2241,
+      "step": 15369
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 15.703153610229492,
+      "learning_rate": 7.140526293770657e-06,
+      "loss": 0.8884,
+      "step": 15370
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.410683631896973,
+      "learning_rate": 7.13968957871397e-06,
+      "loss": 0.4108,
+      "step": 15371
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 7.17274808883667,
+      "learning_rate": 7.138852863657282e-06,
+      "loss": 0.4658,
+      "step": 15372
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 7.382733345031738,
+      "learning_rate": 7.138016148600595e-06,
+      "loss": 0.8086,
+      "step": 15373
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.851573944091797,
+      "learning_rate": 7.137179433543907e-06,
+      "loss": 1.2607,
+      "step": 15374
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.1019930839538574,
+      "learning_rate": 7.13634271848722e-06,
+      "loss": 0.0266,
+      "step": 15375
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 18.079978942871094,
+      "learning_rate": 7.135506003430532e-06,
+      "loss": 2.1212,
+      "step": 15376
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 59.86653518676758,
+      "learning_rate": 7.134669288373845e-06,
+      "loss": 1.968,
+      "step": 15377
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 17.122314453125,
+      "learning_rate": 7.133832573317158e-06,
+      "loss": 0.509,
+      "step": 15378
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 29.334320068359375,
+      "learning_rate": 7.1329958582604696e-06,
+      "loss": 2.0254,
+      "step": 15379
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.485424041748047,
+      "learning_rate": 7.132159143203783e-06,
+      "loss": 0.9514,
+      "step": 15380
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 6.937499046325684,
+      "learning_rate": 7.131322428147095e-06,
+      "loss": 0.4008,
+      "step": 15381
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.3256618976593018,
+      "learning_rate": 7.130485713090408e-06,
+      "loss": 0.1147,
+      "step": 15382
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.174434661865234,
+      "learning_rate": 7.12964899803372e-06,
+      "loss": 0.3715,
+      "step": 15383
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.014540672302246,
+      "learning_rate": 7.128812282977033e-06,
+      "loss": 1.8294,
+      "step": 15384
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 60.12565231323242,
+      "learning_rate": 7.1279755679203455e-06,
+      "loss": 1.9883,
+      "step": 15385
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 15.926053047180176,
+      "learning_rate": 7.1271388528636575e-06,
+      "loss": 0.7368,
+      "step": 15386
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 20.925365447998047,
+      "learning_rate": 7.126302137806971e-06,
+      "loss": 1.8686,
+      "step": 15387
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 12.295696258544922,
+      "learning_rate": 7.125465422750283e-06,
+      "loss": 1.3463,
+      "step": 15388
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 20.267383575439453,
+      "learning_rate": 7.124628707693596e-06,
+      "loss": 0.6291,
+      "step": 15389
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 15.206001281738281,
+      "learning_rate": 7.123791992636908e-06,
+      "loss": 1.6648,
+      "step": 15390
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 50.678199768066406,
+      "learning_rate": 7.12295527758022e-06,
+      "loss": 2.8122,
+      "step": 15391
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 10.255516052246094,
+      "learning_rate": 7.1221185625235334e-06,
+      "loss": 0.7144,
+      "step": 15392
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 18.288196563720703,
+      "learning_rate": 7.121281847466845e-06,
+      "loss": 1.5048,
+      "step": 15393
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 35.55363082885742,
+      "learning_rate": 7.120445132410159e-06,
+      "loss": 1.2026,
+      "step": 15394
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 6.697925090789795,
+      "learning_rate": 7.119608417353471e-06,
+      "loss": 0.4907,
+      "step": 15395
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 18.958486557006836,
+      "learning_rate": 7.118771702296784e-06,
+      "loss": 0.8496,
+      "step": 15396
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 50.062774658203125,
+      "learning_rate": 7.117934987240096e-06,
+      "loss": 1.6842,
+      "step": 15397
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.4984216690063477,
+      "learning_rate": 7.117098272183408e-06,
+      "loss": 0.39,
+      "step": 15398
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 56.26967239379883,
+      "learning_rate": 7.116261557126721e-06,
+      "loss": 0.7481,
+      "step": 15399
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 13.155136108398438,
+      "learning_rate": 7.115424842070033e-06,
+      "loss": 0.7274,
+      "step": 15400
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 8.556282043457031,
+      "learning_rate": 7.114588127013347e-06,
+      "loss": 1.1603,
+      "step": 15401
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 6.849576473236084,
+      "learning_rate": 7.113751411956659e-06,
+      "loss": 0.2008,
+      "step": 15402
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 40.26840591430664,
+      "learning_rate": 7.112914696899972e-06,
+      "loss": 2.0927,
+      "step": 15403
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.60886287689209,
+      "learning_rate": 7.112077981843284e-06,
+      "loss": 0.7664,
+      "step": 15404
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 12.531184196472168,
+      "learning_rate": 7.111241266786596e-06,
+      "loss": 1.0722,
+      "step": 15405
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 17.331567764282227,
+      "learning_rate": 7.110404551729909e-06,
+      "loss": 0.843,
+      "step": 15406
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.1289303302764893,
+      "learning_rate": 7.109567836673221e-06,
+      "loss": 0.2729,
+      "step": 15407
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.882319450378418,
+      "learning_rate": 7.108731121616534e-06,
+      "loss": 0.8114,
+      "step": 15408
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.488268852233887,
+      "learning_rate": 7.107894406559847e-06,
+      "loss": 0.4272,
+      "step": 15409
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 30.07292938232422,
+      "learning_rate": 7.10705769150316e-06,
+      "loss": 1.7169,
+      "step": 15410
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 17.756690979003906,
+      "learning_rate": 7.106220976446472e-06,
+      "loss": 0.7914,
+      "step": 15411
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 68.1766357421875,
+      "learning_rate": 7.1053842613897836e-06,
+      "loss": 0.6702,
+      "step": 15412
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 4.996368408203125,
+      "learning_rate": 7.104547546333097e-06,
+      "loss": 0.3774,
+      "step": 15413
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 12.997795104980469,
+      "learning_rate": 7.103710831276409e-06,
+      "loss": 0.9306,
+      "step": 15414
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 15.791701316833496,
+      "learning_rate": 7.102874116219722e-06,
+      "loss": 0.5919,
+      "step": 15415
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 15.011215209960938,
+      "learning_rate": 7.102037401163035e-06,
+      "loss": 2.4547,
+      "step": 15416
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 12.556661605834961,
+      "learning_rate": 7.1012006861063476e-06,
+      "loss": 1.3745,
+      "step": 15417
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 11.91063404083252,
+      "learning_rate": 7.1003639710496595e-06,
+      "loss": 0.7513,
+      "step": 15418
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 7.577408790588379,
+      "learning_rate": 7.0995272559929715e-06,
+      "loss": 0.9176,
+      "step": 15419
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 13.84679889678955,
+      "learning_rate": 7.098690540936285e-06,
+      "loss": 1.835,
+      "step": 15420
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 35.05377197265625,
+      "learning_rate": 7.097853825879597e-06,
+      "loss": 1.8702,
+      "step": 15421
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 16.646114349365234,
+      "learning_rate": 7.09701711082291e-06,
+      "loss": 1.0236,
+      "step": 15422
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.494402885437012,
+      "learning_rate": 7.096180395766223e-06,
+      "loss": 1.6881,
+      "step": 15423
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 7.679769039154053,
+      "learning_rate": 7.0953436807095355e-06,
+      "loss": 0.4881,
+      "step": 15424
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.879209518432617,
+      "learning_rate": 7.0945069656528474e-06,
+      "loss": 1.7926,
+      "step": 15425
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 16.050445556640625,
+      "learning_rate": 7.093670250596159e-06,
+      "loss": 1.0691,
+      "step": 15426
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 12.42928695678711,
+      "learning_rate": 7.092833535539473e-06,
+      "loss": 0.7106,
+      "step": 15427
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 13.413458824157715,
+      "learning_rate": 7.091996820482785e-06,
+      "loss": 0.751,
+      "step": 15428
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 6.810462474822998,
+      "learning_rate": 7.091160105426098e-06,
+      "loss": 0.3828,
+      "step": 15429
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 30.17489242553711,
+      "learning_rate": 7.090323390369411e-06,
+      "loss": 1.5593,
+      "step": 15430
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 86.79495239257812,
+      "learning_rate": 7.0894866753127226e-06,
+      "loss": 0.6664,
+      "step": 15431
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 20.300350189208984,
+      "learning_rate": 7.088649960256035e-06,
+      "loss": 0.7502,
+      "step": 15432
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 17.606658935546875,
+      "learning_rate": 7.087813245199347e-06,
+      "loss": 1.5995,
+      "step": 15433
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 21.617136001586914,
+      "learning_rate": 7.086976530142661e-06,
+      "loss": 1.7476,
+      "step": 15434
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 22.607784271240234,
+      "learning_rate": 7.086139815085973e-06,
+      "loss": 0.8055,
+      "step": 15435
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 33.39226531982422,
+      "learning_rate": 7.085303100029286e-06,
+      "loss": 1.6133,
+      "step": 15436
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 78.84778594970703,
+      "learning_rate": 7.084466384972598e-06,
+      "loss": 3.1536,
+      "step": 15437
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.63171100616455,
+      "learning_rate": 7.0836296699159105e-06,
+      "loss": 0.9092,
+      "step": 15438
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 26.426563262939453,
+      "learning_rate": 7.082792954859223e-06,
+      "loss": 0.9172,
+      "step": 15439
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.196547031402588,
+      "learning_rate": 7.081956239802535e-06,
+      "loss": 0.6071,
+      "step": 15440
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.708215713500977,
+      "learning_rate": 7.081119524745849e-06,
+      "loss": 1.3831,
+      "step": 15441
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.377507209777832,
+      "learning_rate": 7.080282809689161e-06,
+      "loss": 0.7,
+      "step": 15442
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 27.26991844177246,
+      "learning_rate": 7.079446094632474e-06,
+      "loss": 0.5514,
+      "step": 15443
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.7058744430542,
+      "learning_rate": 7.078609379575786e-06,
+      "loss": 0.8448,
+      "step": 15444
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 27.420000076293945,
+      "learning_rate": 7.077772664519098e-06,
+      "loss": 1.2417,
+      "step": 15445
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 7.791304588317871,
+      "learning_rate": 7.076935949462411e-06,
+      "loss": 0.7116,
+      "step": 15446
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.022967338562012,
+      "learning_rate": 7.076099234405723e-06,
+      "loss": 0.8989,
+      "step": 15447
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 17.566761016845703,
+      "learning_rate": 7.075262519349037e-06,
+      "loss": 0.6213,
+      "step": 15448
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 13.799229621887207,
+      "learning_rate": 7.074425804292349e-06,
+      "loss": 1.849,
+      "step": 15449
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 184.754150390625,
+      "learning_rate": 7.0735890892356616e-06,
+      "loss": 1.5852,
+      "step": 15450
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.525916337966919,
+      "learning_rate": 7.0727523741789735e-06,
+      "loss": 0.0662,
+      "step": 15451
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.680980682373047,
+      "learning_rate": 7.071915659122286e-06,
+      "loss": 0.7585,
+      "step": 15452
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 89.53353881835938,
+      "learning_rate": 7.071078944065599e-06,
+      "loss": 2.3426,
+      "step": 15453
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 18.292312622070312,
+      "learning_rate": 7.070242229008911e-06,
+      "loss": 1.1013,
+      "step": 15454
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.780877113342285,
+      "learning_rate": 7.069405513952225e-06,
+      "loss": 0.9168,
+      "step": 15455
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.118886470794678,
+      "learning_rate": 7.068568798895537e-06,
+      "loss": 0.4283,
+      "step": 15456
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.497488975524902,
+      "learning_rate": 7.0677320838388495e-06,
+      "loss": 0.3166,
+      "step": 15457
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.876434326171875,
+      "learning_rate": 7.0668953687821614e-06,
+      "loss": 0.6062,
+      "step": 15458
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 7.251729488372803,
+      "learning_rate": 7.066058653725474e-06,
+      "loss": 0.2288,
+      "step": 15459
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 16.087635040283203,
+      "learning_rate": 7.065221938668787e-06,
+      "loss": 1.1124,
+      "step": 15460
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 13.209246635437012,
+      "learning_rate": 7.064385223612099e-06,
+      "loss": 1.7922,
+      "step": 15461
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 12.305509567260742,
+      "learning_rate": 7.063548508555413e-06,
+      "loss": 0.9409,
+      "step": 15462
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 7.604883193969727,
+      "learning_rate": 7.062711793498725e-06,
+      "loss": 0.244,
+      "step": 15463
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 60.780635833740234,
+      "learning_rate": 7.061875078442037e-06,
+      "loss": 0.7982,
+      "step": 15464
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 23.57979965209961,
+      "learning_rate": 7.061038363385349e-06,
+      "loss": 1.2766,
+      "step": 15465
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.719807624816895,
+      "learning_rate": 7.060201648328662e-06,
+      "loss": 1.5545,
+      "step": 15466
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.688767433166504,
+      "learning_rate": 7.059364933271975e-06,
+      "loss": 0.461,
+      "step": 15467
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.736171245574951,
+      "learning_rate": 7.058528218215287e-06,
+      "loss": 0.9425,
+      "step": 15468
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.748924255371094,
+      "learning_rate": 7.0576915031586005e-06,
+      "loss": 2.2647,
+      "step": 15469
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 16.284366607666016,
+      "learning_rate": 7.0568547881019125e-06,
+      "loss": 1.0216,
+      "step": 15470
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.771014213562012,
+      "learning_rate": 7.056018073045225e-06,
+      "loss": 0.3961,
+      "step": 15471
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 23.656824111938477,
+      "learning_rate": 7.055181357988537e-06,
+      "loss": 1.489,
+      "step": 15472
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.67492389678955,
+      "learning_rate": 7.054344642931849e-06,
+      "loss": 1.5355,
+      "step": 15473
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 22.204484939575195,
+      "learning_rate": 7.053507927875163e-06,
+      "loss": 1.4844,
+      "step": 15474
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 21.31995391845703,
+      "learning_rate": 7.052671212818475e-06,
+      "loss": 1.2159,
+      "step": 15475
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 6.922398090362549,
+      "learning_rate": 7.0518344977617885e-06,
+      "loss": 0.5267,
+      "step": 15476
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 22.273216247558594,
+      "learning_rate": 7.0509977827051e-06,
+      "loss": 1.3283,
+      "step": 15477
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 17.06501579284668,
+      "learning_rate": 7.050161067648413e-06,
+      "loss": 0.7504,
+      "step": 15478
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.22417163848877,
+      "learning_rate": 7.049324352591725e-06,
+      "loss": 0.4198,
+      "step": 15479
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.979037284851074,
+      "learning_rate": 7.048487637535037e-06,
+      "loss": 0.4843,
+      "step": 15480
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 165.80557250976562,
+      "learning_rate": 7.047650922478351e-06,
+      "loss": 0.6725,
+      "step": 15481
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.835333824157715,
+      "learning_rate": 7.046814207421663e-06,
+      "loss": 0.8138,
+      "step": 15482
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 13.948822021484375,
+      "learning_rate": 7.045977492364976e-06,
+      "loss": 0.1694,
+      "step": 15483
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 13.02929401397705,
+      "learning_rate": 7.045140777308288e-06,
+      "loss": 0.6899,
+      "step": 15484
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.320096492767334,
+      "learning_rate": 7.044304062251601e-06,
+      "loss": 0.1951,
+      "step": 15485
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 15.157958984375,
+      "learning_rate": 7.043467347194913e-06,
+      "loss": 1.6232,
+      "step": 15486
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 23.764209747314453,
+      "learning_rate": 7.042630632138225e-06,
+      "loss": 2.2997,
+      "step": 15487
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 12.446185111999512,
+      "learning_rate": 7.041793917081539e-06,
+      "loss": 0.6686,
+      "step": 15488
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.41409969329834,
+      "learning_rate": 7.040957202024851e-06,
+      "loss": 0.9177,
+      "step": 15489
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 6.236893653869629,
+      "learning_rate": 7.0401204869681635e-06,
+      "loss": 0.255,
+      "step": 15490
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.5406250953674316,
+      "learning_rate": 7.039283771911476e-06,
+      "loss": 0.1584,
+      "step": 15491
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 9.143668174743652,
+      "learning_rate": 7.038447056854789e-06,
+      "loss": 0.6818,
+      "step": 15492
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 13.777385711669922,
+      "learning_rate": 7.037610341798101e-06,
+      "loss": 1.0515,
+      "step": 15493
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 6.954865455627441,
+      "learning_rate": 7.036773626741413e-06,
+      "loss": 0.4036,
+      "step": 15494
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 14.621844291687012,
+      "learning_rate": 7.035936911684727e-06,
+      "loss": 1.7383,
+      "step": 15495
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 19.72504234313965,
+      "learning_rate": 7.0351001966280386e-06,
+      "loss": 0.2982,
+      "step": 15496
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.424211502075195,
+      "learning_rate": 7.034263481571351e-06,
+      "loss": 1.3376,
+      "step": 15497
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 51.879905700683594,
+      "learning_rate": 7.033426766514664e-06,
+      "loss": 0.5933,
+      "step": 15498
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 15.023542404174805,
+      "learning_rate": 7.032590051457977e-06,
+      "loss": 2.1166,
+      "step": 15499
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 25.0118350982666,
+      "learning_rate": 7.031753336401289e-06,
+      "loss": 0.8506,
+      "step": 15500
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 20.47811508178711,
+      "learning_rate": 7.030916621344601e-06,
+      "loss": 1.2114,
+      "step": 15501
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.80174446105957,
+      "learning_rate": 7.0300799062879145e-06,
+      "loss": 0.302,
+      "step": 15502
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 67.72651672363281,
+      "learning_rate": 7.0292431912312265e-06,
+      "loss": 0.9023,
+      "step": 15503
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 17.172203063964844,
+      "learning_rate": 7.028406476174539e-06,
+      "loss": 0.9231,
+      "step": 15504
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 3.053865432739258,
+      "learning_rate": 7.027569761117852e-06,
+      "loss": 0.8876,
+      "step": 15505
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 19.792144775390625,
+      "learning_rate": 7.026733046061165e-06,
+      "loss": 2.6251,
+      "step": 15506
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 13.75473403930664,
+      "learning_rate": 7.025896331004477e-06,
+      "loss": 1.8659,
+      "step": 15507
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.364108085632324,
+      "learning_rate": 7.025059615947789e-06,
+      "loss": 1.7935,
+      "step": 15508
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.748884201049805,
+      "learning_rate": 7.0242229008911025e-06,
+      "loss": 1.4596,
+      "step": 15509
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 4.112232208251953,
+      "learning_rate": 7.023386185834414e-06,
+      "loss": 0.2395,
+      "step": 15510
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 6.716560363769531,
+      "learning_rate": 7.022549470777727e-06,
+      "loss": 0.4791,
+      "step": 15511
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 22.984079360961914,
+      "learning_rate": 7.02171275572104e-06,
+      "loss": 1.3597,
+      "step": 15512
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 17.0230655670166,
+      "learning_rate": 7.020876040664353e-06,
+      "loss": 1.5377,
+      "step": 15513
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 3.351734161376953,
+      "learning_rate": 7.020039325607665e-06,
+      "loss": 0.2617,
+      "step": 15514
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 242.02784729003906,
+      "learning_rate": 7.019202610550977e-06,
+      "loss": 2.3834,
+      "step": 15515
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 4.806171894073486,
+      "learning_rate": 7.01836589549429e-06,
+      "loss": 0.2289,
+      "step": 15516
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 42.979461669921875,
+      "learning_rate": 7.017529180437602e-06,
+      "loss": 1.302,
+      "step": 15517
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.605782508850098,
+      "learning_rate": 7.016692465380915e-06,
+      "loss": 0.9935,
+      "step": 15518
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.702014446258545,
+      "learning_rate": 7.015855750324227e-06,
+      "loss": 0.3763,
+      "step": 15519
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.059732437133789,
+      "learning_rate": 7.015019035267541e-06,
+      "loss": 0.7708,
+      "step": 15520
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 75.51557922363281,
+      "learning_rate": 7.014182320210853e-06,
+      "loss": 1.0951,
+      "step": 15521
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 14.990911483764648,
+      "learning_rate": 7.013345605154165e-06,
+      "loss": 0.8757,
+      "step": 15522
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 26.17129898071289,
+      "learning_rate": 7.012508890097478e-06,
+      "loss": 1.2145,
+      "step": 15523
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.07250690460205,
+      "learning_rate": 7.01167217504079e-06,
+      "loss": 0.9486,
+      "step": 15524
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 95.37599182128906,
+      "learning_rate": 7.010835459984103e-06,
+      "loss": 3.1057,
+      "step": 15525
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 67.44821166992188,
+      "learning_rate": 7.009998744927415e-06,
+      "loss": 1.2477,
+      "step": 15526
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 16.430702209472656,
+      "learning_rate": 7.009162029870729e-06,
+      "loss": 1.0081,
+      "step": 15527
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 11.817269325256348,
+      "learning_rate": 7.008325314814041e-06,
+      "loss": 1.8068,
+      "step": 15528
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.9634690284729004,
+      "learning_rate": 7.0074885997573526e-06,
+      "loss": 0.1884,
+      "step": 15529
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.671106338500977,
+      "learning_rate": 7.006651884700666e-06,
+      "loss": 0.5094,
+      "step": 15530
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 6.831871509552002,
+      "learning_rate": 7.005815169643978e-06,
+      "loss": 0.4738,
+      "step": 15531
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 12.323295593261719,
+      "learning_rate": 7.004978454587291e-06,
+      "loss": 0.2962,
+      "step": 15532
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.955921173095703,
+      "learning_rate": 7.004141739530603e-06,
+      "loss": 0.6173,
+      "step": 15533
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 25.807981491088867,
+      "learning_rate": 7.0033050244739166e-06,
+      "loss": 1.2629,
+      "step": 15534
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 33.78031921386719,
+      "learning_rate": 7.0024683094172285e-06,
+      "loss": 2.3431,
+      "step": 15535
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 24.45186424255371,
+      "learning_rate": 7.0016315943605405e-06,
+      "loss": 0.9759,
+      "step": 15536
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 42.53618240356445,
+      "learning_rate": 7.000794879303854e-06,
+      "loss": 1.7634,
+      "step": 15537
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 7.286814212799072,
+      "learning_rate": 6.999958164247166e-06,
+      "loss": 0.8893,
+      "step": 15538
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 14.286301612854004,
+      "learning_rate": 6.999121449190479e-06,
+      "loss": 0.481,
+      "step": 15539
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 21.10111427307129,
+      "learning_rate": 6.998284734133791e-06,
+      "loss": 0.7561,
+      "step": 15540
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 38.04764175415039,
+      "learning_rate": 6.9974480190771045e-06,
+      "loss": 1.5845,
+      "step": 15541
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 25.33977508544922,
+      "learning_rate": 6.9966113040204164e-06,
+      "loss": 1.8679,
+      "step": 15542
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 26.45148468017578,
+      "learning_rate": 6.995774588963728e-06,
+      "loss": 1.0318,
+      "step": 15543
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 16.335588455200195,
+      "learning_rate": 6.994937873907042e-06,
+      "loss": 0.973,
+      "step": 15544
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 12.946664810180664,
+      "learning_rate": 6.994101158850354e-06,
+      "loss": 1.432,
+      "step": 15545
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 15.203397750854492,
+      "learning_rate": 6.993264443793667e-06,
+      "loss": 0.7775,
+      "step": 15546
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 17.209043502807617,
+      "learning_rate": 6.992427728736979e-06,
+      "loss": 1.6214,
+      "step": 15547
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 4.763405799865723,
+      "learning_rate": 6.991591013680292e-06,
+      "loss": 0.0663,
+      "step": 15548
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 4.411322593688965,
+      "learning_rate": 6.990754298623604e-06,
+      "loss": 0.2267,
+      "step": 15549
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 31.33431053161621,
+      "learning_rate": 6.989917583566916e-06,
+      "loss": 2.5391,
+      "step": 15550
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.03928279876709,
+      "learning_rate": 6.98908086851023e-06,
+      "loss": 0.4988,
+      "step": 15551
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 13.892117500305176,
+      "learning_rate": 6.988244153453542e-06,
+      "loss": 1.6088,
+      "step": 15552
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 333.7330627441406,
+      "learning_rate": 6.987407438396855e-06,
+      "loss": 1.2483,
+      "step": 15553
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.675161361694336,
+      "learning_rate": 6.986570723340167e-06,
+      "loss": 1.1111,
+      "step": 15554
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 17.345584869384766,
+      "learning_rate": 6.98573400828348e-06,
+      "loss": 1.7745,
+      "step": 15555
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 19.145296096801758,
+      "learning_rate": 6.984897293226792e-06,
+      "loss": 1.1672,
+      "step": 15556
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 27.76471519470215,
+      "learning_rate": 6.984060578170104e-06,
+      "loss": 0.8067,
+      "step": 15557
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 14.955503463745117,
+      "learning_rate": 6.983223863113418e-06,
+      "loss": 1.2938,
+      "step": 15558
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 10.937143325805664,
+      "learning_rate": 6.98238714805673e-06,
+      "loss": 0.2304,
+      "step": 15559
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 16.57361602783203,
+      "learning_rate": 6.981550433000043e-06,
+      "loss": 1.6325,
+      "step": 15560
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 139.50778198242188,
+      "learning_rate": 6.980713717943355e-06,
+      "loss": 0.9943,
+      "step": 15561
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 8.616129875183105,
+      "learning_rate": 6.979877002886668e-06,
+      "loss": 0.3728,
+      "step": 15562
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 7.047756195068359,
+      "learning_rate": 6.97904028782998e-06,
+      "loss": 0.9883,
+      "step": 15563
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 15.066789627075195,
+      "learning_rate": 6.978203572773292e-06,
+      "loss": 0.3011,
+      "step": 15564
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.801277160644531,
+      "learning_rate": 6.977366857716605e-06,
+      "loss": 1.6511,
+      "step": 15565
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 32.564884185791016,
+      "learning_rate": 6.976530142659918e-06,
+      "loss": 1.3165,
+      "step": 15566
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 47.5469970703125,
+      "learning_rate": 6.9756934276032306e-06,
+      "loss": 2.0164,
+      "step": 15567
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 21.765548706054688,
+      "learning_rate": 6.9748567125465425e-06,
+      "loss": 1.201,
+      "step": 15568
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 27.66904067993164,
+      "learning_rate": 6.974019997489856e-06,
+      "loss": 1.2243,
+      "step": 15569
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 16.696565628051758,
+      "learning_rate": 6.973183282433168e-06,
+      "loss": 2.2312,
+      "step": 15570
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 26.649904251098633,
+      "learning_rate": 6.97234656737648e-06,
+      "loss": 1.092,
+      "step": 15571
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 24.76673126220703,
+      "learning_rate": 6.971509852319793e-06,
+      "loss": 2.7237,
+      "step": 15572
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 14.912168502807617,
+      "learning_rate": 6.970673137263106e-06,
+      "loss": 0.9972,
+      "step": 15573
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 14.813760757446289,
+      "learning_rate": 6.9698364222064185e-06,
+      "loss": 0.5662,
+      "step": 15574
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 19.716306686401367,
+      "learning_rate": 6.9689997071497304e-06,
+      "loss": 1.6188,
+      "step": 15575
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 11.275886535644531,
+      "learning_rate": 6.968162992093044e-06,
+      "loss": 1.7687,
+      "step": 15576
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 122.64093017578125,
+      "learning_rate": 6.967326277036356e-06,
+      "loss": 3.0464,
+      "step": 15577
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 17.656211853027344,
+      "learning_rate": 6.966489561979668e-06,
+      "loss": 0.9186,
+      "step": 15578
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 21.526639938354492,
+      "learning_rate": 6.965652846922981e-06,
+      "loss": 0.6122,
+      "step": 15579
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 15.666875839233398,
+      "learning_rate": 6.964816131866294e-06,
+      "loss": 0.8211,
+      "step": 15580
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 6.567947864532471,
+      "learning_rate": 6.963979416809606e-06,
+      "loss": 0.6006,
+      "step": 15581
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 11.897137641906738,
+      "learning_rate": 6.963142701752918e-06,
+      "loss": 0.2065,
+      "step": 15582
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 35.2044677734375,
+      "learning_rate": 6.962305986696232e-06,
+      "loss": 0.8952,
+      "step": 15583
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 36.7293815612793,
+      "learning_rate": 6.961469271639544e-06,
+      "loss": 0.5825,
+      "step": 15584
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 4.57098913192749,
+      "learning_rate": 6.960632556582856e-06,
+      "loss": 0.2379,
+      "step": 15585
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 12.477896690368652,
+      "learning_rate": 6.959795841526169e-06,
+      "loss": 1.7507,
+      "step": 15586
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 25.72513771057129,
+      "learning_rate": 6.9589591264694815e-06,
+      "loss": 0.7155,
+      "step": 15587
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 12.6240234375,
+      "learning_rate": 6.958122411412794e-06,
+      "loss": 0.4072,
+      "step": 15588
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 10.934189796447754,
+      "learning_rate": 6.957285696356106e-06,
+      "loss": 0.5685,
+      "step": 15589
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 169.14804077148438,
+      "learning_rate": 6.95644898129942e-06,
+      "loss": 2.121,
+      "step": 15590
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 20.99358367919922,
+      "learning_rate": 6.955612266242732e-06,
+      "loss": 1.0553,
+      "step": 15591
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.8657612800598145,
+      "learning_rate": 6.954775551186044e-06,
+      "loss": 0.6709,
+      "step": 15592
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 17.866134643554688,
+      "learning_rate": 6.953938836129357e-06,
+      "loss": 1.0087,
+      "step": 15593
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 45.2334098815918,
+      "learning_rate": 6.9531021210726694e-06,
+      "loss": 2.2569,
+      "step": 15594
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 20.872684478759766,
+      "learning_rate": 6.952265406015982e-06,
+      "loss": 1.8624,
+      "step": 15595
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.320897102355957,
+      "learning_rate": 6.951428690959294e-06,
+      "loss": 0.7625,
+      "step": 15596
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 21.983694076538086,
+      "learning_rate": 6.950591975902608e-06,
+      "loss": 1.1616,
+      "step": 15597
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 31.263404846191406,
+      "learning_rate": 6.94975526084592e-06,
+      "loss": 0.454,
+      "step": 15598
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 22.783296585083008,
+      "learning_rate": 6.948918545789232e-06,
+      "loss": 0.932,
+      "step": 15599
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 45.86775207519531,
+      "learning_rate": 6.9480818307325446e-06,
+      "loss": 1.7207,
+      "step": 15600
+    },
+    {
+      "epoch": 1.96,
+      "eval_loss": 0.08063507825136185,
+      "eval_runtime": 96.7731,
+      "eval_samples_per_second": 36.601,
+      "eval_steps_per_second": 36.601,
+      "step": 15600
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 55.06128692626953,
+      "learning_rate": 6.9472451156758565e-06,
+      "loss": 2.4919,
+      "step": 15601
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 12.562663078308105,
+      "learning_rate": 6.94640840061917e-06,
+      "loss": 0.6102,
+      "step": 15602
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 18.10681915283203,
+      "learning_rate": 6.945571685562482e-06,
+      "loss": 0.9363,
+      "step": 15603
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 10.809121131896973,
+      "learning_rate": 6.944734970505796e-06,
+      "loss": 0.9123,
+      "step": 15604
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 6.625796318054199,
+      "learning_rate": 6.943898255449108e-06,
+      "loss": 1.3045,
+      "step": 15605
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.727932453155518,
+      "learning_rate": 6.94306154039242e-06,
+      "loss": 1.7243,
+      "step": 15606
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 57.7634391784668,
+      "learning_rate": 6.9422248253357325e-06,
+      "loss": 4.2435,
+      "step": 15607
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.309063911437988,
+      "learning_rate": 6.9413881102790444e-06,
+      "loss": 0.6944,
+      "step": 15608
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 25.379261016845703,
+      "learning_rate": 6.940551395222358e-06,
+      "loss": 1.1329,
+      "step": 15609
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 13.84262752532959,
+      "learning_rate": 6.93971468016567e-06,
+      "loss": 1.0546,
+      "step": 15610
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 3.004401922225952,
+      "learning_rate": 6.938877965108982e-06,
+      "loss": 0.1816,
+      "step": 15611
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.947330951690674,
+      "learning_rate": 6.938041250052296e-06,
+      "loss": 0.3381,
+      "step": 15612
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 15.322650909423828,
+      "learning_rate": 6.937204534995608e-06,
+      "loss": 0.3763,
+      "step": 15613
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 26.744220733642578,
+      "learning_rate": 6.93636781993892e-06,
+      "loss": 1.0411,
+      "step": 15614
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 35.896907806396484,
+      "learning_rate": 6.935531104882232e-06,
+      "loss": 3.2921,
+      "step": 15615
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 17.5107479095459,
+      "learning_rate": 6.934694389825546e-06,
+      "loss": 2.53,
+      "step": 15616
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 13.85987377166748,
+      "learning_rate": 6.933857674768858e-06,
+      "loss": 1.5989,
+      "step": 15617
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 26.768171310424805,
+      "learning_rate": 6.93302095971217e-06,
+      "loss": 3.0122,
+      "step": 15618
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 4.744687080383301,
+      "learning_rate": 6.9321842446554835e-06,
+      "loss": 0.2583,
+      "step": 15619
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 11.324332237243652,
+      "learning_rate": 6.9313475295987955e-06,
+      "loss": 1.3009,
+      "step": 15620
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 4.5160698890686035,
+      "learning_rate": 6.930510814542108e-06,
+      "loss": 0.26,
+      "step": 15621
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.876502990722656,
+      "learning_rate": 6.92967409948542e-06,
+      "loss": 2.2473,
+      "step": 15622
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 22.673030853271484,
+      "learning_rate": 6.928837384428734e-06,
+      "loss": 1.5846,
+      "step": 15623
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 5.811694622039795,
+      "learning_rate": 6.928000669372046e-06,
+      "loss": 0.3244,
+      "step": 15624
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 18.304227828979492,
+      "learning_rate": 6.927163954315358e-06,
+      "loss": 0.5696,
+      "step": 15625
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 20.676616668701172,
+      "learning_rate": 6.9263272392586715e-06,
+      "loss": 1.7732,
+      "step": 15626
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 6.337315559387207,
+      "learning_rate": 6.925490524201983e-06,
+      "loss": 1.3696,
+      "step": 15627
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 16.525117874145508,
+      "learning_rate": 6.924653809145296e-06,
+      "loss": 1.1781,
+      "step": 15628
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 39.423526763916016,
+      "learning_rate": 6.923817094088608e-06,
+      "loss": 1.7619,
+      "step": 15629
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 15.524224281311035,
+      "learning_rate": 6.922980379031922e-06,
+      "loss": 1.4073,
+      "step": 15630
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 10.538905143737793,
+      "learning_rate": 6.922143663975234e-06,
+      "loss": 1.2483,
+      "step": 15631
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 18.400815963745117,
+      "learning_rate": 6.921306948918546e-06,
+      "loss": 0.7544,
+      "step": 15632
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 88.06678771972656,
+      "learning_rate": 6.920470233861859e-06,
+      "loss": 2.307,
+      "step": 15633
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 6.407633304595947,
+      "learning_rate": 6.919633518805171e-06,
+      "loss": 1.6535,
+      "step": 15634
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 11.588862419128418,
+      "learning_rate": 6.918796803748484e-06,
+      "loss": 0.3925,
+      "step": 15635
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.575936317443848,
+      "learning_rate": 6.917960088691796e-06,
+      "loss": 1.0875,
+      "step": 15636
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.416591167449951,
+      "learning_rate": 6.91712337363511e-06,
+      "loss": 0.4413,
+      "step": 15637
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 19.250289916992188,
+      "learning_rate": 6.916286658578422e-06,
+      "loss": 0.6748,
+      "step": 15638
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 155.1161651611328,
+      "learning_rate": 6.915449943521734e-06,
+      "loss": 1.7139,
+      "step": 15639
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.13044261932373,
+      "learning_rate": 6.914613228465047e-06,
+      "loss": 0.7617,
+      "step": 15640
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 10.329883575439453,
+      "learning_rate": 6.913776513408359e-06,
+      "loss": 1.1411,
+      "step": 15641
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 11.309494972229004,
+      "learning_rate": 6.912939798351672e-06,
+      "loss": 0.5356,
+      "step": 15642
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 71.26300811767578,
+      "learning_rate": 6.912103083294984e-06,
+      "loss": 0.5755,
+      "step": 15643
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 18.788625717163086,
+      "learning_rate": 6.911266368238298e-06,
+      "loss": 1.2436,
+      "step": 15644
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 12.986571311950684,
+      "learning_rate": 6.91042965318161e-06,
+      "loss": 0.4995,
+      "step": 15645
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 10.278299331665039,
+      "learning_rate": 6.9095929381249216e-06,
+      "loss": 1.0472,
+      "step": 15646
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 9.618620872497559,
+      "learning_rate": 6.908756223068234e-06,
+      "loss": 0.1579,
+      "step": 15647
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 19.59689712524414,
+      "learning_rate": 6.907919508011547e-06,
+      "loss": 1.0557,
+      "step": 15648
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.157246112823486,
+      "learning_rate": 6.90708279295486e-06,
+      "loss": 0.4243,
+      "step": 15649
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 12.754986763000488,
+      "learning_rate": 6.906246077898172e-06,
+      "loss": 0.4012,
+      "step": 15650
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 33.26277542114258,
+      "learning_rate": 6.905409362841486e-06,
+      "loss": 0.9033,
+      "step": 15651
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 22.752958297729492,
+      "learning_rate": 6.9045726477847975e-06,
+      "loss": 0.71,
+      "step": 15652
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 10.783706665039062,
+      "learning_rate": 6.9037359327281095e-06,
+      "loss": 0.4343,
+      "step": 15653
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 20.73919677734375,
+      "learning_rate": 6.902899217671422e-06,
+      "loss": 0.7173,
+      "step": 15654
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 50.49082946777344,
+      "learning_rate": 6.902062502614735e-06,
+      "loss": 2.1645,
+      "step": 15655
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 14.673543930053711,
+      "learning_rate": 6.901225787558048e-06,
+      "loss": 0.9353,
+      "step": 15656
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 13.010211944580078,
+      "learning_rate": 6.90038907250136e-06,
+      "loss": 1.5859,
+      "step": 15657
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 17.193904876708984,
+      "learning_rate": 6.8995523574446735e-06,
+      "loss": 0.8288,
+      "step": 15658
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 22.396930694580078,
+      "learning_rate": 6.8987156423879855e-06,
+      "loss": 1.4121,
+      "step": 15659
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 22.452795028686523,
+      "learning_rate": 6.897878927331297e-06,
+      "loss": 2.8046,
+      "step": 15660
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 12.525009155273438,
+      "learning_rate": 6.89704221227461e-06,
+      "loss": 0.4185,
+      "step": 15661
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 17.753753662109375,
+      "learning_rate": 6.896205497217923e-06,
+      "loss": 0.9865,
+      "step": 15662
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 75.84577178955078,
+      "learning_rate": 6.895368782161236e-06,
+      "loss": 0.7734,
+      "step": 15663
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 16.62454605102539,
+      "learning_rate": 6.894532067104548e-06,
+      "loss": 1.01,
+      "step": 15664
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 6.9103569984436035,
+      "learning_rate": 6.893695352047861e-06,
+      "loss": 0.5124,
+      "step": 15665
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 34.67379379272461,
+      "learning_rate": 6.892858636991173e-06,
+      "loss": 1.1893,
+      "step": 15666
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 104.24536895751953,
+      "learning_rate": 6.892021921934485e-06,
+      "loss": 1.6502,
+      "step": 15667
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 13.472212791442871,
+      "learning_rate": 6.891185206877798e-06,
+      "loss": 0.8939,
+      "step": 15668
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 11.587250709533691,
+      "learning_rate": 6.890348491821111e-06,
+      "loss": 0.6514,
+      "step": 15669
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 6.328396797180176,
+      "learning_rate": 6.889511776764424e-06,
+      "loss": 1.2435,
+      "step": 15670
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.189692974090576,
+      "learning_rate": 6.888675061707736e-06,
+      "loss": 0.4131,
+      "step": 15671
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 2.220883369445801,
+      "learning_rate": 6.887838346651049e-06,
+      "loss": 0.1098,
+      "step": 15672
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.100465774536133,
+      "learning_rate": 6.887001631594361e-06,
+      "loss": 0.8029,
+      "step": 15673
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 33.33985900878906,
+      "learning_rate": 6.886164916537673e-06,
+      "loss": 2.2551,
+      "step": 15674
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 18.337589263916016,
+      "learning_rate": 6.885328201480986e-06,
+      "loss": 0.5092,
+      "step": 15675
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 7.437816143035889,
+      "learning_rate": 6.884491486424299e-06,
+      "loss": 0.6743,
+      "step": 15676
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 39.140586853027344,
+      "learning_rate": 6.883654771367612e-06,
+      "loss": 1.7836,
+      "step": 15677
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 21.392349243164062,
+      "learning_rate": 6.882818056310924e-06,
+      "loss": 0.5786,
+      "step": 15678
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 13.871636390686035,
+      "learning_rate": 6.881981341254237e-06,
+      "loss": 1.7352,
+      "step": 15679
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.367588996887207,
+      "learning_rate": 6.881144626197549e-06,
+      "loss": 0.5013,
+      "step": 15680
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 21.83454704284668,
+      "learning_rate": 6.880307911140861e-06,
+      "loss": 0.7217,
+      "step": 15681
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 9.30942440032959,
+      "learning_rate": 6.879471196084174e-06,
+      "loss": 1.1288,
+      "step": 15682
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 29.0285701751709,
+      "learning_rate": 6.878634481027486e-06,
+      "loss": 0.9151,
+      "step": 15683
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 102.32872772216797,
+      "learning_rate": 6.8777977659707996e-06,
+      "loss": 2.2491,
+      "step": 15684
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 12.969430923461914,
+      "learning_rate": 6.8769610509141115e-06,
+      "loss": 0.6064,
+      "step": 15685
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 16.655744552612305,
+      "learning_rate": 6.876124335857425e-06,
+      "loss": 0.6201,
+      "step": 15686
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.896740913391113,
+      "learning_rate": 6.875287620800737e-06,
+      "loss": 1.1829,
+      "step": 15687
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 12.075094223022461,
+      "learning_rate": 6.874450905744049e-06,
+      "loss": 1.008,
+      "step": 15688
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 10.580779075622559,
+      "learning_rate": 6.873614190687362e-06,
+      "loss": 1.104,
+      "step": 15689
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 6.3209547996521,
+      "learning_rate": 6.872777475630674e-06,
+      "loss": 0.51,
+      "step": 15690
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.2571667432785034,
+      "learning_rate": 6.8719407605739875e-06,
+      "loss": 0.0328,
+      "step": 15691
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 14.199670791625977,
+      "learning_rate": 6.8711040455172994e-06,
+      "loss": 0.7279,
+      "step": 15692
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 17.16489028930664,
+      "learning_rate": 6.870267330460613e-06,
+      "loss": 0.9808,
+      "step": 15693
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 13.66822338104248,
+      "learning_rate": 6.869430615403925e-06,
+      "loss": 0.88,
+      "step": 15694
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 17.813467025756836,
+      "learning_rate": 6.868593900347237e-06,
+      "loss": 1.3307,
+      "step": 15695
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.129831314086914,
+      "learning_rate": 6.86775718529055e-06,
+      "loss": 2.2649,
+      "step": 15696
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 7.627436637878418,
+      "learning_rate": 6.866920470233862e-06,
+      "loss": 0.3787,
+      "step": 15697
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 23.57646369934082,
+      "learning_rate": 6.866083755177175e-06,
+      "loss": 1.2059,
+      "step": 15698
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 23.84362030029297,
+      "learning_rate": 6.865247040120487e-06,
+      "loss": 0.3332,
+      "step": 15699
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 3.6035149097442627,
+      "learning_rate": 6.8644103250638e-06,
+      "loss": 0.2212,
+      "step": 15700
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.199368000030518,
+      "learning_rate": 6.863573610007113e-06,
+      "loss": 0.5067,
+      "step": 15701
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 16.549041748046875,
+      "learning_rate": 6.862736894950425e-06,
+      "loss": 1.5231,
+      "step": 15702
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.173349380493164,
+      "learning_rate": 6.861900179893738e-06,
+      "loss": 0.7989,
+      "step": 15703
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 43.922786712646484,
+      "learning_rate": 6.86106346483705e-06,
+      "loss": 0.707,
+      "step": 15704
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 105.08285522460938,
+      "learning_rate": 6.860226749780363e-06,
+      "loss": 1.2973,
+      "step": 15705
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 23.7647647857666,
+      "learning_rate": 6.859390034723675e-06,
+      "loss": 0.8263,
+      "step": 15706
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 17.59117889404297,
+      "learning_rate": 6.858553319666988e-06,
+      "loss": 0.6021,
+      "step": 15707
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 38.81248092651367,
+      "learning_rate": 6.857716604610301e-06,
+      "loss": 1.6105,
+      "step": 15708
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 20.170320510864258,
+      "learning_rate": 6.856879889553613e-06,
+      "loss": 1.2641,
+      "step": 15709
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 36.23456954956055,
+      "learning_rate": 6.856043174496926e-06,
+      "loss": 1.802,
+      "step": 15710
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 31.516807556152344,
+      "learning_rate": 6.855206459440238e-06,
+      "loss": 1.4221,
+      "step": 15711
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 60.37530517578125,
+      "learning_rate": 6.854369744383551e-06,
+      "loss": 1.0894,
+      "step": 15712
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.5870620012283325,
+      "learning_rate": 6.853533029326863e-06,
+      "loss": 0.0508,
+      "step": 15713
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.285735130310059,
+      "learning_rate": 6.852696314270176e-06,
+      "loss": 1.4598,
+      "step": 15714
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.626489162445068,
+      "learning_rate": 6.851859599213489e-06,
+      "loss": 0.3816,
+      "step": 15715
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 242.55050659179688,
+      "learning_rate": 6.851022884156801e-06,
+      "loss": 1.5678,
+      "step": 15716
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 12.18159294128418,
+      "learning_rate": 6.8501861691001136e-06,
+      "loss": 0.7714,
+      "step": 15717
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 22.286861419677734,
+      "learning_rate": 6.8493494540434255e-06,
+      "loss": 0.6816,
+      "step": 15718
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.448339462280273,
+      "learning_rate": 6.848512738986739e-06,
+      "loss": 0.4246,
+      "step": 15719
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 18.41815757751465,
+      "learning_rate": 6.847676023930051e-06,
+      "loss": 1.3264,
+      "step": 15720
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 48.002986907958984,
+      "learning_rate": 6.846839308873364e-06,
+      "loss": 1.2621,
+      "step": 15721
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 7.736732482910156,
+      "learning_rate": 6.846002593816677e-06,
+      "loss": 0.6592,
+      "step": 15722
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.901252746582031,
+      "learning_rate": 6.845165878759989e-06,
+      "loss": 0.6618,
+      "step": 15723
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 15.029723167419434,
+      "learning_rate": 6.8443291637033015e-06,
+      "loss": 1.0123,
+      "step": 15724
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 11.73349380493164,
+      "learning_rate": 6.8434924486466134e-06,
+      "loss": 0.7331,
+      "step": 15725
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 40.74237823486328,
+      "learning_rate": 6.842655733589927e-06,
+      "loss": 2.2732,
+      "step": 15726
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 21.932228088378906,
+      "learning_rate": 6.841819018533239e-06,
+      "loss": 1.0555,
+      "step": 15727
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 15.126953125,
+      "learning_rate": 6.840982303476552e-06,
+      "loss": 1.2464,
+      "step": 15728
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 17.180967330932617,
+      "learning_rate": 6.840145588419864e-06,
+      "loss": 2.1003,
+      "step": 15729
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 6.244019985198975,
+      "learning_rate": 6.839308873363177e-06,
+      "loss": 0.3997,
+      "step": 15730
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 13.408403396606445,
+      "learning_rate": 6.838472158306489e-06,
+      "loss": 0.7182,
+      "step": 15731
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 17.05916976928711,
+      "learning_rate": 6.837635443249801e-06,
+      "loss": 0.8322,
+      "step": 15732
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.671708106994629,
+      "learning_rate": 6.836798728193115e-06,
+      "loss": 0.5849,
+      "step": 15733
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 17.564306259155273,
+      "learning_rate": 6.835962013136427e-06,
+      "loss": 1.3005,
+      "step": 15734
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 30.314828872680664,
+      "learning_rate": 6.83512529807974e-06,
+      "loss": 1.777,
+      "step": 15735
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 15.86076545715332,
+      "learning_rate": 6.834288583023052e-06,
+      "loss": 1.0505,
+      "step": 15736
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.155749320983887,
+      "learning_rate": 6.8334518679663645e-06,
+      "loss": 0.2771,
+      "step": 15737
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.9560956954956055,
+      "learning_rate": 6.832615152909677e-06,
+      "loss": 0.8509,
+      "step": 15738
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 26.87178611755371,
+      "learning_rate": 6.831778437852989e-06,
+      "loss": 1.9133,
+      "step": 15739
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 12.6340913772583,
+      "learning_rate": 6.830941722796303e-06,
+      "loss": 2.5513,
+      "step": 15740
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 13.080458641052246,
+      "learning_rate": 6.830105007739615e-06,
+      "loss": 0.8845,
+      "step": 15741
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.604981899261475,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.5194,
+      "step": 15742
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 18.114355087280273,
+      "learning_rate": 6.82843157762624e-06,
+      "loss": 0.8396,
+      "step": 15743
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.608214378356934,
+      "learning_rate": 6.8275948625695524e-06,
+      "loss": 0.5554,
+      "step": 15744
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 24.107295989990234,
+      "learning_rate": 6.826758147512865e-06,
+      "loss": 2.9412,
+      "step": 15745
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 14.742103576660156,
+      "learning_rate": 6.825921432456177e-06,
+      "loss": 2.0741,
+      "step": 15746
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 15.706936836242676,
+      "learning_rate": 6.825084717399491e-06,
+      "loss": 1.4164,
+      "step": 15747
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 27.102779388427734,
+      "learning_rate": 6.824248002342803e-06,
+      "loss": 1.9597,
+      "step": 15748
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 10.980995178222656,
+      "learning_rate": 6.823411287286116e-06,
+      "loss": 1.5035,
+      "step": 15749
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 52.42707443237305,
+      "learning_rate": 6.8225745722294276e-06,
+      "loss": 2.2065,
+      "step": 15750
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 3.3362693786621094,
+      "learning_rate": 6.82173785717274e-06,
+      "loss": 0.1472,
+      "step": 15751
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 25.937246322631836,
+      "learning_rate": 6.820901142116053e-06,
+      "loss": 2.0787,
+      "step": 15752
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 17.075586318969727,
+      "learning_rate": 6.820064427059365e-06,
+      "loss": 2.0738,
+      "step": 15753
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 24.563770294189453,
+      "learning_rate": 6.819227712002679e-06,
+      "loss": 1.5253,
+      "step": 15754
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 26.96422004699707,
+      "learning_rate": 6.818390996945991e-06,
+      "loss": 0.6918,
+      "step": 15755
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.356304168701172,
+      "learning_rate": 6.8175542818893035e-06,
+      "loss": 1.0219,
+      "step": 15756
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.026080131530762,
+      "learning_rate": 6.8167175668326155e-06,
+      "loss": 0.6225,
+      "step": 15757
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 14.464239120483398,
+      "learning_rate": 6.8158808517759274e-06,
+      "loss": 1.389,
+      "step": 15758
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.517858982086182,
+      "learning_rate": 6.815044136719241e-06,
+      "loss": 1.3605,
+      "step": 15759
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 16.6304874420166,
+      "learning_rate": 6.814207421662553e-06,
+      "loss": 0.8383,
+      "step": 15760
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 13.514803886413574,
+      "learning_rate": 6.813370706605867e-06,
+      "loss": 1.129,
+      "step": 15761
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.15904426574707,
+      "learning_rate": 6.812533991549179e-06,
+      "loss": 0.0998,
+      "step": 15762
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.209280014038086,
+      "learning_rate": 6.8116972764924914e-06,
+      "loss": 0.2516,
+      "step": 15763
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 3.692831516265869,
+      "learning_rate": 6.810860561435803e-06,
+      "loss": 0.3186,
+      "step": 15764
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 45.93081283569336,
+      "learning_rate": 6.810023846379115e-06,
+      "loss": 5.1503,
+      "step": 15765
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 7.090846538543701,
+      "learning_rate": 6.809187131322429e-06,
+      "loss": 0.5046,
+      "step": 15766
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 18.20060920715332,
+      "learning_rate": 6.808350416265741e-06,
+      "loss": 1.0815,
+      "step": 15767
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 17.6390323638916,
+      "learning_rate": 6.807513701209055e-06,
+      "loss": 0.9999,
+      "step": 15768
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 52.04750061035156,
+      "learning_rate": 6.8066769861523665e-06,
+      "loss": 3.1283,
+      "step": 15769
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 15.587772369384766,
+      "learning_rate": 6.805840271095679e-06,
+      "loss": 0.5959,
+      "step": 15770
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.01774263381958,
+      "learning_rate": 6.805003556038991e-06,
+      "loss": 0.337,
+      "step": 15771
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 12.749217987060547,
+      "learning_rate": 6.804166840982303e-06,
+      "loss": 0.6262,
+      "step": 15772
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 13.583781242370605,
+      "learning_rate": 6.803330125925617e-06,
+      "loss": 0.8778,
+      "step": 15773
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 152.0609130859375,
+      "learning_rate": 6.802493410868929e-06,
+      "loss": 1.3635,
+      "step": 15774
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.898676872253418,
+      "learning_rate": 6.801656695812242e-06,
+      "loss": 0.7382,
+      "step": 15775
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 14.59203052520752,
+      "learning_rate": 6.8008199807555545e-06,
+      "loss": 0.7811,
+      "step": 15776
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 31.98999786376953,
+      "learning_rate": 6.799983265698867e-06,
+      "loss": 1.3975,
+      "step": 15777
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 21.140865325927734,
+      "learning_rate": 6.799146550642179e-06,
+      "loss": 1.5799,
+      "step": 15778
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 14.374770164489746,
+      "learning_rate": 6.798309835585491e-06,
+      "loss": 0.7239,
+      "step": 15779
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 12.85692024230957,
+      "learning_rate": 6.797473120528805e-06,
+      "loss": 1.3584,
+      "step": 15780
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.9161810874938965,
+      "learning_rate": 6.796636405472117e-06,
+      "loss": 0.9429,
+      "step": 15781
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 18.36745262145996,
+      "learning_rate": 6.79579969041543e-06,
+      "loss": 2.0932,
+      "step": 15782
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 13.395533561706543,
+      "learning_rate": 6.794962975358742e-06,
+      "loss": 1.2362,
+      "step": 15783
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 31.147062301635742,
+      "learning_rate": 6.794126260302055e-06,
+      "loss": 2.2719,
+      "step": 15784
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 15.50479793548584,
+      "learning_rate": 6.793289545245367e-06,
+      "loss": 1.2375,
+      "step": 15785
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 12.041942596435547,
+      "learning_rate": 6.792452830188679e-06,
+      "loss": 0.2188,
+      "step": 15786
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 16.76938819885254,
+      "learning_rate": 6.791616115131993e-06,
+      "loss": 1.1379,
+      "step": 15787
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 38.72340393066406,
+      "learning_rate": 6.790779400075305e-06,
+      "loss": 0.9639,
+      "step": 15788
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 10.512739181518555,
+      "learning_rate": 6.7899426850186175e-06,
+      "loss": 0.7662,
+      "step": 15789
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 30.250303268432617,
+      "learning_rate": 6.78910596996193e-06,
+      "loss": 1.3182,
+      "step": 15790
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 15.59507942199707,
+      "learning_rate": 6.788269254905242e-06,
+      "loss": 1.2405,
+      "step": 15791
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 19.40499496459961,
+      "learning_rate": 6.787432539848555e-06,
+      "loss": 0.6822,
+      "step": 15792
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 14.11445140838623,
+      "learning_rate": 6.786595824791867e-06,
+      "loss": 1.0599,
+      "step": 15793
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 34.05891799926758,
+      "learning_rate": 6.785759109735181e-06,
+      "loss": 2.9596,
+      "step": 15794
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.87421178817749,
+      "learning_rate": 6.784922394678493e-06,
+      "loss": 0.3535,
+      "step": 15795
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.5683294534683228,
+      "learning_rate": 6.784085679621805e-06,
+      "loss": 0.0357,
+      "step": 15796
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.782618522644043,
+      "learning_rate": 6.783248964565118e-06,
+      "loss": 0.2633,
+      "step": 15797
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 46.601898193359375,
+      "learning_rate": 6.78241224950843e-06,
+      "loss": 1.0305,
+      "step": 15798
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 19.762128829956055,
+      "learning_rate": 6.781575534451743e-06,
+      "loss": 1.0901,
+      "step": 15799
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 4.58148717880249,
+      "learning_rate": 6.780738819395055e-06,
+      "loss": 0.2555,
+      "step": 15800
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 114.76978302001953,
+      "learning_rate": 6.779902104338369e-06,
+      "loss": 1.4787,
+      "step": 15801
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 10.883560180664062,
+      "learning_rate": 6.7790653892816805e-06,
+      "loss": 0.4775,
+      "step": 15802
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 35.43465042114258,
+      "learning_rate": 6.778228674224993e-06,
+      "loss": 3.3315,
+      "step": 15803
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.82397174835205,
+      "learning_rate": 6.777391959168306e-06,
+      "loss": 0.6702,
+      "step": 15804
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.601953506469727,
+      "learning_rate": 6.776555244111618e-06,
+      "loss": 0.4739,
+      "step": 15805
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 9.222111701965332,
+      "learning_rate": 6.775718529054931e-06,
+      "loss": 0.3269,
+      "step": 15806
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 10.697507858276367,
+      "learning_rate": 6.774881813998243e-06,
+      "loss": 1.3201,
+      "step": 15807
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 16.113651275634766,
+      "learning_rate": 6.7740450989415565e-06,
+      "loss": 1.4599,
+      "step": 15808
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 21.23736572265625,
+      "learning_rate": 6.7732083838848685e-06,
+      "loss": 0.8539,
+      "step": 15809
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 22.37204933166504,
+      "learning_rate": 6.772371668828181e-06,
+      "loss": 0.8251,
+      "step": 15810
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 52.310489654541016,
+      "learning_rate": 6.771534953771493e-06,
+      "loss": 2.5391,
+      "step": 15811
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 49.94077682495117,
+      "learning_rate": 6.770698238714806e-06,
+      "loss": 1.2796,
+      "step": 15812
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.511701583862305,
+      "learning_rate": 6.769861523658119e-06,
+      "loss": 0.714,
+      "step": 15813
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 11.521330833435059,
+      "learning_rate": 6.769024808601431e-06,
+      "loss": 1.385,
+      "step": 15814
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 15.310072898864746,
+      "learning_rate": 6.768188093544744e-06,
+      "loss": 0.913,
+      "step": 15815
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 31.162324905395508,
+      "learning_rate": 6.767351378488056e-06,
+      "loss": 1.6257,
+      "step": 15816
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 56.60767364501953,
+      "learning_rate": 6.766514663431369e-06,
+      "loss": 2.2295,
+      "step": 15817
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 6.426440238952637,
+      "learning_rate": 6.765677948374681e-06,
+      "loss": 0.5198,
+      "step": 15818
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 7.248169422149658,
+      "learning_rate": 6.764841233317994e-06,
+      "loss": 0.739,
+      "step": 15819
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 22.613384246826172,
+      "learning_rate": 6.764004518261307e-06,
+      "loss": 0.5279,
+      "step": 15820
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 14.047466278076172,
+      "learning_rate": 6.763167803204619e-06,
+      "loss": 1.7756,
+      "step": 15821
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 103.04438018798828,
+      "learning_rate": 6.762331088147932e-06,
+      "loss": 1.9127,
+      "step": 15822
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 12.794999122619629,
+      "learning_rate": 6.761494373091244e-06,
+      "loss": 0.5107,
+      "step": 15823
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 31.93400764465332,
+      "learning_rate": 6.760657658034557e-06,
+      "loss": 0.9843,
+      "step": 15824
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 18.88462257385254,
+      "learning_rate": 6.759820942977869e-06,
+      "loss": 0.3994,
+      "step": 15825
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 19.0087833404541,
+      "learning_rate": 6.758984227921182e-06,
+      "loss": 0.7096,
+      "step": 15826
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 15.642922401428223,
+      "learning_rate": 6.758147512864495e-06,
+      "loss": 0.4583,
+      "step": 15827
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 83.87149810791016,
+      "learning_rate": 6.757310797807807e-06,
+      "loss": 1.4162,
+      "step": 15828
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 4.853623390197754,
+      "learning_rate": 6.75647408275112e-06,
+      "loss": 0.8932,
+      "step": 15829
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 53.79916000366211,
+      "learning_rate": 6.755637367694432e-06,
+      "loss": 0.9648,
+      "step": 15830
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 77.17891693115234,
+      "learning_rate": 6.754800652637745e-06,
+      "loss": 1.7642,
+      "step": 15831
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.801971435546875,
+      "learning_rate": 6.753963937581057e-06,
+      "loss": 0.5852,
+      "step": 15832
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 7.212321758270264,
+      "learning_rate": 6.75312722252437e-06,
+      "loss": 0.5346,
+      "step": 15833
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 6.900275707244873,
+      "learning_rate": 6.7522905074676826e-06,
+      "loss": 1.6755,
+      "step": 15834
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 18.999773025512695,
+      "learning_rate": 6.7514537924109945e-06,
+      "loss": 2.0601,
+      "step": 15835
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 22.585071563720703,
+      "learning_rate": 6.750617077354308e-06,
+      "loss": 0.6447,
+      "step": 15836
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.195775985717773,
+      "learning_rate": 6.74978036229762e-06,
+      "loss": 0.3813,
+      "step": 15837
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 25.38553237915039,
+      "learning_rate": 6.748943647240933e-06,
+      "loss": 0.5727,
+      "step": 15838
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 17.144113540649414,
+      "learning_rate": 6.748106932184245e-06,
+      "loss": 0.3331,
+      "step": 15839
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 13.481681823730469,
+      "learning_rate": 6.747270217127557e-06,
+      "loss": 1.0861,
+      "step": 15840
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 31.40520477294922,
+      "learning_rate": 6.7464335020708705e-06,
+      "loss": 1.6721,
+      "step": 15841
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.16893482208252,
+      "learning_rate": 6.7455967870141824e-06,
+      "loss": 0.2693,
+      "step": 15842
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 36.15472412109375,
+      "learning_rate": 6.744760071957496e-06,
+      "loss": 1.2241,
+      "step": 15843
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.867559432983398,
+      "learning_rate": 6.743923356900808e-06,
+      "loss": 0.4183,
+      "step": 15844
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 12.108124732971191,
+      "learning_rate": 6.743086641844121e-06,
+      "loss": 0.4826,
+      "step": 15845
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 18.097091674804688,
+      "learning_rate": 6.742249926787433e-06,
+      "loss": 1.4556,
+      "step": 15846
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.807405471801758,
+      "learning_rate": 6.741413211730745e-06,
+      "loss": 0.637,
+      "step": 15847
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 14.497922897338867,
+      "learning_rate": 6.740576496674058e-06,
+      "loss": 0.9905,
+      "step": 15848
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.1741943359375,
+      "learning_rate": 6.73973978161737e-06,
+      "loss": 0.9797,
+      "step": 15849
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 92.6616439819336,
+      "learning_rate": 6.738903066560684e-06,
+      "loss": 1.8077,
+      "step": 15850
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 37.04857635498047,
+      "learning_rate": 6.738066351503996e-06,
+      "loss": 1.5059,
+      "step": 15851
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 3.5521743297576904,
+      "learning_rate": 6.737229636447309e-06,
+      "loss": 0.1421,
+      "step": 15852
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.347033500671387,
+      "learning_rate": 6.736392921390621e-06,
+      "loss": 0.1225,
+      "step": 15853
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 15.23282527923584,
+      "learning_rate": 6.735556206333933e-06,
+      "loss": 1.318,
+      "step": 15854
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 36.05027389526367,
+      "learning_rate": 6.734719491277246e-06,
+      "loss": 2.4698,
+      "step": 15855
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 20.172067642211914,
+      "learning_rate": 6.733882776220558e-06,
+      "loss": 1.0434,
+      "step": 15856
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 12.372947692871094,
+      "learning_rate": 6.733046061163871e-06,
+      "loss": 1.7002,
+      "step": 15857
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 32.12907409667969,
+      "learning_rate": 6.732209346107184e-06,
+      "loss": 1.9306,
+      "step": 15858
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 3.9602692127227783,
+      "learning_rate": 6.731372631050497e-06,
+      "loss": 0.0323,
+      "step": 15859
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.315047264099121,
+      "learning_rate": 6.730535915993809e-06,
+      "loss": 0.4447,
+      "step": 15860
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.079931259155273,
+      "learning_rate": 6.729699200937121e-06,
+      "loss": 0.6453,
+      "step": 15861
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 28.420324325561523,
+      "learning_rate": 6.728862485880434e-06,
+      "loss": 0.652,
+      "step": 15862
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.298340797424316,
+      "learning_rate": 6.728025770823746e-06,
+      "loss": 1.6646,
+      "step": 15863
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.225053787231445,
+      "learning_rate": 6.727189055767059e-06,
+      "loss": 0.1328,
+      "step": 15864
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 21.30705451965332,
+      "learning_rate": 6.726352340710372e-06,
+      "loss": 1.5404,
+      "step": 15865
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.347716331481934,
+      "learning_rate": 6.725515625653685e-06,
+      "loss": 0.2698,
+      "step": 15866
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7319647073745728,
+      "learning_rate": 6.7246789105969966e-06,
+      "loss": 0.0144,
+      "step": 15867
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.512052536010742,
+      "learning_rate": 6.7238421955403085e-06,
+      "loss": 1.2367,
+      "step": 15868
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.839876174926758,
+      "learning_rate": 6.723005480483622e-06,
+      "loss": 0.6775,
+      "step": 15869
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 13.457022666931152,
+      "learning_rate": 6.722168765426934e-06,
+      "loss": 0.4371,
+      "step": 15870
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 16.29427146911621,
+      "learning_rate": 6.721332050370247e-06,
+      "loss": 0.644,
+      "step": 15871
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.478692054748535,
+      "learning_rate": 6.72049533531356e-06,
+      "loss": 0.5189,
+      "step": 15872
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 20.0339412689209,
+      "learning_rate": 6.7196586202568725e-06,
+      "loss": 1.0351,
+      "step": 15873
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 13.582653999328613,
+      "learning_rate": 6.7188219052001845e-06,
+      "loss": 2.4049,
+      "step": 15874
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 19.97344398498535,
+      "learning_rate": 6.7179851901434964e-06,
+      "loss": 1.2458,
+      "step": 15875
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 4.125348091125488,
+      "learning_rate": 6.71714847508681e-06,
+      "loss": 0.2057,
+      "step": 15876
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 7.991709232330322,
+      "learning_rate": 6.716311760030122e-06,
+      "loss": 0.7323,
+      "step": 15877
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.625768661499023,
+      "learning_rate": 6.715475044973435e-06,
+      "loss": 0.8304,
+      "step": 15878
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 15.126842498779297,
+      "learning_rate": 6.714638329916748e-06,
+      "loss": 1.3284,
+      "step": 15879
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 14.254463195800781,
+      "learning_rate": 6.7138016148600604e-06,
+      "loss": 1.4844,
+      "step": 15880
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.407301902770996,
+      "learning_rate": 6.712964899803372e-06,
+      "loss": 0.7541,
+      "step": 15881
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 17.281606674194336,
+      "learning_rate": 6.712128184746684e-06,
+      "loss": 2.9646,
+      "step": 15882
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.113269805908203,
+      "learning_rate": 6.711291469689998e-06,
+      "loss": 1.4294,
+      "step": 15883
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 17.159299850463867,
+      "learning_rate": 6.71045475463331e-06,
+      "loss": 2.8152,
+      "step": 15884
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 40.33321762084961,
+      "learning_rate": 6.709618039576623e-06,
+      "loss": 2.357,
+      "step": 15885
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 50.981689453125,
+      "learning_rate": 6.708781324519935e-06,
+      "loss": 1.0512,
+      "step": 15886
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 25.388002395629883,
+      "learning_rate": 6.707944609463248e-06,
+      "loss": 1.5439,
+      "step": 15887
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 12.173666954040527,
+      "learning_rate": 6.70710789440656e-06,
+      "loss": 1.3675,
+      "step": 15888
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 43.078392028808594,
+      "learning_rate": 6.706271179349872e-06,
+      "loss": 1.0155,
+      "step": 15889
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 10.333561897277832,
+      "learning_rate": 6.705434464293186e-06,
+      "loss": 1.494,
+      "step": 15890
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 9.58069133758545,
+      "learning_rate": 6.704597749236498e-06,
+      "loss": 1.607,
+      "step": 15891
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.508028030395508,
+      "learning_rate": 6.703761034179811e-06,
+      "loss": 0.87,
+      "step": 15892
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 30.451860427856445,
+      "learning_rate": 6.702924319123123e-06,
+      "loss": 1.4955,
+      "step": 15893
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 12.09868335723877,
+      "learning_rate": 6.702087604066436e-06,
+      "loss": 0.5746,
+      "step": 15894
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 42.80390930175781,
+      "learning_rate": 6.701250889009748e-06,
+      "loss": 1.0079,
+      "step": 15895
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 24.87275505065918,
+      "learning_rate": 6.70041417395306e-06,
+      "loss": 2.8248,
+      "step": 15896
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 11.7715482711792,
+      "learning_rate": 6.699577458896374e-06,
+      "loss": 0.6132,
+      "step": 15897
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 9.595332145690918,
+      "learning_rate": 6.698740743839686e-06,
+      "loss": 1.7415,
+      "step": 15898
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 16.042381286621094,
+      "learning_rate": 6.697904028782999e-06,
+      "loss": 1.001,
+      "step": 15899
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 14.5777587890625,
+      "learning_rate": 6.6970673137263106e-06,
+      "loss": 1.7448,
+      "step": 15900
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.866508483886719,
+      "learning_rate": 6.696230598669624e-06,
+      "loss": 0.6962,
+      "step": 15901
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 14.958714485168457,
+      "learning_rate": 6.695393883612936e-06,
+      "loss": 1.2525,
+      "step": 15902
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 6.288279056549072,
+      "learning_rate": 6.694557168556248e-06,
+      "loss": 0.7352,
+      "step": 15903
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 29.31307601928711,
+      "learning_rate": 6.693720453499562e-06,
+      "loss": 1.1016,
+      "step": 15904
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 43.20933532714844,
+      "learning_rate": 6.692883738442874e-06,
+      "loss": 0.5855,
+      "step": 15905
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.682262420654297,
+      "learning_rate": 6.6920470233861865e-06,
+      "loss": 0.4505,
+      "step": 15906
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.800933361053467,
+      "learning_rate": 6.6912103083294985e-06,
+      "loss": 0.1849,
+      "step": 15907
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 11.816986083984375,
+      "learning_rate": 6.690373593272812e-06,
+      "loss": 0.8428,
+      "step": 15908
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 33.56275939941406,
+      "learning_rate": 6.689536878216124e-06,
+      "loss": 0.7461,
+      "step": 15909
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 15.040940284729004,
+      "learning_rate": 6.688700163159436e-06,
+      "loss": 0.7987,
+      "step": 15910
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 11.767467498779297,
+      "learning_rate": 6.68786344810275e-06,
+      "loss": 1.5834,
+      "step": 15911
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 7.761677265167236,
+      "learning_rate": 6.687026733046062e-06,
+      "loss": 1.5374,
+      "step": 15912
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 16.573644638061523,
+      "learning_rate": 6.6861900179893744e-06,
+      "loss": 1.1204,
+      "step": 15913
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 12.538320541381836,
+      "learning_rate": 6.685353302932686e-06,
+      "loss": 0.524,
+      "step": 15914
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 17.518587112426758,
+      "learning_rate": 6.684516587876e-06,
+      "loss": 0.9166,
+      "step": 15915
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 11.541840553283691,
+      "learning_rate": 6.683679872819312e-06,
+      "loss": 0.6606,
+      "step": 15916
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 80.66234588623047,
+      "learning_rate": 6.682843157762624e-06,
+      "loss": 3.8966,
+      "step": 15917
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 21.442481994628906,
+      "learning_rate": 6.682006442705938e-06,
+      "loss": 1.0088,
+      "step": 15918
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 12.824055671691895,
+      "learning_rate": 6.6811697276492495e-06,
+      "loss": 0.9655,
+      "step": 15919
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 21.428932189941406,
+      "learning_rate": 6.680333012592562e-06,
+      "loss": 1.3878,
+      "step": 15920
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 122.69747924804688,
+      "learning_rate": 6.679496297535874e-06,
+      "loss": 2.1404,
+      "step": 15921
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.174544334411621,
+      "learning_rate": 6.678659582479188e-06,
+      "loss": 0.6177,
+      "step": 15922
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 10.728100776672363,
+      "learning_rate": 6.6778228674225e-06,
+      "loss": 1.957,
+      "step": 15923
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 11.2260103225708,
+      "learning_rate": 6.676986152365812e-06,
+      "loss": 0.4722,
+      "step": 15924
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 14.812954902648926,
+      "learning_rate": 6.6761494373091255e-06,
+      "loss": 1.6674,
+      "step": 15925
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 20.64780044555664,
+      "learning_rate": 6.6753127222524375e-06,
+      "loss": 0.8833,
+      "step": 15926
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.755573272705078,
+      "learning_rate": 6.67447600719575e-06,
+      "loss": 0.4229,
+      "step": 15927
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 16.041831970214844,
+      "learning_rate": 6.673639292139062e-06,
+      "loss": 0.8347,
+      "step": 15928
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 17.680612564086914,
+      "learning_rate": 6.672802577082376e-06,
+      "loss": 1.5199,
+      "step": 15929
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.9551849365234375,
+      "learning_rate": 6.671965862025688e-06,
+      "loss": 0.1124,
+      "step": 15930
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 64.43891906738281,
+      "learning_rate": 6.671129146969e-06,
+      "loss": 0.7415,
+      "step": 15931
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 7.296156883239746,
+      "learning_rate": 6.6702924319123134e-06,
+      "loss": 0.6088,
+      "step": 15932
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 21.48379898071289,
+      "learning_rate": 6.669455716855625e-06,
+      "loss": 1.7839,
+      "step": 15933
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.890734672546387,
+      "learning_rate": 6.668619001798938e-06,
+      "loss": 2.0624,
+      "step": 15934
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 87.4519271850586,
+      "learning_rate": 6.66778228674225e-06,
+      "loss": 1.0414,
+      "step": 15935
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 20.54428482055664,
+      "learning_rate": 6.666945571685564e-06,
+      "loss": 2.0353,
+      "step": 15936
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 15.328015327453613,
+      "learning_rate": 6.666108856628876e-06,
+      "loss": 0.8762,
+      "step": 15937
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 390.0461120605469,
+      "learning_rate": 6.665272141572188e-06,
+      "loss": 1.416,
+      "step": 15938
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 14.488633155822754,
+      "learning_rate": 6.6644354265155005e-06,
+      "loss": 2.1592,
+      "step": 15939
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 25.96225357055664,
+      "learning_rate": 6.663598711458813e-06,
+      "loss": 1.4226,
+      "step": 15940
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 12.014482498168945,
+      "learning_rate": 6.662761996402126e-06,
+      "loss": 1.2372,
+      "step": 15941
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 15.766691207885742,
+      "learning_rate": 6.661925281345438e-06,
+      "loss": 1.0284,
+      "step": 15942
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 20.78445053100586,
+      "learning_rate": 6.661088566288752e-06,
+      "loss": 2.3436,
+      "step": 15943
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 22.455188751220703,
+      "learning_rate": 6.660251851232064e-06,
+      "loss": 0.9648,
+      "step": 15944
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 16.59456443786621,
+      "learning_rate": 6.659415136175376e-06,
+      "loss": 0.7957,
+      "step": 15945
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 10.40251350402832,
+      "learning_rate": 6.658578421118688e-06,
+      "loss": 0.9505,
+      "step": 15946
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 20.96028709411621,
+      "learning_rate": 6.657741706062001e-06,
+      "loss": 1.5892,
+      "step": 15947
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 23.296125411987305,
+      "learning_rate": 6.656904991005314e-06,
+      "loss": 2.7671,
+      "step": 15948
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 10.336874008178711,
+      "learning_rate": 6.656068275948626e-06,
+      "loss": 1.5377,
+      "step": 15949
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 6.621896266937256,
+      "learning_rate": 6.65523156089194e-06,
+      "loss": 0.4338,
+      "step": 15950
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.4018852412700653,
+      "learning_rate": 6.654394845835252e-06,
+      "loss": 0.0091,
+      "step": 15951
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.823291778564453,
+      "learning_rate": 6.6535581307785635e-06,
+      "loss": 1.597,
+      "step": 15952
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 47.25034713745117,
+      "learning_rate": 6.652721415721876e-06,
+      "loss": 1.5957,
+      "step": 15953
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 14.00681209564209,
+      "learning_rate": 6.651884700665189e-06,
+      "loss": 0.5162,
+      "step": 15954
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 50.61946487426758,
+      "learning_rate": 6.651047985608502e-06,
+      "loss": 1.0697,
+      "step": 15955
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 7.162833213806152,
+      "learning_rate": 6.650211270551814e-06,
+      "loss": 1.5073,
+      "step": 15956
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 22.026241302490234,
+      "learning_rate": 6.6493745554951275e-06,
+      "loss": 0.8263,
+      "step": 15957
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.92219877243042,
+      "learning_rate": 6.6485378404384395e-06,
+      "loss": 0.2425,
+      "step": 15958
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.8470804691314697,
+      "learning_rate": 6.6477011253817515e-06,
+      "loss": 0.3319,
+      "step": 15959
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 14.090774536132812,
+      "learning_rate": 6.646864410325064e-06,
+      "loss": 1.4975,
+      "step": 15960
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 20.857500076293945,
+      "learning_rate": 6.646027695268377e-06,
+      "loss": 0.9667,
+      "step": 15961
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.1130306720733643,
+      "learning_rate": 6.64519098021169e-06,
+      "loss": 0.0662,
+      "step": 15962
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 25.573955535888672,
+      "learning_rate": 6.644354265155002e-06,
+      "loss": 0.8767,
+      "step": 15963
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 14.740697860717773,
+      "learning_rate": 6.6435175500983155e-06,
+      "loss": 1.8486,
+      "step": 15964
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 10.59947681427002,
+      "learning_rate": 6.642680835041627e-06,
+      "loss": 0.6213,
+      "step": 15965
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 14.851434707641602,
+      "learning_rate": 6.641844119984939e-06,
+      "loss": 0.6295,
+      "step": 15966
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 16.237768173217773,
+      "learning_rate": 6.641007404928252e-06,
+      "loss": 1.1565,
+      "step": 15967
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 10.23038387298584,
+      "learning_rate": 6.640170689871564e-06,
+      "loss": 0.4564,
+      "step": 15968
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 10.985434532165527,
+      "learning_rate": 6.639333974814878e-06,
+      "loss": 0.7974,
+      "step": 15969
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 67.11539459228516,
+      "learning_rate": 6.63849725975819e-06,
+      "loss": 1.8991,
+      "step": 15970
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 74.15109252929688,
+      "learning_rate": 6.637660544701502e-06,
+      "loss": 0.4018,
+      "step": 15971
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 21.274198532104492,
+      "learning_rate": 6.636823829644815e-06,
+      "loss": 1.4911,
+      "step": 15972
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 26.153928756713867,
+      "learning_rate": 6.635987114588127e-06,
+      "loss": 2.2189,
+      "step": 15973
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 46.78422546386719,
+      "learning_rate": 6.63515039953144e-06,
+      "loss": 2.7653,
+      "step": 15974
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 11.140388488769531,
+      "learning_rate": 6.634313684474752e-06,
+      "loss": 0.5892,
+      "step": 15975
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 9.493162155151367,
+      "learning_rate": 6.633476969418066e-06,
+      "loss": 0.8846,
+      "step": 15976
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 8.969060897827148,
+      "learning_rate": 6.632640254361378e-06,
+      "loss": 1.0295,
+      "step": 15977
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 24.31886100769043,
+      "learning_rate": 6.63180353930469e-06,
+      "loss": 1.8822,
+      "step": 15978
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 18.47286605834961,
+      "learning_rate": 6.630966824248003e-06,
+      "loss": 1.128,
+      "step": 15979
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 8.595629692077637,
+      "learning_rate": 6.630130109191315e-06,
+      "loss": 1.3727,
+      "step": 15980
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 10.602871894836426,
+      "learning_rate": 6.629293394134628e-06,
+      "loss": 2.0915,
+      "step": 15981
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.2133660316467285,
+      "learning_rate": 6.62845667907794e-06,
+      "loss": 0.2119,
+      "step": 15982
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 20.664487838745117,
+      "learning_rate": 6.627619964021254e-06,
+      "loss": 1.0672,
+      "step": 15983
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 15.849124908447266,
+      "learning_rate": 6.6267832489645656e-06,
+      "loss": 3.5109,
+      "step": 15984
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.37014627456665,
+      "learning_rate": 6.6259465339078775e-06,
+      "loss": 0.2146,
+      "step": 15985
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 16.57113265991211,
+      "learning_rate": 6.625109818851191e-06,
+      "loss": 1.0359,
+      "step": 15986
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 14.121919631958008,
+      "learning_rate": 6.624273103794503e-06,
+      "loss": 0.7774,
+      "step": 15987
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.6017680168151855,
+      "learning_rate": 6.623436388737816e-06,
+      "loss": 0.3663,
+      "step": 15988
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 11.929272651672363,
+      "learning_rate": 6.622599673681128e-06,
+      "loss": 1.2326,
+      "step": 15989
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 24.052127838134766,
+      "learning_rate": 6.6217629586244415e-06,
+      "loss": 1.5549,
+      "step": 15990
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 10.859208106994629,
+      "learning_rate": 6.6209262435677535e-06,
+      "loss": 0.5273,
+      "step": 15991
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 17.65408706665039,
+      "learning_rate": 6.6200895285110654e-06,
+      "loss": 0.4234,
+      "step": 15992
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 11.088404655456543,
+      "learning_rate": 6.619252813454379e-06,
+      "loss": 0.4453,
+      "step": 15993
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 23.68683433532715,
+      "learning_rate": 6.618416098397691e-06,
+      "loss": 0.4521,
+      "step": 15994
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 9.509688377380371,
+      "learning_rate": 6.617579383341004e-06,
+      "loss": 0.5721,
+      "step": 15995
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 11.537019729614258,
+      "learning_rate": 6.616742668284316e-06,
+      "loss": 0.7805,
+      "step": 15996
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 16.062152862548828,
+      "learning_rate": 6.6159059532276295e-06,
+      "loss": 0.8361,
+      "step": 15997
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 8.554738998413086,
+      "learning_rate": 6.615069238170941e-06,
+      "loss": 1.0527,
+      "step": 15998
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 27.2183895111084,
+      "learning_rate": 6.614232523114253e-06,
+      "loss": 0.4897,
+      "step": 15999
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 8.592169761657715,
+      "learning_rate": 6.613395808057567e-06,
+      "loss": 0.9188,
+      "step": 16000
+    },
+    {
+      "epoch": 2.01,
+      "eval_loss": 0.08059647679328918,
+      "eval_runtime": 95.2297,
+      "eval_samples_per_second": 37.194,
+      "eval_steps_per_second": 37.194,
+      "step": 16000
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 19.152080535888672,
+      "learning_rate": 6.612559093000879e-06,
+      "loss": 0.5195,
+      "step": 16001
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 22.343149185180664,
+      "learning_rate": 6.611722377944192e-06,
+      "loss": 1.7367,
+      "step": 16002
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 8.451386451721191,
+      "learning_rate": 6.610885662887504e-06,
+      "loss": 1.1314,
+      "step": 16003
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 13.740487098693848,
+      "learning_rate": 6.610048947830817e-06,
+      "loss": 1.0626,
+      "step": 16004
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 8.336570739746094,
+      "learning_rate": 6.609212232774129e-06,
+      "loss": 1.1301,
+      "step": 16005
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 25.304929733276367,
+      "learning_rate": 6.608375517717441e-06,
+      "loss": 1.0049,
+      "step": 16006
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 9.847472190856934,
+      "learning_rate": 6.607538802660755e-06,
+      "loss": 0.7213,
+      "step": 16007
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 9.159112930297852,
+      "learning_rate": 6.606702087604067e-06,
+      "loss": 1.4592,
+      "step": 16008
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.6362738609313965,
+      "learning_rate": 6.60586537254738e-06,
+      "loss": 0.2961,
+      "step": 16009
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 117.4923095703125,
+      "learning_rate": 6.605028657490692e-06,
+      "loss": 1.3204,
+      "step": 16010
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 18.80195426940918,
+      "learning_rate": 6.604191942434005e-06,
+      "loss": 1.7741,
+      "step": 16011
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 11.38399887084961,
+      "learning_rate": 6.603355227377317e-06,
+      "loss": 0.3152,
+      "step": 16012
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 10.227343559265137,
+      "learning_rate": 6.602518512320629e-06,
+      "loss": 0.4564,
+      "step": 16013
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 17.74940299987793,
+      "learning_rate": 6.601681797263942e-06,
+      "loss": 1.1158,
+      "step": 16014
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 10.877930641174316,
+      "learning_rate": 6.600845082207255e-06,
+      "loss": 0.3915,
+      "step": 16015
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 8.845476150512695,
+      "learning_rate": 6.600008367150568e-06,
+      "loss": 1.0693,
+      "step": 16016
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 13.181829452514648,
+      "learning_rate": 6.5991716520938796e-06,
+      "loss": 0.8126,
+      "step": 16017
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 27.765037536621094,
+      "learning_rate": 6.598334937037193e-06,
+      "loss": 1.3239,
+      "step": 16018
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 12.452431678771973,
+      "learning_rate": 6.597498221980505e-06,
+      "loss": 0.5701,
+      "step": 16019
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 15.18510627746582,
+      "learning_rate": 6.596661506923817e-06,
+      "loss": 1.2167,
+      "step": 16020
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 2.89917254447937,
+      "learning_rate": 6.59582479186713e-06,
+      "loss": 0.079,
+      "step": 16021
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 45.87964630126953,
+      "learning_rate": 6.594988076810443e-06,
+      "loss": 0.3751,
+      "step": 16022
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 6.7308173179626465,
+      "learning_rate": 6.5941513617537555e-06,
+      "loss": 0.2547,
+      "step": 16023
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 6.030806064605713,
+      "learning_rate": 6.5933146466970675e-06,
+      "loss": 0.4948,
+      "step": 16024
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 20.833017349243164,
+      "learning_rate": 6.592477931640381e-06,
+      "loss": 1.7271,
+      "step": 16025
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.376020431518555,
+      "learning_rate": 6.591641216583693e-06,
+      "loss": 1.2232,
+      "step": 16026
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 12.35608196258545,
+      "learning_rate": 6.590804501527005e-06,
+      "loss": 0.3531,
+      "step": 16027
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 29.041461944580078,
+      "learning_rate": 6.589967786470318e-06,
+      "loss": 1.8665,
+      "step": 16028
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 24.14327049255371,
+      "learning_rate": 6.589131071413631e-06,
+      "loss": 1.0578,
+      "step": 16029
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 6.061190605163574,
+      "learning_rate": 6.5882943563569434e-06,
+      "loss": 0.237,
+      "step": 16030
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 8.493118286132812,
+      "learning_rate": 6.587457641300255e-06,
+      "loss": 0.6232,
+      "step": 16031
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 46.43001937866211,
+      "learning_rate": 6.586620926243569e-06,
+      "loss": 2.7755,
+      "step": 16032
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 20.77440643310547,
+      "learning_rate": 6.585784211186881e-06,
+      "loss": 2.6907,
+      "step": 16033
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 18.526092529296875,
+      "learning_rate": 6.584947496130193e-06,
+      "loss": 1.2859,
+      "step": 16034
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 16.359085083007812,
+      "learning_rate": 6.584110781073506e-06,
+      "loss": 0.5416,
+      "step": 16035
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 13.703718185424805,
+      "learning_rate": 6.5832740660168186e-06,
+      "loss": 2.1605,
+      "step": 16036
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 26.011127471923828,
+      "learning_rate": 6.582437350960131e-06,
+      "loss": 1.2199,
+      "step": 16037
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 12.94687557220459,
+      "learning_rate": 6.581600635903443e-06,
+      "loss": 1.2709,
+      "step": 16038
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.392760276794434,
+      "learning_rate": 6.580763920846757e-06,
+      "loss": 0.3088,
+      "step": 16039
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 28.460010528564453,
+      "learning_rate": 6.579927205790069e-06,
+      "loss": 0.3051,
+      "step": 16040
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 6.393258094787598,
+      "learning_rate": 6.579090490733381e-06,
+      "loss": 0.2671,
+      "step": 16041
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 19.735218048095703,
+      "learning_rate": 6.578253775676694e-06,
+      "loss": 0.9901,
+      "step": 16042
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 7.071054458618164,
+      "learning_rate": 6.5774170606200065e-06,
+      "loss": 0.5651,
+      "step": 16043
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 102.0082015991211,
+      "learning_rate": 6.576580345563319e-06,
+      "loss": 2.522,
+      "step": 16044
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 40.931800842285156,
+      "learning_rate": 6.575743630506631e-06,
+      "loss": 0.9531,
+      "step": 16045
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 10.436454772949219,
+      "learning_rate": 6.574906915449945e-06,
+      "loss": 0.5927,
+      "step": 16046
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 10.416304588317871,
+      "learning_rate": 6.574070200393257e-06,
+      "loss": 0.6314,
+      "step": 16047
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 13.067681312561035,
+      "learning_rate": 6.573233485336569e-06,
+      "loss": 0.65,
+      "step": 16048
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 15.980208396911621,
+      "learning_rate": 6.572396770279882e-06,
+      "loss": 2.3473,
+      "step": 16049
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 16.0473690032959,
+      "learning_rate": 6.5715600552231936e-06,
+      "loss": 1.1509,
+      "step": 16050
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 8.895906448364258,
+      "learning_rate": 6.570723340166507e-06,
+      "loss": 0.8638,
+      "step": 16051
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 8.34984016418457,
+      "learning_rate": 6.569886625109819e-06,
+      "loss": 0.5495,
+      "step": 16052
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 9.876906394958496,
+      "learning_rate": 6.569049910053133e-06,
+      "loss": 1.6489,
+      "step": 16053
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 107.1396713256836,
+      "learning_rate": 6.568213194996445e-06,
+      "loss": 1.6347,
+      "step": 16054
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 55.78226089477539,
+      "learning_rate": 6.567376479939757e-06,
+      "loss": 1.3686,
+      "step": 16055
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 17.969043731689453,
+      "learning_rate": 6.5665397648830695e-06,
+      "loss": 1.5541,
+      "step": 16056
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 8.708976745605469,
+      "learning_rate": 6.5657030498263815e-06,
+      "loss": 1.0385,
+      "step": 16057
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 9.779619216918945,
+      "learning_rate": 6.564866334769695e-06,
+      "loss": 0.4743,
+      "step": 16058
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 6.712967395782471,
+      "learning_rate": 6.564029619713007e-06,
+      "loss": 0.381,
+      "step": 16059
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 6.852048873901367,
+      "learning_rate": 6.563192904656321e-06,
+      "loss": 1.7019,
+      "step": 16060
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 7.393848896026611,
+      "learning_rate": 6.562356189599633e-06,
+      "loss": 0.4451,
+      "step": 16061
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 10.314332008361816,
+      "learning_rate": 6.561519474542945e-06,
+      "loss": 0.8183,
+      "step": 16062
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 24.86186408996582,
+      "learning_rate": 6.5606827594862574e-06,
+      "loss": 1.3703,
+      "step": 16063
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.029012680053711,
+      "learning_rate": 6.559846044429569e-06,
+      "loss": 0.0475,
+      "step": 16064
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 12.916887283325195,
+      "learning_rate": 6.559009329372883e-06,
+      "loss": 0.6797,
+      "step": 16065
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 7.170773029327393,
+      "learning_rate": 6.558172614316195e-06,
+      "loss": 0.7314,
+      "step": 16066
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 11.910627365112305,
+      "learning_rate": 6.557335899259508e-06,
+      "loss": 0.3921,
+      "step": 16067
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 40.78810501098633,
+      "learning_rate": 6.556499184202821e-06,
+      "loss": 1.2843,
+      "step": 16068
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 35.77200698852539,
+      "learning_rate": 6.5556624691461325e-06,
+      "loss": 1.6284,
+      "step": 16069
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 11.998297691345215,
+      "learning_rate": 6.554825754089445e-06,
+      "loss": 1.7564,
+      "step": 16070
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 79.4481201171875,
+      "learning_rate": 6.553989039032757e-06,
+      "loss": 1.7479,
+      "step": 16071
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 21.77023696899414,
+      "learning_rate": 6.553152323976071e-06,
+      "loss": 2.0174,
+      "step": 16072
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 44.96331787109375,
+      "learning_rate": 6.552315608919383e-06,
+      "loss": 0.5996,
+      "step": 16073
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 11.881548881530762,
+      "learning_rate": 6.551478893862696e-06,
+      "loss": 0.5137,
+      "step": 16074
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 25.157161712646484,
+      "learning_rate": 6.5506421788060085e-06,
+      "loss": 1.3631,
+      "step": 16075
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 7.388939380645752,
+      "learning_rate": 6.5498054637493205e-06,
+      "loss": 0.5749,
+      "step": 16076
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 13.814678192138672,
+      "learning_rate": 6.548968748692633e-06,
+      "loss": 0.5423,
+      "step": 16077
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 12.471866607666016,
+      "learning_rate": 6.548132033635945e-06,
+      "loss": 1.0481,
+      "step": 16078
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.258547306060791,
+      "learning_rate": 6.547295318579259e-06,
+      "loss": 0.262,
+      "step": 16079
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 14.908158302307129,
+      "learning_rate": 6.546458603522571e-06,
+      "loss": 1.0873,
+      "step": 16080
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 11.934511184692383,
+      "learning_rate": 6.545621888465884e-06,
+      "loss": 0.7848,
+      "step": 16081
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 19.13489532470703,
+      "learning_rate": 6.5447851734091964e-06,
+      "loss": 1.4351,
+      "step": 16082
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 20.378049850463867,
+      "learning_rate": 6.543948458352508e-06,
+      "loss": 1.1822,
+      "step": 16083
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 11.956856727600098,
+      "learning_rate": 6.543111743295821e-06,
+      "loss": 0.5044,
+      "step": 16084
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 11.374585151672363,
+      "learning_rate": 6.542275028239133e-06,
+      "loss": 1.4749,
+      "step": 16085
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 12.470864295959473,
+      "learning_rate": 6.541438313182447e-06,
+      "loss": 0.4487,
+      "step": 16086
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 11.731393814086914,
+      "learning_rate": 6.540601598125759e-06,
+      "loss": 1.7252,
+      "step": 16087
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 19.13770866394043,
+      "learning_rate": 6.5397648830690715e-06,
+      "loss": 0.5294,
+      "step": 16088
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 9.0031156539917,
+      "learning_rate": 6.538928168012384e-06,
+      "loss": 0.3526,
+      "step": 16089
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 8.142091751098633,
+      "learning_rate": 6.538091452955696e-06,
+      "loss": 0.4971,
+      "step": 16090
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 10.265229225158691,
+      "learning_rate": 6.537254737899009e-06,
+      "loss": 0.6881,
+      "step": 16091
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 9.502202987670898,
+      "learning_rate": 6.536418022842321e-06,
+      "loss": 0.3887,
+      "step": 16092
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 11.04644775390625,
+      "learning_rate": 6.535581307785635e-06,
+      "loss": 0.411,
+      "step": 16093
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 127.50508117675781,
+      "learning_rate": 6.534744592728947e-06,
+      "loss": 1.6468,
+      "step": 16094
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 21.08706283569336,
+      "learning_rate": 6.5339078776722595e-06,
+      "loss": 1.0562,
+      "step": 16095
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 22.2269344329834,
+      "learning_rate": 6.533071162615571e-06,
+      "loss": 0.7309,
+      "step": 16096
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 16.021883010864258,
+      "learning_rate": 6.532234447558884e-06,
+      "loss": 0.8293,
+      "step": 16097
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 22.984682083129883,
+      "learning_rate": 6.531397732502197e-06,
+      "loss": 1.6267,
+      "step": 16098
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 8.846332550048828,
+      "learning_rate": 6.530561017445509e-06,
+      "loss": 0.7814,
+      "step": 16099
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 13.226693153381348,
+      "learning_rate": 6.529724302388823e-06,
+      "loss": 0.393,
+      "step": 16100
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 29.962265014648438,
+      "learning_rate": 6.528887587332135e-06,
+      "loss": 1.4385,
+      "step": 16101
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.677152633666992,
+      "learning_rate": 6.528050872275447e-06,
+      "loss": 0.2706,
+      "step": 16102
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 125.49024963378906,
+      "learning_rate": 6.527214157218759e-06,
+      "loss": 4.2126,
+      "step": 16103
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 9.701367378234863,
+      "learning_rate": 6.526377442162072e-06,
+      "loss": 0.3329,
+      "step": 16104
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.124175548553467,
+      "learning_rate": 6.525540727105385e-06,
+      "loss": 0.5724,
+      "step": 16105
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 19.317773818969727,
+      "learning_rate": 6.524704012048697e-06,
+      "loss": 1.0071,
+      "step": 16106
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 32.84058380126953,
+      "learning_rate": 6.5238672969920105e-06,
+      "loss": 1.9674,
+      "step": 16107
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 60.97601318359375,
+      "learning_rate": 6.5230305819353225e-06,
+      "loss": 0.9619,
+      "step": 16108
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 9.599262237548828,
+      "learning_rate": 6.522193866878635e-06,
+      "loss": 0.5553,
+      "step": 16109
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.903003215789795,
+      "learning_rate": 6.521357151821947e-06,
+      "loss": 0.5936,
+      "step": 16110
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 19.20045280456543,
+      "learning_rate": 6.52052043676526e-06,
+      "loss": 1.0966,
+      "step": 16111
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 32.570472717285156,
+      "learning_rate": 6.519683721708573e-06,
+      "loss": 0.6197,
+      "step": 16112
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 16.27768325805664,
+      "learning_rate": 6.518847006651885e-06,
+      "loss": 0.7894,
+      "step": 16113
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 27.70846176147461,
+      "learning_rate": 6.5180102915951985e-06,
+      "loss": 1.7573,
+      "step": 16114
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 96.64350128173828,
+      "learning_rate": 6.51717357653851e-06,
+      "loss": 1.8776,
+      "step": 16115
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 10.895058631896973,
+      "learning_rate": 6.516336861481823e-06,
+      "loss": 2.8214,
+      "step": 16116
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 23.313756942749023,
+      "learning_rate": 6.515500146425135e-06,
+      "loss": 2.5583,
+      "step": 16117
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 30.129369735717773,
+      "learning_rate": 6.514663431368448e-06,
+      "loss": 2.4679,
+      "step": 16118
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 18.207834243774414,
+      "learning_rate": 6.513826716311761e-06,
+      "loss": 1.3371,
+      "step": 16119
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 17.993852615356445,
+      "learning_rate": 6.512990001255073e-06,
+      "loss": 1.3252,
+      "step": 16120
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 13.623716354370117,
+      "learning_rate": 6.512153286198386e-06,
+      "loss": 0.4988,
+      "step": 16121
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.604955196380615,
+      "learning_rate": 6.511316571141698e-06,
+      "loss": 0.1058,
+      "step": 16122
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 9.774563789367676,
+      "learning_rate": 6.510479856085011e-06,
+      "loss": 0.8261,
+      "step": 16123
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 9.117918014526367,
+      "learning_rate": 6.509643141028323e-06,
+      "loss": 1.2245,
+      "step": 16124
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 14.01237964630127,
+      "learning_rate": 6.508806425971636e-06,
+      "loss": 0.8158,
+      "step": 16125
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 259.6254577636719,
+      "learning_rate": 6.507969710914949e-06,
+      "loss": 1.7877,
+      "step": 16126
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 63.59457778930664,
+      "learning_rate": 6.507132995858261e-06,
+      "loss": 1.2412,
+      "step": 16127
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 10.758515357971191,
+      "learning_rate": 6.506296280801574e-06,
+      "loss": 0.8092,
+      "step": 16128
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 19.523914337158203,
+      "learning_rate": 6.505459565744886e-06,
+      "loss": 1.5905,
+      "step": 16129
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.093863010406494,
+      "learning_rate": 6.504622850688199e-06,
+      "loss": 0.2513,
+      "step": 16130
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.924009799957275,
+      "learning_rate": 6.503786135631511e-06,
+      "loss": 0.279,
+      "step": 16131
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 8.36949634552002,
+      "learning_rate": 6.502949420574823e-06,
+      "loss": 0.1637,
+      "step": 16132
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 9.036155700683594,
+      "learning_rate": 6.502112705518137e-06,
+      "loss": 1.0372,
+      "step": 16133
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 13.477315902709961,
+      "learning_rate": 6.5012759904614486e-06,
+      "loss": 1.1656,
+      "step": 16134
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.973839282989502,
+      "learning_rate": 6.500439275404762e-06,
+      "loss": 0.16,
+      "step": 16135
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 54.16847610473633,
+      "learning_rate": 6.499602560348074e-06,
+      "loss": 2.3118,
+      "step": 16136
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 110.89966583251953,
+      "learning_rate": 6.498765845291387e-06,
+      "loss": 0.6806,
+      "step": 16137
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 11.091172218322754,
+      "learning_rate": 6.497929130234699e-06,
+      "loss": 0.5446,
+      "step": 16138
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 13.850701332092285,
+      "learning_rate": 6.497092415178011e-06,
+      "loss": 1.1616,
+      "step": 16139
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 24.403318405151367,
+      "learning_rate": 6.4962557001213245e-06,
+      "loss": 1.1152,
+      "step": 16140
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.9210470914840698,
+      "learning_rate": 6.4954189850646365e-06,
+      "loss": 0.0214,
+      "step": 16141
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 6.3548784255981445,
+      "learning_rate": 6.49458227000795e-06,
+      "loss": 0.9924,
+      "step": 16142
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 30.173181533813477,
+      "learning_rate": 6.493745554951262e-06,
+      "loss": 1.5111,
+      "step": 16143
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 398.6379089355469,
+      "learning_rate": 6.492908839894575e-06,
+      "loss": 3.3468,
+      "step": 16144
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 11.111824989318848,
+      "learning_rate": 6.492072124837887e-06,
+      "loss": 0.7169,
+      "step": 16145
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 15.077082633972168,
+      "learning_rate": 6.491235409781199e-06,
+      "loss": 0.8733,
+      "step": 16146
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.078538417816162,
+      "learning_rate": 6.4903986947245125e-06,
+      "loss": 0.2718,
+      "step": 16147
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 7.101211071014404,
+      "learning_rate": 6.489561979667824e-06,
+      "loss": 0.5178,
+      "step": 16148
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.255901336669922,
+      "learning_rate": 6.488725264611137e-06,
+      "loss": 0.8095,
+      "step": 16149
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 8.656655311584473,
+      "learning_rate": 6.48788854955445e-06,
+      "loss": 0.6101,
+      "step": 16150
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 14.041111946105957,
+      "learning_rate": 6.487051834497762e-06,
+      "loss": 1.4424,
+      "step": 16151
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 9.727947235107422,
+      "learning_rate": 6.486215119441075e-06,
+      "loss": 2.4062,
+      "step": 16152
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 19.479961395263672,
+      "learning_rate": 6.485378404384387e-06,
+      "loss": 2.2615,
+      "step": 16153
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 6.5831732749938965,
+      "learning_rate": 6.4845416893277e-06,
+      "loss": 0.4318,
+      "step": 16154
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.607565402984619,
+      "learning_rate": 6.483704974271012e-06,
+      "loss": 0.3844,
+      "step": 16155
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 8.164737701416016,
+      "learning_rate": 6.482868259214325e-06,
+      "loss": 0.362,
+      "step": 16156
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 9.221551895141602,
+      "learning_rate": 6.482031544157638e-06,
+      "loss": 0.8428,
+      "step": 16157
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 9.317448616027832,
+      "learning_rate": 6.48119482910095e-06,
+      "loss": 0.5997,
+      "step": 16158
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 19.246498107910156,
+      "learning_rate": 6.480358114044263e-06,
+      "loss": 1.2734,
+      "step": 16159
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 8.3816499710083,
+      "learning_rate": 6.479521398987575e-06,
+      "loss": 0.4112,
+      "step": 16160
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.023574352264404,
+      "learning_rate": 6.478684683930888e-06,
+      "loss": 0.1879,
+      "step": 16161
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 21.623899459838867,
+      "learning_rate": 6.4778479688742e-06,
+      "loss": 1.7094,
+      "step": 16162
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 12.556909561157227,
+      "learning_rate": 6.477011253817513e-06,
+      "loss": 1.462,
+      "step": 16163
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 18.379074096679688,
+      "learning_rate": 6.476174538760826e-06,
+      "loss": 1.4132,
+      "step": 16164
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 6.25236177444458,
+      "learning_rate": 6.475337823704138e-06,
+      "loss": 0.8684,
+      "step": 16165
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.661009311676025,
+      "learning_rate": 6.474501108647451e-06,
+      "loss": 0.3523,
+      "step": 16166
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 20.906024932861328,
+      "learning_rate": 6.4736643935907626e-06,
+      "loss": 1.8841,
+      "step": 16167
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 20.141586303710938,
+      "learning_rate": 6.472827678534076e-06,
+      "loss": 0.3814,
+      "step": 16168
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 11.476037979125977,
+      "learning_rate": 6.471990963477388e-06,
+      "loss": 0.5897,
+      "step": 16169
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 8.756049156188965,
+      "learning_rate": 6.471154248420701e-06,
+      "loss": 0.6763,
+      "step": 16170
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 11.013794898986816,
+      "learning_rate": 6.470317533364014e-06,
+      "loss": 0.8109,
+      "step": 16171
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 3.914527177810669,
+      "learning_rate": 6.469480818307326e-06,
+      "loss": 0.1943,
+      "step": 16172
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 10.756383895874023,
+      "learning_rate": 6.4686441032506385e-06,
+      "loss": 0.267,
+      "step": 16173
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 12.76132869720459,
+      "learning_rate": 6.4678073881939505e-06,
+      "loss": 0.7225,
+      "step": 16174
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 8.393716812133789,
+      "learning_rate": 6.466970673137264e-06,
+      "loss": 1.6324,
+      "step": 16175
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 32.774295806884766,
+      "learning_rate": 6.466133958080576e-06,
+      "loss": 1.6458,
+      "step": 16176
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 16.518552780151367,
+      "learning_rate": 6.465297243023889e-06,
+      "loss": 1.0371,
+      "step": 16177
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 18.92826271057129,
+      "learning_rate": 6.464460527967201e-06,
+      "loss": 0.6249,
+      "step": 16178
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 17.560800552368164,
+      "learning_rate": 6.463623812910514e-06,
+      "loss": 0.7807,
+      "step": 16179
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 24.438623428344727,
+      "learning_rate": 6.4627870978538264e-06,
+      "loss": 1.0112,
+      "step": 16180
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 6.223598957061768,
+      "learning_rate": 6.461950382797138e-06,
+      "loss": 0.3599,
+      "step": 16181
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 7.577473163604736,
+      "learning_rate": 6.461113667740452e-06,
+      "loss": 0.3013,
+      "step": 16182
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 20.38938331604004,
+      "learning_rate": 6.460276952683764e-06,
+      "loss": 0.8726,
+      "step": 16183
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 124.90731811523438,
+      "learning_rate": 6.459440237627077e-06,
+      "loss": 3.3834,
+      "step": 16184
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 77.51597595214844,
+      "learning_rate": 6.458603522570389e-06,
+      "loss": 2.6717,
+      "step": 16185
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.382199764251709,
+      "learning_rate": 6.4577668075137016e-06,
+      "loss": 0.0814,
+      "step": 16186
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 14.492742538452148,
+      "learning_rate": 6.456930092457014e-06,
+      "loss": 0.9572,
+      "step": 16187
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 8.434473037719727,
+      "learning_rate": 6.456093377400326e-06,
+      "loss": 0.3047,
+      "step": 16188
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 6.0500569343566895,
+      "learning_rate": 6.45525666234364e-06,
+      "loss": 0.4383,
+      "step": 16189
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.993636608123779,
+      "learning_rate": 6.454419947286952e-06,
+      "loss": 1.2767,
+      "step": 16190
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 14.795984268188477,
+      "learning_rate": 6.453583232230265e-06,
+      "loss": 0.6547,
+      "step": 16191
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 27.816919326782227,
+      "learning_rate": 6.452746517173577e-06,
+      "loss": 3.139,
+      "step": 16192
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 13.740135192871094,
+      "learning_rate": 6.4519098021168895e-06,
+      "loss": 1.9539,
+      "step": 16193
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 22.25745391845703,
+      "learning_rate": 6.451073087060202e-06,
+      "loss": 0.6469,
+      "step": 16194
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 20.07583236694336,
+      "learning_rate": 6.450236372003514e-06,
+      "loss": 1.3022,
+      "step": 16195
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 36.75003433227539,
+      "learning_rate": 6.449399656946828e-06,
+      "loss": 3.4121,
+      "step": 16196
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 9.228172302246094,
+      "learning_rate": 6.44856294189014e-06,
+      "loss": 0.6371,
+      "step": 16197
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 34.50830078125,
+      "learning_rate": 6.447726226833453e-06,
+      "loss": 1.3297,
+      "step": 16198
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 11.33741569519043,
+      "learning_rate": 6.446889511776765e-06,
+      "loss": 0.3887,
+      "step": 16199
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 12.4901123046875,
+      "learning_rate": 6.446052796720077e-06,
+      "loss": 1.7242,
+      "step": 16200
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 7.714563846588135,
+      "learning_rate": 6.44521608166339e-06,
+      "loss": 0.3376,
+      "step": 16201
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 9.506186485290527,
+      "learning_rate": 6.444379366606702e-06,
+      "loss": 0.7857,
+      "step": 16202
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 162.11468505859375,
+      "learning_rate": 6.443542651550016e-06,
+      "loss": 2.5083,
+      "step": 16203
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 23.76825714111328,
+      "learning_rate": 6.442705936493328e-06,
+      "loss": 1.1915,
+      "step": 16204
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 21.995790481567383,
+      "learning_rate": 6.4418692214366406e-06,
+      "loss": 0.8841,
+      "step": 16205
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 21.455398559570312,
+      "learning_rate": 6.4410325063799525e-06,
+      "loss": 1.2645,
+      "step": 16206
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 6.0819830894470215,
+      "learning_rate": 6.4401957913232645e-06,
+      "loss": 1.8939,
+      "step": 16207
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 21.00604820251465,
+      "learning_rate": 6.439359076266578e-06,
+      "loss": 1.5358,
+      "step": 16208
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 19.687803268432617,
+      "learning_rate": 6.43852236120989e-06,
+      "loss": 2.1922,
+      "step": 16209
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 18.701154708862305,
+      "learning_rate": 6.437685646153204e-06,
+      "loss": 0.8948,
+      "step": 16210
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 17.60835075378418,
+      "learning_rate": 6.436848931096516e-06,
+      "loss": 0.7196,
+      "step": 16211
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 16.8372859954834,
+      "learning_rate": 6.4360122160398285e-06,
+      "loss": 0.7158,
+      "step": 16212
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.943900108337402,
+      "learning_rate": 6.4351755009831404e-06,
+      "loss": 0.1508,
+      "step": 16213
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 68.16217803955078,
+      "learning_rate": 6.434338785926452e-06,
+      "loss": 1.7712,
+      "step": 16214
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 16.084169387817383,
+      "learning_rate": 6.433502070869766e-06,
+      "loss": 0.2979,
+      "step": 16215
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 51.093727111816406,
+      "learning_rate": 6.432665355813078e-06,
+      "loss": 1.1893,
+      "step": 16216
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 10.752718925476074,
+      "learning_rate": 6.431828640756392e-06,
+      "loss": 0.8647,
+      "step": 16217
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 13.057160377502441,
+      "learning_rate": 6.430991925699704e-06,
+      "loss": 1.4833,
+      "step": 16218
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 13.10116195678711,
+      "learning_rate": 6.430155210643016e-06,
+      "loss": 2.2656,
+      "step": 16219
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 8.28217601776123,
+      "learning_rate": 6.429318495586328e-06,
+      "loss": 0.629,
+      "step": 16220
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 9.03894329071045,
+      "learning_rate": 6.42848178052964e-06,
+      "loss": 1.5399,
+      "step": 16221
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.5642266273498535,
+      "learning_rate": 6.427645065472954e-06,
+      "loss": 0.0701,
+      "step": 16222
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 14.697591781616211,
+      "learning_rate": 6.426808350416266e-06,
+      "loss": 1.04,
+      "step": 16223
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 11.729602813720703,
+      "learning_rate": 6.425971635359579e-06,
+      "loss": 1.0618,
+      "step": 16224
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 51.51275634765625,
+      "learning_rate": 6.4251349203028915e-06,
+      "loss": 3.2666,
+      "step": 16225
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 16.761516571044922,
+      "learning_rate": 6.424298205246204e-06,
+      "loss": 1.0147,
+      "step": 16226
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 10.326380729675293,
+      "learning_rate": 6.423461490189516e-06,
+      "loss": 1.275,
+      "step": 16227
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 6.452572822570801,
+      "learning_rate": 6.422624775132828e-06,
+      "loss": 0.7351,
+      "step": 16228
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 37.921119689941406,
+      "learning_rate": 6.421788060076142e-06,
+      "loss": 1.6725,
+      "step": 16229
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 14.596298217773438,
+      "learning_rate": 6.420951345019454e-06,
+      "loss": 2.1811,
+      "step": 16230
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 69.2381820678711,
+      "learning_rate": 6.420114629962767e-06,
+      "loss": 1.6541,
+      "step": 16231
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.9978963136672974,
+      "learning_rate": 6.4192779149060794e-06,
+      "loss": 0.0773,
+      "step": 16232
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 18.001026153564453,
+      "learning_rate": 6.418441199849392e-06,
+      "loss": 1.02,
+      "step": 16233
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 17.160947799682617,
+      "learning_rate": 6.417604484792704e-06,
+      "loss": 0.6446,
+      "step": 16234
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 13.894946098327637,
+      "learning_rate": 6.416767769736016e-06,
+      "loss": 1.539,
+      "step": 16235
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 325.5096435546875,
+      "learning_rate": 6.41593105467933e-06,
+      "loss": 1.2966,
+      "step": 16236
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 6.154231071472168,
+      "learning_rate": 6.415094339622642e-06,
+      "loss": 0.6995,
+      "step": 16237
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 14.135480880737305,
+      "learning_rate": 6.4142576245659545e-06,
+      "loss": 1.6384,
+      "step": 16238
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 7.413065433502197,
+      "learning_rate": 6.413420909509267e-06,
+      "loss": 0.8911,
+      "step": 16239
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 12.196990013122559,
+      "learning_rate": 6.41258419445258e-06,
+      "loss": 1.6185,
+      "step": 16240
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 16.50975227355957,
+      "learning_rate": 6.411747479395892e-06,
+      "loss": 2.359,
+      "step": 16241
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 15.77778148651123,
+      "learning_rate": 6.410910764339204e-06,
+      "loss": 1.4863,
+      "step": 16242
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 13.257022857666016,
+      "learning_rate": 6.410074049282518e-06,
+      "loss": 0.2781,
+      "step": 16243
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 27.746564865112305,
+      "learning_rate": 6.40923733422583e-06,
+      "loss": 1.7557,
+      "step": 16244
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 16.584228515625,
+      "learning_rate": 6.4084006191691425e-06,
+      "loss": 0.9563,
+      "step": 16245
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 13.610138893127441,
+      "learning_rate": 6.407563904112455e-06,
+      "loss": 0.6585,
+      "step": 16246
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 9.752397537231445,
+      "learning_rate": 6.406727189055768e-06,
+      "loss": 0.2989,
+      "step": 16247
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 9.259270668029785,
+      "learning_rate": 6.40589047399908e-06,
+      "loss": 0.2724,
+      "step": 16248
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 17.941730499267578,
+      "learning_rate": 6.405053758942392e-06,
+      "loss": 0.6028,
+      "step": 16249
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 7.572376728057861,
+      "learning_rate": 6.404217043885706e-06,
+      "loss": 0.273,
+      "step": 16250
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 110.7977066040039,
+      "learning_rate": 6.403380328829018e-06,
+      "loss": 1.3644,
+      "step": 16251
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.024694442749023,
+      "learning_rate": 6.40254361377233e-06,
+      "loss": 0.5423,
+      "step": 16252
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 36.58980178833008,
+      "learning_rate": 6.401706898715643e-06,
+      "loss": 1.0159,
+      "step": 16253
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 30.15936851501465,
+      "learning_rate": 6.400870183658956e-06,
+      "loss": 0.9333,
+      "step": 16254
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 21.83393669128418,
+      "learning_rate": 6.400033468602268e-06,
+      "loss": 2.7222,
+      "step": 16255
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 20.019332885742188,
+      "learning_rate": 6.39919675354558e-06,
+      "loss": 1.6089,
+      "step": 16256
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 16.850595474243164,
+      "learning_rate": 6.3983600384888935e-06,
+      "loss": 1.8224,
+      "step": 16257
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 9.293508529663086,
+      "learning_rate": 6.3975233234322055e-06,
+      "loss": 0.6111,
+      "step": 16258
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 7.602249622344971,
+      "learning_rate": 6.396686608375518e-06,
+      "loss": 0.5085,
+      "step": 16259
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 11.73695182800293,
+      "learning_rate": 6.39584989331883e-06,
+      "loss": 1.1122,
+      "step": 16260
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 24.734926223754883,
+      "learning_rate": 6.395013178262144e-06,
+      "loss": 0.6938,
+      "step": 16261
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 28.150827407836914,
+      "learning_rate": 6.394176463205456e-06,
+      "loss": 0.8636,
+      "step": 16262
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 21.256317138671875,
+      "learning_rate": 6.393339748148768e-06,
+      "loss": 0.9196,
+      "step": 16263
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 46.79572296142578,
+      "learning_rate": 6.3925030330920815e-06,
+      "loss": 1.6798,
+      "step": 16264
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 8.550714492797852,
+      "learning_rate": 6.391666318035393e-06,
+      "loss": 0.7154,
+      "step": 16265
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.677887439727783,
+      "learning_rate": 6.390829602978706e-06,
+      "loss": 0.3291,
+      "step": 16266
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 8.84570026397705,
+      "learning_rate": 6.389992887922018e-06,
+      "loss": 0.6039,
+      "step": 16267
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 14.745457649230957,
+      "learning_rate": 6.389156172865332e-06,
+      "loss": 1.8272,
+      "step": 16268
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 13.759382247924805,
+      "learning_rate": 6.388319457808644e-06,
+      "loss": 0.6211,
+      "step": 16269
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 19.65152931213379,
+      "learning_rate": 6.387482742751956e-06,
+      "loss": 0.6657,
+      "step": 16270
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 16.69057846069336,
+      "learning_rate": 6.386646027695269e-06,
+      "loss": 0.7727,
+      "step": 16271
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 11.009669303894043,
+      "learning_rate": 6.385809312638581e-06,
+      "loss": 0.8503,
+      "step": 16272
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 14.906228065490723,
+      "learning_rate": 6.384972597581894e-06,
+      "loss": 1.5865,
+      "step": 16273
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 52.21102523803711,
+      "learning_rate": 6.384135882525206e-06,
+      "loss": 1.8425,
+      "step": 16274
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 14.800554275512695,
+      "learning_rate": 6.38329916746852e-06,
+      "loss": 0.6739,
+      "step": 16275
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 11.220170974731445,
+      "learning_rate": 6.382462452411832e-06,
+      "loss": 0.4152,
+      "step": 16276
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 26.87870216369629,
+      "learning_rate": 6.381625737355144e-06,
+      "loss": 0.6924,
+      "step": 16277
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 14.723917007446289,
+      "learning_rate": 6.380789022298457e-06,
+      "loss": 1.1726,
+      "step": 16278
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 17.075979232788086,
+      "learning_rate": 6.379952307241769e-06,
+      "loss": 0.7418,
+      "step": 16279
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.230533599853516,
+      "learning_rate": 6.379115592185082e-06,
+      "loss": 0.4674,
+      "step": 16280
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 10.00238037109375,
+      "learning_rate": 6.378278877128394e-06,
+      "loss": 0.2023,
+      "step": 16281
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 10.267596244812012,
+      "learning_rate": 6.377442162071708e-06,
+      "loss": 1.499,
+      "step": 16282
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 7.407226085662842,
+      "learning_rate": 6.37660544701502e-06,
+      "loss": 0.361,
+      "step": 16283
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 14.427005767822266,
+      "learning_rate": 6.3757687319583316e-06,
+      "loss": 1.1869,
+      "step": 16284
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 24.61405372619629,
+      "learning_rate": 6.374932016901645e-06,
+      "loss": 0.8257,
+      "step": 16285
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 9.572637557983398,
+      "learning_rate": 6.374095301844957e-06,
+      "loss": 0.8062,
+      "step": 16286
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 167.83787536621094,
+      "learning_rate": 6.37325858678827e-06,
+      "loss": 1.2678,
+      "step": 16287
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 11.481160163879395,
+      "learning_rate": 6.372421871731582e-06,
+      "loss": 1.8921,
+      "step": 16288
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 58.42168045043945,
+      "learning_rate": 6.371585156674896e-06,
+      "loss": 1.0874,
+      "step": 16289
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 12.236416816711426,
+      "learning_rate": 6.3707484416182075e-06,
+      "loss": 0.5991,
+      "step": 16290
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.3818658590316772,
+      "learning_rate": 6.3699117265615195e-06,
+      "loss": 0.0391,
+      "step": 16291
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 38.054527282714844,
+      "learning_rate": 6.369075011504833e-06,
+      "loss": 1.0768,
+      "step": 16292
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 11.2073974609375,
+      "learning_rate": 6.368238296448145e-06,
+      "loss": 0.9246,
+      "step": 16293
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 12.246166229248047,
+      "learning_rate": 6.367401581391458e-06,
+      "loss": 0.547,
+      "step": 16294
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 12.671717643737793,
+      "learning_rate": 6.36656486633477e-06,
+      "loss": 1.6035,
+      "step": 16295
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 7.1969380378723145,
+      "learning_rate": 6.3657281512780835e-06,
+      "loss": 0.483,
+      "step": 16296
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 40.141963958740234,
+      "learning_rate": 6.3648914362213955e-06,
+      "loss": 1.3982,
+      "step": 16297
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 38.6901741027832,
+      "learning_rate": 6.364054721164707e-06,
+      "loss": 1.5976,
+      "step": 16298
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 12.444235801696777,
+      "learning_rate": 6.363218006108021e-06,
+      "loss": 0.8923,
+      "step": 16299
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 33.04055404663086,
+      "learning_rate": 6.362381291051333e-06,
+      "loss": 1.8274,
+      "step": 16300
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 17.85938262939453,
+      "learning_rate": 6.361544575994646e-06,
+      "loss": 2.887,
+      "step": 16301
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 15.372467041015625,
+      "learning_rate": 6.360707860937958e-06,
+      "loss": 1.3028,
+      "step": 16302
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 19.73614501953125,
+      "learning_rate": 6.359871145881271e-06,
+      "loss": 1.0338,
+      "step": 16303
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 13.654969215393066,
+      "learning_rate": 6.359034430824583e-06,
+      "loss": 2.9795,
+      "step": 16304
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 19.09703826904297,
+      "learning_rate": 6.358197715767895e-06,
+      "loss": 1.8351,
+      "step": 16305
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.372680187225342,
+      "learning_rate": 6.357361000711208e-06,
+      "loss": 1.0544,
+      "step": 16306
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 6.331582546234131,
+      "learning_rate": 6.356524285654521e-06,
+      "loss": 0.7461,
+      "step": 16307
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 13.04987907409668,
+      "learning_rate": 6.355687570597834e-06,
+      "loss": 0.7377,
+      "step": 16308
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 13.980428695678711,
+      "learning_rate": 6.354850855541146e-06,
+      "loss": 0.9329,
+      "step": 16309
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 19.196767807006836,
+      "learning_rate": 6.354014140484459e-06,
+      "loss": 0.8305,
+      "step": 16310
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 18.332805633544922,
+      "learning_rate": 6.353177425427771e-06,
+      "loss": 1.2122,
+      "step": 16311
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 13.444842338562012,
+      "learning_rate": 6.352340710371083e-06,
+      "loss": 1.1295,
+      "step": 16312
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 67.86763000488281,
+      "learning_rate": 6.351503995314396e-06,
+      "loss": 1.1505,
+      "step": 16313
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 17.4975643157959,
+      "learning_rate": 6.350667280257709e-06,
+      "loss": 1.8383,
+      "step": 16314
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 8.981825828552246,
+      "learning_rate": 6.349830565201022e-06,
+      "loss": 0.3658,
+      "step": 16315
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 9.776616096496582,
+      "learning_rate": 6.348993850144334e-06,
+      "loss": 0.5528,
+      "step": 16316
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 63.22365188598633,
+      "learning_rate": 6.348157135087647e-06,
+      "loss": 1.9502,
+      "step": 16317
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 77.63703155517578,
+      "learning_rate": 6.347320420030959e-06,
+      "loss": 0.7079,
+      "step": 16318
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.89634370803833,
+      "learning_rate": 6.346483704974271e-06,
+      "loss": 0.2386,
+      "step": 16319
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 15.108169555664062,
+      "learning_rate": 6.345646989917584e-06,
+      "loss": 0.5373,
+      "step": 16320
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 12.114225387573242,
+      "learning_rate": 6.344810274860897e-06,
+      "loss": 0.7184,
+      "step": 16321
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 6.531534671783447,
+      "learning_rate": 6.3439735598042096e-06,
+      "loss": 0.3666,
+      "step": 16322
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 24.10402488708496,
+      "learning_rate": 6.3431368447475215e-06,
+      "loss": 0.578,
+      "step": 16323
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 25.831274032592773,
+      "learning_rate": 6.3423001296908335e-06,
+      "loss": 0.9635,
+      "step": 16324
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 20.153762817382812,
+      "learning_rate": 6.341463414634147e-06,
+      "loss": 0.6292,
+      "step": 16325
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 21.914087295532227,
+      "learning_rate": 6.340626699577459e-06,
+      "loss": 0.7087,
+      "step": 16326
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 19.947664260864258,
+      "learning_rate": 6.339789984520772e-06,
+      "loss": 0.5721,
+      "step": 16327
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 7.1366376876831055,
+      "learning_rate": 6.338953269464085e-06,
+      "loss": 0.5614,
+      "step": 16328
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 179.075927734375,
+      "learning_rate": 6.3381165544073975e-06,
+      "loss": 4.9287,
+      "step": 16329
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 20.651126861572266,
+      "learning_rate": 6.3372798393507094e-06,
+      "loss": 1.4351,
+      "step": 16330
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 26.97290802001953,
+      "learning_rate": 6.336443124294021e-06,
+      "loss": 2.8531,
+      "step": 16331
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 8.706089973449707,
+      "learning_rate": 6.335606409237335e-06,
+      "loss": 0.8824,
+      "step": 16332
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.200913429260254,
+      "learning_rate": 6.334769694180647e-06,
+      "loss": 0.1274,
+      "step": 16333
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 28.518836975097656,
+      "learning_rate": 6.33393297912396e-06,
+      "loss": 1.2004,
+      "step": 16334
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 7.6518025398254395,
+      "learning_rate": 6.333096264067272e-06,
+      "loss": 0.528,
+      "step": 16335
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 39.421993255615234,
+      "learning_rate": 6.332259549010585e-06,
+      "loss": 3.1023,
+      "step": 16336
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 9.441398620605469,
+      "learning_rate": 6.331422833953897e-06,
+      "loss": 1.2269,
+      "step": 16337
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 10.061051368713379,
+      "learning_rate": 6.330586118897209e-06,
+      "loss": 1.3231,
+      "step": 16338
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 7.994561195373535,
+      "learning_rate": 6.329749403840523e-06,
+      "loss": 0.4762,
+      "step": 16339
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 30.543291091918945,
+      "learning_rate": 6.328912688783835e-06,
+      "loss": 1.2392,
+      "step": 16340
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 11.687193870544434,
+      "learning_rate": 6.328075973727148e-06,
+      "loss": 2.0838,
+      "step": 16341
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 11.96678352355957,
+      "learning_rate": 6.32723925867046e-06,
+      "loss": 0.4478,
+      "step": 16342
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.491357803344727,
+      "learning_rate": 6.326402543613773e-06,
+      "loss": 0.9369,
+      "step": 16343
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 9.527076721191406,
+      "learning_rate": 6.325565828557085e-06,
+      "loss": 0.6563,
+      "step": 16344
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 31.517492294311523,
+      "learning_rate": 6.324729113500397e-06,
+      "loss": 1.1611,
+      "step": 16345
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 9.010356903076172,
+      "learning_rate": 6.323892398443711e-06,
+      "loss": 0.5052,
+      "step": 16346
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 28.53644561767578,
+      "learning_rate": 6.323055683387023e-06,
+      "loss": 2.9225,
+      "step": 16347
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 35.999961853027344,
+      "learning_rate": 6.322218968330336e-06,
+      "loss": 1.7751,
+      "step": 16348
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 19.59505844116211,
+      "learning_rate": 6.321382253273648e-06,
+      "loss": 0.9192,
+      "step": 16349
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 19.588050842285156,
+      "learning_rate": 6.320545538216961e-06,
+      "loss": 0.5096,
+      "step": 16350
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 6.057505130767822,
+      "learning_rate": 6.319708823160273e-06,
+      "loss": 0.2356,
+      "step": 16351
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.174546718597412,
+      "learning_rate": 6.318872108103585e-06,
+      "loss": 0.3658,
+      "step": 16352
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 11.382641792297363,
+      "learning_rate": 6.318035393046899e-06,
+      "loss": 1.3706,
+      "step": 16353
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 8.718911170959473,
+      "learning_rate": 6.317198677990211e-06,
+      "loss": 0.762,
+      "step": 16354
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 15.025772094726562,
+      "learning_rate": 6.3163619629335236e-06,
+      "loss": 0.956,
+      "step": 16355
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 20.396451950073242,
+      "learning_rate": 6.3155252478768355e-06,
+      "loss": 2.0995,
+      "step": 16356
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 16.324954986572266,
+      "learning_rate": 6.314688532820149e-06,
+      "loss": 0.6291,
+      "step": 16357
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 11.061717987060547,
+      "learning_rate": 6.313851817763461e-06,
+      "loss": 1.262,
+      "step": 16358
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 12.975667953491211,
+      "learning_rate": 6.313015102706773e-06,
+      "loss": 0.5564,
+      "step": 16359
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 53.27537155151367,
+      "learning_rate": 6.312178387650087e-06,
+      "loss": 0.4769,
+      "step": 16360
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.7657904624938965,
+      "learning_rate": 6.311341672593399e-06,
+      "loss": 0.3185,
+      "step": 16361
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 20.796558380126953,
+      "learning_rate": 6.3105049575367115e-06,
+      "loss": 1.8968,
+      "step": 16362
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 49.964481353759766,
+      "learning_rate": 6.3096682424800234e-06,
+      "loss": 0.6971,
+      "step": 16363
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 17.00904655456543,
+      "learning_rate": 6.308831527423337e-06,
+      "loss": 0.9146,
+      "step": 16364
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 13.459504127502441,
+      "learning_rate": 6.307994812366649e-06,
+      "loss": 2.1822,
+      "step": 16365
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 22.19635772705078,
+      "learning_rate": 6.307158097309961e-06,
+      "loss": 1.0316,
+      "step": 16366
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 17.777423858642578,
+      "learning_rate": 6.306321382253275e-06,
+      "loss": 0.6407,
+      "step": 16367
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 29.012802124023438,
+      "learning_rate": 6.305484667196587e-06,
+      "loss": 0.9657,
+      "step": 16368
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 14.237957954406738,
+      "learning_rate": 6.304647952139899e-06,
+      "loss": 0.9197,
+      "step": 16369
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 12.470968246459961,
+      "learning_rate": 6.303811237083211e-06,
+      "loss": 0.5394,
+      "step": 16370
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 14.259575843811035,
+      "learning_rate": 6.302974522026525e-06,
+      "loss": 0.7527,
+      "step": 16371
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 166.35455322265625,
+      "learning_rate": 6.302137806969837e-06,
+      "loss": 1.5979,
+      "step": 16372
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 7.62042760848999,
+      "learning_rate": 6.301301091913149e-06,
+      "loss": 0.6344,
+      "step": 16373
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 16.454174041748047,
+      "learning_rate": 6.3004643768564626e-06,
+      "loss": 1.7882,
+      "step": 16374
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 27.83538818359375,
+      "learning_rate": 6.2996276617997745e-06,
+      "loss": 1.9177,
+      "step": 16375
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 270.7811584472656,
+      "learning_rate": 6.298790946743087e-06,
+      "loss": 1.335,
+      "step": 16376
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 27.523265838623047,
+      "learning_rate": 6.297954231686399e-06,
+      "loss": 1.2153,
+      "step": 16377
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 47.53352737426758,
+      "learning_rate": 6.297117516629713e-06,
+      "loss": 1.9278,
+      "step": 16378
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 30.873390197753906,
+      "learning_rate": 6.296280801573025e-06,
+      "loss": 0.7341,
+      "step": 16379
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 14.739265441894531,
+      "learning_rate": 6.295444086516337e-06,
+      "loss": 1.216,
+      "step": 16380
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 23.66485595703125,
+      "learning_rate": 6.2946073714596505e-06,
+      "loss": 4.2067,
+      "step": 16381
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 11.664165496826172,
+      "learning_rate": 6.2937706564029624e-06,
+      "loss": 0.6,
+      "step": 16382
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 44.6564826965332,
+      "learning_rate": 6.292933941346275e-06,
+      "loss": 1.4229,
+      "step": 16383
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.448591232299805,
+      "learning_rate": 6.292097226289587e-06,
+      "loss": 0.4441,
+      "step": 16384
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 18.735206604003906,
+      "learning_rate": 6.291260511232901e-06,
+      "loss": 0.7372,
+      "step": 16385
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 55.1180534362793,
+      "learning_rate": 6.290423796176213e-06,
+      "loss": 2.4001,
+      "step": 16386
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 13.985992431640625,
+      "learning_rate": 6.289587081119525e-06,
+      "loss": 0.9724,
+      "step": 16387
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 12.777236938476562,
+      "learning_rate": 6.2887503660628375e-06,
+      "loss": 0.6601,
+      "step": 16388
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 9.22414493560791,
+      "learning_rate": 6.28791365100615e-06,
+      "loss": 0.85,
+      "step": 16389
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.828909873962402,
+      "learning_rate": 6.287076935949463e-06,
+      "loss": 0.2522,
+      "step": 16390
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 6.063150405883789,
+      "learning_rate": 6.286240220892775e-06,
+      "loss": 0.3322,
+      "step": 16391
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 24.44740104675293,
+      "learning_rate": 6.285403505836089e-06,
+      "loss": 1.1274,
+      "step": 16392
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.613438129425049,
+      "learning_rate": 6.284566790779401e-06,
+      "loss": 0.5376,
+      "step": 16393
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 6.197526454925537,
+      "learning_rate": 6.283730075722713e-06,
+      "loss": 0.3193,
+      "step": 16394
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 12.503471374511719,
+      "learning_rate": 6.2828933606660255e-06,
+      "loss": 0.7365,
+      "step": 16395
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.639339923858643,
+      "learning_rate": 6.282056645609338e-06,
+      "loss": 0.3307,
+      "step": 16396
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 9.989636421203613,
+      "learning_rate": 6.281219930552651e-06,
+      "loss": 0.6942,
+      "step": 16397
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 8.763270378112793,
+      "learning_rate": 6.280383215495963e-06,
+      "loss": 0.5278,
+      "step": 16398
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 28.691089630126953,
+      "learning_rate": 6.279546500439277e-06,
+      "loss": 2.196,
+      "step": 16399
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 24.95621681213379,
+      "learning_rate": 6.278709785382589e-06,
+      "loss": 1.4303,
+      "step": 16400
+    },
+    {
+      "epoch": 2.06,
+      "eval_loss": 0.08141389489173889,
+      "eval_runtime": 95.1344,
+      "eval_samples_per_second": 37.232,
+      "eval_steps_per_second": 37.232,
+      "step": 16400
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 85.77784729003906,
+      "learning_rate": 6.277873070325901e-06,
+      "loss": 5.2265,
+      "step": 16401
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 18.586429595947266,
+      "learning_rate": 6.277036355269213e-06,
+      "loss": 1.6389,
+      "step": 16402
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 26.477685928344727,
+      "learning_rate": 6.276199640212526e-06,
+      "loss": 1.7303,
+      "step": 16403
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 8.007719039916992,
+      "learning_rate": 6.275362925155839e-06,
+      "loss": 1.2385,
+      "step": 16404
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 13.247702598571777,
+      "learning_rate": 6.274526210099151e-06,
+      "loss": 2.375,
+      "step": 16405
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.825837135314941,
+      "learning_rate": 6.273689495042465e-06,
+      "loss": 0.2625,
+      "step": 16406
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 15.234128952026367,
+      "learning_rate": 6.2728527799857765e-06,
+      "loss": 1.317,
+      "step": 16407
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 13.170249938964844,
+      "learning_rate": 6.2720160649290885e-06,
+      "loss": 1.2648,
+      "step": 16408
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.402313232421875,
+      "learning_rate": 6.271179349872401e-06,
+      "loss": 0.1905,
+      "step": 16409
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 16.913536071777344,
+      "learning_rate": 6.270342634815714e-06,
+      "loss": 0.4488,
+      "step": 16410
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 10.359716415405273,
+      "learning_rate": 6.269505919759027e-06,
+      "loss": 0.7147,
+      "step": 16411
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 27.842947006225586,
+      "learning_rate": 6.268669204702339e-06,
+      "loss": 0.3549,
+      "step": 16412
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 47.443660736083984,
+      "learning_rate": 6.2678324896456525e-06,
+      "loss": 1.9468,
+      "step": 16413
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 20.700239181518555,
+      "learning_rate": 6.2669957745889645e-06,
+      "loss": 0.3877,
+      "step": 16414
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 9.135885238647461,
+      "learning_rate": 6.266159059532276e-06,
+      "loss": 0.5151,
+      "step": 16415
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.4493727684021,
+      "learning_rate": 6.265322344475589e-06,
+      "loss": 0.8757,
+      "step": 16416
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 17.44921875,
+      "learning_rate": 6.264485629418901e-06,
+      "loss": 1.7656,
+      "step": 16417
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.254001617431641,
+      "learning_rate": 6.263648914362215e-06,
+      "loss": 0.4605,
+      "step": 16418
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.9415082931518555,
+      "learning_rate": 6.262812199305527e-06,
+      "loss": 0.2035,
+      "step": 16419
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 27.944520950317383,
+      "learning_rate": 6.2619754842488404e-06,
+      "loss": 2.6041,
+      "step": 16420
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 6.755660533905029,
+      "learning_rate": 6.261138769192152e-06,
+      "loss": 0.1938,
+      "step": 16421
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 20.80733299255371,
+      "learning_rate": 6.260302054135464e-06,
+      "loss": 1.573,
+      "step": 16422
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 16.19023895263672,
+      "learning_rate": 6.259465339078777e-06,
+      "loss": 1.1644,
+      "step": 16423
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 6.79762601852417,
+      "learning_rate": 6.258628624022089e-06,
+      "loss": 0.312,
+      "step": 16424
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 14.102092742919922,
+      "learning_rate": 6.257791908965403e-06,
+      "loss": 1.3472,
+      "step": 16425
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 9.547575950622559,
+      "learning_rate": 6.256955193908715e-06,
+      "loss": 0.4151,
+      "step": 16426
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 7.325801849365234,
+      "learning_rate": 6.256118478852028e-06,
+      "loss": 1.101,
+      "step": 16427
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 8.511363983154297,
+      "learning_rate": 6.25528176379534e-06,
+      "loss": 0.5515,
+      "step": 16428
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 34.52033233642578,
+      "learning_rate": 6.254445048738652e-06,
+      "loss": 2.2967,
+      "step": 16429
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 15.332392692565918,
+      "learning_rate": 6.253608333681965e-06,
+      "loss": 1.3402,
+      "step": 16430
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 43.08570098876953,
+      "learning_rate": 6.252771618625277e-06,
+      "loss": 0.66,
+      "step": 16431
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 10.303796768188477,
+      "learning_rate": 6.251934903568591e-06,
+      "loss": 0.4777,
+      "step": 16432
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 13.234012603759766,
+      "learning_rate": 6.251098188511903e-06,
+      "loss": 0.8158,
+      "step": 16433
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 14.474245071411133,
+      "learning_rate": 6.250261473455215e-06,
+      "loss": 0.6216,
+      "step": 16434
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 80.25108337402344,
+      "learning_rate": 6.249424758398528e-06,
+      "loss": 1.3335,
+      "step": 16435
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.7391889095306396,
+      "learning_rate": 6.24858804334184e-06,
+      "loss": 0.0587,
+      "step": 16436
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 8.63875961303711,
+      "learning_rate": 6.247751328285153e-06,
+      "loss": 0.7776,
+      "step": 16437
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 19.064735412597656,
+      "learning_rate": 6.246914613228465e-06,
+      "loss": 0.6267,
+      "step": 16438
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 28.71219825744629,
+      "learning_rate": 6.246077898171779e-06,
+      "loss": 0.9808,
+      "step": 16439
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 132.15008544921875,
+      "learning_rate": 6.2452411831150905e-06,
+      "loss": 3.8898,
+      "step": 16440
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 67.4249496459961,
+      "learning_rate": 6.244404468058403e-06,
+      "loss": 1.7181,
+      "step": 16441
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 27.999494552612305,
+      "learning_rate": 6.243567753001716e-06,
+      "loss": 0.3562,
+      "step": 16442
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.9176435470581055,
+      "learning_rate": 6.242731037945028e-06,
+      "loss": 0.2938,
+      "step": 16443
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 15.687005996704102,
+      "learning_rate": 6.241894322888341e-06,
+      "loss": 1.2571,
+      "step": 16444
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 7.77381706237793,
+      "learning_rate": 6.241057607831653e-06,
+      "loss": 1.7553,
+      "step": 16445
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.529983043670654,
+      "learning_rate": 6.2402208927749665e-06,
+      "loss": 0.2274,
+      "step": 16446
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 13.678400039672852,
+      "learning_rate": 6.2393841777182785e-06,
+      "loss": 1.8998,
+      "step": 16447
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 13.9041748046875,
+      "learning_rate": 6.238547462661591e-06,
+      "loss": 2.8395,
+      "step": 16448
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 31.396278381347656,
+      "learning_rate": 6.237710747604904e-06,
+      "loss": 0.3692,
+      "step": 16449
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 8.526172637939453,
+      "learning_rate": 6.236874032548216e-06,
+      "loss": 0.8206,
+      "step": 16450
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 7.275291442871094,
+      "learning_rate": 6.236037317491529e-06,
+      "loss": 0.3681,
+      "step": 16451
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 15.358269691467285,
+      "learning_rate": 6.235200602434841e-06,
+      "loss": 1.2854,
+      "step": 16452
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.6213921308517456,
+      "learning_rate": 6.234363887378154e-06,
+      "loss": 0.0466,
+      "step": 16453
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 9.204558372497559,
+      "learning_rate": 6.233527172321466e-06,
+      "loss": 0.4216,
+      "step": 16454
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 17.904741287231445,
+      "learning_rate": 6.232690457264779e-06,
+      "loss": 0.8174,
+      "step": 16455
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 31.63105583190918,
+      "learning_rate": 6.231853742208092e-06,
+      "loss": 1.1253,
+      "step": 16456
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 12.756241798400879,
+      "learning_rate": 6.231017027151404e-06,
+      "loss": 0.3802,
+      "step": 16457
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 12.611263275146484,
+      "learning_rate": 6.230180312094717e-06,
+      "loss": 1.9548,
+      "step": 16458
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 27.46123695373535,
+      "learning_rate": 6.229343597038029e-06,
+      "loss": 2.2183,
+      "step": 16459
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 28.820478439331055,
+      "learning_rate": 6.228506881981342e-06,
+      "loss": 3.7846,
+      "step": 16460
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 13.83780574798584,
+      "learning_rate": 6.227670166924654e-06,
+      "loss": 0.4726,
+      "step": 16461
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 11.4133939743042,
+      "learning_rate": 6.226833451867967e-06,
+      "loss": 0.9803,
+      "step": 16462
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 10.623127937316895,
+      "learning_rate": 6.22599673681128e-06,
+      "loss": 1.0674,
+      "step": 16463
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 47.73965072631836,
+      "learning_rate": 6.225160021754592e-06,
+      "loss": 0.5572,
+      "step": 16464
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 26.700315475463867,
+      "learning_rate": 6.224323306697905e-06,
+      "loss": 1.035,
+      "step": 16465
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 9.991043090820312,
+      "learning_rate": 6.223486591641217e-06,
+      "loss": 0.4435,
+      "step": 16466
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 7.870398044586182,
+      "learning_rate": 6.22264987658453e-06,
+      "loss": 0.3138,
+      "step": 16467
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 45.63023376464844,
+      "learning_rate": 6.221813161527842e-06,
+      "loss": 0.3552,
+      "step": 16468
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 49.17428970336914,
+      "learning_rate": 6.220976446471155e-06,
+      "loss": 3.2582,
+      "step": 16469
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 10.028911590576172,
+      "learning_rate": 6.220139731414467e-06,
+      "loss": 1.1186,
+      "step": 16470
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 12.706177711486816,
+      "learning_rate": 6.21930301635778e-06,
+      "loss": 1.1401,
+      "step": 16471
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.1864635944366455,
+      "learning_rate": 6.2184663013010926e-06,
+      "loss": 0.1269,
+      "step": 16472
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 14.817275047302246,
+      "learning_rate": 6.2176295862444045e-06,
+      "loss": 0.8975,
+      "step": 16473
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 18.514623641967773,
+      "learning_rate": 6.216792871187718e-06,
+      "loss": 0.542,
+      "step": 16474
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 21.610883712768555,
+      "learning_rate": 6.21595615613103e-06,
+      "loss": 2.3149,
+      "step": 16475
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 21.595613479614258,
+      "learning_rate": 6.215119441074343e-06,
+      "loss": 2.1623,
+      "step": 16476
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 18.9511661529541,
+      "learning_rate": 6.214282726017655e-06,
+      "loss": 0.5788,
+      "step": 16477
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 8.170537948608398,
+      "learning_rate": 6.213446010960968e-06,
+      "loss": 1.1928,
+      "step": 16478
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 10.969879150390625,
+      "learning_rate": 6.2126092959042805e-06,
+      "loss": 0.5808,
+      "step": 16479
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 22.910619735717773,
+      "learning_rate": 6.2117725808475924e-06,
+      "loss": 1.5677,
+      "step": 16480
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 11.913558006286621,
+      "learning_rate": 6.210935865790906e-06,
+      "loss": 0.3098,
+      "step": 16481
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 13.480531692504883,
+      "learning_rate": 6.210099150734218e-06,
+      "loss": 1.3747,
+      "step": 16482
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 14.577484130859375,
+      "learning_rate": 6.209262435677531e-06,
+      "loss": 0.4669,
+      "step": 16483
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 6.817378044128418,
+      "learning_rate": 6.208425720620843e-06,
+      "loss": 0.2599,
+      "step": 16484
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.815977096557617,
+      "learning_rate": 6.207589005564156e-06,
+      "loss": 0.6524,
+      "step": 16485
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 44.233131408691406,
+      "learning_rate": 6.206752290507468e-06,
+      "loss": 1.8451,
+      "step": 16486
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 15.90046215057373,
+      "learning_rate": 6.20591557545078e-06,
+      "loss": 2.2867,
+      "step": 16487
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 54.03504180908203,
+      "learning_rate": 6.205078860394094e-06,
+      "loss": 1.2996,
+      "step": 16488
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 17.677053451538086,
+      "learning_rate": 6.204242145337406e-06,
+      "loss": 1.2908,
+      "step": 16489
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 7.055785655975342,
+      "learning_rate": 6.203405430280719e-06,
+      "loss": 0.3546,
+      "step": 16490
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 27.47056007385254,
+      "learning_rate": 6.202568715224031e-06,
+      "loss": 1.6954,
+      "step": 16491
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 11.763664245605469,
+      "learning_rate": 6.2017320001673435e-06,
+      "loss": 0.8556,
+      "step": 16492
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 14.156816482543945,
+      "learning_rate": 6.200895285110656e-06,
+      "loss": 1.5798,
+      "step": 16493
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.979343891143799,
+      "learning_rate": 6.200058570053968e-06,
+      "loss": 0.1109,
+      "step": 16494
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 11.397286415100098,
+      "learning_rate": 6.199221854997282e-06,
+      "loss": 1.6852,
+      "step": 16495
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 8.399728775024414,
+      "learning_rate": 6.198385139940594e-06,
+      "loss": 0.4494,
+      "step": 16496
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 11.568540573120117,
+      "learning_rate": 6.197548424883907e-06,
+      "loss": 0.4989,
+      "step": 16497
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 6.6547112464904785,
+      "learning_rate": 6.196711709827219e-06,
+      "loss": 0.2911,
+      "step": 16498
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 8.213935852050781,
+      "learning_rate": 6.195874994770531e-06,
+      "loss": 0.4342,
+      "step": 16499
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 63.703330993652344,
+      "learning_rate": 6.195038279713844e-06,
+      "loss": 1.4025,
+      "step": 16500
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 41.80888366699219,
+      "learning_rate": 6.194201564657156e-06,
+      "loss": 0.7504,
+      "step": 16501
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 14.639147758483887,
+      "learning_rate": 6.19336484960047e-06,
+      "loss": 0.9503,
+      "step": 16502
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 11.767823219299316,
+      "learning_rate": 6.192528134543782e-06,
+      "loss": 0.492,
+      "step": 16503
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 13.004864692687988,
+      "learning_rate": 6.191691419487094e-06,
+      "loss": 1.4874,
+      "step": 16504
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 7.3194074630737305,
+      "learning_rate": 6.1908547044304066e-06,
+      "loss": 0.3932,
+      "step": 16505
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.1155903339385986,
+      "learning_rate": 6.1900179893737185e-06,
+      "loss": 0.0421,
+      "step": 16506
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 10.562871932983398,
+      "learning_rate": 6.189181274317032e-06,
+      "loss": 0.2833,
+      "step": 16507
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 51.60620880126953,
+      "learning_rate": 6.188344559260344e-06,
+      "loss": 0.9758,
+      "step": 16508
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 22.08502960205078,
+      "learning_rate": 6.187507844203658e-06,
+      "loss": 1.843,
+      "step": 16509
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 21.697383880615234,
+      "learning_rate": 6.18667112914697e-06,
+      "loss": 2.187,
+      "step": 16510
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 23.2861328125,
+      "learning_rate": 6.185834414090282e-06,
+      "loss": 0.6558,
+      "step": 16511
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 10.984660148620605,
+      "learning_rate": 6.1849976990335945e-06,
+      "loss": 0.549,
+      "step": 16512
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 17.949138641357422,
+      "learning_rate": 6.1841609839769064e-06,
+      "loss": 1.1071,
+      "step": 16513
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 18.823471069335938,
+      "learning_rate": 6.18332426892022e-06,
+      "loss": 0.9926,
+      "step": 16514
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 18.417369842529297,
+      "learning_rate": 6.182487553863532e-06,
+      "loss": 1.7676,
+      "step": 16515
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 55.80198669433594,
+      "learning_rate": 6.181650838806845e-06,
+      "loss": 1.1355,
+      "step": 16516
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 14.4066743850708,
+      "learning_rate": 6.180814123750158e-06,
+      "loss": 2.8706,
+      "step": 16517
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 30.233991622924805,
+      "learning_rate": 6.17997740869347e-06,
+      "loss": 1.1455,
+      "step": 16518
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 19.445171356201172,
+      "learning_rate": 6.179140693636782e-06,
+      "loss": 1.1413,
+      "step": 16519
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 24.63876724243164,
+      "learning_rate": 6.178303978580094e-06,
+      "loss": 0.8298,
+      "step": 16520
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 18.23326301574707,
+      "learning_rate": 6.177467263523408e-06,
+      "loss": 2.423,
+      "step": 16521
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 100.80715942382812,
+      "learning_rate": 6.17663054846672e-06,
+      "loss": 1.3618,
+      "step": 16522
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 15.919983863830566,
+      "learning_rate": 6.175793833410033e-06,
+      "loss": 0.9049,
+      "step": 16523
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 20.7061710357666,
+      "learning_rate": 6.1749571183533456e-06,
+      "loss": 0.5704,
+      "step": 16524
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 29.360403060913086,
+      "learning_rate": 6.1741204032966575e-06,
+      "loss": 1.7679,
+      "step": 16525
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 22.493640899658203,
+      "learning_rate": 6.17328368823997e-06,
+      "loss": 1.6815,
+      "step": 16526
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 26.56145477294922,
+      "learning_rate": 6.172446973183282e-06,
+      "loss": 0.378,
+      "step": 16527
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 18.64031410217285,
+      "learning_rate": 6.171610258126596e-06,
+      "loss": 0.4292,
+      "step": 16528
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 6.236285209655762,
+      "learning_rate": 6.170773543069908e-06,
+      "loss": 0.7544,
+      "step": 16529
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 54.25901412963867,
+      "learning_rate": 6.169936828013221e-06,
+      "loss": 0.6793,
+      "step": 16530
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 6.192501068115234,
+      "learning_rate": 6.1691001129565335e-06,
+      "loss": 1.6053,
+      "step": 16531
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.473605632781982,
+      "learning_rate": 6.1682633978998454e-06,
+      "loss": 1.059,
+      "step": 16532
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 38.38534164428711,
+      "learning_rate": 6.167426682843158e-06,
+      "loss": 2.0188,
+      "step": 16533
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.745693206787109,
+      "learning_rate": 6.16658996778647e-06,
+      "loss": 2.0387,
+      "step": 16534
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 9.371700286865234,
+      "learning_rate": 6.165753252729784e-06,
+      "loss": 2.7258,
+      "step": 16535
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 14.374646186828613,
+      "learning_rate": 6.164916537673096e-06,
+      "loss": 1.3601,
+      "step": 16536
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 45.54530334472656,
+      "learning_rate": 6.164079822616409e-06,
+      "loss": 3.1562,
+      "step": 16537
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 12.899748802185059,
+      "learning_rate": 6.163243107559721e-06,
+      "loss": 0.9303,
+      "step": 16538
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 8.409440040588379,
+      "learning_rate": 6.162406392503033e-06,
+      "loss": 1.0218,
+      "step": 16539
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 16.851974487304688,
+      "learning_rate": 6.161569677446346e-06,
+      "loss": 0.3433,
+      "step": 16540
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 15.8661470413208,
+      "learning_rate": 6.160732962389658e-06,
+      "loss": 1.2453,
+      "step": 16541
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 33.72930908203125,
+      "learning_rate": 6.159896247332972e-06,
+      "loss": 1.4199,
+      "step": 16542
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 99.52818298339844,
+      "learning_rate": 6.159059532276284e-06,
+      "loss": 2.0069,
+      "step": 16543
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 6.886445045471191,
+      "learning_rate": 6.1582228172195965e-06,
+      "loss": 0.8724,
+      "step": 16544
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 12.840018272399902,
+      "learning_rate": 6.1573861021629085e-06,
+      "loss": 1.1763,
+      "step": 16545
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 15.798508644104004,
+      "learning_rate": 6.156549387106221e-06,
+      "loss": 0.8332,
+      "step": 16546
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 13.992999076843262,
+      "learning_rate": 6.155712672049534e-06,
+      "loss": 0.3119,
+      "step": 16547
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 16.383399963378906,
+      "learning_rate": 6.154875956992846e-06,
+      "loss": 0.6404,
+      "step": 16548
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 12.821876525878906,
+      "learning_rate": 6.15403924193616e-06,
+      "loss": 1.9272,
+      "step": 16549
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 11.50905704498291,
+      "learning_rate": 6.153202526879472e-06,
+      "loss": 0.5468,
+      "step": 16550
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 72.6382827758789,
+      "learning_rate": 6.1523658118227844e-06,
+      "loss": 1.3661,
+      "step": 16551
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 12.334877967834473,
+      "learning_rate": 6.151529096766096e-06,
+      "loss": 1.9228,
+      "step": 16552
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 14.781194686889648,
+      "learning_rate": 6.150692381709409e-06,
+      "loss": 0.684,
+      "step": 16553
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 48.85089111328125,
+      "learning_rate": 6.149855666652722e-06,
+      "loss": 2.3371,
+      "step": 16554
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 18.68092155456543,
+      "learning_rate": 6.149018951596034e-06,
+      "loss": 0.6029,
+      "step": 16555
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 10.330491065979004,
+      "learning_rate": 6.148182236539348e-06,
+      "loss": 0.611,
+      "step": 16556
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 14.230433464050293,
+      "learning_rate": 6.1473455214826595e-06,
+      "loss": 0.798,
+      "step": 16557
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 8.532289505004883,
+      "learning_rate": 6.146508806425972e-06,
+      "loss": 1.5137,
+      "step": 16558
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 10.542137145996094,
+      "learning_rate": 6.145672091369284e-06,
+      "loss": 0.5654,
+      "step": 16559
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 16.967517852783203,
+      "learning_rate": 6.144835376312597e-06,
+      "loss": 0.8737,
+      "step": 16560
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 23.06812286376953,
+      "learning_rate": 6.14399866125591e-06,
+      "loss": 1.518,
+      "step": 16561
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 47.58055114746094,
+      "learning_rate": 6.143161946199222e-06,
+      "loss": 1.056,
+      "step": 16562
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.630557060241699,
+      "learning_rate": 6.1423252311425355e-06,
+      "loss": 0.1811,
+      "step": 16563
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 17.748130798339844,
+      "learning_rate": 6.1414885160858475e-06,
+      "loss": 0.6627,
+      "step": 16564
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 14.95833969116211,
+      "learning_rate": 6.14065180102916e-06,
+      "loss": 0.5508,
+      "step": 16565
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 10.208039283752441,
+      "learning_rate": 6.139815085972472e-06,
+      "loss": 1.4903,
+      "step": 16566
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 13.743385314941406,
+      "learning_rate": 6.138978370915785e-06,
+      "loss": 0.672,
+      "step": 16567
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 9.239084243774414,
+      "learning_rate": 6.138141655859098e-06,
+      "loss": 0.3315,
+      "step": 16568
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 41.48976135253906,
+      "learning_rate": 6.13730494080241e-06,
+      "loss": 1.9272,
+      "step": 16569
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 95.03357696533203,
+      "learning_rate": 6.1364682257457234e-06,
+      "loss": 1.5247,
+      "step": 16570
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 63.016475677490234,
+      "learning_rate": 6.135631510689035e-06,
+      "loss": 1.0691,
+      "step": 16571
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 6.475004196166992,
+      "learning_rate": 6.134794795632348e-06,
+      "loss": 0.4216,
+      "step": 16572
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 15.221724510192871,
+      "learning_rate": 6.13395808057566e-06,
+      "loss": 0.7941,
+      "step": 16573
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 17.43982696533203,
+      "learning_rate": 6.133121365518973e-06,
+      "loss": 2.5072,
+      "step": 16574
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 8.709810256958008,
+      "learning_rate": 6.132284650462286e-06,
+      "loss": 0.5433,
+      "step": 16575
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 22.730981826782227,
+      "learning_rate": 6.131447935405598e-06,
+      "loss": 1.9044,
+      "step": 16576
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 28.671415328979492,
+      "learning_rate": 6.130611220348911e-06,
+      "loss": 1.8637,
+      "step": 16577
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 6.897902011871338,
+      "learning_rate": 6.129774505292223e-06,
+      "loss": 0.3145,
+      "step": 16578
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 57.18122863769531,
+      "learning_rate": 6.128937790235536e-06,
+      "loss": 1.7443,
+      "step": 16579
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 38.12550735473633,
+      "learning_rate": 6.128101075178848e-06,
+      "loss": 2.8525,
+      "step": 16580
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 9.102025985717773,
+      "learning_rate": 6.12726436012216e-06,
+      "loss": 0.8997,
+      "step": 16581
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 14.379570960998535,
+      "learning_rate": 6.126427645065474e-06,
+      "loss": 1.4295,
+      "step": 16582
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 21.59740447998047,
+      "learning_rate": 6.125590930008786e-06,
+      "loss": 2.4355,
+      "step": 16583
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 9.414067268371582,
+      "learning_rate": 6.124754214952099e-06,
+      "loss": 0.7416,
+      "step": 16584
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 9.886018753051758,
+      "learning_rate": 6.123917499895411e-06,
+      "loss": 1.4918,
+      "step": 16585
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 14.882733345031738,
+      "learning_rate": 6.123080784838724e-06,
+      "loss": 1.1041,
+      "step": 16586
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 17.257169723510742,
+      "learning_rate": 6.122244069782036e-06,
+      "loss": 0.7118,
+      "step": 16587
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 16.37677574157715,
+      "learning_rate": 6.121407354725348e-06,
+      "loss": 1.1891,
+      "step": 16588
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 9.790069580078125,
+      "learning_rate": 6.120570639668662e-06,
+      "loss": 0.5618,
+      "step": 16589
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 27.5288143157959,
+      "learning_rate": 6.1197339246119735e-06,
+      "loss": 1.4248,
+      "step": 16590
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 13.253369331359863,
+      "learning_rate": 6.118897209555287e-06,
+      "loss": 0.355,
+      "step": 16591
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.6081290245056152,
+      "learning_rate": 6.118060494498599e-06,
+      "loss": 0.0221,
+      "step": 16592
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 11.606402397155762,
+      "learning_rate": 6.117223779441912e-06,
+      "loss": 0.6794,
+      "step": 16593
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 25.089847564697266,
+      "learning_rate": 6.116387064385224e-06,
+      "loss": 0.7521,
+      "step": 16594
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 9.566123008728027,
+      "learning_rate": 6.115550349328536e-06,
+      "loss": 1.3233,
+      "step": 16595
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 8.801780700683594,
+      "learning_rate": 6.1147136342718495e-06,
+      "loss": 0.8277,
+      "step": 16596
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 36.346614837646484,
+      "learning_rate": 6.1138769192151615e-06,
+      "loss": 2.524,
+      "step": 16597
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 14.323728561401367,
+      "learning_rate": 6.113040204158474e-06,
+      "loss": 1.3023,
+      "step": 16598
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 13.32658576965332,
+      "learning_rate": 6.112203489101787e-06,
+      "loss": 0.8885,
+      "step": 16599
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 8.776103019714355,
+      "learning_rate": 6.1113667740451e-06,
+      "loss": 1.3184,
+      "step": 16600
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 5.627133846282959,
+      "learning_rate": 6.110530058988412e-06,
+      "loss": 0.9762,
+      "step": 16601
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 25.990951538085938,
+      "learning_rate": 6.109693343931724e-06,
+      "loss": 1.3927,
+      "step": 16602
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 13.433871269226074,
+      "learning_rate": 6.108856628875037e-06,
+      "loss": 0.6496,
+      "step": 16603
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 11.072653770446777,
+      "learning_rate": 6.108019913818349e-06,
+      "loss": 0.7007,
+      "step": 16604
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 15.921674728393555,
+      "learning_rate": 6.107183198761662e-06,
+      "loss": 0.774,
+      "step": 16605
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.8428282737731934,
+      "learning_rate": 6.106346483704975e-06,
+      "loss": 0.6958,
+      "step": 16606
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 10.573344230651855,
+      "learning_rate": 6.105509768648288e-06,
+      "loss": 0.8632,
+      "step": 16607
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 13.61486530303955,
+      "learning_rate": 6.1046730535916e-06,
+      "loss": 0.2684,
+      "step": 16608
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.86008620262146,
+      "learning_rate": 6.103836338534912e-06,
+      "loss": 0.0893,
+      "step": 16609
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 12.860692024230957,
+      "learning_rate": 6.102999623478225e-06,
+      "loss": 2.0543,
+      "step": 16610
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 13.402647018432617,
+      "learning_rate": 6.102162908421537e-06,
+      "loss": 1.267,
+      "step": 16611
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 11.44622802734375,
+      "learning_rate": 6.10132619336485e-06,
+      "loss": 1.4332,
+      "step": 16612
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 6.49962854385376,
+      "learning_rate": 6.100489478308163e-06,
+      "loss": 0.2258,
+      "step": 16613
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 6.8343987464904785,
+      "learning_rate": 6.099652763251476e-06,
+      "loss": 0.4795,
+      "step": 16614
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 19.482908248901367,
+      "learning_rate": 6.098816048194788e-06,
+      "loss": 0.9212,
+      "step": 16615
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 10.905943870544434,
+      "learning_rate": 6.0979793331381e-06,
+      "loss": 1.0125,
+      "step": 16616
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 13.229659080505371,
+      "learning_rate": 6.097142618081413e-06,
+      "loss": 0.8366,
+      "step": 16617
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 28.90032386779785,
+      "learning_rate": 6.096305903024725e-06,
+      "loss": 2.8562,
+      "step": 16618
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 11.121617317199707,
+      "learning_rate": 6.095469187968038e-06,
+      "loss": 0.6452,
+      "step": 16619
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 14.468332290649414,
+      "learning_rate": 6.094632472911351e-06,
+      "loss": 0.642,
+      "step": 16620
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 17.694711685180664,
+      "learning_rate": 6.093795757854664e-06,
+      "loss": 1.0717,
+      "step": 16621
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 17.3740234375,
+      "learning_rate": 6.0929590427979756e-06,
+      "loss": 0.4831,
+      "step": 16622
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 9.99255084991455,
+      "learning_rate": 6.0921223277412875e-06,
+      "loss": 0.2145,
+      "step": 16623
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 13.872404098510742,
+      "learning_rate": 6.091285612684601e-06,
+      "loss": 1.1057,
+      "step": 16624
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 16.23759651184082,
+      "learning_rate": 6.090448897627913e-06,
+      "loss": 1.049,
+      "step": 16625
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 8.336773872375488,
+      "learning_rate": 6.089612182571226e-06,
+      "loss": 1.2625,
+      "step": 16626
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 22.109317779541016,
+      "learning_rate": 6.088775467514538e-06,
+      "loss": 1.6373,
+      "step": 16627
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.8382434844970703,
+      "learning_rate": 6.0879387524578515e-06,
+      "loss": 0.2619,
+      "step": 16628
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 13.31667709350586,
+      "learning_rate": 6.0871020374011635e-06,
+      "loss": 0.43,
+      "step": 16629
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 16.8389892578125,
+      "learning_rate": 6.0862653223444754e-06,
+      "loss": 2.5586,
+      "step": 16630
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 32.07686996459961,
+      "learning_rate": 6.085428607287789e-06,
+      "loss": 1.9517,
+      "step": 16631
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 98.96281433105469,
+      "learning_rate": 6.084591892231101e-06,
+      "loss": 2.1242,
+      "step": 16632
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 28.124950408935547,
+      "learning_rate": 6.083755177174414e-06,
+      "loss": 1.0739,
+      "step": 16633
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 11.089404106140137,
+      "learning_rate": 6.082918462117726e-06,
+      "loss": 0.789,
+      "step": 16634
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 63.404911041259766,
+      "learning_rate": 6.0820817470610395e-06,
+      "loss": 2.2756,
+      "step": 16635
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 26.401927947998047,
+      "learning_rate": 6.081245032004351e-06,
+      "loss": 2.0859,
+      "step": 16636
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 10.558520317077637,
+      "learning_rate": 6.080408316947663e-06,
+      "loss": 0.891,
+      "step": 16637
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 33.86064910888672,
+      "learning_rate": 6.079571601890977e-06,
+      "loss": 1.4469,
+      "step": 16638
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 17.923425674438477,
+      "learning_rate": 6.078734886834289e-06,
+      "loss": 0.7863,
+      "step": 16639
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 16.607006072998047,
+      "learning_rate": 6.077898171777602e-06,
+      "loss": 0.5759,
+      "step": 16640
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 18.633817672729492,
+      "learning_rate": 6.077061456720914e-06,
+      "loss": 1.5332,
+      "step": 16641
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 16.57147789001465,
+      "learning_rate": 6.076224741664227e-06,
+      "loss": 1.6609,
+      "step": 16642
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 14.349488258361816,
+      "learning_rate": 6.075388026607539e-06,
+      "loss": 2.4245,
+      "step": 16643
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 10.078313827514648,
+      "learning_rate": 6.074551311550851e-06,
+      "loss": 0.6327,
+      "step": 16644
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 36.451690673828125,
+      "learning_rate": 6.073714596494165e-06,
+      "loss": 3.0654,
+      "step": 16645
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 80.59728240966797,
+      "learning_rate": 6.072877881437477e-06,
+      "loss": 1.1205,
+      "step": 16646
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 36.045108795166016,
+      "learning_rate": 6.07204116638079e-06,
+      "loss": 1.4382,
+      "step": 16647
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 9.137932777404785,
+      "learning_rate": 6.071204451324102e-06,
+      "loss": 0.7327,
+      "step": 16648
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 17.748319625854492,
+      "learning_rate": 6.070367736267415e-06,
+      "loss": 0.7783,
+      "step": 16649
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 21.247251510620117,
+      "learning_rate": 6.069531021210727e-06,
+      "loss": 1.3124,
+      "step": 16650
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 26.821157455444336,
+      "learning_rate": 6.068694306154039e-06,
+      "loss": 1.7285,
+      "step": 16651
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 6.4322285652160645,
+      "learning_rate": 6.067857591097353e-06,
+      "loss": 1.6197,
+      "step": 16652
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 6.7288055419921875,
+      "learning_rate": 6.067020876040665e-06,
+      "loss": 0.6719,
+      "step": 16653
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 97.98516845703125,
+      "learning_rate": 6.066184160983978e-06,
+      "loss": 2.367,
+      "step": 16654
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 10.283223152160645,
+      "learning_rate": 6.0653474459272896e-06,
+      "loss": 1.7968,
+      "step": 16655
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 6.09956169128418,
+      "learning_rate": 6.064510730870603e-06,
+      "loss": 0.6225,
+      "step": 16656
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 18.431461334228516,
+      "learning_rate": 6.063674015813915e-06,
+      "loss": 0.7375,
+      "step": 16657
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 8.284697532653809,
+      "learning_rate": 6.062837300757227e-06,
+      "loss": 1.4054,
+      "step": 16658
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 9.18532943725586,
+      "learning_rate": 6.062000585700541e-06,
+      "loss": 0.6537,
+      "step": 16659
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 8.965328216552734,
+      "learning_rate": 6.061163870643853e-06,
+      "loss": 0.6235,
+      "step": 16660
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 100.433837890625,
+      "learning_rate": 6.0603271555871655e-06,
+      "loss": 2.0296,
+      "step": 16661
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 17.930927276611328,
+      "learning_rate": 6.0594904405304775e-06,
+      "loss": 1.2623,
+      "step": 16662
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.940665245056152,
+      "learning_rate": 6.058653725473791e-06,
+      "loss": 0.1586,
+      "step": 16663
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 26.91118621826172,
+      "learning_rate": 6.057817010417103e-06,
+      "loss": 1.7924,
+      "step": 16664
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 23.53251838684082,
+      "learning_rate": 6.056980295360415e-06,
+      "loss": 1.4451,
+      "step": 16665
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 10.728267669677734,
+      "learning_rate": 6.056143580303729e-06,
+      "loss": 0.7361,
+      "step": 16666
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.9758715629577637,
+      "learning_rate": 6.055306865247041e-06,
+      "loss": 0.1412,
+      "step": 16667
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 18.519758224487305,
+      "learning_rate": 6.0544701501903534e-06,
+      "loss": 0.8254,
+      "step": 16668
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 6.1864519119262695,
+      "learning_rate": 6.053633435133665e-06,
+      "loss": 0.5522,
+      "step": 16669
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 22.56398582458496,
+      "learning_rate": 6.052796720076979e-06,
+      "loss": 2.1626,
+      "step": 16670
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.48407506942749,
+      "learning_rate": 6.051960005020291e-06,
+      "loss": 0.2158,
+      "step": 16671
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 29.247690200805664,
+      "learning_rate": 6.051123289963603e-06,
+      "loss": 1.0602,
+      "step": 16672
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.660470962524414,
+      "learning_rate": 6.050286574906916e-06,
+      "loss": 0.1487,
+      "step": 16673
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 17.81192970275879,
+      "learning_rate": 6.0494498598502286e-06,
+      "loss": 1.1854,
+      "step": 16674
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 11.670249938964844,
+      "learning_rate": 6.048613144793541e-06,
+      "loss": 1.5447,
+      "step": 16675
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 7.811960697174072,
+      "learning_rate": 6.047776429736853e-06,
+      "loss": 0.7241,
+      "step": 16676
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 13.082968711853027,
+      "learning_rate": 6.046939714680167e-06,
+      "loss": 0.6176,
+      "step": 16677
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 13.379081726074219,
+      "learning_rate": 6.046102999623479e-06,
+      "loss": 0.5995,
+      "step": 16678
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 10.199109077453613,
+      "learning_rate": 6.045266284566791e-06,
+      "loss": 0.6131,
+      "step": 16679
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 11.826645851135254,
+      "learning_rate": 6.044429569510104e-06,
+      "loss": 0.6941,
+      "step": 16680
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 20.454917907714844,
+      "learning_rate": 6.0435928544534165e-06,
+      "loss": 0.8341,
+      "step": 16681
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 13.764884948730469,
+      "learning_rate": 6.042756139396729e-06,
+      "loss": 1.596,
+      "step": 16682
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 37.32571792602539,
+      "learning_rate": 6.041919424340041e-06,
+      "loss": 2.5734,
+      "step": 16683
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.9186553955078125,
+      "learning_rate": 6.041082709283353e-06,
+      "loss": 0.3749,
+      "step": 16684
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 7.147922039031982,
+      "learning_rate": 6.040245994226667e-06,
+      "loss": 1.8393,
+      "step": 16685
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 13.041387557983398,
+      "learning_rate": 6.039409279169979e-06,
+      "loss": 0.3931,
+      "step": 16686
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 12.86512279510498,
+      "learning_rate": 6.038572564113292e-06,
+      "loss": 0.3173,
+      "step": 16687
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.978394031524658,
+      "learning_rate": 6.037735849056604e-06,
+      "loss": 0.4025,
+      "step": 16688
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 58.6747932434082,
+      "learning_rate": 6.036899133999917e-06,
+      "loss": 1.4673,
+      "step": 16689
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 35.88507080078125,
+      "learning_rate": 6.036062418943229e-06,
+      "loss": 0.612,
+      "step": 16690
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 10.869979858398438,
+      "learning_rate": 6.035225703886541e-06,
+      "loss": 1.1676,
+      "step": 16691
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 63.22040557861328,
+      "learning_rate": 6.034388988829855e-06,
+      "loss": 1.4376,
+      "step": 16692
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 10.475271224975586,
+      "learning_rate": 6.033552273773167e-06,
+      "loss": 1.3598,
+      "step": 16693
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.576467514038086,
+      "learning_rate": 6.0327155587164795e-06,
+      "loss": 0.9442,
+      "step": 16694
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 23.366558074951172,
+      "learning_rate": 6.031878843659792e-06,
+      "loss": 2.5683,
+      "step": 16695
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 18.892393112182617,
+      "learning_rate": 6.031042128603105e-06,
+      "loss": 0.7045,
+      "step": 16696
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 23.35431671142578,
+      "learning_rate": 6.030205413546417e-06,
+      "loss": 1.4328,
+      "step": 16697
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 36.365665435791016,
+      "learning_rate": 6.029368698489729e-06,
+      "loss": 1.9751,
+      "step": 16698
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 7.835363388061523,
+      "learning_rate": 6.028531983433043e-06,
+      "loss": 0.2088,
+      "step": 16699
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 9.498435974121094,
+      "learning_rate": 6.027695268376355e-06,
+      "loss": 1.686,
+      "step": 16700
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 22.374313354492188,
+      "learning_rate": 6.0268585533196674e-06,
+      "loss": 1.1463,
+      "step": 16701
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 11.154014587402344,
+      "learning_rate": 6.02602183826298e-06,
+      "loss": 0.7849,
+      "step": 16702
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 14.308752059936523,
+      "learning_rate": 6.025185123206293e-06,
+      "loss": 1.6134,
+      "step": 16703
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.799342632293701,
+      "learning_rate": 6.024348408149605e-06,
+      "loss": 0.2157,
+      "step": 16704
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 16.234066009521484,
+      "learning_rate": 6.023511693092917e-06,
+      "loss": 0.759,
+      "step": 16705
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 10.211417198181152,
+      "learning_rate": 6.022674978036231e-06,
+      "loss": 0.4717,
+      "step": 16706
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 15.918299674987793,
+      "learning_rate": 6.0218382629795425e-06,
+      "loss": 0.7842,
+      "step": 16707
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.734321594238281,
+      "learning_rate": 6.021001547922855e-06,
+      "loss": 1.2201,
+      "step": 16708
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 7.47617244720459,
+      "learning_rate": 6.020164832866167e-06,
+      "loss": 0.505,
+      "step": 16709
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 10.15281867980957,
+      "learning_rate": 6.019328117809481e-06,
+      "loss": 0.5723,
+      "step": 16710
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.6108498573303223,
+      "learning_rate": 6.018491402752793e-06,
+      "loss": 0.0858,
+      "step": 16711
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 17.00050926208496,
+      "learning_rate": 6.017654687696105e-06,
+      "loss": 0.789,
+      "step": 16712
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 11.504976272583008,
+      "learning_rate": 6.0168179726394185e-06,
+      "loss": 0.2506,
+      "step": 16713
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 12.045175552368164,
+      "learning_rate": 6.0159812575827305e-06,
+      "loss": 0.8217,
+      "step": 16714
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 8.361842155456543,
+      "learning_rate": 6.015144542526043e-06,
+      "loss": 1.6557,
+      "step": 16715
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 27.015836715698242,
+      "learning_rate": 6.014307827469355e-06,
+      "loss": 1.4325,
+      "step": 16716
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 11.400221824645996,
+      "learning_rate": 6.013471112412669e-06,
+      "loss": 0.731,
+      "step": 16717
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 13.500144958496094,
+      "learning_rate": 6.012634397355981e-06,
+      "loss": 1.4767,
+      "step": 16718
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 32.673580169677734,
+      "learning_rate": 6.011797682299293e-06,
+      "loss": 2.2346,
+      "step": 16719
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 61.020599365234375,
+      "learning_rate": 6.0109609672426064e-06,
+      "loss": 1.4349,
+      "step": 16720
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 20.405357360839844,
+      "learning_rate": 6.010124252185918e-06,
+      "loss": 0.7874,
+      "step": 16721
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 24.97215461730957,
+      "learning_rate": 6.009287537129231e-06,
+      "loss": 1.1234,
+      "step": 16722
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 18.511730194091797,
+      "learning_rate": 6.008450822072543e-06,
+      "loss": 2.2126,
+      "step": 16723
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 20.64064598083496,
+      "learning_rate": 6.007614107015857e-06,
+      "loss": 1.0869,
+      "step": 16724
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 7.057304859161377,
+      "learning_rate": 6.006777391959169e-06,
+      "loss": 0.5425,
+      "step": 16725
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 25.64223289489746,
+      "learning_rate": 6.005940676902481e-06,
+      "loss": 1.1163,
+      "step": 16726
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 21.548139572143555,
+      "learning_rate": 6.005103961845794e-06,
+      "loss": 2.5731,
+      "step": 16727
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 35.97419738769531,
+      "learning_rate": 6.004267246789106e-06,
+      "loss": 0.6331,
+      "step": 16728
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 18.08547592163086,
+      "learning_rate": 6.003430531732419e-06,
+      "loss": 0.988,
+      "step": 16729
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 16.485050201416016,
+      "learning_rate": 6.002593816675731e-06,
+      "loss": 2.8046,
+      "step": 16730
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 32.443904876708984,
+      "learning_rate": 6.001757101619045e-06,
+      "loss": 1.4081,
+      "step": 16731
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 14.452461242675781,
+      "learning_rate": 6.000920386562357e-06,
+      "loss": 0.4725,
+      "step": 16732
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 15.986902236938477,
+      "learning_rate": 6.000083671505669e-06,
+      "loss": 2.028,
+      "step": 16733
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 15.833672523498535,
+      "learning_rate": 5.999246956448982e-06,
+      "loss": 3.3999,
+      "step": 16734
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 9.893026351928711,
+      "learning_rate": 5.998410241392294e-06,
+      "loss": 0.439,
+      "step": 16735
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 19.52991485595703,
+      "learning_rate": 5.997573526335607e-06,
+      "loss": 0.9038,
+      "step": 16736
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 16.582826614379883,
+      "learning_rate": 5.996736811278919e-06,
+      "loss": 0.4433,
+      "step": 16737
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 12.043107032775879,
+      "learning_rate": 5.995900096222233e-06,
+      "loss": 0.7207,
+      "step": 16738
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 17.766197204589844,
+      "learning_rate": 5.995063381165545e-06,
+      "loss": 1.7718,
+      "step": 16739
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 35.01610565185547,
+      "learning_rate": 5.9942266661088565e-06,
+      "loss": 1.124,
+      "step": 16740
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 157.57369995117188,
+      "learning_rate": 5.99338995105217e-06,
+      "loss": 0.9192,
+      "step": 16741
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 8.94036865234375,
+      "learning_rate": 5.992553235995482e-06,
+      "loss": 0.2372,
+      "step": 16742
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 10.568174362182617,
+      "learning_rate": 5.991716520938795e-06,
+      "loss": 0.3193,
+      "step": 16743
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 22.63112449645996,
+      "learning_rate": 5.990879805882107e-06,
+      "loss": 2.0692,
+      "step": 16744
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 11.302915573120117,
+      "learning_rate": 5.9900430908254205e-06,
+      "loss": 0.8339,
+      "step": 16745
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 23.451662063598633,
+      "learning_rate": 5.9892063757687325e-06,
+      "loss": 0.9134,
+      "step": 16746
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 20.594402313232422,
+      "learning_rate": 5.9883696607120445e-06,
+      "loss": 1.7236,
+      "step": 16747
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 18.51211929321289,
+      "learning_rate": 5.987532945655358e-06,
+      "loss": 0.8028,
+      "step": 16748
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 93.31168365478516,
+      "learning_rate": 5.98669623059867e-06,
+      "loss": 2.9052,
+      "step": 16749
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 15.28825569152832,
+      "learning_rate": 5.985859515541983e-06,
+      "loss": 2.9189,
+      "step": 16750
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 21.951282501220703,
+      "learning_rate": 5.985022800485295e-06,
+      "loss": 1.1899,
+      "step": 16751
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 21.420368194580078,
+      "learning_rate": 5.9841860854286085e-06,
+      "loss": 0.7361,
+      "step": 16752
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.584796905517578,
+      "learning_rate": 5.98334937037192e-06,
+      "loss": 0.7013,
+      "step": 16753
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 14.52383041381836,
+      "learning_rate": 5.982512655315232e-06,
+      "loss": 0.8589,
+      "step": 16754
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 13.623851776123047,
+      "learning_rate": 5.981675940258545e-06,
+      "loss": 0.4374,
+      "step": 16755
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 17.263608932495117,
+      "learning_rate": 5.980839225201858e-06,
+      "loss": 1.5883,
+      "step": 16756
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 10.26650333404541,
+      "learning_rate": 5.980002510145171e-06,
+      "loss": 0.6201,
+      "step": 16757
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 20.81313705444336,
+      "learning_rate": 5.979165795088483e-06,
+      "loss": 0.9087,
+      "step": 16758
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.574885845184326,
+      "learning_rate": 5.978329080031796e-06,
+      "loss": 0.2811,
+      "step": 16759
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 51.55418014526367,
+      "learning_rate": 5.977492364975108e-06,
+      "loss": 2.3663,
+      "step": 16760
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 7.015112400054932,
+      "learning_rate": 5.97665564991842e-06,
+      "loss": 0.1887,
+      "step": 16761
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 28.71893310546875,
+      "learning_rate": 5.975818934861733e-06,
+      "loss": 2.9956,
+      "step": 16762
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 21.74225425720215,
+      "learning_rate": 5.974982219805046e-06,
+      "loss": 1.2374,
+      "step": 16763
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 12.860550880432129,
+      "learning_rate": 5.974145504748359e-06,
+      "loss": 1.1596,
+      "step": 16764
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 14.887338638305664,
+      "learning_rate": 5.973308789691671e-06,
+      "loss": 1.5552,
+      "step": 16765
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 24.139747619628906,
+      "learning_rate": 5.972472074634984e-06,
+      "loss": 1.1814,
+      "step": 16766
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 18.169158935546875,
+      "learning_rate": 5.971635359578296e-06,
+      "loss": 0.5871,
+      "step": 16767
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 11.368548393249512,
+      "learning_rate": 5.970798644521608e-06,
+      "loss": 1.3178,
+      "step": 16768
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 10.630375862121582,
+      "learning_rate": 5.969961929464921e-06,
+      "loss": 2.2562,
+      "step": 16769
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 23.951719284057617,
+      "learning_rate": 5.969125214408234e-06,
+      "loss": 1.269,
+      "step": 16770
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.449909687042236,
+      "learning_rate": 5.968288499351547e-06,
+      "loss": 1.3628,
+      "step": 16771
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 35.25612258911133,
+      "learning_rate": 5.9674517842948586e-06,
+      "loss": 1.2419,
+      "step": 16772
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.990248680114746,
+      "learning_rate": 5.966615069238172e-06,
+      "loss": 0.1098,
+      "step": 16773
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 23.586139678955078,
+      "learning_rate": 5.965778354181484e-06,
+      "loss": 1.5185,
+      "step": 16774
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 12.270256996154785,
+      "learning_rate": 5.964941639124796e-06,
+      "loss": 0.7521,
+      "step": 16775
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 12.243736267089844,
+      "learning_rate": 5.964104924068109e-06,
+      "loss": 0.8228,
+      "step": 16776
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 31.726364135742188,
+      "learning_rate": 5.963268209011422e-06,
+      "loss": 2.5089,
+      "step": 16777
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 105.982421875,
+      "learning_rate": 5.9624314939547345e-06,
+      "loss": 1.3511,
+      "step": 16778
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 9.604047775268555,
+      "learning_rate": 5.9615947788980465e-06,
+      "loss": 0.3661,
+      "step": 16779
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 13.931992530822754,
+      "learning_rate": 5.96075806384136e-06,
+      "loss": 1.4841,
+      "step": 16780
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 55.37815856933594,
+      "learning_rate": 5.959921348784672e-06,
+      "loss": 1.8066,
+      "step": 16781
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 30.015426635742188,
+      "learning_rate": 5.959084633727984e-06,
+      "loss": 2.854,
+      "step": 16782
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 18.840238571166992,
+      "learning_rate": 5.958247918671297e-06,
+      "loss": 1.0707,
+      "step": 16783
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.141613006591797,
+      "learning_rate": 5.95741120361461e-06,
+      "loss": 0.6381,
+      "step": 16784
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 44.44023132324219,
+      "learning_rate": 5.9565744885579225e-06,
+      "loss": 2.2979,
+      "step": 16785
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 9.82348918914795,
+      "learning_rate": 5.955737773501234e-06,
+      "loss": 0.6119,
+      "step": 16786
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.065723419189453,
+      "learning_rate": 5.954901058444548e-06,
+      "loss": 0.1256,
+      "step": 16787
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 16.968774795532227,
+      "learning_rate": 5.95406434338786e-06,
+      "loss": 1.3901,
+      "step": 16788
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 15.727312088012695,
+      "learning_rate": 5.953227628331172e-06,
+      "loss": 1.1866,
+      "step": 16789
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 25.21003532409668,
+      "learning_rate": 5.952390913274485e-06,
+      "loss": 1.9156,
+      "step": 16790
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 15.855225563049316,
+      "learning_rate": 5.951554198217797e-06,
+      "loss": 0.7767,
+      "step": 16791
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 22.47800636291504,
+      "learning_rate": 5.95071748316111e-06,
+      "loss": 1.3269,
+      "step": 16792
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 18.00648307800293,
+      "learning_rate": 5.949880768104422e-06,
+      "loss": 1.5794,
+      "step": 16793
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 9.49789810180664,
+      "learning_rate": 5.949044053047736e-06,
+      "loss": 0.613,
+      "step": 16794
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.8184410333633423,
+      "learning_rate": 5.948207337991048e-06,
+      "loss": 0.0311,
+      "step": 16795
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 7.746805191040039,
+      "learning_rate": 5.94737062293436e-06,
+      "loss": 0.7932,
+      "step": 16796
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 20.15648651123047,
+      "learning_rate": 5.946533907877673e-06,
+      "loss": 1.8011,
+      "step": 16797
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 38.610687255859375,
+      "learning_rate": 5.945697192820985e-06,
+      "loss": 2.9669,
+      "step": 16798
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 11.16541862487793,
+      "learning_rate": 5.944860477764298e-06,
+      "loss": 0.4904,
+      "step": 16799
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.147881507873535,
+      "learning_rate": 5.94402376270761e-06,
+      "loss": 0.1599,
+      "step": 16800
+    },
+    {
+      "epoch": 2.11,
+      "eval_loss": 0.10721652209758759,
+      "eval_runtime": 95.8916,
+      "eval_samples_per_second": 36.938,
+      "eval_steps_per_second": 36.938,
+      "step": 16800
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 20.84210968017578,
+      "learning_rate": 5.943187047650923e-06,
+      "loss": 1.1602,
+      "step": 16801
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 14.796680450439453,
+      "learning_rate": 5.942350332594236e-06,
+      "loss": 0.8441,
+      "step": 16802
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 18.466554641723633,
+      "learning_rate": 5.941513617537548e-06,
+      "loss": 0.6331,
+      "step": 16803
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.697744846343994,
+      "learning_rate": 5.940676902480861e-06,
+      "loss": 0.2152,
+      "step": 16804
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 46.678199768066406,
+      "learning_rate": 5.9398401874241726e-06,
+      "loss": 1.6034,
+      "step": 16805
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 9.598953247070312,
+      "learning_rate": 5.939003472367486e-06,
+      "loss": 1.4654,
+      "step": 16806
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 7.446468830108643,
+      "learning_rate": 5.938166757310798e-06,
+      "loss": 0.3296,
+      "step": 16807
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 9.030505180358887,
+      "learning_rate": 5.937330042254111e-06,
+      "loss": 1.5311,
+      "step": 16808
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 7.710502624511719,
+      "learning_rate": 5.936493327197424e-06,
+      "loss": 1.4079,
+      "step": 16809
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 21.966482162475586,
+      "learning_rate": 5.935656612140736e-06,
+      "loss": 2.6049,
+      "step": 16810
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 16.916112899780273,
+      "learning_rate": 5.9348198970840485e-06,
+      "loss": 1.845,
+      "step": 16811
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.8744070529937744,
+      "learning_rate": 5.9339831820273605e-06,
+      "loss": 0.0495,
+      "step": 16812
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 20.617137908935547,
+      "learning_rate": 5.933146466970674e-06,
+      "loss": 0.9642,
+      "step": 16813
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 134.33131408691406,
+      "learning_rate": 5.932309751913986e-06,
+      "loss": 1.755,
+      "step": 16814
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.855226516723633,
+      "learning_rate": 5.931473036857299e-06,
+      "loss": 0.2101,
+      "step": 16815
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 10.540989875793457,
+      "learning_rate": 5.930636321800612e-06,
+      "loss": 0.3734,
+      "step": 16816
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.118357181549072,
+      "learning_rate": 5.929799606743924e-06,
+      "loss": 1.5305,
+      "step": 16817
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.998538970947266,
+      "learning_rate": 5.9289628916872364e-06,
+      "loss": 0.4855,
+      "step": 16818
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 9.374042510986328,
+      "learning_rate": 5.928126176630548e-06,
+      "loss": 0.599,
+      "step": 16819
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 21.988372802734375,
+      "learning_rate": 5.927289461573862e-06,
+      "loss": 0.656,
+      "step": 16820
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 19.108642578125,
+      "learning_rate": 5.926452746517174e-06,
+      "loss": 0.8623,
+      "step": 16821
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 9.66044807434082,
+      "learning_rate": 5.925616031460487e-06,
+      "loss": 0.7023,
+      "step": 16822
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 10.207428932189941,
+      "learning_rate": 5.9247793164038e-06,
+      "loss": 0.6425,
+      "step": 16823
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.872533082962036,
+      "learning_rate": 5.9239426013471116e-06,
+      "loss": 0.0445,
+      "step": 16824
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.80725359916687,
+      "learning_rate": 5.923105886290424e-06,
+      "loss": 0.0798,
+      "step": 16825
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 7.395112037658691,
+      "learning_rate": 5.922269171233736e-06,
+      "loss": 0.136,
+      "step": 16826
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 24.68223762512207,
+      "learning_rate": 5.92143245617705e-06,
+      "loss": 1.9311,
+      "step": 16827
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 16.709436416625977,
+      "learning_rate": 5.920595741120362e-06,
+      "loss": 1.2259,
+      "step": 16828
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 37.04315948486328,
+      "learning_rate": 5.919759026063675e-06,
+      "loss": 0.8694,
+      "step": 16829
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 38.659637451171875,
+      "learning_rate": 5.9189223110069875e-06,
+      "loss": 1.38,
+      "step": 16830
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 100.3625717163086,
+      "learning_rate": 5.9180855959502995e-06,
+      "loss": 1.122,
+      "step": 16831
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 84.5021743774414,
+      "learning_rate": 5.917248880893612e-06,
+      "loss": 0.9857,
+      "step": 16832
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 11.296195983886719,
+      "learning_rate": 5.916412165836924e-06,
+      "loss": 0.6122,
+      "step": 16833
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 6.25674295425415,
+      "learning_rate": 5.915575450780238e-06,
+      "loss": 1.6377,
+      "step": 16834
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 12.446991920471191,
+      "learning_rate": 5.91473873572355e-06,
+      "loss": 2.3191,
+      "step": 16835
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 83.3683853149414,
+      "learning_rate": 5.913902020666863e-06,
+      "loss": 2.8401,
+      "step": 16836
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 16.350399017333984,
+      "learning_rate": 5.913065305610175e-06,
+      "loss": 0.758,
+      "step": 16837
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 17.3767032623291,
+      "learning_rate": 5.912228590553487e-06,
+      "loss": 1.6186,
+      "step": 16838
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 21.542068481445312,
+      "learning_rate": 5.9113918754968e-06,
+      "loss": 2.4041,
+      "step": 16839
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 16.57794952392578,
+      "learning_rate": 5.910555160440112e-06,
+      "loss": 0.9186,
+      "step": 16840
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 30.77682113647461,
+      "learning_rate": 5.909718445383426e-06,
+      "loss": 0.3596,
+      "step": 16841
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 22.59938621520996,
+      "learning_rate": 5.908881730326738e-06,
+      "loss": 0.8914,
+      "step": 16842
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 107.36776733398438,
+      "learning_rate": 5.9080450152700506e-06,
+      "loss": 1.1896,
+      "step": 16843
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 65.2825698852539,
+      "learning_rate": 5.9072083002133625e-06,
+      "loss": 0.9863,
+      "step": 16844
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 16.80819320678711,
+      "learning_rate": 5.906371585156675e-06,
+      "loss": 0.5168,
+      "step": 16845
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 41.62934112548828,
+      "learning_rate": 5.905534870099988e-06,
+      "loss": 2.5369,
+      "step": 16846
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 15.434104919433594,
+      "learning_rate": 5.9046981550433e-06,
+      "loss": 2.1379,
+      "step": 16847
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.339593887329102,
+      "learning_rate": 5.903861439986614e-06,
+      "loss": 0.4649,
+      "step": 16848
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 16.37318229675293,
+      "learning_rate": 5.903024724929926e-06,
+      "loss": 0.8275,
+      "step": 16849
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 13.673310279846191,
+      "learning_rate": 5.9021880098732385e-06,
+      "loss": 0.9344,
+      "step": 16850
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 11.147366523742676,
+      "learning_rate": 5.9013512948165504e-06,
+      "loss": 2.0417,
+      "step": 16851
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 12.443256378173828,
+      "learning_rate": 5.900514579759863e-06,
+      "loss": 1.7393,
+      "step": 16852
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 12.487602233886719,
+      "learning_rate": 5.899677864703176e-06,
+      "loss": 1.4971,
+      "step": 16853
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 12.259377479553223,
+      "learning_rate": 5.898841149646488e-06,
+      "loss": 0.9677,
+      "step": 16854
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 17.900497436523438,
+      "learning_rate": 5.898004434589802e-06,
+      "loss": 0.7282,
+      "step": 16855
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 12.962688446044922,
+      "learning_rate": 5.897167719533114e-06,
+      "loss": 1.843,
+      "step": 16856
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 15.524505615234375,
+      "learning_rate": 5.896331004476426e-06,
+      "loss": 0.54,
+      "step": 16857
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 13.852781295776367,
+      "learning_rate": 5.895494289419738e-06,
+      "loss": 0.4138,
+      "step": 16858
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 115.11148071289062,
+      "learning_rate": 5.894657574363051e-06,
+      "loss": 2.0038,
+      "step": 16859
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 24.396728515625,
+      "learning_rate": 5.893820859306364e-06,
+      "loss": 1.4907,
+      "step": 16860
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 17.910776138305664,
+      "learning_rate": 5.892984144249676e-06,
+      "loss": 1.9196,
+      "step": 16861
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 16.473308563232422,
+      "learning_rate": 5.8921474291929896e-06,
+      "loss": 2.9123,
+      "step": 16862
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 9.617788314819336,
+      "learning_rate": 5.8913107141363015e-06,
+      "loss": 0.4149,
+      "step": 16863
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 147.6373291015625,
+      "learning_rate": 5.8904739990796135e-06,
+      "loss": 0.584,
+      "step": 16864
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 7.813459873199463,
+      "learning_rate": 5.889637284022926e-06,
+      "loss": 0.715,
+      "step": 16865
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.4829301834106445,
+      "learning_rate": 5.888800568966238e-06,
+      "loss": 0.4809,
+      "step": 16866
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 26.605653762817383,
+      "learning_rate": 5.887963853909552e-06,
+      "loss": 1.7967,
+      "step": 16867
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 13.526481628417969,
+      "learning_rate": 5.887127138852864e-06,
+      "loss": 0.9315,
+      "step": 16868
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.498909950256348,
+      "learning_rate": 5.8862904237961775e-06,
+      "loss": 0.289,
+      "step": 16869
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 10.260990142822266,
+      "learning_rate": 5.8854537087394894e-06,
+      "loss": 1.0975,
+      "step": 16870
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.1513917446136475,
+      "learning_rate": 5.884616993682801e-06,
+      "loss": 0.172,
+      "step": 16871
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 24.654848098754883,
+      "learning_rate": 5.883780278626114e-06,
+      "loss": 0.7593,
+      "step": 16872
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.956201076507568,
+      "learning_rate": 5.882943563569426e-06,
+      "loss": 1.1053,
+      "step": 16873
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 18.203449249267578,
+      "learning_rate": 5.88210684851274e-06,
+      "loss": 0.6404,
+      "step": 16874
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 18.785099029541016,
+      "learning_rate": 5.881270133456052e-06,
+      "loss": 1.0799,
+      "step": 16875
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 28.527324676513672,
+      "learning_rate": 5.880433418399365e-06,
+      "loss": 1.4479,
+      "step": 16876
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 6.801303386688232,
+      "learning_rate": 5.879596703342677e-06,
+      "loss": 0.8484,
+      "step": 16877
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 15.764060974121094,
+      "learning_rate": 5.878759988285989e-06,
+      "loss": 0.7352,
+      "step": 16878
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 21.629823684692383,
+      "learning_rate": 5.877923273229302e-06,
+      "loss": 0.6906,
+      "step": 16879
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.686356782913208,
+      "learning_rate": 5.877086558172614e-06,
+      "loss": 0.1907,
+      "step": 16880
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 12.804296493530273,
+      "learning_rate": 5.876249843115928e-06,
+      "loss": 0.7735,
+      "step": 16881
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 14.29104995727539,
+      "learning_rate": 5.87541312805924e-06,
+      "loss": 0.9907,
+      "step": 16882
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 95.18588256835938,
+      "learning_rate": 5.8745764130025525e-06,
+      "loss": 2.5333,
+      "step": 16883
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 14.519478797912598,
+      "learning_rate": 5.873739697945865e-06,
+      "loss": 0.5038,
+      "step": 16884
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 110.04188537597656,
+      "learning_rate": 5.872902982889177e-06,
+      "loss": 3.6262,
+      "step": 16885
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 242.27322387695312,
+      "learning_rate": 5.87206626783249e-06,
+      "loss": 0.5165,
+      "step": 16886
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.291407585144043,
+      "learning_rate": 5.871229552775802e-06,
+      "loss": 0.0267,
+      "step": 16887
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 85.1260986328125,
+      "learning_rate": 5.870392837719116e-06,
+      "loss": 0.9073,
+      "step": 16888
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 10.375450134277344,
+      "learning_rate": 5.869556122662428e-06,
+      "loss": 1.2339,
+      "step": 16889
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 5.949151992797852,
+      "learning_rate": 5.86871940760574e-06,
+      "loss": 0.2236,
+      "step": 16890
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 12.326302528381348,
+      "learning_rate": 5.867882692549053e-06,
+      "loss": 1.0115,
+      "step": 16891
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 9.207403182983398,
+      "learning_rate": 5.867045977492365e-06,
+      "loss": 0.4804,
+      "step": 16892
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 7.769590377807617,
+      "learning_rate": 5.866209262435678e-06,
+      "loss": 0.6457,
+      "step": 16893
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 23.980419158935547,
+      "learning_rate": 5.86537254737899e-06,
+      "loss": 1.0284,
+      "step": 16894
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 20.030860900878906,
+      "learning_rate": 5.8645358323223035e-06,
+      "loss": 0.8965,
+      "step": 16895
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 12.609292984008789,
+      "learning_rate": 5.8636991172656155e-06,
+      "loss": 0.9808,
+      "step": 16896
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 9.369558334350586,
+      "learning_rate": 5.862862402208928e-06,
+      "loss": 3.3645,
+      "step": 16897
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 21.751758575439453,
+      "learning_rate": 5.862025687152241e-06,
+      "loss": 2.034,
+      "step": 16898
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 19.85375213623047,
+      "learning_rate": 5.861188972095553e-06,
+      "loss": 1.2635,
+      "step": 16899
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 90.26131439208984,
+      "learning_rate": 5.860352257038866e-06,
+      "loss": 1.43,
+      "step": 16900
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 18.804216384887695,
+      "learning_rate": 5.859515541982178e-06,
+      "loss": 0.8194,
+      "step": 16901
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 9.671285629272461,
+      "learning_rate": 5.8586788269254915e-06,
+      "loss": 0.9071,
+      "step": 16902
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 7.237061977386475,
+      "learning_rate": 5.857842111868803e-06,
+      "loss": 0.4013,
+      "step": 16903
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 29.724605560302734,
+      "learning_rate": 5.857005396812116e-06,
+      "loss": 1.1553,
+      "step": 16904
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 36.172664642333984,
+      "learning_rate": 5.856168681755429e-06,
+      "loss": 0.7432,
+      "step": 16905
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 44.96842575073242,
+      "learning_rate": 5.855331966698741e-06,
+      "loss": 0.6883,
+      "step": 16906
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 65.7833251953125,
+      "learning_rate": 5.854495251642054e-06,
+      "loss": 2.1407,
+      "step": 16907
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 6.288467884063721,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.1448,
+      "step": 16908
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 9.45230484008789,
+      "learning_rate": 5.852821821528679e-06,
+      "loss": 0.4311,
+      "step": 16909
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 23.460378646850586,
+      "learning_rate": 5.851985106471991e-06,
+      "loss": 1.4705,
+      "step": 16910
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 8.106704711914062,
+      "learning_rate": 5.851148391415304e-06,
+      "loss": 2.52,
+      "step": 16911
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 95.96499633789062,
+      "learning_rate": 5.850311676358617e-06,
+      "loss": 1.4573,
+      "step": 16912
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 19.770360946655273,
+      "learning_rate": 5.849474961301929e-06,
+      "loss": 0.7553,
+      "step": 16913
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 10.828676223754883,
+      "learning_rate": 5.848638246245242e-06,
+      "loss": 0.1883,
+      "step": 16914
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 11.930998802185059,
+      "learning_rate": 5.847801531188554e-06,
+      "loss": 0.6466,
+      "step": 16915
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 17.37325096130371,
+      "learning_rate": 5.846964816131867e-06,
+      "loss": 1.3859,
+      "step": 16916
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 5.8114213943481445,
+      "learning_rate": 5.846128101075179e-06,
+      "loss": 0.3087,
+      "step": 16917
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 13.317852973937988,
+      "learning_rate": 5.845291386018492e-06,
+      "loss": 1.6657,
+      "step": 16918
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 15.078089714050293,
+      "learning_rate": 5.844454670961804e-06,
+      "loss": 0.8716,
+      "step": 16919
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 8.380730628967285,
+      "learning_rate": 5.843617955905117e-06,
+      "loss": 0.592,
+      "step": 16920
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 9.53822135925293,
+      "learning_rate": 5.84278124084843e-06,
+      "loss": 1.629,
+      "step": 16921
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 7.400240421295166,
+      "learning_rate": 5.8419445257917416e-06,
+      "loss": 0.4703,
+      "step": 16922
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 11.016195297241211,
+      "learning_rate": 5.841107810735055e-06,
+      "loss": 0.6054,
+      "step": 16923
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 188.6097869873047,
+      "learning_rate": 5.840271095678367e-06,
+      "loss": 1.6023,
+      "step": 16924
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 7.374948978424072,
+      "learning_rate": 5.83943438062168e-06,
+      "loss": 1.0421,
+      "step": 16925
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 11.707968711853027,
+      "learning_rate": 5.838597665564992e-06,
+      "loss": 0.514,
+      "step": 16926
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 10.491926193237305,
+      "learning_rate": 5.837760950508305e-06,
+      "loss": 0.6111,
+      "step": 16927
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 8.48752498626709,
+      "learning_rate": 5.8369242354516175e-06,
+      "loss": 0.2845,
+      "step": 16928
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 23.22374725341797,
+      "learning_rate": 5.8360875203949295e-06,
+      "loss": 0.632,
+      "step": 16929
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 9.480890274047852,
+      "learning_rate": 5.835250805338243e-06,
+      "loss": 0.9876,
+      "step": 16930
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 25.120214462280273,
+      "learning_rate": 5.834414090281555e-06,
+      "loss": 1.0777,
+      "step": 16931
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 18.795116424560547,
+      "learning_rate": 5.833577375224868e-06,
+      "loss": 2.1714,
+      "step": 16932
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 37.169952392578125,
+      "learning_rate": 5.83274066016818e-06,
+      "loss": 1.6353,
+      "step": 16933
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 12.385771751403809,
+      "learning_rate": 5.831903945111493e-06,
+      "loss": 0.9374,
+      "step": 16934
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 7.301438331604004,
+      "learning_rate": 5.8310672300548055e-06,
+      "loss": 0.4351,
+      "step": 16935
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 23.69075584411621,
+      "learning_rate": 5.830230514998117e-06,
+      "loss": 1.438,
+      "step": 16936
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 8.792628288269043,
+      "learning_rate": 5.829393799941431e-06,
+      "loss": 0.4614,
+      "step": 16937
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 35.86687088012695,
+      "learning_rate": 5.828557084884743e-06,
+      "loss": 1.3342,
+      "step": 16938
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 12.236211776733398,
+      "learning_rate": 5.827720369828056e-06,
+      "loss": 0.7526,
+      "step": 16939
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 8.944195747375488,
+      "learning_rate": 5.826883654771368e-06,
+      "loss": 0.834,
+      "step": 16940
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 20.822357177734375,
+      "learning_rate": 5.8260469397146806e-06,
+      "loss": 2.4715,
+      "step": 16941
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 17.76325035095215,
+      "learning_rate": 5.825210224657993e-06,
+      "loss": 1.6494,
+      "step": 16942
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 19.366193771362305,
+      "learning_rate": 5.824373509601305e-06,
+      "loss": 2.7357,
+      "step": 16943
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.908167600631714,
+      "learning_rate": 5.823536794544619e-06,
+      "loss": 0.1785,
+      "step": 16944
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 11.59701156616211,
+      "learning_rate": 5.822700079487931e-06,
+      "loss": 1.4952,
+      "step": 16945
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 15.114361763000488,
+      "learning_rate": 5.821863364431244e-06,
+      "loss": 3.0703,
+      "step": 16946
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 18.187007904052734,
+      "learning_rate": 5.821026649374556e-06,
+      "loss": 0.8608,
+      "step": 16947
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.346529006958008,
+      "learning_rate": 5.820189934317868e-06,
+      "loss": 0.6006,
+      "step": 16948
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 12.202530860900879,
+      "learning_rate": 5.819353219261181e-06,
+      "loss": 0.4576,
+      "step": 16949
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 13.990489959716797,
+      "learning_rate": 5.818516504204493e-06,
+      "loss": 1.1319,
+      "step": 16950
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 11.866337776184082,
+      "learning_rate": 5.817679789147807e-06,
+      "loss": 1.7133,
+      "step": 16951
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 14.335285186767578,
+      "learning_rate": 5.816843074091119e-06,
+      "loss": 0.4969,
+      "step": 16952
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 23.00996971130371,
+      "learning_rate": 5.816006359034432e-06,
+      "loss": 0.8737,
+      "step": 16953
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 29.987459182739258,
+      "learning_rate": 5.815169643977744e-06,
+      "loss": 1.7031,
+      "step": 16954
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 16.34398651123047,
+      "learning_rate": 5.8143329289210556e-06,
+      "loss": 0.7954,
+      "step": 16955
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 8.641459465026855,
+      "learning_rate": 5.813496213864369e-06,
+      "loss": 0.3977,
+      "step": 16956
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 22.90567970275879,
+      "learning_rate": 5.812659498807681e-06,
+      "loss": 1.2081,
+      "step": 16957
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.746809482574463,
+      "learning_rate": 5.811822783750995e-06,
+      "loss": 0.4631,
+      "step": 16958
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.130385398864746,
+      "learning_rate": 5.810986068694307e-06,
+      "loss": 0.3609,
+      "step": 16959
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 10.714170455932617,
+      "learning_rate": 5.8101493536376196e-06,
+      "loss": 0.6854,
+      "step": 16960
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 16.156368255615234,
+      "learning_rate": 5.8093126385809315e-06,
+      "loss": 0.9861,
+      "step": 16961
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 10.297343254089355,
+      "learning_rate": 5.8084759235242435e-06,
+      "loss": 0.4309,
+      "step": 16962
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 62.99007034301758,
+      "learning_rate": 5.807639208467557e-06,
+      "loss": 1.1902,
+      "step": 16963
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 61.34621810913086,
+      "learning_rate": 5.806802493410869e-06,
+      "loss": 2.0277,
+      "step": 16964
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 10.976434707641602,
+      "learning_rate": 5.805965778354182e-06,
+      "loss": 1.5143,
+      "step": 16965
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 24.809417724609375,
+      "learning_rate": 5.805129063297495e-06,
+      "loss": 1.0778,
+      "step": 16966
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 27.914297103881836,
+      "learning_rate": 5.8042923482408075e-06,
+      "loss": 0.6749,
+      "step": 16967
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 11.465100288391113,
+      "learning_rate": 5.8034556331841194e-06,
+      "loss": 0.7415,
+      "step": 16968
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 8.154410362243652,
+      "learning_rate": 5.802618918127431e-06,
+      "loss": 1.5581,
+      "step": 16969
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 15.883102416992188,
+      "learning_rate": 5.801782203070745e-06,
+      "loss": 1.1446,
+      "step": 16970
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 23.598388671875,
+      "learning_rate": 5.800945488014057e-06,
+      "loss": 1.029,
+      "step": 16971
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 7.744470596313477,
+      "learning_rate": 5.80010877295737e-06,
+      "loss": 0.2278,
+      "step": 16972
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 9.633336067199707,
+      "learning_rate": 5.799272057900683e-06,
+      "loss": 0.7205,
+      "step": 16973
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 13.613395690917969,
+      "learning_rate": 5.798435342843995e-06,
+      "loss": 2.3816,
+      "step": 16974
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 27.28310203552246,
+      "learning_rate": 5.797598627787307e-06,
+      "loss": 1.708,
+      "step": 16975
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 23.683673858642578,
+      "learning_rate": 5.796761912730619e-06,
+      "loss": 1.975,
+      "step": 16976
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 17.8934268951416,
+      "learning_rate": 5.795925197673933e-06,
+      "loss": 0.2576,
+      "step": 16977
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 18.01639175415039,
+      "learning_rate": 5.795088482617245e-06,
+      "loss": 1.0333,
+      "step": 16978
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 6.331151485443115,
+      "learning_rate": 5.794251767560558e-06,
+      "loss": 0.6816,
+      "step": 16979
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 21.129030227661133,
+      "learning_rate": 5.7934150525038705e-06,
+      "loss": 1.8913,
+      "step": 16980
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 7.894270896911621,
+      "learning_rate": 5.792578337447183e-06,
+      "loss": 1.8436,
+      "step": 16981
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 22.673084259033203,
+      "learning_rate": 5.791741622390495e-06,
+      "loss": 1.5624,
+      "step": 16982
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.034121990203857,
+      "learning_rate": 5.790904907333807e-06,
+      "loss": 0.2669,
+      "step": 16983
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 10.858423233032227,
+      "learning_rate": 5.790068192277121e-06,
+      "loss": 1.2583,
+      "step": 16984
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 57.927940368652344,
+      "learning_rate": 5.789231477220433e-06,
+      "loss": 3.5892,
+      "step": 16985
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.054805755615234,
+      "learning_rate": 5.788394762163746e-06,
+      "loss": 0.3993,
+      "step": 16986
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.532777786254883,
+      "learning_rate": 5.7875580471070584e-06,
+      "loss": 0.5755,
+      "step": 16987
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 25.14957046508789,
+      "learning_rate": 5.786721332050371e-06,
+      "loss": 2.0354,
+      "step": 16988
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 24.648008346557617,
+      "learning_rate": 5.785884616993683e-06,
+      "loss": 0.8927,
+      "step": 16989
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 19.69295310974121,
+      "learning_rate": 5.785047901936995e-06,
+      "loss": 0.7736,
+      "step": 16990
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.650480031967163,
+      "learning_rate": 5.784211186880309e-06,
+      "loss": 0.1339,
+      "step": 16991
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 8.340292930603027,
+      "learning_rate": 5.783374471823621e-06,
+      "loss": 1.4557,
+      "step": 16992
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 8.41778564453125,
+      "learning_rate": 5.7825377567669336e-06,
+      "loss": 0.9915,
+      "step": 16993
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 8.60283374786377,
+      "learning_rate": 5.7817010417102455e-06,
+      "loss": 0.5233,
+      "step": 16994
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 90.2171859741211,
+      "learning_rate": 5.780864326653559e-06,
+      "loss": 1.8594,
+      "step": 16995
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 14.10474967956543,
+      "learning_rate": 5.780027611596871e-06,
+      "loss": 0.4432,
+      "step": 16996
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 26.008697509765625,
+      "learning_rate": 5.779190896540183e-06,
+      "loss": 1.4974,
+      "step": 16997
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 15.756891250610352,
+      "learning_rate": 5.778354181483497e-06,
+      "loss": 0.9927,
+      "step": 16998
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 17.05091667175293,
+      "learning_rate": 5.777517466426809e-06,
+      "loss": 2.1283,
+      "step": 16999
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.916499137878418,
+      "learning_rate": 5.7766807513701215e-06,
+      "loss": 0.4318,
+      "step": 17000
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 42.44375228881836,
+      "learning_rate": 5.7758440363134334e-06,
+      "loss": 1.2217,
+      "step": 17001
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 15.128802299499512,
+      "learning_rate": 5.775007321256747e-06,
+      "loss": 0.6954,
+      "step": 17002
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 30.88568878173828,
+      "learning_rate": 5.774170606200059e-06,
+      "loss": 1.633,
+      "step": 17003
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 8.705619812011719,
+      "learning_rate": 5.773333891143371e-06,
+      "loss": 0.8603,
+      "step": 17004
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 22.377500534057617,
+      "learning_rate": 5.772497176086685e-06,
+      "loss": 1.6253,
+      "step": 17005
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 6.745636940002441,
+      "learning_rate": 5.771660461029997e-06,
+      "loss": 0.2458,
+      "step": 17006
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 13.070367813110352,
+      "learning_rate": 5.770823745973309e-06,
+      "loss": 1.3823,
+      "step": 17007
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 19.84786033630371,
+      "learning_rate": 5.769987030916621e-06,
+      "loss": 0.6723,
+      "step": 17008
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 10.7432861328125,
+      "learning_rate": 5.769150315859935e-06,
+      "loss": 0.399,
+      "step": 17009
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 27.663814544677734,
+      "learning_rate": 5.768313600803247e-06,
+      "loss": 1.1698,
+      "step": 17010
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 17.132997512817383,
+      "learning_rate": 5.767476885746559e-06,
+      "loss": 0.4798,
+      "step": 17011
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.635779857635498,
+      "learning_rate": 5.7666401706898726e-06,
+      "loss": 0.28,
+      "step": 17012
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 9.170980453491211,
+      "learning_rate": 5.7658034556331845e-06,
+      "loss": 0.5628,
+      "step": 17013
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 11.8899564743042,
+      "learning_rate": 5.764966740576497e-06,
+      "loss": 0.4594,
+      "step": 17014
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 10.861407279968262,
+      "learning_rate": 5.764130025519809e-06,
+      "loss": 2.4474,
+      "step": 17015
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 28.021129608154297,
+      "learning_rate": 5.763293310463123e-06,
+      "loss": 0.9954,
+      "step": 17016
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 10.897455215454102,
+      "learning_rate": 5.762456595406435e-06,
+      "loss": 0.5923,
+      "step": 17017
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 17.00180435180664,
+      "learning_rate": 5.761619880349747e-06,
+      "loss": 1.063,
+      "step": 17018
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 15.839303016662598,
+      "learning_rate": 5.7607831652930605e-06,
+      "loss": 0.5052,
+      "step": 17019
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 426.7574462890625,
+      "learning_rate": 5.7599464502363724e-06,
+      "loss": 3.619,
+      "step": 17020
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 7.838033199310303,
+      "learning_rate": 5.759109735179685e-06,
+      "loss": 0.7132,
+      "step": 17021
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.230981826782227,
+      "learning_rate": 5.758273020122997e-06,
+      "loss": 0.2577,
+      "step": 17022
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 43.44022750854492,
+      "learning_rate": 5.757436305066311e-06,
+      "loss": 1.3863,
+      "step": 17023
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.766050100326538,
+      "learning_rate": 5.756599590009623e-06,
+      "loss": 1.1447,
+      "step": 17024
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 13.0640230178833,
+      "learning_rate": 5.755762874952935e-06,
+      "loss": 1.0206,
+      "step": 17025
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 21.132444381713867,
+      "learning_rate": 5.754926159896248e-06,
+      "loss": 0.5415,
+      "step": 17026
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 87.92571258544922,
+      "learning_rate": 5.75408944483956e-06,
+      "loss": 2.4722,
+      "step": 17027
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 14.061578750610352,
+      "learning_rate": 5.753252729782873e-06,
+      "loss": 1.9392,
+      "step": 17028
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 12.163267135620117,
+      "learning_rate": 5.752416014726185e-06,
+      "loss": 0.4628,
+      "step": 17029
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 15.391785621643066,
+      "learning_rate": 5.751579299669499e-06,
+      "loss": 0.8887,
+      "step": 17030
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 15.223180770874023,
+      "learning_rate": 5.750742584612811e-06,
+      "loss": 0.4689,
+      "step": 17031
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 9.978029251098633,
+      "learning_rate": 5.749905869556123e-06,
+      "loss": 0.4122,
+      "step": 17032
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 13.698570251464844,
+      "learning_rate": 5.749069154499436e-06,
+      "loss": 0.5871,
+      "step": 17033
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 188.66851806640625,
+      "learning_rate": 5.748232439442748e-06,
+      "loss": 1.9903,
+      "step": 17034
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 20.050073623657227,
+      "learning_rate": 5.747395724386061e-06,
+      "loss": 0.6087,
+      "step": 17035
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 23.30031394958496,
+      "learning_rate": 5.746559009329373e-06,
+      "loss": 2.1471,
+      "step": 17036
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 77.1574935913086,
+      "learning_rate": 5.745722294272685e-06,
+      "loss": 0.9452,
+      "step": 17037
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 17.38646697998047,
+      "learning_rate": 5.744885579215999e-06,
+      "loss": 1.301,
+      "step": 17038
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 20.22818946838379,
+      "learning_rate": 5.744048864159311e-06,
+      "loss": 1.7813,
+      "step": 17039
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 9.979143142700195,
+      "learning_rate": 5.743212149102624e-06,
+      "loss": 1.2639,
+      "step": 17040
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 11.882335662841797,
+      "learning_rate": 5.742375434045936e-06,
+      "loss": 0.9601,
+      "step": 17041
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.124966859817505,
+      "learning_rate": 5.741538718989249e-06,
+      "loss": 0.0498,
+      "step": 17042
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 7.987300395965576,
+      "learning_rate": 5.740702003932561e-06,
+      "loss": 0.4838,
+      "step": 17043
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 8.915894508361816,
+      "learning_rate": 5.739865288875873e-06,
+      "loss": 0.2809,
+      "step": 17044
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 239.8693389892578,
+      "learning_rate": 5.7390285738191865e-06,
+      "loss": 1.3925,
+      "step": 17045
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 13.353658676147461,
+      "learning_rate": 5.7381918587624985e-06,
+      "loss": 1.0435,
+      "step": 17046
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 12.716243743896484,
+      "learning_rate": 5.737355143705811e-06,
+      "loss": 0.68,
+      "step": 17047
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 21.51469612121582,
+      "learning_rate": 5.736518428649124e-06,
+      "loss": 1.1867,
+      "step": 17048
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 56.70457077026367,
+      "learning_rate": 5.735681713592437e-06,
+      "loss": 2.0627,
+      "step": 17049
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 23.16199493408203,
+      "learning_rate": 5.734844998535749e-06,
+      "loss": 0.5439,
+      "step": 17050
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 71.07728576660156,
+      "learning_rate": 5.734008283479061e-06,
+      "loss": 2.885,
+      "step": 17051
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 110.95845794677734,
+      "learning_rate": 5.7331715684223745e-06,
+      "loss": 1.0208,
+      "step": 17052
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.101598739624023,
+      "learning_rate": 5.732334853365686e-06,
+      "loss": 0.2971,
+      "step": 17053
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 14.872244834899902,
+      "learning_rate": 5.731498138308999e-06,
+      "loss": 0.692,
+      "step": 17054
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 76.74128723144531,
+      "learning_rate": 5.730661423252312e-06,
+      "loss": 1.8122,
+      "step": 17055
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 12.32666301727295,
+      "learning_rate": 5.729824708195625e-06,
+      "loss": 1.6365,
+      "step": 17056
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 24.103534698486328,
+      "learning_rate": 5.728987993138937e-06,
+      "loss": 0.9405,
+      "step": 17057
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.0472235679626465,
+      "learning_rate": 5.728151278082249e-06,
+      "loss": 0.1581,
+      "step": 17058
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 13.476837158203125,
+      "learning_rate": 5.727314563025562e-06,
+      "loss": 1.5368,
+      "step": 17059
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 13.281248092651367,
+      "learning_rate": 5.726477847968874e-06,
+      "loss": 0.9484,
+      "step": 17060
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 8.271032333374023,
+      "learning_rate": 5.725641132912187e-06,
+      "loss": 0.3287,
+      "step": 17061
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 42.25758743286133,
+      "learning_rate": 5.7248044178555e-06,
+      "loss": 2.0096,
+      "step": 17062
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 11.208483695983887,
+      "learning_rate": 5.723967702798813e-06,
+      "loss": 0.7167,
+      "step": 17063
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.0486040115356445,
+      "learning_rate": 5.723130987742125e-06,
+      "loss": 0.0548,
+      "step": 17064
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 6.347630500793457,
+      "learning_rate": 5.722294272685437e-06,
+      "loss": 1.5378,
+      "step": 17065
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 27.25133514404297,
+      "learning_rate": 5.72145755762875e-06,
+      "loss": 1.6141,
+      "step": 17066
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 22.502073287963867,
+      "learning_rate": 5.720620842572062e-06,
+      "loss": 1.139,
+      "step": 17067
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 16.46257209777832,
+      "learning_rate": 5.719784127515375e-06,
+      "loss": 0.9123,
+      "step": 17068
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 23.177297592163086,
+      "learning_rate": 5.718947412458688e-06,
+      "loss": 0.9117,
+      "step": 17069
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 19.314239501953125,
+      "learning_rate": 5.718110697402001e-06,
+      "loss": 1.452,
+      "step": 17070
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 6.970067024230957,
+      "learning_rate": 5.717273982345313e-06,
+      "loss": 0.3654,
+      "step": 17071
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 35.67385482788086,
+      "learning_rate": 5.7164372672886246e-06,
+      "loss": 0.9929,
+      "step": 17072
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 18.922805786132812,
+      "learning_rate": 5.715600552231938e-06,
+      "loss": 0.4126,
+      "step": 17073
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 33.534156799316406,
+      "learning_rate": 5.71476383717525e-06,
+      "loss": 0.8119,
+      "step": 17074
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 54.796119689941406,
+      "learning_rate": 5.713927122118563e-06,
+      "loss": 1.3977,
+      "step": 17075
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 13.54931354522705,
+      "learning_rate": 5.713090407061875e-06,
+      "loss": 0.8326,
+      "step": 17076
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 10.145544052124023,
+      "learning_rate": 5.712253692005189e-06,
+      "loss": 0.9113,
+      "step": 17077
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 29.392499923706055,
+      "learning_rate": 5.7114169769485005e-06,
+      "loss": 1.2521,
+      "step": 17078
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 6.5416717529296875,
+      "learning_rate": 5.7105802618918125e-06,
+      "loss": 0.8656,
+      "step": 17079
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 9.528194427490234,
+      "learning_rate": 5.709743546835126e-06,
+      "loss": 0.581,
+      "step": 17080
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 18.567523956298828,
+      "learning_rate": 5.708906831778438e-06,
+      "loss": 0.8522,
+      "step": 17081
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 46.99943923950195,
+      "learning_rate": 5.708070116721751e-06,
+      "loss": 0.7091,
+      "step": 17082
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 12.262152671813965,
+      "learning_rate": 5.707233401665063e-06,
+      "loss": 0.6784,
+      "step": 17083
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 7.790750503540039,
+      "learning_rate": 5.7063966866083765e-06,
+      "loss": 1.1704,
+      "step": 17084
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 28.657535552978516,
+      "learning_rate": 5.7055599715516885e-06,
+      "loss": 1.029,
+      "step": 17085
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 15.705316543579102,
+      "learning_rate": 5.704723256495e-06,
+      "loss": 0.5097,
+      "step": 17086
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.930361032485962,
+      "learning_rate": 5.703886541438314e-06,
+      "loss": 0.2663,
+      "step": 17087
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 7.7618608474731445,
+      "learning_rate": 5.703049826381626e-06,
+      "loss": 0.8066,
+      "step": 17088
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 12.091531753540039,
+      "learning_rate": 5.702213111324939e-06,
+      "loss": 0.8412,
+      "step": 17089
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 24.859027862548828,
+      "learning_rate": 5.701376396268251e-06,
+      "loss": 1.9814,
+      "step": 17090
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 15.993396759033203,
+      "learning_rate": 5.700539681211564e-06,
+      "loss": 0.7045,
+      "step": 17091
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 42.554256439208984,
+      "learning_rate": 5.699702966154876e-06,
+      "loss": 1.4175,
+      "step": 17092
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 16.24115753173828,
+      "learning_rate": 5.698866251098188e-06,
+      "loss": 0.5687,
+      "step": 17093
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.3872950077056885,
+      "learning_rate": 5.698029536041502e-06,
+      "loss": 0.1129,
+      "step": 17094
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 10.936417579650879,
+      "learning_rate": 5.697192820984814e-06,
+      "loss": 0.5392,
+      "step": 17095
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 30.877866744995117,
+      "learning_rate": 5.696356105928127e-06,
+      "loss": 1.763,
+      "step": 17096
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 7.140477657318115,
+      "learning_rate": 5.695519390871439e-06,
+      "loss": 0.8474,
+      "step": 17097
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 28.2568359375,
+      "learning_rate": 5.694682675814752e-06,
+      "loss": 1.3709,
+      "step": 17098
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 14.172971725463867,
+      "learning_rate": 5.693845960758064e-06,
+      "loss": 1.1605,
+      "step": 17099
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 10.639191627502441,
+      "learning_rate": 5.693009245701376e-06,
+      "loss": 1.0971,
+      "step": 17100
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 6.369710445404053,
+      "learning_rate": 5.69217253064469e-06,
+      "loss": 0.2289,
+      "step": 17101
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 11.544483184814453,
+      "learning_rate": 5.691335815588002e-06,
+      "loss": 1.5321,
+      "step": 17102
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 11.986268043518066,
+      "learning_rate": 5.690499100531315e-06,
+      "loss": 0.9125,
+      "step": 17103
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 21.226198196411133,
+      "learning_rate": 5.689662385474627e-06,
+      "loss": 1.0037,
+      "step": 17104
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 14.601465225219727,
+      "learning_rate": 5.68882567041794e-06,
+      "loss": 0.6547,
+      "step": 17105
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 6.2679009437561035,
+      "learning_rate": 5.687988955361252e-06,
+      "loss": 0.3051,
+      "step": 17106
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 101.8865737915039,
+      "learning_rate": 5.687152240304564e-06,
+      "loss": 1.8805,
+      "step": 17107
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 14.7133150100708,
+      "learning_rate": 5.686315525247878e-06,
+      "loss": 0.8835,
+      "step": 17108
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 10.83020305633545,
+      "learning_rate": 5.68547881019119e-06,
+      "loss": 0.7251,
+      "step": 17109
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 287.1812438964844,
+      "learning_rate": 5.6846420951345026e-06,
+      "loss": 2.1167,
+      "step": 17110
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 6.349530220031738,
+      "learning_rate": 5.6838053800778145e-06,
+      "loss": 0.516,
+      "step": 17111
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 13.912372589111328,
+      "learning_rate": 5.682968665021128e-06,
+      "loss": 0.9926,
+      "step": 17112
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 17.72711944580078,
+      "learning_rate": 5.68213194996444e-06,
+      "loss": 1.1207,
+      "step": 17113
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 22.058942794799805,
+      "learning_rate": 5.681295234907752e-06,
+      "loss": 1.7815,
+      "step": 17114
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.502975940704346,
+      "learning_rate": 5.680458519851066e-06,
+      "loss": 0.1729,
+      "step": 17115
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.920248985290527,
+      "learning_rate": 5.679621804794378e-06,
+      "loss": 0.261,
+      "step": 17116
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 24.765287399291992,
+      "learning_rate": 5.6787850897376905e-06,
+      "loss": 1.0169,
+      "step": 17117
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 24.200904846191406,
+      "learning_rate": 5.6779483746810024e-06,
+      "loss": 1.4074,
+      "step": 17118
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 19.04697608947754,
+      "learning_rate": 5.677111659624316e-06,
+      "loss": 0.7482,
+      "step": 17119
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 24.989364624023438,
+      "learning_rate": 5.676274944567628e-06,
+      "loss": 1.2341,
+      "step": 17120
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 38.26375961303711,
+      "learning_rate": 5.67543822951094e-06,
+      "loss": 1.3172,
+      "step": 17121
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 22.313762664794922,
+      "learning_rate": 5.674601514454253e-06,
+      "loss": 0.889,
+      "step": 17122
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 16.008052825927734,
+      "learning_rate": 5.673764799397566e-06,
+      "loss": 1.0246,
+      "step": 17123
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 19.361263275146484,
+      "learning_rate": 5.672928084340878e-06,
+      "loss": 0.898,
+      "step": 17124
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 22.887283325195312,
+      "learning_rate": 5.67209136928419e-06,
+      "loss": 1.3125,
+      "step": 17125
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 76.33627319335938,
+      "learning_rate": 5.671254654227504e-06,
+      "loss": 2.9896,
+      "step": 17126
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 8.603480339050293,
+      "learning_rate": 5.670417939170816e-06,
+      "loss": 0.8129,
+      "step": 17127
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.5203114748001099,
+      "learning_rate": 5.669581224114128e-06,
+      "loss": 0.0294,
+      "step": 17128
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 11.878643035888672,
+      "learning_rate": 5.668744509057441e-06,
+      "loss": 2.1819,
+      "step": 17129
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 8.10049057006836,
+      "learning_rate": 5.6679077940007535e-06,
+      "loss": 1.4186,
+      "step": 17130
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 8.806018829345703,
+      "learning_rate": 5.667071078944066e-06,
+      "loss": 0.2814,
+      "step": 17131
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 14.122865676879883,
+      "learning_rate": 5.666234363887378e-06,
+      "loss": 2.386,
+      "step": 17132
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 28.183515548706055,
+      "learning_rate": 5.665397648830692e-06,
+      "loss": 1.6092,
+      "step": 17133
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 26.284934997558594,
+      "learning_rate": 5.664560933774004e-06,
+      "loss": 0.4936,
+      "step": 17134
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 11.293306350708008,
+      "learning_rate": 5.663724218717316e-06,
+      "loss": 0.8182,
+      "step": 17135
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 7.348964214324951,
+      "learning_rate": 5.662887503660629e-06,
+      "loss": 0.8782,
+      "step": 17136
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 25.33659553527832,
+      "learning_rate": 5.6620507886039414e-06,
+      "loss": 0.9993,
+      "step": 17137
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 27.249738693237305,
+      "learning_rate": 5.661214073547254e-06,
+      "loss": 0.1993,
+      "step": 17138
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.9141499996185303,
+      "learning_rate": 5.660377358490566e-06,
+      "loss": 1.3535,
+      "step": 17139
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 20.09246826171875,
+      "learning_rate": 5.65954064343388e-06,
+      "loss": 0.9172,
+      "step": 17140
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 23.448715209960938,
+      "learning_rate": 5.658703928377192e-06,
+      "loss": 0.6893,
+      "step": 17141
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 11.247332572937012,
+      "learning_rate": 5.657867213320504e-06,
+      "loss": 0.3365,
+      "step": 17142
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 21.369384765625,
+      "learning_rate": 5.6570304982638166e-06,
+      "loss": 1.2574,
+      "step": 17143
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 21.29892349243164,
+      "learning_rate": 5.656193783207129e-06,
+      "loss": 0.6715,
+      "step": 17144
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 13.750541687011719,
+      "learning_rate": 5.655357068150442e-06,
+      "loss": 1.0391,
+      "step": 17145
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 10.424272537231445,
+      "learning_rate": 5.654520353093754e-06,
+      "loss": 1.0453,
+      "step": 17146
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 27.254045486450195,
+      "learning_rate": 5.653683638037068e-06,
+      "loss": 3.0949,
+      "step": 17147
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 14.978679656982422,
+      "learning_rate": 5.65284692298038e-06,
+      "loss": 0.7781,
+      "step": 17148
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 11.079421997070312,
+      "learning_rate": 5.652010207923692e-06,
+      "loss": 0.5054,
+      "step": 17149
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 106.8862075805664,
+      "learning_rate": 5.6511734928670045e-06,
+      "loss": 2.5155,
+      "step": 17150
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 27.096290588378906,
+      "learning_rate": 5.650336777810317e-06,
+      "loss": 0.6781,
+      "step": 17151
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 16.702634811401367,
+      "learning_rate": 5.64950006275363e-06,
+      "loss": 0.9907,
+      "step": 17152
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 15.890658378601074,
+      "learning_rate": 5.648663347696942e-06,
+      "loss": 0.7768,
+      "step": 17153
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 14.421810150146484,
+      "learning_rate": 5.647826632640256e-06,
+      "loss": 0.846,
+      "step": 17154
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 8.938249588012695,
+      "learning_rate": 5.646989917583568e-06,
+      "loss": 0.4093,
+      "step": 17155
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 84.1568374633789,
+      "learning_rate": 5.64615320252688e-06,
+      "loss": 1.7657,
+      "step": 17156
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 6.703531742095947,
+      "learning_rate": 5.645316487470192e-06,
+      "loss": 0.3169,
+      "step": 17157
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 24.118864059448242,
+      "learning_rate": 5.644479772413504e-06,
+      "loss": 0.5048,
+      "step": 17158
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 18.195344924926758,
+      "learning_rate": 5.643643057356818e-06,
+      "loss": 1.7614,
+      "step": 17159
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 40.21952438354492,
+      "learning_rate": 5.64280634230013e-06,
+      "loss": 0.9294,
+      "step": 17160
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 16.129159927368164,
+      "learning_rate": 5.641969627243444e-06,
+      "loss": 0.6936,
+      "step": 17161
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 14.008243560791016,
+      "learning_rate": 5.6411329121867556e-06,
+      "loss": 1.0347,
+      "step": 17162
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 10.446682929992676,
+      "learning_rate": 5.6402961971300675e-06,
+      "loss": 0.3626,
+      "step": 17163
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 36.29135513305664,
+      "learning_rate": 5.63945948207338e-06,
+      "loss": 1.0444,
+      "step": 17164
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 15.276895523071289,
+      "learning_rate": 5.638622767016692e-06,
+      "loss": 1.23,
+      "step": 17165
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 17.33664321899414,
+      "learning_rate": 5.637786051960006e-06,
+      "loss": 0.7524,
+      "step": 17166
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 12.592365264892578,
+      "learning_rate": 5.636949336903318e-06,
+      "loss": 0.2947,
+      "step": 17167
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 21.139570236206055,
+      "learning_rate": 5.6361126218466315e-06,
+      "loss": 1.3471,
+      "step": 17168
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 14.71281623840332,
+      "learning_rate": 5.6352759067899435e-06,
+      "loss": 1.2693,
+      "step": 17169
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.269510269165039,
+      "learning_rate": 5.6344391917332554e-06,
+      "loss": 0.3714,
+      "step": 17170
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 18.213468551635742,
+      "learning_rate": 5.633602476676568e-06,
+      "loss": 0.9179,
+      "step": 17171
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 87.11209869384766,
+      "learning_rate": 5.63276576161988e-06,
+      "loss": 3.7237,
+      "step": 17172
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 11.030548095703125,
+      "learning_rate": 5.631929046563194e-06,
+      "loss": 1.569,
+      "step": 17173
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 17.495464324951172,
+      "learning_rate": 5.631092331506506e-06,
+      "loss": 0.5882,
+      "step": 17174
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 25.288028717041016,
+      "learning_rate": 5.630255616449819e-06,
+      "loss": 0.7862,
+      "step": 17175
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 17.80689239501953,
+      "learning_rate": 5.629418901393131e-06,
+      "loss": 1.5696,
+      "step": 17176
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 12.747581481933594,
+      "learning_rate": 5.628582186336443e-06,
+      "loss": 1.9478,
+      "step": 17177
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 18.356338500976562,
+      "learning_rate": 5.627745471279756e-06,
+      "loss": 2.3311,
+      "step": 17178
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 13.22213077545166,
+      "learning_rate": 5.626908756223068e-06,
+      "loss": 1.0209,
+      "step": 17179
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 27.774328231811523,
+      "learning_rate": 5.626072041166382e-06,
+      "loss": 0.961,
+      "step": 17180
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 41.390079498291016,
+      "learning_rate": 5.625235326109694e-06,
+      "loss": 1.0548,
+      "step": 17181
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 10.147668838500977,
+      "learning_rate": 5.6243986110530065e-06,
+      "loss": 0.8387,
+      "step": 17182
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 6.634181022644043,
+      "learning_rate": 5.623561895996319e-06,
+      "loss": 0.5521,
+      "step": 17183
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 13.955337524414062,
+      "learning_rate": 5.622725180939631e-06,
+      "loss": 1.1473,
+      "step": 17184
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 7.0592360496521,
+      "learning_rate": 5.621888465882944e-06,
+      "loss": 0.7489,
+      "step": 17185
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 16.544206619262695,
+      "learning_rate": 5.621051750826256e-06,
+      "loss": 2.2619,
+      "step": 17186
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 24.937488555908203,
+      "learning_rate": 5.62021503576957e-06,
+      "loss": 0.9906,
+      "step": 17187
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 51.04780960083008,
+      "learning_rate": 5.619378320712882e-06,
+      "loss": 1.937,
+      "step": 17188
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 13.549792289733887,
+      "learning_rate": 5.6185416056561944e-06,
+      "loss": 1.9272,
+      "step": 17189
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 13.317730903625488,
+      "learning_rate": 5.617704890599507e-06,
+      "loss": 0.5433,
+      "step": 17190
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.867985248565674,
+      "learning_rate": 5.616868175542819e-06,
+      "loss": 0.8244,
+      "step": 17191
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 17.57651710510254,
+      "learning_rate": 5.616031460486132e-06,
+      "loss": 2.1001,
+      "step": 17192
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 7.705465793609619,
+      "learning_rate": 5.615194745429444e-06,
+      "loss": 0.5644,
+      "step": 17193
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 142.6583251953125,
+      "learning_rate": 5.614358030372758e-06,
+      "loss": 1.5906,
+      "step": 17194
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 13.02302074432373,
+      "learning_rate": 5.6135213153160695e-06,
+      "loss": 0.8627,
+      "step": 17195
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 12.016786575317383,
+      "learning_rate": 5.612684600259382e-06,
+      "loss": 0.4129,
+      "step": 17196
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 168.90357971191406,
+      "learning_rate": 5.611847885202695e-06,
+      "loss": 1.415,
+      "step": 17197
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 13.086783409118652,
+      "learning_rate": 5.611011170146007e-06,
+      "loss": 0.6233,
+      "step": 17198
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 11.101323127746582,
+      "learning_rate": 5.61017445508932e-06,
+      "loss": 0.536,
+      "step": 17199
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 12.286422729492188,
+      "learning_rate": 5.609337740032632e-06,
+      "loss": 0.1976,
+      "step": 17200
+    },
+    {
+      "epoch": 2.16,
+      "eval_loss": 0.0822504535317421,
+      "eval_runtime": 94.4976,
+      "eval_samples_per_second": 37.482,
+      "eval_steps_per_second": 37.482,
+      "step": 17200
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.7932496070861816,
+      "learning_rate": 5.6085010249759455e-06,
+      "loss": 0.0928,
+      "step": 17201
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.5041267275810242,
+      "learning_rate": 5.6076643099192575e-06,
+      "loss": 0.0213,
+      "step": 17202
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 16.08133888244629,
+      "learning_rate": 5.60682759486257e-06,
+      "loss": 1.0277,
+      "step": 17203
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 21.73778533935547,
+      "learning_rate": 5.605990879805882e-06,
+      "loss": 0.7484,
+      "step": 17204
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.9961888790130615,
+      "learning_rate": 5.605154164749195e-06,
+      "loss": 1.2175,
+      "step": 17205
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.718421936035156,
+      "learning_rate": 5.604317449692508e-06,
+      "loss": 0.3244,
+      "step": 17206
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 12.62595272064209,
+      "learning_rate": 5.60348073463582e-06,
+      "loss": 0.6958,
+      "step": 17207
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 10.217425346374512,
+      "learning_rate": 5.6026440195791334e-06,
+      "loss": 1.3796,
+      "step": 17208
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 13.226073265075684,
+      "learning_rate": 5.601807304522445e-06,
+      "loss": 0.8957,
+      "step": 17209
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 30.363792419433594,
+      "learning_rate": 5.600970589465758e-06,
+      "loss": 2.2451,
+      "step": 17210
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 37.012794494628906,
+      "learning_rate": 5.60013387440907e-06,
+      "loss": 1.0801,
+      "step": 17211
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 60.06344223022461,
+      "learning_rate": 5.599297159352383e-06,
+      "loss": 1.6978,
+      "step": 17212
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 14.471539497375488,
+      "learning_rate": 5.598460444295696e-06,
+      "loss": 1.596,
+      "step": 17213
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 7.093859672546387,
+      "learning_rate": 5.597623729239008e-06,
+      "loss": 0.2557,
+      "step": 17214
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 8.392356872558594,
+      "learning_rate": 5.596787014182321e-06,
+      "loss": 0.5757,
+      "step": 17215
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 28.59241485595703,
+      "learning_rate": 5.595950299125633e-06,
+      "loss": 1.7856,
+      "step": 17216
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 10.944947242736816,
+      "learning_rate": 5.595113584068945e-06,
+      "loss": 0.5933,
+      "step": 17217
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 15.047503471374512,
+      "learning_rate": 5.594276869012258e-06,
+      "loss": 1.4373,
+      "step": 17218
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.249769687652588,
+      "learning_rate": 5.593440153955571e-06,
+      "loss": 0.2011,
+      "step": 17219
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.669090747833252,
+      "learning_rate": 5.592603438898884e-06,
+      "loss": 0.322,
+      "step": 17220
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 24.267330169677734,
+      "learning_rate": 5.591766723842196e-06,
+      "loss": 0.2359,
+      "step": 17221
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 10.642965316772461,
+      "learning_rate": 5.590930008785509e-06,
+      "loss": 0.4666,
+      "step": 17222
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 9.365653991699219,
+      "learning_rate": 5.590093293728821e-06,
+      "loss": 0.564,
+      "step": 17223
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 7.013499736785889,
+      "learning_rate": 5.589256578672133e-06,
+      "loss": 0.869,
+      "step": 17224
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 14.563632011413574,
+      "learning_rate": 5.588419863615446e-06,
+      "loss": 0.8464,
+      "step": 17225
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 9.125322341918945,
+      "learning_rate": 5.587583148558759e-06,
+      "loss": 0.4213,
+      "step": 17226
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.3125064373016357,
+      "learning_rate": 5.586746433502072e-06,
+      "loss": 0.207,
+      "step": 17227
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 27.964567184448242,
+      "learning_rate": 5.5859097184453835e-06,
+      "loss": 1.5827,
+      "step": 17228
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 19.723154067993164,
+      "learning_rate": 5.585073003388697e-06,
+      "loss": 0.8522,
+      "step": 17229
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 26.846416473388672,
+      "learning_rate": 5.584236288332009e-06,
+      "loss": 0.9918,
+      "step": 17230
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 11.809608459472656,
+      "learning_rate": 5.583399573275321e-06,
+      "loss": 0.2451,
+      "step": 17231
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 11.435138702392578,
+      "learning_rate": 5.582562858218634e-06,
+      "loss": 0.4537,
+      "step": 17232
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 31.543373107910156,
+      "learning_rate": 5.581726143161947e-06,
+      "loss": 0.5097,
+      "step": 17233
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 28.25628662109375,
+      "learning_rate": 5.5808894281052595e-06,
+      "loss": 1.1043,
+      "step": 17234
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 14.437735557556152,
+      "learning_rate": 5.5800527130485715e-06,
+      "loss": 0.6009,
+      "step": 17235
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 14.932991027832031,
+      "learning_rate": 5.579215997991885e-06,
+      "loss": 0.5173,
+      "step": 17236
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 11.054858207702637,
+      "learning_rate": 5.578379282935197e-06,
+      "loss": 1.342,
+      "step": 17237
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 103.68341064453125,
+      "learning_rate": 5.577542567878509e-06,
+      "loss": 2.8162,
+      "step": 17238
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 15.472814559936523,
+      "learning_rate": 5.576705852821822e-06,
+      "loss": 0.7346,
+      "step": 17239
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 13.45470905303955,
+      "learning_rate": 5.575869137765134e-06,
+      "loss": 0.7036,
+      "step": 17240
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 13.154472351074219,
+      "learning_rate": 5.575032422708447e-06,
+      "loss": 0.6328,
+      "step": 17241
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 8.994661331176758,
+      "learning_rate": 5.574195707651759e-06,
+      "loss": 1.4348,
+      "step": 17242
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 15.41472053527832,
+      "learning_rate": 5.573358992595073e-06,
+      "loss": 1.0158,
+      "step": 17243
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 9.832098007202148,
+      "learning_rate": 5.572522277538385e-06,
+      "loss": 1.2464,
+      "step": 17244
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.16098165512085,
+      "learning_rate": 5.571685562481697e-06,
+      "loss": 0.83,
+      "step": 17245
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 18.821264266967773,
+      "learning_rate": 5.57084884742501e-06,
+      "loss": 0.6399,
+      "step": 17246
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 27.851112365722656,
+      "learning_rate": 5.570012132368322e-06,
+      "loss": 1.2835,
+      "step": 17247
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 6.089595794677734,
+      "learning_rate": 5.569175417311635e-06,
+      "loss": 0.1637,
+      "step": 17248
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 16.148115158081055,
+      "learning_rate": 5.568338702254947e-06,
+      "loss": 1.0256,
+      "step": 17249
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 12.849231719970703,
+      "learning_rate": 5.567501987198261e-06,
+      "loss": 0.2856,
+      "step": 17250
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 12.162538528442383,
+      "learning_rate": 5.566665272141573e-06,
+      "loss": 1.8148,
+      "step": 17251
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 22.895511627197266,
+      "learning_rate": 5.565828557084885e-06,
+      "loss": 1.0507,
+      "step": 17252
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 96.24250030517578,
+      "learning_rate": 5.564991842028198e-06,
+      "loss": 1.7526,
+      "step": 17253
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 18.24812126159668,
+      "learning_rate": 5.56415512697151e-06,
+      "loss": 1.153,
+      "step": 17254
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 9.374454498291016,
+      "learning_rate": 5.563318411914823e-06,
+      "loss": 0.4764,
+      "step": 17255
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 17.744022369384766,
+      "learning_rate": 5.562481696858135e-06,
+      "loss": 1.8648,
+      "step": 17256
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 17.501155853271484,
+      "learning_rate": 5.561644981801448e-06,
+      "loss": 1.7762,
+      "step": 17257
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 36.74434280395508,
+      "learning_rate": 5.560808266744761e-06,
+      "loss": 1.3477,
+      "step": 17258
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 53.19886779785156,
+      "learning_rate": 5.559971551688073e-06,
+      "loss": 1.1854,
+      "step": 17259
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 19.726200103759766,
+      "learning_rate": 5.5591348366313856e-06,
+      "loss": 1.6473,
+      "step": 17260
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 10.460951805114746,
+      "learning_rate": 5.5582981215746975e-06,
+      "loss": 0.827,
+      "step": 17261
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 8.402612686157227,
+      "learning_rate": 5.557461406518011e-06,
+      "loss": 0.9577,
+      "step": 17262
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 51.408695220947266,
+      "learning_rate": 5.556624691461323e-06,
+      "loss": 2.2755,
+      "step": 17263
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 11.783794403076172,
+      "learning_rate": 5.555787976404636e-06,
+      "loss": 1.1531,
+      "step": 17264
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 19.813913345336914,
+      "learning_rate": 5.554951261347949e-06,
+      "loss": 1.2154,
+      "step": 17265
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 9.960100173950195,
+      "learning_rate": 5.554114546291261e-06,
+      "loss": 0.9847,
+      "step": 17266
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.53121817111969,
+      "learning_rate": 5.5532778312345735e-06,
+      "loss": 0.1649,
+      "step": 17267
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 16.12785530090332,
+      "learning_rate": 5.5524411161778854e-06,
+      "loss": 0.6851,
+      "step": 17268
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 6.951560020446777,
+      "learning_rate": 5.551604401121199e-06,
+      "loss": 0.2635,
+      "step": 17269
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 48.68133544921875,
+      "learning_rate": 5.550767686064511e-06,
+      "loss": 0.7961,
+      "step": 17270
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 13.477615356445312,
+      "learning_rate": 5.549930971007824e-06,
+      "loss": 0.4971,
+      "step": 17271
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 24.526586532592773,
+      "learning_rate": 5.549094255951137e-06,
+      "loss": 1.4679,
+      "step": 17272
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 29.198915481567383,
+      "learning_rate": 5.548257540894449e-06,
+      "loss": 1.934,
+      "step": 17273
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 20.456193923950195,
+      "learning_rate": 5.547420825837761e-06,
+      "loss": 1.0854,
+      "step": 17274
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 12.702293395996094,
+      "learning_rate": 5.546584110781073e-06,
+      "loss": 1.004,
+      "step": 17275
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 13.598925590515137,
+      "learning_rate": 5.545747395724387e-06,
+      "loss": 0.5497,
+      "step": 17276
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 14.103580474853516,
+      "learning_rate": 5.544910680667699e-06,
+      "loss": 1.1607,
+      "step": 17277
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 16.528247833251953,
+      "learning_rate": 5.544073965611012e-06,
+      "loss": 0.8895,
+      "step": 17278
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 14.104270935058594,
+      "learning_rate": 5.5432372505543246e-06,
+      "loss": 0.6044,
+      "step": 17279
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 17.39735984802246,
+      "learning_rate": 5.5424005354976365e-06,
+      "loss": 0.4925,
+      "step": 17280
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 40.398956298828125,
+      "learning_rate": 5.541563820440949e-06,
+      "loss": 1.3486,
+      "step": 17281
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 10.429101943969727,
+      "learning_rate": 5.540727105384261e-06,
+      "loss": 1.1617,
+      "step": 17282
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 50.06691360473633,
+      "learning_rate": 5.539890390327575e-06,
+      "loss": 4.6265,
+      "step": 17283
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 8.183061599731445,
+      "learning_rate": 5.539053675270887e-06,
+      "loss": 0.5522,
+      "step": 17284
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 25.18187141418457,
+      "learning_rate": 5.5382169602142e-06,
+      "loss": 1.5083,
+      "step": 17285
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 15.119938850402832,
+      "learning_rate": 5.537380245157512e-06,
+      "loss": 1.7149,
+      "step": 17286
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 17.999061584472656,
+      "learning_rate": 5.5365435301008244e-06,
+      "loss": 1.1504,
+      "step": 17287
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 12.407395362854004,
+      "learning_rate": 5.535706815044137e-06,
+      "loss": 0.9227,
+      "step": 17288
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.739668369293213,
+      "learning_rate": 5.534870099987449e-06,
+      "loss": 1.2882,
+      "step": 17289
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 14.490335464477539,
+      "learning_rate": 5.534033384930763e-06,
+      "loss": 0.5932,
+      "step": 17290
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 13.677287101745605,
+      "learning_rate": 5.533196669874075e-06,
+      "loss": 0.4711,
+      "step": 17291
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 229.60195922851562,
+      "learning_rate": 5.532359954817388e-06,
+      "loss": 2.2706,
+      "step": 17292
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 64.56965637207031,
+      "learning_rate": 5.5315232397606996e-06,
+      "loss": 1.1746,
+      "step": 17293
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 11.953279495239258,
+      "learning_rate": 5.530686524704012e-06,
+      "loss": 0.78,
+      "step": 17294
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 14.518604278564453,
+      "learning_rate": 5.529849809647325e-06,
+      "loss": 0.371,
+      "step": 17295
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 12.54118537902832,
+      "learning_rate": 5.529013094590637e-06,
+      "loss": 1.3565,
+      "step": 17296
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 6.692623615264893,
+      "learning_rate": 5.528176379533951e-06,
+      "loss": 0.305,
+      "step": 17297
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 9.086627960205078,
+      "learning_rate": 5.527339664477263e-06,
+      "loss": 0.0801,
+      "step": 17298
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 23.496551513671875,
+      "learning_rate": 5.5265029494205755e-06,
+      "loss": 0.6855,
+      "step": 17299
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 20.957077026367188,
+      "learning_rate": 5.5256662343638875e-06,
+      "loss": 0.8958,
+      "step": 17300
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 42.834800720214844,
+      "learning_rate": 5.5248295193072e-06,
+      "loss": 2.9654,
+      "step": 17301
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 19.34640884399414,
+      "learning_rate": 5.523992804250513e-06,
+      "loss": 0.8642,
+      "step": 17302
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 8.827037811279297,
+      "learning_rate": 5.523156089193825e-06,
+      "loss": 0.4299,
+      "step": 17303
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 10.81065559387207,
+      "learning_rate": 5.522319374137139e-06,
+      "loss": 0.3272,
+      "step": 17304
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 11.338500022888184,
+      "learning_rate": 5.521482659080451e-06,
+      "loss": 0.2851,
+      "step": 17305
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 10.40788745880127,
+      "learning_rate": 5.5206459440237634e-06,
+      "loss": 0.6665,
+      "step": 17306
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 128.7415008544922,
+      "learning_rate": 5.519809228967075e-06,
+      "loss": 0.7273,
+      "step": 17307
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.43458890914917,
+      "learning_rate": 5.518972513910388e-06,
+      "loss": 0.0861,
+      "step": 17308
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 17.122840881347656,
+      "learning_rate": 5.518135798853701e-06,
+      "loss": 1.4625,
+      "step": 17309
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 12.97182846069336,
+      "learning_rate": 5.517299083797013e-06,
+      "loss": 0.6766,
+      "step": 17310
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 9.539743423461914,
+      "learning_rate": 5.516462368740327e-06,
+      "loss": 1.0235,
+      "step": 17311
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 13.798757553100586,
+      "learning_rate": 5.5156256536836386e-06,
+      "loss": 0.4069,
+      "step": 17312
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 9.476639747619629,
+      "learning_rate": 5.514788938626951e-06,
+      "loss": 1.3685,
+      "step": 17313
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.449814796447754,
+      "learning_rate": 5.513952223570263e-06,
+      "loss": 0.1509,
+      "step": 17314
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 11.458866119384766,
+      "learning_rate": 5.513115508513575e-06,
+      "loss": 1.1127,
+      "step": 17315
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 86.14989471435547,
+      "learning_rate": 5.512278793456889e-06,
+      "loss": 0.8236,
+      "step": 17316
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 74.04230499267578,
+      "learning_rate": 5.511442078400201e-06,
+      "loss": 1.3761,
+      "step": 17317
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 15.484674453735352,
+      "learning_rate": 5.5106053633435145e-06,
+      "loss": 0.9134,
+      "step": 17318
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 6.356900215148926,
+      "learning_rate": 5.5097686482868265e-06,
+      "loss": 0.3291,
+      "step": 17319
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 6.78397274017334,
+      "learning_rate": 5.508931933230139e-06,
+      "loss": 0.1679,
+      "step": 17320
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 33.1566047668457,
+      "learning_rate": 5.508095218173451e-06,
+      "loss": 0.9174,
+      "step": 17321
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 11.877245903015137,
+      "learning_rate": 5.507258503116763e-06,
+      "loss": 0.6691,
+      "step": 17322
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 321.0672607421875,
+      "learning_rate": 5.506421788060077e-06,
+      "loss": 2.0765,
+      "step": 17323
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 7.010623455047607,
+      "learning_rate": 5.505585073003389e-06,
+      "loss": 0.6944,
+      "step": 17324
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 31.49948501586914,
+      "learning_rate": 5.5047483579467024e-06,
+      "loss": 0.5722,
+      "step": 17325
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 9.569120407104492,
+      "learning_rate": 5.503911642890014e-06,
+      "loss": 0.3115,
+      "step": 17326
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 17.101133346557617,
+      "learning_rate": 5.503074927833327e-06,
+      "loss": 0.7327,
+      "step": 17327
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 34.0203971862793,
+      "learning_rate": 5.502238212776639e-06,
+      "loss": 2.2211,
+      "step": 17328
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 61.76817321777344,
+      "learning_rate": 5.501401497719951e-06,
+      "loss": 1.3252,
+      "step": 17329
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.967968702316284,
+      "learning_rate": 5.500564782663265e-06,
+      "loss": 0.3913,
+      "step": 17330
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 8.685111045837402,
+      "learning_rate": 5.499728067606577e-06,
+      "loss": 0.3842,
+      "step": 17331
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 9.978775978088379,
+      "learning_rate": 5.4988913525498895e-06,
+      "loss": 1.1715,
+      "step": 17332
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 8.293410301208496,
+      "learning_rate": 5.498054637493202e-06,
+      "loss": 0.1429,
+      "step": 17333
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 27.3328800201416,
+      "learning_rate": 5.497217922436515e-06,
+      "loss": 0.9815,
+      "step": 17334
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 13.956252098083496,
+      "learning_rate": 5.496381207379827e-06,
+      "loss": 0.8331,
+      "step": 17335
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 52.58840560913086,
+      "learning_rate": 5.495544492323139e-06,
+      "loss": 1.2963,
+      "step": 17336
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 11.63193416595459,
+      "learning_rate": 5.494707777266453e-06,
+      "loss": 0.4902,
+      "step": 17337
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 12.473057746887207,
+      "learning_rate": 5.493871062209765e-06,
+      "loss": 1.2166,
+      "step": 17338
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 14.674174308776855,
+      "learning_rate": 5.4930343471530774e-06,
+      "loss": 1.8472,
+      "step": 17339
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.343213081359863,
+      "learning_rate": 5.49219763209639e-06,
+      "loss": 0.3098,
+      "step": 17340
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 16.50578498840332,
+      "learning_rate": 5.491360917039703e-06,
+      "loss": 0.3704,
+      "step": 17341
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 11.50082015991211,
+      "learning_rate": 5.490524201983015e-06,
+      "loss": 0.9004,
+      "step": 17342
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 9.233156204223633,
+      "learning_rate": 5.489687486926327e-06,
+      "loss": 0.4855,
+      "step": 17343
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 17.008365631103516,
+      "learning_rate": 5.488850771869641e-06,
+      "loss": 0.5787,
+      "step": 17344
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 16.309478759765625,
+      "learning_rate": 5.4880140568129525e-06,
+      "loss": 0.9927,
+      "step": 17345
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 9.710187911987305,
+      "learning_rate": 5.487177341756265e-06,
+      "loss": 1.9782,
+      "step": 17346
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 6.9360270500183105,
+      "learning_rate": 5.486340626699578e-06,
+      "loss": 0.4631,
+      "step": 17347
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 16.3384952545166,
+      "learning_rate": 5.485503911642891e-06,
+      "loss": 1.7151,
+      "step": 17348
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 8.947735786437988,
+      "learning_rate": 5.484667196586203e-06,
+      "loss": 1.7752,
+      "step": 17349
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 6.5993499755859375,
+      "learning_rate": 5.483830481529515e-06,
+      "loss": 0.6455,
+      "step": 17350
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.6965012550354004,
+      "learning_rate": 5.4829937664728285e-06,
+      "loss": 0.8186,
+      "step": 17351
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 8.805519104003906,
+      "learning_rate": 5.4821570514161405e-06,
+      "loss": 0.3716,
+      "step": 17352
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 11.217713356018066,
+      "learning_rate": 5.481320336359453e-06,
+      "loss": 1.1276,
+      "step": 17353
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 37.654449462890625,
+      "learning_rate": 5.480483621302766e-06,
+      "loss": 1.6106,
+      "step": 17354
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 55.929996490478516,
+      "learning_rate": 5.479646906246079e-06,
+      "loss": 2.1045,
+      "step": 17355
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.489141464233398,
+      "learning_rate": 5.478810191189391e-06,
+      "loss": 0.2208,
+      "step": 17356
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 12.1909761428833,
+      "learning_rate": 5.477973476132703e-06,
+      "loss": 1.5717,
+      "step": 17357
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 50.50084686279297,
+      "learning_rate": 5.4771367610760164e-06,
+      "loss": 2.6516,
+      "step": 17358
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 13.93696117401123,
+      "learning_rate": 5.476300046019328e-06,
+      "loss": 1.3454,
+      "step": 17359
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 115.9477767944336,
+      "learning_rate": 5.475463330962641e-06,
+      "loss": 1.1183,
+      "step": 17360
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 13.79046630859375,
+      "learning_rate": 5.474626615905954e-06,
+      "loss": 0.9186,
+      "step": 17361
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 9.239429473876953,
+      "learning_rate": 5.473789900849267e-06,
+      "loss": 1.5267,
+      "step": 17362
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 17.92961311340332,
+      "learning_rate": 5.472953185792579e-06,
+      "loss": 0.2789,
+      "step": 17363
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 23.084224700927734,
+      "learning_rate": 5.472116470735891e-06,
+      "loss": 2.3164,
+      "step": 17364
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 12.765448570251465,
+      "learning_rate": 5.471279755679204e-06,
+      "loss": 0.9762,
+      "step": 17365
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 23.09999656677246,
+      "learning_rate": 5.470443040622516e-06,
+      "loss": 1.0109,
+      "step": 17366
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 51.35440444946289,
+      "learning_rate": 5.469606325565829e-06,
+      "loss": 1.6294,
+      "step": 17367
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 15.2879056930542,
+      "learning_rate": 5.468769610509141e-06,
+      "loss": 1.4256,
+      "step": 17368
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 43.24333190917969,
+      "learning_rate": 5.467932895452455e-06,
+      "loss": 1.2839,
+      "step": 17369
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.293132781982422,
+      "learning_rate": 5.467096180395767e-06,
+      "loss": 0.2765,
+      "step": 17370
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 14.49522590637207,
+      "learning_rate": 5.466259465339079e-06,
+      "loss": 1.0694,
+      "step": 17371
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 11.335780143737793,
+      "learning_rate": 5.465422750282392e-06,
+      "loss": 1.1652,
+      "step": 17372
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.1716794967651367,
+      "learning_rate": 5.464586035225704e-06,
+      "loss": 0.0202,
+      "step": 17373
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.579696178436279,
+      "learning_rate": 5.463749320169017e-06,
+      "loss": 0.9418,
+      "step": 17374
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 21.71388816833496,
+      "learning_rate": 5.462912605112329e-06,
+      "loss": 0.6225,
+      "step": 17375
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 13.785713195800781,
+      "learning_rate": 5.462075890055643e-06,
+      "loss": 0.7838,
+      "step": 17376
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 10.466874122619629,
+      "learning_rate": 5.461239174998955e-06,
+      "loss": 0.9145,
+      "step": 17377
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 21.933841705322266,
+      "learning_rate": 5.4604024599422665e-06,
+      "loss": 0.9588,
+      "step": 17378
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 17.930667877197266,
+      "learning_rate": 5.45956574488558e-06,
+      "loss": 1.2568,
+      "step": 17379
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 15.766815185546875,
+      "learning_rate": 5.458729029828892e-06,
+      "loss": 0.9218,
+      "step": 17380
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 20.991121292114258,
+      "learning_rate": 5.457892314772205e-06,
+      "loss": 1.9213,
+      "step": 17381
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 71.60786437988281,
+      "learning_rate": 5.457055599715517e-06,
+      "loss": 1.5991,
+      "step": 17382
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.674159526824951,
+      "learning_rate": 5.4562188846588305e-06,
+      "loss": 0.277,
+      "step": 17383
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.543520450592041,
+      "learning_rate": 5.4553821696021425e-06,
+      "loss": 0.0572,
+      "step": 17384
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 10.390192031860352,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 1.2012,
+      "step": 17385
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 20.592607498168945,
+      "learning_rate": 5.453708739488768e-06,
+      "loss": 1.5867,
+      "step": 17386
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 9.117387771606445,
+      "learning_rate": 5.45287202443208e-06,
+      "loss": 0.5777,
+      "step": 17387
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 14.631564140319824,
+      "learning_rate": 5.452035309375393e-06,
+      "loss": 1.2054,
+      "step": 17388
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 9.758777618408203,
+      "learning_rate": 5.451198594318705e-06,
+      "loss": 0.2893,
+      "step": 17389
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 17.93175506591797,
+      "learning_rate": 5.4503618792620185e-06,
+      "loss": 0.7353,
+      "step": 17390
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 18.551061630249023,
+      "learning_rate": 5.44952516420533e-06,
+      "loss": 0.6877,
+      "step": 17391
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 24.413114547729492,
+      "learning_rate": 5.448688449148642e-06,
+      "loss": 0.6687,
+      "step": 17392
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.8335249423980713,
+      "learning_rate": 5.447851734091956e-06,
+      "loss": 0.2867,
+      "step": 17393
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 7.5022807121276855,
+      "learning_rate": 5.447015019035268e-06,
+      "loss": 0.4303,
+      "step": 17394
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 10.752012252807617,
+      "learning_rate": 5.446178303978581e-06,
+      "loss": 1.0519,
+      "step": 17395
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 29.8737850189209,
+      "learning_rate": 5.445341588921893e-06,
+      "loss": 1.8048,
+      "step": 17396
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 31.07291603088379,
+      "learning_rate": 5.444504873865205e-06,
+      "loss": 1.9485,
+      "step": 17397
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 122.77910614013672,
+      "learning_rate": 5.443668158808518e-06,
+      "loss": 1.9164,
+      "step": 17398
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 17.61583709716797,
+      "learning_rate": 5.44283144375183e-06,
+      "loss": 1.5101,
+      "step": 17399
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 19.916318893432617,
+      "learning_rate": 5.441994728695144e-06,
+      "loss": 1.4863,
+      "step": 17400
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 14.747823715209961,
+      "learning_rate": 5.441158013638456e-06,
+      "loss": 0.4247,
+      "step": 17401
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 30.967308044433594,
+      "learning_rate": 5.440321298581769e-06,
+      "loss": 3.5009,
+      "step": 17402
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 6.669802188873291,
+      "learning_rate": 5.439484583525081e-06,
+      "loss": 1.0505,
+      "step": 17403
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 14.642730712890625,
+      "learning_rate": 5.438647868468393e-06,
+      "loss": 0.7301,
+      "step": 17404
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 10.343039512634277,
+      "learning_rate": 5.437811153411706e-06,
+      "loss": 0.3258,
+      "step": 17405
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 21.614656448364258,
+      "learning_rate": 5.436974438355018e-06,
+      "loss": 1.0137,
+      "step": 17406
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 9.189238548278809,
+      "learning_rate": 5.436137723298332e-06,
+      "loss": 2.0193,
+      "step": 17407
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 6.540914535522461,
+      "learning_rate": 5.435301008241644e-06,
+      "loss": 0.9241,
+      "step": 17408
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 15.18000316619873,
+      "learning_rate": 5.434464293184957e-06,
+      "loss": 0.7895,
+      "step": 17409
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 26.84891128540039,
+      "learning_rate": 5.4336275781282686e-06,
+      "loss": 1.0056,
+      "step": 17410
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 6.153435707092285,
+      "learning_rate": 5.4327908630715805e-06,
+      "loss": 1.6114,
+      "step": 17411
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 11.218341827392578,
+      "learning_rate": 5.431954148014894e-06,
+      "loss": 1.0399,
+      "step": 17412
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 13.591835975646973,
+      "learning_rate": 5.431117432958206e-06,
+      "loss": 0.7951,
+      "step": 17413
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 15.405440330505371,
+      "learning_rate": 5.430280717901519e-06,
+      "loss": 0.9816,
+      "step": 17414
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 20.23955726623535,
+      "learning_rate": 5.429444002844832e-06,
+      "loss": 0.9346,
+      "step": 17415
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 9.765039443969727,
+      "learning_rate": 5.4286072877881445e-06,
+      "loss": 1.1874,
+      "step": 17416
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 39.12038803100586,
+      "learning_rate": 5.4277705727314565e-06,
+      "loss": 1.787,
+      "step": 17417
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 7.175165176391602,
+      "learning_rate": 5.4269338576747684e-06,
+      "loss": 0.451,
+      "step": 17418
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 12.465749740600586,
+      "learning_rate": 5.426097142618082e-06,
+      "loss": 0.9885,
+      "step": 17419
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 6.003023624420166,
+      "learning_rate": 5.425260427561394e-06,
+      "loss": 1.206,
+      "step": 17420
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 29.761905670166016,
+      "learning_rate": 5.424423712504707e-06,
+      "loss": 1.4167,
+      "step": 17421
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 13.826894760131836,
+      "learning_rate": 5.42358699744802e-06,
+      "loss": 0.4789,
+      "step": 17422
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.056317329406738,
+      "learning_rate": 5.4227502823913325e-06,
+      "loss": 0.2607,
+      "step": 17423
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 76.80038452148438,
+      "learning_rate": 5.421913567334644e-06,
+      "loss": 2.3984,
+      "step": 17424
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 6.621973037719727,
+      "learning_rate": 5.421076852277956e-06,
+      "loss": 0.7779,
+      "step": 17425
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 12.780646324157715,
+      "learning_rate": 5.42024013722127e-06,
+      "loss": 0.6478,
+      "step": 17426
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 25.02349281311035,
+      "learning_rate": 5.419403422164582e-06,
+      "loss": 1.3122,
+      "step": 17427
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 21.12056541442871,
+      "learning_rate": 5.418566707107895e-06,
+      "loss": 0.5195,
+      "step": 17428
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 14.527588844299316,
+      "learning_rate": 5.4177299920512076e-06,
+      "loss": 0.8056,
+      "step": 17429
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 28.591569900512695,
+      "learning_rate": 5.41689327699452e-06,
+      "loss": 1.4283,
+      "step": 17430
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 11.90042495727539,
+      "learning_rate": 5.416056561937832e-06,
+      "loss": 0.9807,
+      "step": 17431
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 26.2853946685791,
+      "learning_rate": 5.415219846881144e-06,
+      "loss": 0.7286,
+      "step": 17432
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 8.017561912536621,
+      "learning_rate": 5.414383131824458e-06,
+      "loss": 0.5506,
+      "step": 17433
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 13.73519515991211,
+      "learning_rate": 5.41354641676777e-06,
+      "loss": 0.7781,
+      "step": 17434
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 13.892803192138672,
+      "learning_rate": 5.412709701711083e-06,
+      "loss": 0.9187,
+      "step": 17435
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 10.17711353302002,
+      "learning_rate": 5.4118729866543955e-06,
+      "loss": 1.3602,
+      "step": 17436
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 22.67192268371582,
+      "learning_rate": 5.411036271597708e-06,
+      "loss": 1.5603,
+      "step": 17437
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 23.222631454467773,
+      "learning_rate": 5.41019955654102e-06,
+      "loss": 0.5541,
+      "step": 17438
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 17.52680015563965,
+      "learning_rate": 5.409362841484332e-06,
+      "loss": 0.9239,
+      "step": 17439
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 9.070002555847168,
+      "learning_rate": 5.408526126427646e-06,
+      "loss": 0.6678,
+      "step": 17440
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 13.469550132751465,
+      "learning_rate": 5.407689411370958e-06,
+      "loss": 0.7127,
+      "step": 17441
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 26.821613311767578,
+      "learning_rate": 5.406852696314271e-06,
+      "loss": 0.5464,
+      "step": 17442
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 20.423538208007812,
+      "learning_rate": 5.4060159812575826e-06,
+      "loss": 0.87,
+      "step": 17443
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.490708827972412,
+      "learning_rate": 5.405179266200896e-06,
+      "loss": 0.4314,
+      "step": 17444
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 26.50490379333496,
+      "learning_rate": 5.404342551144208e-06,
+      "loss": 2.3217,
+      "step": 17445
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 33.80739974975586,
+      "learning_rate": 5.40350583608752e-06,
+      "loss": 3.0031,
+      "step": 17446
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 24.16927146911621,
+      "learning_rate": 5.402669121030834e-06,
+      "loss": 0.7531,
+      "step": 17447
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 16.82111930847168,
+      "learning_rate": 5.401832405974146e-06,
+      "loss": 1.4441,
+      "step": 17448
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 7.376926898956299,
+      "learning_rate": 5.4009956909174585e-06,
+      "loss": 1.3377,
+      "step": 17449
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 32.68447494506836,
+      "learning_rate": 5.4001589758607705e-06,
+      "loss": 1.5825,
+      "step": 17450
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 10.641036987304688,
+      "learning_rate": 5.399322260804084e-06,
+      "loss": 0.5053,
+      "step": 17451
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 9.782011032104492,
+      "learning_rate": 5.398485545747396e-06,
+      "loss": 0.1486,
+      "step": 17452
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 24.293102264404297,
+      "learning_rate": 5.397648830690708e-06,
+      "loss": 0.4749,
+      "step": 17453
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 33.50642776489258,
+      "learning_rate": 5.396812115634022e-06,
+      "loss": 0.8247,
+      "step": 17454
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 29.112720489501953,
+      "learning_rate": 5.395975400577334e-06,
+      "loss": 0.6069,
+      "step": 17455
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 32.35807800292969,
+      "learning_rate": 5.3951386855206464e-06,
+      "loss": 1.8802,
+      "step": 17456
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 6.860402584075928,
+      "learning_rate": 5.394301970463958e-06,
+      "loss": 0.2353,
+      "step": 17457
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 11.936412811279297,
+      "learning_rate": 5.393465255407272e-06,
+      "loss": 0.1448,
+      "step": 17458
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 12.567363739013672,
+      "learning_rate": 5.392628540350584e-06,
+      "loss": 0.9206,
+      "step": 17459
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 7.77496337890625,
+      "learning_rate": 5.391791825293896e-06,
+      "loss": 0.8399,
+      "step": 17460
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 19.330772399902344,
+      "learning_rate": 5.39095511023721e-06,
+      "loss": 0.9187,
+      "step": 17461
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 6.970808982849121,
+      "learning_rate": 5.3901183951805216e-06,
+      "loss": 0.8056,
+      "step": 17462
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 14.46152400970459,
+      "learning_rate": 5.389281680123834e-06,
+      "loss": 0.4848,
+      "step": 17463
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 10.946375846862793,
+      "learning_rate": 5.388444965067146e-06,
+      "loss": 1.0824,
+      "step": 17464
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 10.016998291015625,
+      "learning_rate": 5.38760825001046e-06,
+      "loss": 0.3515,
+      "step": 17465
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 7.684912204742432,
+      "learning_rate": 5.386771534953772e-06,
+      "loss": 0.689,
+      "step": 17466
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 13.379766464233398,
+      "learning_rate": 5.385934819897084e-06,
+      "loss": 1.5413,
+      "step": 17467
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 7.916872024536133,
+      "learning_rate": 5.3850981048403975e-06,
+      "loss": 1.6151,
+      "step": 17468
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 6.842104911804199,
+      "learning_rate": 5.3842613897837095e-06,
+      "loss": 0.2652,
+      "step": 17469
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 14.729266166687012,
+      "learning_rate": 5.383424674727022e-06,
+      "loss": 1.7612,
+      "step": 17470
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 14.622649192810059,
+      "learning_rate": 5.382587959670334e-06,
+      "loss": 1.2312,
+      "step": 17471
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 14.882375717163086,
+      "learning_rate": 5.381751244613648e-06,
+      "loss": 1.3312,
+      "step": 17472
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 11.30505084991455,
+      "learning_rate": 5.38091452955696e-06,
+      "loss": 0.4109,
+      "step": 17473
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 14.341401100158691,
+      "learning_rate": 5.380077814500272e-06,
+      "loss": 0.5819,
+      "step": 17474
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 71.64712524414062,
+      "learning_rate": 5.3792410994435854e-06,
+      "loss": 1.3767,
+      "step": 17475
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 11.888653755187988,
+      "learning_rate": 5.378404384386897e-06,
+      "loss": 0.8234,
+      "step": 17476
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 12.765649795532227,
+      "learning_rate": 5.37756766933021e-06,
+      "loss": 0.7687,
+      "step": 17477
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 18.29014778137207,
+      "learning_rate": 5.376730954273522e-06,
+      "loss": 1.1736,
+      "step": 17478
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 11.075541496276855,
+      "learning_rate": 5.375894239216836e-06,
+      "loss": 0.4156,
+      "step": 17479
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 11.490434646606445,
+      "learning_rate": 5.375057524160148e-06,
+      "loss": 0.8648,
+      "step": 17480
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 24.799287796020508,
+      "learning_rate": 5.37422080910346e-06,
+      "loss": 0.7148,
+      "step": 17481
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 122.64232635498047,
+      "learning_rate": 5.373384094046773e-06,
+      "loss": 1.5835,
+      "step": 17482
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 27.34222412109375,
+      "learning_rate": 5.372547378990085e-06,
+      "loss": 1.2789,
+      "step": 17483
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 23.286502838134766,
+      "learning_rate": 5.371710663933398e-06,
+      "loss": 2.4069,
+      "step": 17484
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 7.177731037139893,
+      "learning_rate": 5.37087394887671e-06,
+      "loss": 0.8277,
+      "step": 17485
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 19.814857482910156,
+      "learning_rate": 5.370037233820024e-06,
+      "loss": 0.2308,
+      "step": 17486
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 34.60707473754883,
+      "learning_rate": 5.369200518763336e-06,
+      "loss": 3.0668,
+      "step": 17487
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 11.37707805633545,
+      "learning_rate": 5.368363803706648e-06,
+      "loss": 0.7361,
+      "step": 17488
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.2040646076202393,
+      "learning_rate": 5.367527088649961e-06,
+      "loss": 0.2547,
+      "step": 17489
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 11.337788581848145,
+      "learning_rate": 5.366690373593273e-06,
+      "loss": 1.5259,
+      "step": 17490
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 15.41157341003418,
+      "learning_rate": 5.365853658536586e-06,
+      "loss": 1.224,
+      "step": 17491
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 39.75604248046875,
+      "learning_rate": 5.365016943479898e-06,
+      "loss": 0.8128,
+      "step": 17492
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 6.115261554718018,
+      "learning_rate": 5.364180228423212e-06,
+      "loss": 0.162,
+      "step": 17493
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 12.996819496154785,
+      "learning_rate": 5.363343513366524e-06,
+      "loss": 0.4829,
+      "step": 17494
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 8.252242088317871,
+      "learning_rate": 5.3625067983098355e-06,
+      "loss": 0.4253,
+      "step": 17495
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 7.422755718231201,
+      "learning_rate": 5.361670083253148e-06,
+      "loss": 0.3621,
+      "step": 17496
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 10.536725044250488,
+      "learning_rate": 5.360833368196461e-06,
+      "loss": 1.104,
+      "step": 17497
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 65.46198272705078,
+      "learning_rate": 5.359996653139774e-06,
+      "loss": 1.338,
+      "step": 17498
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.54756498336792,
+      "learning_rate": 5.359159938083086e-06,
+      "loss": 0.0467,
+      "step": 17499
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 6.500692367553711,
+      "learning_rate": 5.3583232230263996e-06,
+      "loss": 0.3971,
+      "step": 17500
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 34.247474670410156,
+      "learning_rate": 5.3574865079697115e-06,
+      "loss": 1.277,
+      "step": 17501
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 18.990402221679688,
+      "learning_rate": 5.3566497929130235e-06,
+      "loss": 2.9791,
+      "step": 17502
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 14.820756912231445,
+      "learning_rate": 5.355813077856336e-06,
+      "loss": 2.2124,
+      "step": 17503
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 22.829708099365234,
+      "learning_rate": 5.354976362799649e-06,
+      "loss": 1.2131,
+      "step": 17504
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 19.45446014404297,
+      "learning_rate": 5.354139647742962e-06,
+      "loss": 1.1271,
+      "step": 17505
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 9.499173164367676,
+      "learning_rate": 5.353302932686274e-06,
+      "loss": 1.1174,
+      "step": 17506
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 15.749574661254883,
+      "learning_rate": 5.3524662176295875e-06,
+      "loss": 2.0691,
+      "step": 17507
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 10.936659812927246,
+      "learning_rate": 5.3516295025728994e-06,
+      "loss": 0.8516,
+      "step": 17508
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.4799952507019043,
+      "learning_rate": 5.350792787516211e-06,
+      "loss": 0.0853,
+      "step": 17509
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 17.480205535888672,
+      "learning_rate": 5.349956072459524e-06,
+      "loss": 1.037,
+      "step": 17510
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 23.715269088745117,
+      "learning_rate": 5.349119357402837e-06,
+      "loss": 2.084,
+      "step": 17511
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 39.498802185058594,
+      "learning_rate": 5.34828264234615e-06,
+      "loss": 1.6885,
+      "step": 17512
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 8.803202629089355,
+      "learning_rate": 5.347445927289462e-06,
+      "loss": 0.3146,
+      "step": 17513
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 8.69205379486084,
+      "learning_rate": 5.346609212232775e-06,
+      "loss": 0.3413,
+      "step": 17514
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 6.984821319580078,
+      "learning_rate": 5.345772497176087e-06,
+      "loss": 0.4525,
+      "step": 17515
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 66.03707122802734,
+      "learning_rate": 5.344935782119399e-06,
+      "loss": 0.8255,
+      "step": 17516
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 14.11879825592041,
+      "learning_rate": 5.344099067062712e-06,
+      "loss": 0.8665,
+      "step": 17517
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 33.14910888671875,
+      "learning_rate": 5.343262352006025e-06,
+      "loss": 2.288,
+      "step": 17518
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 14.559138298034668,
+      "learning_rate": 5.342425636949338e-06,
+      "loss": 3.0581,
+      "step": 17519
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 15.356159210205078,
+      "learning_rate": 5.34158892189265e-06,
+      "loss": 1.2524,
+      "step": 17520
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 37.47621154785156,
+      "learning_rate": 5.340752206835963e-06,
+      "loss": 0.5242,
+      "step": 17521
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 87.72489929199219,
+      "learning_rate": 5.339915491779275e-06,
+      "loss": 1.4852,
+      "step": 17522
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 36.96324920654297,
+      "learning_rate": 5.339078776722587e-06,
+      "loss": 0.9385,
+      "step": 17523
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 7.783929347991943,
+      "learning_rate": 5.3382420616659e-06,
+      "loss": 0.7675,
+      "step": 17524
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 15.189329147338867,
+      "learning_rate": 5.337405346609212e-06,
+      "loss": 1.3922,
+      "step": 17525
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 9.003801345825195,
+      "learning_rate": 5.336568631552526e-06,
+      "loss": 1.2127,
+      "step": 17526
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 13.577678680419922,
+      "learning_rate": 5.335731916495838e-06,
+      "loss": 1.5522,
+      "step": 17527
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 45.36927032470703,
+      "learning_rate": 5.334895201439151e-06,
+      "loss": 2.2272,
+      "step": 17528
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 18.816207885742188,
+      "learning_rate": 5.334058486382463e-06,
+      "loss": 0.8083,
+      "step": 17529
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 9.518589973449707,
+      "learning_rate": 5.333221771325775e-06,
+      "loss": 0.3954,
+      "step": 17530
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 143.4003143310547,
+      "learning_rate": 5.332385056269088e-06,
+      "loss": 2.8706,
+      "step": 17531
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 7.963112831115723,
+      "learning_rate": 5.3315483412124e-06,
+      "loss": 1.2719,
+      "step": 17532
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 20.16355323791504,
+      "learning_rate": 5.3307116261557135e-06,
+      "loss": 2.0285,
+      "step": 17533
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 10.042006492614746,
+      "learning_rate": 5.3298749110990255e-06,
+      "loss": 1.1726,
+      "step": 17534
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.8463993072509766,
+      "learning_rate": 5.329038196042339e-06,
+      "loss": 0.4411,
+      "step": 17535
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 11.920524597167969,
+      "learning_rate": 5.328201480985651e-06,
+      "loss": 0.5937,
+      "step": 17536
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 27.405885696411133,
+      "learning_rate": 5.327364765928963e-06,
+      "loss": 0.2532,
+      "step": 17537
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 10.97453498840332,
+      "learning_rate": 5.326528050872276e-06,
+      "loss": 0.5277,
+      "step": 17538
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 35.17599105834961,
+      "learning_rate": 5.325691335815588e-06,
+      "loss": 0.7843,
+      "step": 17539
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 13.560148239135742,
+      "learning_rate": 5.3248546207589015e-06,
+      "loss": 0.7429,
+      "step": 17540
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 23.572628021240234,
+      "learning_rate": 5.324017905702213e-06,
+      "loss": 1.2264,
+      "step": 17541
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 12.191099166870117,
+      "learning_rate": 5.323181190645526e-06,
+      "loss": 0.8501,
+      "step": 17542
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 13.795705795288086,
+      "learning_rate": 5.322344475588839e-06,
+      "loss": 1.266,
+      "step": 17543
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 11.90030288696289,
+      "learning_rate": 5.321507760532151e-06,
+      "loss": 0.7567,
+      "step": 17544
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 34.31806182861328,
+      "learning_rate": 5.320671045475464e-06,
+      "loss": 1.7403,
+      "step": 17545
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 11.005486488342285,
+      "learning_rate": 5.319834330418776e-06,
+      "loss": 0.6046,
+      "step": 17546
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 11.976456642150879,
+      "learning_rate": 5.318997615362089e-06,
+      "loss": 0.4688,
+      "step": 17547
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.7449660301208496,
+      "learning_rate": 5.318160900305401e-06,
+      "loss": 0.1778,
+      "step": 17548
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 10.648322105407715,
+      "learning_rate": 5.317324185248714e-06,
+      "loss": 1.4392,
+      "step": 17549
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 9.711552619934082,
+      "learning_rate": 5.316487470192027e-06,
+      "loss": 0.5422,
+      "step": 17550
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 11.51738166809082,
+      "learning_rate": 5.315650755135339e-06,
+      "loss": 1.7181,
+      "step": 17551
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 31.752824783325195,
+      "learning_rate": 5.314814040078652e-06,
+      "loss": 1.9601,
+      "step": 17552
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 20.286109924316406,
+      "learning_rate": 5.313977325021964e-06,
+      "loss": 0.5984,
+      "step": 17553
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.763519763946533,
+      "learning_rate": 5.313140609965277e-06,
+      "loss": 0.147,
+      "step": 17554
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 153.0485076904297,
+      "learning_rate": 5.312303894908589e-06,
+      "loss": 1.3208,
+      "step": 17555
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 6.014166831970215,
+      "learning_rate": 5.311467179851902e-06,
+      "loss": 0.7034,
+      "step": 17556
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 13.632169723510742,
+      "learning_rate": 5.310630464795215e-06,
+      "loss": 2.0591,
+      "step": 17557
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 40.078006744384766,
+      "learning_rate": 5.309793749738527e-06,
+      "loss": 0.7908,
+      "step": 17558
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 51.911190032958984,
+      "learning_rate": 5.30895703468184e-06,
+      "loss": 1.8859,
+      "step": 17559
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 12.481959342956543,
+      "learning_rate": 5.3081203196251516e-06,
+      "loss": 1.0484,
+      "step": 17560
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 14.264859199523926,
+      "learning_rate": 5.307283604568465e-06,
+      "loss": 1.6794,
+      "step": 17561
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 11.537360191345215,
+      "learning_rate": 5.306446889511777e-06,
+      "loss": 1.0252,
+      "step": 17562
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 16.290449142456055,
+      "learning_rate": 5.30561017445509e-06,
+      "loss": 1.8328,
+      "step": 17563
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 33.80059814453125,
+      "learning_rate": 5.304773459398403e-06,
+      "loss": 2.6643,
+      "step": 17564
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 13.9462251663208,
+      "learning_rate": 5.303936744341715e-06,
+      "loss": 2.2787,
+      "step": 17565
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 14.095157623291016,
+      "learning_rate": 5.3031000292850275e-06,
+      "loss": 0.6708,
+      "step": 17566
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 6.371540546417236,
+      "learning_rate": 5.3022633142283395e-06,
+      "loss": 0.2943,
+      "step": 17567
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 23.037803649902344,
+      "learning_rate": 5.301426599171653e-06,
+      "loss": 1.0293,
+      "step": 17568
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 13.298918724060059,
+      "learning_rate": 5.300589884114965e-06,
+      "loss": 0.9281,
+      "step": 17569
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 9.514727592468262,
+      "learning_rate": 5.299753169058278e-06,
+      "loss": 0.5065,
+      "step": 17570
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 10.597125053405762,
+      "learning_rate": 5.298916454001591e-06,
+      "loss": 0.7589,
+      "step": 17571
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 13.349027633666992,
+      "learning_rate": 5.298079738944903e-06,
+      "loss": 2.7605,
+      "step": 17572
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.783844470977783,
+      "learning_rate": 5.2972430238882155e-06,
+      "loss": 0.5558,
+      "step": 17573
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 21.072872161865234,
+      "learning_rate": 5.296406308831527e-06,
+      "loss": 0.6579,
+      "step": 17574
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.221395254135132,
+      "learning_rate": 5.295569593774841e-06,
+      "loss": 0.2594,
+      "step": 17575
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 12.586833000183105,
+      "learning_rate": 5.294732878718153e-06,
+      "loss": 1.1632,
+      "step": 17576
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 16.660640716552734,
+      "learning_rate": 5.293896163661465e-06,
+      "loss": 0.4993,
+      "step": 17577
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 7.333919525146484,
+      "learning_rate": 5.293059448604778e-06,
+      "loss": 0.4903,
+      "step": 17578
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 31.236549377441406,
+      "learning_rate": 5.2922227335480906e-06,
+      "loss": 0.9915,
+      "step": 17579
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 14.862787246704102,
+      "learning_rate": 5.291386018491403e-06,
+      "loss": 0.4041,
+      "step": 17580
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 13.543603897094727,
+      "learning_rate": 5.290549303434715e-06,
+      "loss": 0.9671,
+      "step": 17581
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 23.977313995361328,
+      "learning_rate": 5.289712588378029e-06,
+      "loss": 1.0142,
+      "step": 17582
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 24.36107635498047,
+      "learning_rate": 5.288875873321341e-06,
+      "loss": 1.347,
+      "step": 17583
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.181324005126953,
+      "learning_rate": 5.288039158264653e-06,
+      "loss": 0.8222,
+      "step": 17584
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 11.00633716583252,
+      "learning_rate": 5.287202443207966e-06,
+      "loss": 0.3454,
+      "step": 17585
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 97.79345703125,
+      "learning_rate": 5.2863657281512785e-06,
+      "loss": 1.1273,
+      "step": 17586
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.67331600189209,
+      "learning_rate": 5.285529013094591e-06,
+      "loss": 0.4736,
+      "step": 17587
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 14.97656536102295,
+      "learning_rate": 5.284692298037903e-06,
+      "loss": 0.8189,
+      "step": 17588
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 13.51713752746582,
+      "learning_rate": 5.283855582981217e-06,
+      "loss": 1.57,
+      "step": 17589
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 17.975173950195312,
+      "learning_rate": 5.283018867924529e-06,
+      "loss": 1.2526,
+      "step": 17590
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 18.56690216064453,
+      "learning_rate": 5.282182152867841e-06,
+      "loss": 1.7809,
+      "step": 17591
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 9.219497680664062,
+      "learning_rate": 5.281345437811154e-06,
+      "loss": 0.4182,
+      "step": 17592
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 24.483352661132812,
+      "learning_rate": 5.280508722754466e-06,
+      "loss": 2.3645,
+      "step": 17593
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 10.796198844909668,
+      "learning_rate": 5.279672007697779e-06,
+      "loss": 0.3318,
+      "step": 17594
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 27.02687644958496,
+      "learning_rate": 5.278835292641091e-06,
+      "loss": 1.1065,
+      "step": 17595
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.5153303146362305,
+      "learning_rate": 5.277998577584405e-06,
+      "loss": 0.1791,
+      "step": 17596
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 12.836956024169922,
+      "learning_rate": 5.277161862527717e-06,
+      "loss": 0.3242,
+      "step": 17597
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 57.1956787109375,
+      "learning_rate": 5.276325147471029e-06,
+      "loss": 0.6991,
+      "step": 17598
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 18.503223419189453,
+      "learning_rate": 5.2754884324143415e-06,
+      "loss": 1.3607,
+      "step": 17599
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 13.974534034729004,
+      "learning_rate": 5.274651717357654e-06,
+      "loss": 0.7077,
+      "step": 17600
+    },
+    {
+      "epoch": 2.21,
+      "eval_loss": 0.08298930525779724,
+      "eval_runtime": 95.116,
+      "eval_samples_per_second": 37.239,
+      "eval_steps_per_second": 37.239,
+      "step": 17600
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 33.60075759887695,
+      "learning_rate": 5.273815002300967e-06,
+      "loss": 2.1778,
+      "step": 17601
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 20.371700286865234,
+      "learning_rate": 5.272978287244279e-06,
+      "loss": 1.0122,
+      "step": 17602
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 7.336877346038818,
+      "learning_rate": 5.272141572187593e-06,
+      "loss": 0.3176,
+      "step": 17603
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 232.3487091064453,
+      "learning_rate": 5.271304857130905e-06,
+      "loss": 1.5995,
+      "step": 17604
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 12.699992179870605,
+      "learning_rate": 5.270468142074217e-06,
+      "loss": 0.5626,
+      "step": 17605
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 15.630974769592285,
+      "learning_rate": 5.2696314270175294e-06,
+      "loss": 1.003,
+      "step": 17606
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 8.735376358032227,
+      "learning_rate": 5.268794711960841e-06,
+      "loss": 0.6007,
+      "step": 17607
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 12.420937538146973,
+      "learning_rate": 5.267957996904155e-06,
+      "loss": 0.4516,
+      "step": 17608
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 20.599655151367188,
+      "learning_rate": 5.267121281847467e-06,
+      "loss": 0.724,
+      "step": 17609
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.0221431255340576,
+      "learning_rate": 5.266284566790781e-06,
+      "loss": 0.0848,
+      "step": 17610
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 26.17032814025879,
+      "learning_rate": 5.265447851734093e-06,
+      "loss": 1.3508,
+      "step": 17611
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 30.255577087402344,
+      "learning_rate": 5.2646111366774046e-06,
+      "loss": 1.1052,
+      "step": 17612
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 18.89253044128418,
+      "learning_rate": 5.263774421620717e-06,
+      "loss": 0.9781,
+      "step": 17613
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 11.845908164978027,
+      "learning_rate": 5.262937706564029e-06,
+      "loss": 0.4968,
+      "step": 17614
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 142.40744018554688,
+      "learning_rate": 5.262100991507343e-06,
+      "loss": 2.5384,
+      "step": 17615
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 11.632862091064453,
+      "learning_rate": 5.261264276450655e-06,
+      "loss": 0.7456,
+      "step": 17616
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 49.94683074951172,
+      "learning_rate": 5.2604275613939686e-06,
+      "loss": 4.7063,
+      "step": 17617
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 13.744133949279785,
+      "learning_rate": 5.2595908463372805e-06,
+      "loss": 0.8506,
+      "step": 17618
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 6.055446147918701,
+      "learning_rate": 5.2587541312805925e-06,
+      "loss": 1.0562,
+      "step": 17619
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 18.613239288330078,
+      "learning_rate": 5.257917416223905e-06,
+      "loss": 0.6748,
+      "step": 17620
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 49.51359176635742,
+      "learning_rate": 5.257080701167217e-06,
+      "loss": 1.5565,
+      "step": 17621
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 23.81951141357422,
+      "learning_rate": 5.256243986110531e-06,
+      "loss": 0.8403,
+      "step": 17622
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 18.800947189331055,
+      "learning_rate": 5.255407271053843e-06,
+      "loss": 2.0553,
+      "step": 17623
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 14.106256484985352,
+      "learning_rate": 5.254570555997156e-06,
+      "loss": 2.1392,
+      "step": 17624
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 6.436473846435547,
+      "learning_rate": 5.2537338409404684e-06,
+      "loss": 0.068,
+      "step": 17625
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 47.873111724853516,
+      "learning_rate": 5.25289712588378e-06,
+      "loss": 2.3184,
+      "step": 17626
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 15.037662506103516,
+      "learning_rate": 5.252060410827093e-06,
+      "loss": 1.1386,
+      "step": 17627
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 199.6589813232422,
+      "learning_rate": 5.251223695770405e-06,
+      "loss": 1.5881,
+      "step": 17628
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 10.401008605957031,
+      "learning_rate": 5.250386980713719e-06,
+      "loss": 1.2039,
+      "step": 17629
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 11.683107376098633,
+      "learning_rate": 5.249550265657031e-06,
+      "loss": 0.7879,
+      "step": 17630
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 7.70859956741333,
+      "learning_rate": 5.2487135506003436e-06,
+      "loss": 0.6255,
+      "step": 17631
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 24.480422973632812,
+      "learning_rate": 5.247876835543656e-06,
+      "loss": 1.4999,
+      "step": 17632
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 20.174907684326172,
+      "learning_rate": 5.247040120486968e-06,
+      "loss": 1.5308,
+      "step": 17633
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 26.942628860473633,
+      "learning_rate": 5.246203405430281e-06,
+      "loss": 1.1784,
+      "step": 17634
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 13.76342487335205,
+      "learning_rate": 5.245366690373593e-06,
+      "loss": 0.4631,
+      "step": 17635
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 38.779319763183594,
+      "learning_rate": 5.244529975316907e-06,
+      "loss": 0.6271,
+      "step": 17636
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 6.188003063201904,
+      "learning_rate": 5.243693260260219e-06,
+      "loss": 0.9849,
+      "step": 17637
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 18.986454010009766,
+      "learning_rate": 5.2428565452035315e-06,
+      "loss": 0.9632,
+      "step": 17638
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 88.78659057617188,
+      "learning_rate": 5.242019830146844e-06,
+      "loss": 0.7307,
+      "step": 17639
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.771511554718018,
+      "learning_rate": 5.241183115090156e-06,
+      "loss": 0.8864,
+      "step": 17640
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 10.675668716430664,
+      "learning_rate": 5.240346400033469e-06,
+      "loss": 1.0148,
+      "step": 17641
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 32.972877502441406,
+      "learning_rate": 5.239509684976781e-06,
+      "loss": 1.5309,
+      "step": 17642
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 7.496189117431641,
+      "learning_rate": 5.238672969920095e-06,
+      "loss": 0.5997,
+      "step": 17643
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 21.612586975097656,
+      "learning_rate": 5.237836254863407e-06,
+      "loss": 1.1054,
+      "step": 17644
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 31.007505416870117,
+      "learning_rate": 5.236999539806719e-06,
+      "loss": 1.1644,
+      "step": 17645
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 34.03743362426758,
+      "learning_rate": 5.236162824750032e-06,
+      "loss": 1.6796,
+      "step": 17646
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 6.421183109283447,
+      "learning_rate": 5.235326109693344e-06,
+      "loss": 0.2818,
+      "step": 17647
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 13.530191421508789,
+      "learning_rate": 5.234489394636657e-06,
+      "loss": 1.8073,
+      "step": 17648
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 7.906088352203369,
+      "learning_rate": 5.233652679579969e-06,
+      "loss": 1.7872,
+      "step": 17649
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 7.971576690673828,
+      "learning_rate": 5.2328159645232826e-06,
+      "loss": 0.1354,
+      "step": 17650
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.973159313201904,
+      "learning_rate": 5.2319792494665945e-06,
+      "loss": 1.2619,
+      "step": 17651
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 7.155309200286865,
+      "learning_rate": 5.231142534409907e-06,
+      "loss": 0.8706,
+      "step": 17652
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 14.330678939819336,
+      "learning_rate": 5.230305819353219e-06,
+      "loss": 0.3922,
+      "step": 17653
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 10.194661140441895,
+      "learning_rate": 5.229469104296532e-06,
+      "loss": 0.5062,
+      "step": 17654
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 29.81911277770996,
+      "learning_rate": 5.228632389239845e-06,
+      "loss": 1.8285,
+      "step": 17655
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 34.0993537902832,
+      "learning_rate": 5.227795674183157e-06,
+      "loss": 1.3139,
+      "step": 17656
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 10.460409164428711,
+      "learning_rate": 5.2269589591264705e-06,
+      "loss": 1.3534,
+      "step": 17657
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 28.992599487304688,
+      "learning_rate": 5.2261222440697824e-06,
+      "loss": 3.2598,
+      "step": 17658
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 15.127599716186523,
+      "learning_rate": 5.225285529013095e-06,
+      "loss": 1.5536,
+      "step": 17659
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 11.926405906677246,
+      "learning_rate": 5.224448813956407e-06,
+      "loss": 0.5888,
+      "step": 17660
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 67.72642517089844,
+      "learning_rate": 5.22361209889972e-06,
+      "loss": 2.6624,
+      "step": 17661
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 13.176055908203125,
+      "learning_rate": 5.222775383843033e-06,
+      "loss": 0.5765,
+      "step": 17662
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 6.374279499053955,
+      "learning_rate": 5.221938668786345e-06,
+      "loss": 0.5205,
+      "step": 17663
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 64.94546508789062,
+      "learning_rate": 5.221101953729658e-06,
+      "loss": 1.7056,
+      "step": 17664
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 40.390602111816406,
+      "learning_rate": 5.22026523867297e-06,
+      "loss": 2.0192,
+      "step": 17665
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 15.01516342163086,
+      "learning_rate": 5.219428523616283e-06,
+      "loss": 0.5252,
+      "step": 17666
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 22.53778648376465,
+      "learning_rate": 5.218591808559595e-06,
+      "loss": 0.9188,
+      "step": 17667
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.551015853881836,
+      "learning_rate": 5.217755093502908e-06,
+      "loss": 0.4476,
+      "step": 17668
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.424520969390869,
+      "learning_rate": 5.216918378446221e-06,
+      "loss": 0.7938,
+      "step": 17669
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 8.108823776245117,
+      "learning_rate": 5.216081663389533e-06,
+      "loss": 0.407,
+      "step": 17670
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 7.024099349975586,
+      "learning_rate": 5.215244948332846e-06,
+      "loss": 1.005,
+      "step": 17671
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 17.924938201904297,
+      "learning_rate": 5.214408233276158e-06,
+      "loss": 1.6325,
+      "step": 17672
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 26.36313247680664,
+      "learning_rate": 5.213571518219471e-06,
+      "loss": 1.2318,
+      "step": 17673
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 6.06789493560791,
+      "learning_rate": 5.212734803162783e-06,
+      "loss": 0.6525,
+      "step": 17674
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 14.966464042663574,
+      "learning_rate": 5.211898088106096e-06,
+      "loss": 0.9419,
+      "step": 17675
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 8.155410766601562,
+      "learning_rate": 5.211061373049409e-06,
+      "loss": 1.016,
+      "step": 17676
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 49.047706604003906,
+      "learning_rate": 5.210224657992721e-06,
+      "loss": 1.491,
+      "step": 17677
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 40.98280334472656,
+      "learning_rate": 5.209387942936034e-06,
+      "loss": 1.6621,
+      "step": 17678
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 98.17125701904297,
+      "learning_rate": 5.208551227879346e-06,
+      "loss": 1.1052,
+      "step": 17679
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 14.517447471618652,
+      "learning_rate": 5.207714512822659e-06,
+      "loss": 0.3564,
+      "step": 17680
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.574953079223633,
+      "learning_rate": 5.206877797765971e-06,
+      "loss": 0.2258,
+      "step": 17681
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 16.248891830444336,
+      "learning_rate": 5.206041082709284e-06,
+      "loss": 0.6977,
+      "step": 17682
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 9.710600852966309,
+      "learning_rate": 5.2052043676525965e-06,
+      "loss": 0.8049,
+      "step": 17683
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 18.749849319458008,
+      "learning_rate": 5.2043676525959085e-06,
+      "loss": 0.4875,
+      "step": 17684
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 14.403364181518555,
+      "learning_rate": 5.203530937539222e-06,
+      "loss": 0.4326,
+      "step": 17685
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 25.39562225341797,
+      "learning_rate": 5.202694222482534e-06,
+      "loss": 1.1199,
+      "step": 17686
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 21.403207778930664,
+      "learning_rate": 5.201857507425847e-06,
+      "loss": 0.6541,
+      "step": 17687
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 10.751093864440918,
+      "learning_rate": 5.201020792369159e-06,
+      "loss": 0.2954,
+      "step": 17688
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 22.85277557373047,
+      "learning_rate": 5.200184077312471e-06,
+      "loss": 2.2193,
+      "step": 17689
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.862344741821289,
+      "learning_rate": 5.1993473622557845e-06,
+      "loss": 1.3405,
+      "step": 17690
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.9521684646606445,
+      "learning_rate": 5.198510647199096e-06,
+      "loss": 0.3263,
+      "step": 17691
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 9.945138931274414,
+      "learning_rate": 5.19767393214241e-06,
+      "loss": 0.4844,
+      "step": 17692
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 7.673474311828613,
+      "learning_rate": 5.196837217085722e-06,
+      "loss": 0.04,
+      "step": 17693
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 9.783452033996582,
+      "learning_rate": 5.196000502029035e-06,
+      "loss": 0.9584,
+      "step": 17694
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 23.981342315673828,
+      "learning_rate": 5.195163786972347e-06,
+      "loss": 1.8661,
+      "step": 17695
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 18.916805267333984,
+      "learning_rate": 5.194327071915659e-06,
+      "loss": 0.9652,
+      "step": 17696
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 76.17505645751953,
+      "learning_rate": 5.193490356858972e-06,
+      "loss": 1.6803,
+      "step": 17697
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 20.964216232299805,
+      "learning_rate": 5.192653641802284e-06,
+      "loss": 0.6796,
+      "step": 17698
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 11.345860481262207,
+      "learning_rate": 5.191816926745598e-06,
+      "loss": 1.497,
+      "step": 17699
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 6.6955180168151855,
+      "learning_rate": 5.19098021168891e-06,
+      "loss": 0.6932,
+      "step": 17700
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 14.110368728637695,
+      "learning_rate": 5.190143496632223e-06,
+      "loss": 0.8485,
+      "step": 17701
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 12.59630298614502,
+      "learning_rate": 5.189306781575535e-06,
+      "loss": 1.4359,
+      "step": 17702
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 7.114568710327148,
+      "learning_rate": 5.188470066518847e-06,
+      "loss": 0.2758,
+      "step": 17703
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 11.192057609558105,
+      "learning_rate": 5.18763335146216e-06,
+      "loss": 1.1341,
+      "step": 17704
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 8.494654655456543,
+      "learning_rate": 5.186796636405472e-06,
+      "loss": 1.4699,
+      "step": 17705
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 11.067359924316406,
+      "learning_rate": 5.185959921348785e-06,
+      "loss": 0.2085,
+      "step": 17706
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 49.78376388549805,
+      "learning_rate": 5.185123206292098e-06,
+      "loss": 1.5447,
+      "step": 17707
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 13.176912307739258,
+      "learning_rate": 5.184286491235411e-06,
+      "loss": 0.234,
+      "step": 17708
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 14.371654510498047,
+      "learning_rate": 5.183449776178723e-06,
+      "loss": 0.5464,
+      "step": 17709
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 15.566927909851074,
+      "learning_rate": 5.1826130611220346e-06,
+      "loss": 1.1125,
+      "step": 17710
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 7.577420234680176,
+      "learning_rate": 5.181776346065348e-06,
+      "loss": 1.8091,
+      "step": 17711
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.659139633178711,
+      "learning_rate": 5.18093963100866e-06,
+      "loss": 0.5024,
+      "step": 17712
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 12.561866760253906,
+      "learning_rate": 5.180102915951973e-06,
+      "loss": 0.2913,
+      "step": 17713
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 39.97332763671875,
+      "learning_rate": 5.179266200895286e-06,
+      "loss": 1.2649,
+      "step": 17714
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 106.55900573730469,
+      "learning_rate": 5.178429485838599e-06,
+      "loss": 0.8508,
+      "step": 17715
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 32.46993637084961,
+      "learning_rate": 5.1775927707819105e-06,
+      "loss": 1.5124,
+      "step": 17716
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 10.00876235961914,
+      "learning_rate": 5.1767560557252225e-06,
+      "loss": 0.9794,
+      "step": 17717
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 8.991241455078125,
+      "learning_rate": 5.175919340668536e-06,
+      "loss": 1.698,
+      "step": 17718
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.391348361968994,
+      "learning_rate": 5.175082625611848e-06,
+      "loss": 0.1494,
+      "step": 17719
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 15.718477249145508,
+      "learning_rate": 5.174245910555161e-06,
+      "loss": 0.5882,
+      "step": 17720
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.732717037200928,
+      "learning_rate": 5.173409195498474e-06,
+      "loss": 0.9163,
+      "step": 17721
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 14.677023887634277,
+      "learning_rate": 5.1725724804417865e-06,
+      "loss": 0.4184,
+      "step": 17722
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 44.430641174316406,
+      "learning_rate": 5.1717357653850985e-06,
+      "loss": 3.6664,
+      "step": 17723
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.236386775970459,
+      "learning_rate": 5.17089905032841e-06,
+      "loss": 0.3657,
+      "step": 17724
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 12.09508991241455,
+      "learning_rate": 5.170062335271724e-06,
+      "loss": 2.1726,
+      "step": 17725
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 12.473302841186523,
+      "learning_rate": 5.169225620215036e-06,
+      "loss": 0.3477,
+      "step": 17726
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 13.235149383544922,
+      "learning_rate": 5.168388905158349e-06,
+      "loss": 0.4917,
+      "step": 17727
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 7.2687811851501465,
+      "learning_rate": 5.167552190101662e-06,
+      "loss": 2.1986,
+      "step": 17728
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 46.85900115966797,
+      "learning_rate": 5.166715475044974e-06,
+      "loss": 2.0231,
+      "step": 17729
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 13.293231010437012,
+      "learning_rate": 5.165878759988286e-06,
+      "loss": 0.201,
+      "step": 17730
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 13.189783096313477,
+      "learning_rate": 5.165042044931598e-06,
+      "loss": 0.8895,
+      "step": 17731
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 16.408214569091797,
+      "learning_rate": 5.164205329874912e-06,
+      "loss": 1.075,
+      "step": 17732
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 17.93597984313965,
+      "learning_rate": 5.163368614818224e-06,
+      "loss": 1.2993,
+      "step": 17733
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 17.551984786987305,
+      "learning_rate": 5.162531899761537e-06,
+      "loss": 0.2072,
+      "step": 17734
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 15.895017623901367,
+      "learning_rate": 5.161695184704849e-06,
+      "loss": 0.9385,
+      "step": 17735
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 42.44586181640625,
+      "learning_rate": 5.160858469648162e-06,
+      "loss": 0.6849,
+      "step": 17736
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 14.050836563110352,
+      "learning_rate": 5.160021754591474e-06,
+      "loss": 0.5258,
+      "step": 17737
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 21.7032413482666,
+      "learning_rate": 5.159185039534786e-06,
+      "loss": 2.3616,
+      "step": 17738
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 21.918964385986328,
+      "learning_rate": 5.1583483244781e-06,
+      "loss": 0.6126,
+      "step": 17739
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 13.94185733795166,
+      "learning_rate": 5.157511609421412e-06,
+      "loss": 0.7284,
+      "step": 17740
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 9.73373794555664,
+      "learning_rate": 5.156674894364725e-06,
+      "loss": 0.623,
+      "step": 17741
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 13.341050148010254,
+      "learning_rate": 5.155838179308037e-06,
+      "loss": 0.3461,
+      "step": 17742
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 11.477652549743652,
+      "learning_rate": 5.15500146425135e-06,
+      "loss": 1.3524,
+      "step": 17743
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 15.867574691772461,
+      "learning_rate": 5.154164749194662e-06,
+      "loss": 1.4677,
+      "step": 17744
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 17.404987335205078,
+      "learning_rate": 5.153328034137974e-06,
+      "loss": 0.4331,
+      "step": 17745
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 12.849403381347656,
+      "learning_rate": 5.152491319081288e-06,
+      "loss": 0.5099,
+      "step": 17746
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 18.763734817504883,
+      "learning_rate": 5.1516546040246e-06,
+      "loss": 1.9163,
+      "step": 17747
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 24.120838165283203,
+      "learning_rate": 5.1508178889679126e-06,
+      "loss": 1.3214,
+      "step": 17748
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 41.783485412597656,
+      "learning_rate": 5.1499811739112245e-06,
+      "loss": 2.4696,
+      "step": 17749
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 17.785659790039062,
+      "learning_rate": 5.149144458854537e-06,
+      "loss": 0.6855,
+      "step": 17750
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 12.047317504882812,
+      "learning_rate": 5.14830774379785e-06,
+      "loss": 0.4267,
+      "step": 17751
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.101709842681885,
+      "learning_rate": 5.147471028741162e-06,
+      "loss": 0.2058,
+      "step": 17752
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 17.78318977355957,
+      "learning_rate": 5.146634313684476e-06,
+      "loss": 0.8566,
+      "step": 17753
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 15.483389854431152,
+      "learning_rate": 5.145797598627788e-06,
+      "loss": 1.5509,
+      "step": 17754
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 30.174823760986328,
+      "learning_rate": 5.1449608835711005e-06,
+      "loss": 1.3108,
+      "step": 17755
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.915126323699951,
+      "learning_rate": 5.1441241685144124e-06,
+      "loss": 0.4768,
+      "step": 17756
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 15.084676742553711,
+      "learning_rate": 5.143287453457725e-06,
+      "loss": 1.3511,
+      "step": 17757
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 12.236443519592285,
+      "learning_rate": 5.142450738401038e-06,
+      "loss": 0.8682,
+      "step": 17758
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 8.270609855651855,
+      "learning_rate": 5.14161402334435e-06,
+      "loss": 0.4339,
+      "step": 17759
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 39.991859436035156,
+      "learning_rate": 5.140777308287664e-06,
+      "loss": 2.3828,
+      "step": 17760
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 41.814022064208984,
+      "learning_rate": 5.139940593230976e-06,
+      "loss": 1.1213,
+      "step": 17761
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 7.52501916885376,
+      "learning_rate": 5.139103878174288e-06,
+      "loss": 0.3014,
+      "step": 17762
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 20.50660514831543,
+      "learning_rate": 5.1382671631176e-06,
+      "loss": 0.9175,
+      "step": 17763
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 18.19256591796875,
+      "learning_rate": 5.137430448060912e-06,
+      "loss": 0.7378,
+      "step": 17764
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.9844133853912354,
+      "learning_rate": 5.136593733004226e-06,
+      "loss": 0.0584,
+      "step": 17765
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 14.300190925598145,
+      "learning_rate": 5.135757017947538e-06,
+      "loss": 0.7404,
+      "step": 17766
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 38.13930130004883,
+      "learning_rate": 5.1349203028908516e-06,
+      "loss": 0.8283,
+      "step": 17767
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 11.909294128417969,
+      "learning_rate": 5.1340835878341635e-06,
+      "loss": 0.8189,
+      "step": 17768
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 11.108964920043945,
+      "learning_rate": 5.133246872777476e-06,
+      "loss": 0.8144,
+      "step": 17769
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 8.476258277893066,
+      "learning_rate": 5.132410157720788e-06,
+      "loss": 0.5914,
+      "step": 17770
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 6.253983497619629,
+      "learning_rate": 5.1315734426641e-06,
+      "loss": 0.4603,
+      "step": 17771
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 33.24333572387695,
+      "learning_rate": 5.130736727607414e-06,
+      "loss": 0.9329,
+      "step": 17772
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 16.08575439453125,
+      "learning_rate": 5.129900012550726e-06,
+      "loss": 0.5251,
+      "step": 17773
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 115.9076156616211,
+      "learning_rate": 5.1290632974940395e-06,
+      "loss": 2.309,
+      "step": 17774
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 19.51165008544922,
+      "learning_rate": 5.1282265824373514e-06,
+      "loss": 0.4296,
+      "step": 17775
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 17.818742752075195,
+      "learning_rate": 5.127389867380664e-06,
+      "loss": 0.9087,
+      "step": 17776
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 16.055883407592773,
+      "learning_rate": 5.126553152323976e-06,
+      "loss": 0.6511,
+      "step": 17777
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 24.93146324157715,
+      "learning_rate": 5.125716437267288e-06,
+      "loss": 1.779,
+      "step": 17778
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 29.532642364501953,
+      "learning_rate": 5.124879722210602e-06,
+      "loss": 0.9848,
+      "step": 17779
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 10.471334457397461,
+      "learning_rate": 5.124043007153914e-06,
+      "loss": 0.8294,
+      "step": 17780
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 22.57203483581543,
+      "learning_rate": 5.1232062920972266e-06,
+      "loss": 1.5461,
+      "step": 17781
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 19.7887020111084,
+      "learning_rate": 5.122369577040539e-06,
+      "loss": 0.9483,
+      "step": 17782
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 6.547120094299316,
+      "learning_rate": 5.121532861983852e-06,
+      "loss": 0.4218,
+      "step": 17783
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 11.875927925109863,
+      "learning_rate": 5.120696146927164e-06,
+      "loss": 0.7022,
+      "step": 17784
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 8.516631126403809,
+      "learning_rate": 5.119859431870476e-06,
+      "loss": 0.8346,
+      "step": 17785
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 24.55098533630371,
+      "learning_rate": 5.11902271681379e-06,
+      "loss": 0.9406,
+      "step": 17786
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 39.46234893798828,
+      "learning_rate": 5.118186001757102e-06,
+      "loss": 0.9281,
+      "step": 17787
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 10.087285995483398,
+      "learning_rate": 5.1173492867004145e-06,
+      "loss": 0.6703,
+      "step": 17788
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 15.53774356842041,
+      "learning_rate": 5.116512571643727e-06,
+      "loss": 1.8845,
+      "step": 17789
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 18.932775497436523,
+      "learning_rate": 5.11567585658704e-06,
+      "loss": 1.1073,
+      "step": 17790
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 12.49699592590332,
+      "learning_rate": 5.114839141530352e-06,
+      "loss": 0.9165,
+      "step": 17791
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 19.66526222229004,
+      "learning_rate": 5.114002426473664e-06,
+      "loss": 1.3602,
+      "step": 17792
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 12.239690780639648,
+      "learning_rate": 5.113165711416978e-06,
+      "loss": 1.5688,
+      "step": 17793
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 24.5960750579834,
+      "learning_rate": 5.11232899636029e-06,
+      "loss": 1.767,
+      "step": 17794
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 9.939472198486328,
+      "learning_rate": 5.111492281303602e-06,
+      "loss": 0.7966,
+      "step": 17795
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 9.753925323486328,
+      "learning_rate": 5.110655566246915e-06,
+      "loss": 1.5595,
+      "step": 17796
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 6.685403347015381,
+      "learning_rate": 5.109818851190228e-06,
+      "loss": 1.0454,
+      "step": 17797
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 22.421031951904297,
+      "learning_rate": 5.10898213613354e-06,
+      "loss": 2.2144,
+      "step": 17798
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 15.689521789550781,
+      "learning_rate": 5.108145421076852e-06,
+      "loss": 1.0526,
+      "step": 17799
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 16.362619400024414,
+      "learning_rate": 5.1073087060201656e-06,
+      "loss": 0.7085,
+      "step": 17800
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 20.69801139831543,
+      "learning_rate": 5.1064719909634775e-06,
+      "loss": 1.1504,
+      "step": 17801
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 23.02520179748535,
+      "learning_rate": 5.10563527590679e-06,
+      "loss": 1.2343,
+      "step": 17802
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 22.365503311157227,
+      "learning_rate": 5.104798560850103e-06,
+      "loss": 1.7095,
+      "step": 17803
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 9.561039924621582,
+      "learning_rate": 5.103961845793416e-06,
+      "loss": 1.0846,
+      "step": 17804
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 6.607800006866455,
+      "learning_rate": 5.103125130736728e-06,
+      "loss": 0.1923,
+      "step": 17805
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.634875774383545,
+      "learning_rate": 5.10228841568004e-06,
+      "loss": 0.085,
+      "step": 17806
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 23.58032989501953,
+      "learning_rate": 5.1014517006233535e-06,
+      "loss": 0.8119,
+      "step": 17807
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.756401538848877,
+      "learning_rate": 5.1006149855666654e-06,
+      "loss": 0.4115,
+      "step": 17808
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 22.091915130615234,
+      "learning_rate": 5.099778270509978e-06,
+      "loss": 0.7643,
+      "step": 17809
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 26.409635543823242,
+      "learning_rate": 5.098941555453291e-06,
+      "loss": 0.7003,
+      "step": 17810
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 12.126300811767578,
+      "learning_rate": 5.098104840396604e-06,
+      "loss": 0.2088,
+      "step": 17811
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 10.77668571472168,
+      "learning_rate": 5.097268125339916e-06,
+      "loss": 1.3343,
+      "step": 17812
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 21.273651123046875,
+      "learning_rate": 5.096431410283228e-06,
+      "loss": 0.7169,
+      "step": 17813
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 6.164846420288086,
+      "learning_rate": 5.095594695226541e-06,
+      "loss": 1.7545,
+      "step": 17814
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 33.088409423828125,
+      "learning_rate": 5.094757980169853e-06,
+      "loss": 1.7557,
+      "step": 17815
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 12.935946464538574,
+      "learning_rate": 5.093921265113166e-06,
+      "loss": 0.72,
+      "step": 17816
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 9.345338821411133,
+      "learning_rate": 5.093084550056478e-06,
+      "loss": 0.3246,
+      "step": 17817
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 13.312705039978027,
+      "learning_rate": 5.092247834999792e-06,
+      "loss": 0.7129,
+      "step": 17818
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.6184065341949463,
+      "learning_rate": 5.091411119943104e-06,
+      "loss": 0.2902,
+      "step": 17819
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 21.613859176635742,
+      "learning_rate": 5.090574404886416e-06,
+      "loss": 0.3501,
+      "step": 17820
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 7.444402694702148,
+      "learning_rate": 5.089737689829729e-06,
+      "loss": 0.1674,
+      "step": 17821
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 102.95565795898438,
+      "learning_rate": 5.088900974773041e-06,
+      "loss": 0.5143,
+      "step": 17822
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.1632742881774902,
+      "learning_rate": 5.088064259716354e-06,
+      "loss": 0.0502,
+      "step": 17823
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 54.36214065551758,
+      "learning_rate": 5.087227544659666e-06,
+      "loss": 2.7835,
+      "step": 17824
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 10.562241554260254,
+      "learning_rate": 5.08639082960298e-06,
+      "loss": 0.9218,
+      "step": 17825
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 81.18700408935547,
+      "learning_rate": 5.085554114546292e-06,
+      "loss": 0.4029,
+      "step": 17826
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 14.44067096710205,
+      "learning_rate": 5.084717399489604e-06,
+      "loss": 0.9139,
+      "step": 17827
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 18.229602813720703,
+      "learning_rate": 5.083880684432917e-06,
+      "loss": 0.9942,
+      "step": 17828
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 15.732873916625977,
+      "learning_rate": 5.083043969376229e-06,
+      "loss": 4.3611,
+      "step": 17829
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 8.699427604675293,
+      "learning_rate": 5.082207254319542e-06,
+      "loss": 0.2181,
+      "step": 17830
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 9.435542106628418,
+      "learning_rate": 5.081370539262854e-06,
+      "loss": 0.6497,
+      "step": 17831
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.771695137023926,
+      "learning_rate": 5.080533824206168e-06,
+      "loss": 0.2273,
+      "step": 17832
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 10.254509925842285,
+      "learning_rate": 5.0796971091494795e-06,
+      "loss": 1.5389,
+      "step": 17833
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 25.043521881103516,
+      "learning_rate": 5.0788603940927915e-06,
+      "loss": 1.1398,
+      "step": 17834
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 10.375333786010742,
+      "learning_rate": 5.078023679036105e-06,
+      "loss": 0.6474,
+      "step": 17835
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 16.78009605407715,
+      "learning_rate": 5.077186963979417e-06,
+      "loss": 0.5516,
+      "step": 17836
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 82.75853729248047,
+      "learning_rate": 5.07635024892273e-06,
+      "loss": 1.4797,
+      "step": 17837
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 7.424945831298828,
+      "learning_rate": 5.075513533866042e-06,
+      "loss": 0.9605,
+      "step": 17838
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 19.120899200439453,
+      "learning_rate": 5.0746768188093555e-06,
+      "loss": 1.8178,
+      "step": 17839
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 26.920454025268555,
+      "learning_rate": 5.0738401037526675e-06,
+      "loss": 1.1617,
+      "step": 17840
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 13.276774406433105,
+      "learning_rate": 5.073003388695979e-06,
+      "loss": 0.5867,
+      "step": 17841
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 12.754081726074219,
+      "learning_rate": 5.072166673639293e-06,
+      "loss": 0.6544,
+      "step": 17842
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 6.249405384063721,
+      "learning_rate": 5.071329958582605e-06,
+      "loss": 0.3336,
+      "step": 17843
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 15.704106330871582,
+      "learning_rate": 5.070493243525918e-06,
+      "loss": 0.538,
+      "step": 17844
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 11.90325927734375,
+      "learning_rate": 5.06965652846923e-06,
+      "loss": 1.2049,
+      "step": 17845
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 29.243885040283203,
+      "learning_rate": 5.0688198134125434e-06,
+      "loss": 1.0439,
+      "step": 17846
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 7.157782077789307,
+      "learning_rate": 5.067983098355855e-06,
+      "loss": 0.6835,
+      "step": 17847
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 13.187911033630371,
+      "learning_rate": 5.067146383299167e-06,
+      "loss": 2.1351,
+      "step": 17848
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.006730079650879,
+      "learning_rate": 5.066309668242481e-06,
+      "loss": 0.4027,
+      "step": 17849
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 109.41995239257812,
+      "learning_rate": 5.065472953185793e-06,
+      "loss": 1.1113,
+      "step": 17850
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 9.678338050842285,
+      "learning_rate": 5.064636238129106e-06,
+      "loss": 0.3227,
+      "step": 17851
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 8.335624694824219,
+      "learning_rate": 5.063799523072418e-06,
+      "loss": 1.5265,
+      "step": 17852
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 18.686553955078125,
+      "learning_rate": 5.062962808015731e-06,
+      "loss": 0.4259,
+      "step": 17853
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 60.7696533203125,
+      "learning_rate": 5.062126092959043e-06,
+      "loss": 2.018,
+      "step": 17854
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 10.57986831665039,
+      "learning_rate": 5.061289377902355e-06,
+      "loss": 1.7964,
+      "step": 17855
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 7.243568420410156,
+      "learning_rate": 5.060452662845669e-06,
+      "loss": 0.5711,
+      "step": 17856
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 28.93546485900879,
+      "learning_rate": 5.059615947788981e-06,
+      "loss": 1.0337,
+      "step": 17857
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 7.0018815994262695,
+      "learning_rate": 5.058779232732294e-06,
+      "loss": 0.2486,
+      "step": 17858
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.1246564388275146,
+      "learning_rate": 5.057942517675606e-06,
+      "loss": 0.2019,
+      "step": 17859
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 7.868934154510498,
+      "learning_rate": 5.057105802618919e-06,
+      "loss": 0.3886,
+      "step": 17860
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 177.0337677001953,
+      "learning_rate": 5.056269087562231e-06,
+      "loss": 1.9114,
+      "step": 17861
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 28.44269561767578,
+      "learning_rate": 5.055432372505543e-06,
+      "loss": 0.7514,
+      "step": 17862
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 23.441205978393555,
+      "learning_rate": 5.054595657448856e-06,
+      "loss": 1.4996,
+      "step": 17863
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 19.890857696533203,
+      "learning_rate": 5.053758942392169e-06,
+      "loss": 2.2758,
+      "step": 17864
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 29.43320655822754,
+      "learning_rate": 5.052922227335482e-06,
+      "loss": 2.0113,
+      "step": 17865
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 43.13685989379883,
+      "learning_rate": 5.0520855122787935e-06,
+      "loss": 2.4186,
+      "step": 17866
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 14.734983444213867,
+      "learning_rate": 5.051248797222107e-06,
+      "loss": 0.7916,
+      "step": 17867
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 9.410062789916992,
+      "learning_rate": 5.050412082165419e-06,
+      "loss": 0.6375,
+      "step": 17868
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 50.64573669433594,
+      "learning_rate": 5.049575367108731e-06,
+      "loss": 1.6407,
+      "step": 17869
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 20.77255630493164,
+      "learning_rate": 5.048738652052044e-06,
+      "loss": 1.1511,
+      "step": 17870
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 54.63801956176758,
+      "learning_rate": 5.047901936995357e-06,
+      "loss": 2.2643,
+      "step": 17871
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 8.90231704711914,
+      "learning_rate": 5.0470652219386695e-06,
+      "loss": 0.6662,
+      "step": 17872
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 9.863375663757324,
+      "learning_rate": 5.0462285068819815e-06,
+      "loss": 2.183,
+      "step": 17873
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 9.324417114257812,
+      "learning_rate": 5.045391791825295e-06,
+      "loss": 1.1831,
+      "step": 17874
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 41.506832122802734,
+      "learning_rate": 5.044555076768607e-06,
+      "loss": 1.467,
+      "step": 17875
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.597385406494141,
+      "learning_rate": 5.043718361711919e-06,
+      "loss": 0.6364,
+      "step": 17876
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 11.700260162353516,
+      "learning_rate": 5.042881646655232e-06,
+      "loss": 0.211,
+      "step": 17877
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 9.609248161315918,
+      "learning_rate": 5.042044931598545e-06,
+      "loss": 0.5709,
+      "step": 17878
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 44.30210876464844,
+      "learning_rate": 5.041208216541857e-06,
+      "loss": 2.3704,
+      "step": 17879
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 39.628448486328125,
+      "learning_rate": 5.040371501485169e-06,
+      "loss": 2.082,
+      "step": 17880
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 6.445079326629639,
+      "learning_rate": 5.039534786428483e-06,
+      "loss": 0.2678,
+      "step": 17881
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.163549900054932,
+      "learning_rate": 5.038698071371795e-06,
+      "loss": 0.198,
+      "step": 17882
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 7.012538433074951,
+      "learning_rate": 5.037861356315107e-06,
+      "loss": 0.2496,
+      "step": 17883
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.5808498859405518,
+      "learning_rate": 5.03702464125842e-06,
+      "loss": 0.2042,
+      "step": 17884
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 18.245594024658203,
+      "learning_rate": 5.0361879262017325e-06,
+      "loss": 0.9452,
+      "step": 17885
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.119741916656494,
+      "learning_rate": 5.035351211145045e-06,
+      "loss": 0.9425,
+      "step": 17886
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 50.56310272216797,
+      "learning_rate": 5.034514496088357e-06,
+      "loss": 3.0142,
+      "step": 17887
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 173.828125,
+      "learning_rate": 5.033677781031671e-06,
+      "loss": 1.614,
+      "step": 17888
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 38.569576263427734,
+      "learning_rate": 5.032841065974983e-06,
+      "loss": 1.5961,
+      "step": 17889
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 25.682310104370117,
+      "learning_rate": 5.032004350918295e-06,
+      "loss": 2.4878,
+      "step": 17890
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 6.593906879425049,
+      "learning_rate": 5.031167635861608e-06,
+      "loss": 0.8402,
+      "step": 17891
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 12.642677307128906,
+      "learning_rate": 5.03033092080492e-06,
+      "loss": 0.8541,
+      "step": 17892
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 7.8724365234375,
+      "learning_rate": 5.029494205748233e-06,
+      "loss": 0.5848,
+      "step": 17893
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 21.33145523071289,
+      "learning_rate": 5.028657490691545e-06,
+      "loss": 1.9971,
+      "step": 17894
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 38.787960052490234,
+      "learning_rate": 5.027820775634859e-06,
+      "loss": 1.665,
+      "step": 17895
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 11.145357131958008,
+      "learning_rate": 5.026984060578171e-06,
+      "loss": 0.8777,
+      "step": 17896
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 14.004895210266113,
+      "learning_rate": 5.026147345521483e-06,
+      "loss": 0.5583,
+      "step": 17897
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 68.97599029541016,
+      "learning_rate": 5.0253106304647956e-06,
+      "loss": 0.5841,
+      "step": 17898
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 7.229062557220459,
+      "learning_rate": 5.0244739154081075e-06,
+      "loss": 0.452,
+      "step": 17899
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 22.491193771362305,
+      "learning_rate": 5.023637200351421e-06,
+      "loss": 1.6259,
+      "step": 17900
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 94.6976318359375,
+      "learning_rate": 5.022800485294733e-06,
+      "loss": 2.652,
+      "step": 17901
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 13.50717830657959,
+      "learning_rate": 5.021963770238047e-06,
+      "loss": 1.7198,
+      "step": 17902
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.883352279663086,
+      "learning_rate": 5.021127055181359e-06,
+      "loss": 0.2975,
+      "step": 17903
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 6.9274001121521,
+      "learning_rate": 5.020290340124671e-06,
+      "loss": 0.4971,
+      "step": 17904
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 23.261295318603516,
+      "learning_rate": 5.0194536250679835e-06,
+      "loss": 1.2112,
+      "step": 17905
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 10.829798698425293,
+      "learning_rate": 5.0186169100112954e-06,
+      "loss": 1.3173,
+      "step": 17906
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.5662546157836914,
+      "learning_rate": 5.017780194954609e-06,
+      "loss": 0.1248,
+      "step": 17907
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 24.686922073364258,
+      "learning_rate": 5.016943479897921e-06,
+      "loss": 2.8531,
+      "step": 17908
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 19.223573684692383,
+      "learning_rate": 5.016106764841234e-06,
+      "loss": 1.7334,
+      "step": 17909
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 12.839527130126953,
+      "learning_rate": 5.015270049784547e-06,
+      "loss": 0.7207,
+      "step": 17910
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.787776947021484,
+      "learning_rate": 5.014433334727859e-06,
+      "loss": 0.3833,
+      "step": 17911
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.333447217941284,
+      "learning_rate": 5.013596619671171e-06,
+      "loss": 0.1528,
+      "step": 17912
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 19.80486297607422,
+      "learning_rate": 5.012759904614483e-06,
+      "loss": 0.7569,
+      "step": 17913
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 10.29593276977539,
+      "learning_rate": 5.011923189557797e-06,
+      "loss": 0.3161,
+      "step": 17914
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 9.504191398620605,
+      "learning_rate": 5.011086474501109e-06,
+      "loss": 0.5831,
+      "step": 17915
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 16.78968620300293,
+      "learning_rate": 5.010249759444422e-06,
+      "loss": 0.9673,
+      "step": 17916
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 9.07922649383545,
+      "learning_rate": 5.0094130443877346e-06,
+      "loss": 0.533,
+      "step": 17917
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 8.133056640625,
+      "learning_rate": 5.0085763293310465e-06,
+      "loss": 0.5087,
+      "step": 17918
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 60.7887077331543,
+      "learning_rate": 5.007739614274359e-06,
+      "loss": 1.2224,
+      "step": 17919
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 16.089384078979492,
+      "learning_rate": 5.006902899217671e-06,
+      "loss": 2.1645,
+      "step": 17920
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 10.402376174926758,
+      "learning_rate": 5.006066184160985e-06,
+      "loss": 0.9864,
+      "step": 17921
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.572499990463257,
+      "learning_rate": 5.005229469104297e-06,
+      "loss": 0.348,
+      "step": 17922
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 11.185772895812988,
+      "learning_rate": 5.00439275404761e-06,
+      "loss": 1.6682,
+      "step": 17923
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 13.980674743652344,
+      "learning_rate": 5.0035560389909225e-06,
+      "loss": 0.6962,
+      "step": 17924
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 19.253629684448242,
+      "learning_rate": 5.0027193239342344e-06,
+      "loss": 1.5223,
+      "step": 17925
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 9.644561767578125,
+      "learning_rate": 5.001882608877547e-06,
+      "loss": 0.3699,
+      "step": 17926
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 19.637935638427734,
+      "learning_rate": 5.001045893820859e-06,
+      "loss": 0.4528,
+      "step": 17927
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 47.7040901184082,
+      "learning_rate": 5.000209178764173e-06,
+      "loss": 1.1264,
+      "step": 17928
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 6.8891191482543945,
+      "learning_rate": 4.999372463707485e-06,
+      "loss": 0.5625,
+      "step": 17929
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 26.81926155090332,
+      "learning_rate": 4.998535748650798e-06,
+      "loss": 2.1966,
+      "step": 17930
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 29.25107765197754,
+      "learning_rate": 4.99769903359411e-06,
+      "loss": 3.0911,
+      "step": 17931
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.49907922744751,
+      "learning_rate": 4.996862318537422e-06,
+      "loss": 1.1366,
+      "step": 17932
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 7.054262161254883,
+      "learning_rate": 4.996025603480735e-06,
+      "loss": 0.2225,
+      "step": 17933
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 36.04946517944336,
+      "learning_rate": 4.995188888424047e-06,
+      "loss": 1.2687,
+      "step": 17934
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 11.313431739807129,
+      "learning_rate": 4.99435217336736e-06,
+      "loss": 0.4309,
+      "step": 17935
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.431252956390381,
+      "learning_rate": 4.993515458310673e-06,
+      "loss": 0.3986,
+      "step": 17936
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 8.059343338012695,
+      "learning_rate": 4.9926787432539855e-06,
+      "loss": 0.7262,
+      "step": 17937
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 25.366069793701172,
+      "learning_rate": 4.991842028197298e-06,
+      "loss": 0.6772,
+      "step": 17938
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.172663688659668,
+      "learning_rate": 4.99100531314061e-06,
+      "loss": 0.1793,
+      "step": 17939
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 39.05305862426758,
+      "learning_rate": 4.990168598083923e-06,
+      "loss": 0.9276,
+      "step": 17940
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 6.8393964767456055,
+      "learning_rate": 4.989331883027235e-06,
+      "loss": 0.678,
+      "step": 17941
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 11.347168922424316,
+      "learning_rate": 4.988495167970548e-06,
+      "loss": 0.9141,
+      "step": 17942
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 20.5839900970459,
+      "learning_rate": 4.987658452913861e-06,
+      "loss": 1.2812,
+      "step": 17943
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.2079124450683594,
+      "learning_rate": 4.9868217378571734e-06,
+      "loss": 0.2727,
+      "step": 17944
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 25.391883850097656,
+      "learning_rate": 4.985985022800485e-06,
+      "loss": 0.7583,
+      "step": 17945
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 6.606420040130615,
+      "learning_rate": 4.985148307743798e-06,
+      "loss": 0.329,
+      "step": 17946
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 16.52248764038086,
+      "learning_rate": 4.984311592687111e-06,
+      "loss": 0.7837,
+      "step": 17947
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 7.127424240112305,
+      "learning_rate": 4.983474877630423e-06,
+      "loss": 0.43,
+      "step": 17948
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 8.710345268249512,
+      "learning_rate": 4.982638162573736e-06,
+      "loss": 1.8954,
+      "step": 17949
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 11.40097427368164,
+      "learning_rate": 4.9818014475170486e-06,
+      "loss": 0.3691,
+      "step": 17950
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 96.28443145751953,
+      "learning_rate": 4.980964732460361e-06,
+      "loss": 1.2264,
+      "step": 17951
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 57.5449333190918,
+      "learning_rate": 4.980128017403673e-06,
+      "loss": 1.812,
+      "step": 17952
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 24.60512351989746,
+      "learning_rate": 4.979291302346986e-06,
+      "loss": 2.532,
+      "step": 17953
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 18.563039779663086,
+      "learning_rate": 4.978454587290299e-06,
+      "loss": 1.1142,
+      "step": 17954
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 27.25714683532715,
+      "learning_rate": 4.977617872233611e-06,
+      "loss": 0.6761,
+      "step": 17955
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 18.81361961364746,
+      "learning_rate": 4.976781157176924e-06,
+      "loss": 0.9452,
+      "step": 17956
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.5752861499786377,
+      "learning_rate": 4.9759444421202365e-06,
+      "loss": 0.4309,
+      "step": 17957
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 10.082378387451172,
+      "learning_rate": 4.975107727063549e-06,
+      "loss": 0.7022,
+      "step": 17958
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 22.070344924926758,
+      "learning_rate": 4.974271012006861e-06,
+      "loss": 1.0242,
+      "step": 17959
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 28.86927604675293,
+      "learning_rate": 4.973434296950174e-06,
+      "loss": 2.25,
+      "step": 17960
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 13.019451141357422,
+      "learning_rate": 4.972597581893487e-06,
+      "loss": 0.7717,
+      "step": 17961
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 10.226906776428223,
+      "learning_rate": 4.971760866836799e-06,
+      "loss": 1.1278,
+      "step": 17962
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 8.096986770629883,
+      "learning_rate": 4.970924151780112e-06,
+      "loss": 0.7276,
+      "step": 17963
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.092785596847534,
+      "learning_rate": 4.970087436723424e-06,
+      "loss": 0.0536,
+      "step": 17964
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 10.456246376037598,
+      "learning_rate": 4.969250721666737e-06,
+      "loss": 0.5507,
+      "step": 17965
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 17.064895629882812,
+      "learning_rate": 4.968414006610049e-06,
+      "loss": 1.1995,
+      "step": 17966
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 113.00723266601562,
+      "learning_rate": 4.967577291553362e-06,
+      "loss": 1.3712,
+      "step": 17967
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 10.568375587463379,
+      "learning_rate": 4.966740576496675e-06,
+      "loss": 0.4062,
+      "step": 17968
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 12.843826293945312,
+      "learning_rate": 4.965903861439987e-06,
+      "loss": 1.7527,
+      "step": 17969
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 18.707082748413086,
+      "learning_rate": 4.9650671463832995e-06,
+      "loss": 0.6466,
+      "step": 17970
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 8.290353775024414,
+      "learning_rate": 4.964230431326612e-06,
+      "loss": 0.3877,
+      "step": 17971
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 7.575691223144531,
+      "learning_rate": 4.963393716269925e-06,
+      "loss": 0.4287,
+      "step": 17972
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.154853820800781,
+      "learning_rate": 4.962557001213237e-06,
+      "loss": 0.2588,
+      "step": 17973
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 23.596227645874023,
+      "learning_rate": 4.96172028615655e-06,
+      "loss": 1.1944,
+      "step": 17974
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 54.577293395996094,
+      "learning_rate": 4.960883571099862e-06,
+      "loss": 2.0033,
+      "step": 17975
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 17.267406463623047,
+      "learning_rate": 4.960046856043175e-06,
+      "loss": 1.2169,
+      "step": 17976
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 8.010673522949219,
+      "learning_rate": 4.9592101409864874e-06,
+      "loss": 0.35,
+      "step": 17977
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 22.85257911682129,
+      "learning_rate": 4.9583734259298e-06,
+      "loss": 1.5214,
+      "step": 17978
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 14.447754859924316,
+      "learning_rate": 4.957536710873113e-06,
+      "loss": 0.3605,
+      "step": 17979
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 18.63277244567871,
+      "learning_rate": 4.956699995816425e-06,
+      "loss": 0.969,
+      "step": 17980
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 7.11646032333374,
+      "learning_rate": 4.955863280759738e-06,
+      "loss": 0.2212,
+      "step": 17981
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 167.06617736816406,
+      "learning_rate": 4.95502656570305e-06,
+      "loss": 0.998,
+      "step": 17982
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 17.073394775390625,
+      "learning_rate": 4.9541898506463625e-06,
+      "loss": 2.1347,
+      "step": 17983
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 31.762836456298828,
+      "learning_rate": 4.953353135589675e-06,
+      "loss": 1.343,
+      "step": 17984
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 7.8832173347473145,
+      "learning_rate": 4.952516420532988e-06,
+      "loss": 0.7458,
+      "step": 17985
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.58601713180542,
+      "learning_rate": 4.9516797054763e-06,
+      "loss": 0.5316,
+      "step": 17986
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 29.055683135986328,
+      "learning_rate": 4.950842990419613e-06,
+      "loss": 0.6991,
+      "step": 17987
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.409537315368652,
+      "learning_rate": 4.950006275362926e-06,
+      "loss": 0.2481,
+      "step": 17988
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 10.621453285217285,
+      "learning_rate": 4.949169560306238e-06,
+      "loss": 1.452,
+      "step": 17989
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 174.6796112060547,
+      "learning_rate": 4.9483328452495505e-06,
+      "loss": 1.3904,
+      "step": 17990
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 14.097359657287598,
+      "learning_rate": 4.947496130192863e-06,
+      "loss": 1.0402,
+      "step": 17991
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 18.349971771240234,
+      "learning_rate": 4.946659415136176e-06,
+      "loss": 1.0192,
+      "step": 17992
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 12.767474174499512,
+      "learning_rate": 4.945822700079488e-06,
+      "loss": 0.7705,
+      "step": 17993
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 14.867411613464355,
+      "learning_rate": 4.944985985022801e-06,
+      "loss": 1.523,
+      "step": 17994
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 97.9448471069336,
+      "learning_rate": 4.944149269966114e-06,
+      "loss": 2.2483,
+      "step": 17995
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 8.423328399658203,
+      "learning_rate": 4.943312554909426e-06,
+      "loss": 0.2358,
+      "step": 17996
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 13.076930046081543,
+      "learning_rate": 4.942475839852738e-06,
+      "loss": 0.1316,
+      "step": 17997
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 13.969695091247559,
+      "learning_rate": 4.941639124796051e-06,
+      "loss": 1.0452,
+      "step": 17998
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.603994846343994,
+      "learning_rate": 4.940802409739364e-06,
+      "loss": 0.2369,
+      "step": 17999
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 25.964096069335938,
+      "learning_rate": 4.939965694682676e-06,
+      "loss": 1.8896,
+      "step": 18000
+    },
+    {
+      "epoch": 2.26,
+      "eval_loss": 0.07683395594358444,
+      "eval_runtime": 94.3131,
+      "eval_samples_per_second": 37.556,
+      "eval_steps_per_second": 37.556,
+      "step": 18000
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 11.208454132080078,
+      "learning_rate": 4.939128979625989e-06,
+      "loss": 0.6867,
+      "step": 18001
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 17.421865463256836,
+      "learning_rate": 4.9382922645693015e-06,
+      "loss": 1.2619,
+      "step": 18002
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 10.385871887207031,
+      "learning_rate": 4.9374555495126135e-06,
+      "loss": 0.9416,
+      "step": 18003
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 12.817423820495605,
+      "learning_rate": 4.936618834455926e-06,
+      "loss": 2.2117,
+      "step": 18004
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 35.08500671386719,
+      "learning_rate": 4.935782119399239e-06,
+      "loss": 2.569,
+      "step": 18005
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 22.860774993896484,
+      "learning_rate": 4.934945404342552e-06,
+      "loss": 0.8472,
+      "step": 18006
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 7.265113830566406,
+      "learning_rate": 4.934108689285864e-06,
+      "loss": 2.1218,
+      "step": 18007
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 11.792815208435059,
+      "learning_rate": 4.933271974229177e-06,
+      "loss": 1.3434,
+      "step": 18008
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.4329118728637695,
+      "learning_rate": 4.9324352591724895e-06,
+      "loss": 0.4935,
+      "step": 18009
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 26.49026107788086,
+      "learning_rate": 4.931598544115801e-06,
+      "loss": 1.1913,
+      "step": 18010
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 20.838491439819336,
+      "learning_rate": 4.930761829059114e-06,
+      "loss": 1.0779,
+      "step": 18011
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.645506858825684,
+      "learning_rate": 4.929925114002427e-06,
+      "loss": 0.3584,
+      "step": 18012
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.891331195831299,
+      "learning_rate": 4.92908839894574e-06,
+      "loss": 0.1227,
+      "step": 18013
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.6347407102584839,
+      "learning_rate": 4.928251683889052e-06,
+      "loss": 0.0455,
+      "step": 18014
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 16.91248321533203,
+      "learning_rate": 4.927414968832365e-06,
+      "loss": 0.8666,
+      "step": 18015
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 33.073455810546875,
+      "learning_rate": 4.926578253775677e-06,
+      "loss": 1.8501,
+      "step": 18016
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 14.698752403259277,
+      "learning_rate": 4.925741538718989e-06,
+      "loss": 0.5226,
+      "step": 18017
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 41.15288162231445,
+      "learning_rate": 4.924904823662302e-06,
+      "loss": 0.4485,
+      "step": 18018
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 33.41960525512695,
+      "learning_rate": 4.924068108605615e-06,
+      "loss": 0.9904,
+      "step": 18019
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 28.272079467773438,
+      "learning_rate": 4.923231393548928e-06,
+      "loss": 0.3973,
+      "step": 18020
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 12.279377937316895,
+      "learning_rate": 4.92239467849224e-06,
+      "loss": 0.4734,
+      "step": 18021
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 13.807548522949219,
+      "learning_rate": 4.9215579634355525e-06,
+      "loss": 0.1609,
+      "step": 18022
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.790491580963135,
+      "learning_rate": 4.920721248378865e-06,
+      "loss": 0.2628,
+      "step": 18023
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 30.30073356628418,
+      "learning_rate": 4.919884533322177e-06,
+      "loss": 1.6022,
+      "step": 18024
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 21.043136596679688,
+      "learning_rate": 4.91904781826549e-06,
+      "loss": 2.1917,
+      "step": 18025
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 15.728610038757324,
+      "learning_rate": 4.918211103208803e-06,
+      "loss": 0.8905,
+      "step": 18026
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 9.309231758117676,
+      "learning_rate": 4.917374388152115e-06,
+      "loss": 0.5748,
+      "step": 18027
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 18.681480407714844,
+      "learning_rate": 4.916537673095428e-06,
+      "loss": 1.3546,
+      "step": 18028
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 9.547511100769043,
+      "learning_rate": 4.91570095803874e-06,
+      "loss": 0.7077,
+      "step": 18029
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 16.273372650146484,
+      "learning_rate": 4.914864242982053e-06,
+      "loss": 1.1192,
+      "step": 18030
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 17.370267868041992,
+      "learning_rate": 4.914027527925365e-06,
+      "loss": 0.7352,
+      "step": 18031
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 7.943282127380371,
+      "learning_rate": 4.913190812868678e-06,
+      "loss": 1.6347,
+      "step": 18032
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 11.803362846374512,
+      "learning_rate": 4.912354097811991e-06,
+      "loss": 0.6748,
+      "step": 18033
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 11.606377601623535,
+      "learning_rate": 4.911517382755303e-06,
+      "loss": 1.2015,
+      "step": 18034
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 36.61336135864258,
+      "learning_rate": 4.9106806676986155e-06,
+      "loss": 1.6687,
+      "step": 18035
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 43.68743896484375,
+      "learning_rate": 4.909843952641928e-06,
+      "loss": 1.9454,
+      "step": 18036
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 18.23581886291504,
+      "learning_rate": 4.909007237585241e-06,
+      "loss": 0.4798,
+      "step": 18037
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 15.893588066101074,
+      "learning_rate": 4.908170522528553e-06,
+      "loss": 1.0806,
+      "step": 18038
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 18.41132164001465,
+      "learning_rate": 4.907333807471866e-06,
+      "loss": 1.9558,
+      "step": 18039
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 17.931055068969727,
+      "learning_rate": 4.906497092415179e-06,
+      "loss": 0.9484,
+      "step": 18040
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 15.66031265258789,
+      "learning_rate": 4.905660377358491e-06,
+      "loss": 0.4607,
+      "step": 18041
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 14.757575035095215,
+      "learning_rate": 4.9048236623018034e-06,
+      "loss": 1.4596,
+      "step": 18042
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 17.42569351196289,
+      "learning_rate": 4.903986947245116e-06,
+      "loss": 1.3227,
+      "step": 18043
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 28.305334091186523,
+      "learning_rate": 4.903150232188429e-06,
+      "loss": 1.1253,
+      "step": 18044
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 15.629451751708984,
+      "learning_rate": 4.902313517131741e-06,
+      "loss": 1.5697,
+      "step": 18045
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 11.656510353088379,
+      "learning_rate": 4.901476802075054e-06,
+      "loss": 1.5753,
+      "step": 18046
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 12.482386589050293,
+      "learning_rate": 4.900640087018367e-06,
+      "loss": 0.7014,
+      "step": 18047
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 7.773298263549805,
+      "learning_rate": 4.8998033719616786e-06,
+      "loss": 0.4116,
+      "step": 18048
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 9.567391395568848,
+      "learning_rate": 4.898966656904991e-06,
+      "loss": 0.2018,
+      "step": 18049
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 724.5193481445312,
+      "learning_rate": 4.898129941848304e-06,
+      "loss": 0.6798,
+      "step": 18050
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 6.985085487365723,
+      "learning_rate": 4.897293226791617e-06,
+      "loss": 0.6674,
+      "step": 18051
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 20.335052490234375,
+      "learning_rate": 4.896456511734929e-06,
+      "loss": 0.6555,
+      "step": 18052
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 16.140939712524414,
+      "learning_rate": 4.895619796678242e-06,
+      "loss": 1.3869,
+      "step": 18053
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 13.487977027893066,
+      "learning_rate": 4.8947830816215545e-06,
+      "loss": 0.9246,
+      "step": 18054
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 31.696489334106445,
+      "learning_rate": 4.8939463665648665e-06,
+      "loss": 1.5,
+      "step": 18055
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 7.162998676300049,
+      "learning_rate": 4.893109651508179e-06,
+      "loss": 1.1,
+      "step": 18056
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 17.811870574951172,
+      "learning_rate": 4.892272936451492e-06,
+      "loss": 1.477,
+      "step": 18057
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 14.948601722717285,
+      "learning_rate": 4.891436221394805e-06,
+      "loss": 0.6962,
+      "step": 18058
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 20.97901725769043,
+      "learning_rate": 4.890599506338117e-06,
+      "loss": 2.3629,
+      "step": 18059
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 38.151309967041016,
+      "learning_rate": 4.88976279128143e-06,
+      "loss": 0.5859,
+      "step": 18060
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 29.360761642456055,
+      "learning_rate": 4.888926076224742e-06,
+      "loss": 2.9119,
+      "step": 18061
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.248826503753662,
+      "learning_rate": 4.888089361168054e-06,
+      "loss": 0.2653,
+      "step": 18062
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 18.261001586914062,
+      "learning_rate": 4.887252646111367e-06,
+      "loss": 1.687,
+      "step": 18063
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.85291862487793,
+      "learning_rate": 4.88641593105468e-06,
+      "loss": 0.2157,
+      "step": 18064
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 8.904922485351562,
+      "learning_rate": 4.885579215997992e-06,
+      "loss": 0.2832,
+      "step": 18065
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 36.32822799682617,
+      "learning_rate": 4.884742500941305e-06,
+      "loss": 1.5748,
+      "step": 18066
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 7.538064479827881,
+      "learning_rate": 4.8839057858846176e-06,
+      "loss": 1.3308,
+      "step": 18067
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 6.013849258422852,
+      "learning_rate": 4.8830690708279295e-06,
+      "loss": 0.3816,
+      "step": 18068
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 53.89246368408203,
+      "learning_rate": 4.882232355771242e-06,
+      "loss": 4.6914,
+      "step": 18069
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 8.806126594543457,
+      "learning_rate": 4.881395640714555e-06,
+      "loss": 0.7824,
+      "step": 18070
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 49.080078125,
+      "learning_rate": 4.880558925657868e-06,
+      "loss": 1.7653,
+      "step": 18071
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 51.791839599609375,
+      "learning_rate": 4.87972221060118e-06,
+      "loss": 1.6891,
+      "step": 18072
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 18.03871726989746,
+      "learning_rate": 4.878885495544493e-06,
+      "loss": 1.208,
+      "step": 18073
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 12.533004760742188,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 2.1779,
+      "step": 18074
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 13.060806274414062,
+      "learning_rate": 4.8772120654311174e-06,
+      "loss": 2.3596,
+      "step": 18075
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 83.15618896484375,
+      "learning_rate": 4.87637535037443e-06,
+      "loss": 0.9586,
+      "step": 18076
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 48.380252838134766,
+      "learning_rate": 4.875538635317743e-06,
+      "loss": 1.1134,
+      "step": 18077
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 11.484421730041504,
+      "learning_rate": 4.874701920261056e-06,
+      "loss": 0.6175,
+      "step": 18078
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.373561382293701,
+      "learning_rate": 4.873865205204368e-06,
+      "loss": 0.5831,
+      "step": 18079
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 140.86215209960938,
+      "learning_rate": 4.873028490147681e-06,
+      "loss": 1.3971,
+      "step": 18080
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 27.262065887451172,
+      "learning_rate": 4.872191775090993e-06,
+      "loss": 1.3046,
+      "step": 18081
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 17.172199249267578,
+      "learning_rate": 4.871355060034305e-06,
+      "loss": 0.6865,
+      "step": 18082
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 13.459043502807617,
+      "learning_rate": 4.870518344977618e-06,
+      "loss": 0.4893,
+      "step": 18083
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 13.085651397705078,
+      "learning_rate": 4.869681629920931e-06,
+      "loss": 1.2103,
+      "step": 18084
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 14.714868545532227,
+      "learning_rate": 4.868844914864244e-06,
+      "loss": 0.2914,
+      "step": 18085
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 13.607343673706055,
+      "learning_rate": 4.868008199807556e-06,
+      "loss": 0.7138,
+      "step": 18086
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 11.629634857177734,
+      "learning_rate": 4.8671714847508685e-06,
+      "loss": 0.7351,
+      "step": 18087
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 6.835043907165527,
+      "learning_rate": 4.866334769694181e-06,
+      "loss": 0.704,
+      "step": 18088
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 11.061873435974121,
+      "learning_rate": 4.865498054637493e-06,
+      "loss": 1.0773,
+      "step": 18089
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 10.615815162658691,
+      "learning_rate": 4.864661339580806e-06,
+      "loss": 1.0865,
+      "step": 18090
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 23.319171905517578,
+      "learning_rate": 4.863824624524119e-06,
+      "loss": 2.2188,
+      "step": 18091
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.291536569595337,
+      "learning_rate": 4.862987909467432e-06,
+      "loss": 0.1462,
+      "step": 18092
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 18.689645767211914,
+      "learning_rate": 4.862151194410744e-06,
+      "loss": 0.9393,
+      "step": 18093
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 11.912179946899414,
+      "learning_rate": 4.8613144793540564e-06,
+      "loss": 1.312,
+      "step": 18094
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 11.882221221923828,
+      "learning_rate": 4.860477764297369e-06,
+      "loss": 0.436,
+      "step": 18095
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 31.453899383544922,
+      "learning_rate": 4.859641049240681e-06,
+      "loss": 3.0691,
+      "step": 18096
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 39.959293365478516,
+      "learning_rate": 4.858804334183994e-06,
+      "loss": 2.2514,
+      "step": 18097
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 11.969451904296875,
+      "learning_rate": 4.857967619127307e-06,
+      "loss": 0.5848,
+      "step": 18098
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 19.55484390258789,
+      "learning_rate": 4.85713090407062e-06,
+      "loss": 1.8364,
+      "step": 18099
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 27.38494300842285,
+      "learning_rate": 4.8562941890139316e-06,
+      "loss": 1.1832,
+      "step": 18100
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 24.023799896240234,
+      "learning_rate": 4.855457473957244e-06,
+      "loss": 1.5347,
+      "step": 18101
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.8433549404144287,
+      "learning_rate": 4.854620758900556e-06,
+      "loss": 0.0844,
+      "step": 18102
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 7.295028209686279,
+      "learning_rate": 4.853784043843869e-06,
+      "loss": 2.2588,
+      "step": 18103
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 18.740617752075195,
+      "learning_rate": 4.852947328787182e-06,
+      "loss": 1.6035,
+      "step": 18104
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 16.479990005493164,
+      "learning_rate": 4.852110613730495e-06,
+      "loss": 1.7165,
+      "step": 18105
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 11.476269721984863,
+      "learning_rate": 4.8512738986738075e-06,
+      "loss": 0.7912,
+      "step": 18106
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 17.679611206054688,
+      "learning_rate": 4.8504371836171195e-06,
+      "loss": 1.3959,
+      "step": 18107
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 20.310611724853516,
+      "learning_rate": 4.849600468560432e-06,
+      "loss": 2.0444,
+      "step": 18108
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 6.861479759216309,
+      "learning_rate": 4.848763753503744e-06,
+      "loss": 0.1372,
+      "step": 18109
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 8.993551254272461,
+      "learning_rate": 4.847927038447057e-06,
+      "loss": 2.1999,
+      "step": 18110
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 12.034762382507324,
+      "learning_rate": 4.84709032339037e-06,
+      "loss": 0.7087,
+      "step": 18111
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.3666305541992188,
+      "learning_rate": 4.846253608333683e-06,
+      "loss": 0.0713,
+      "step": 18112
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 18.19227409362793,
+      "learning_rate": 4.8454168932769954e-06,
+      "loss": 1.681,
+      "step": 18113
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 13.807527542114258,
+      "learning_rate": 4.844580178220307e-06,
+      "loss": 0.3026,
+      "step": 18114
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 22.87261199951172,
+      "learning_rate": 4.84374346316362e-06,
+      "loss": 0.8078,
+      "step": 18115
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 657.290283203125,
+      "learning_rate": 4.842906748106932e-06,
+      "loss": 1.5545,
+      "step": 18116
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 12.585835456848145,
+      "learning_rate": 4.842070033050245e-06,
+      "loss": 1.2678,
+      "step": 18117
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 24.29072380065918,
+      "learning_rate": 4.841233317993558e-06,
+      "loss": 1.8188,
+      "step": 18118
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 9.583120346069336,
+      "learning_rate": 4.8403966029368706e-06,
+      "loss": 0.9413,
+      "step": 18119
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 21.73714256286621,
+      "learning_rate": 4.839559887880183e-06,
+      "loss": 2.7989,
+      "step": 18120
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.3437893390655518,
+      "learning_rate": 4.838723172823495e-06,
+      "loss": 0.0507,
+      "step": 18121
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 41.34368896484375,
+      "learning_rate": 4.837886457766808e-06,
+      "loss": 0.5975,
+      "step": 18122
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 29.665691375732422,
+      "learning_rate": 4.83704974271012e-06,
+      "loss": 2.3017,
+      "step": 18123
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.171306133270264,
+      "learning_rate": 4.836213027653433e-06,
+      "loss": 0.8568,
+      "step": 18124
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 8.918929100036621,
+      "learning_rate": 4.835376312596746e-06,
+      "loss": 0.4179,
+      "step": 18125
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 100.70255279541016,
+      "learning_rate": 4.8345395975400585e-06,
+      "loss": 1.7599,
+      "step": 18126
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 19.132160186767578,
+      "learning_rate": 4.833702882483371e-06,
+      "loss": 2.4708,
+      "step": 18127
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.666585445404053,
+      "learning_rate": 4.832866167426683e-06,
+      "loss": 0.2085,
+      "step": 18128
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 19.80629539489746,
+      "learning_rate": 4.832029452369996e-06,
+      "loss": 0.4924,
+      "step": 18129
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 58.904666900634766,
+      "learning_rate": 4.831192737313308e-06,
+      "loss": 0.7678,
+      "step": 18130
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 10.534483909606934,
+      "learning_rate": 4.830356022256621e-06,
+      "loss": 0.9355,
+      "step": 18131
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 18.848360061645508,
+      "learning_rate": 4.829519307199934e-06,
+      "loss": 0.8069,
+      "step": 18132
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 14.515966415405273,
+      "learning_rate": 4.828682592143246e-06,
+      "loss": 1.9142,
+      "step": 18133
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 17.976545333862305,
+      "learning_rate": 4.827845877086559e-06,
+      "loss": 2.0127,
+      "step": 18134
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 31.835487365722656,
+      "learning_rate": 4.827009162029871e-06,
+      "loss": 1.7525,
+      "step": 18135
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 94.18576049804688,
+      "learning_rate": 4.826172446973184e-06,
+      "loss": 3.045,
+      "step": 18136
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 12.619370460510254,
+      "learning_rate": 4.825335731916496e-06,
+      "loss": 0.5362,
+      "step": 18137
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.5252768993377686,
+      "learning_rate": 4.824499016859809e-06,
+      "loss": 0.1731,
+      "step": 18138
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 6.68712043762207,
+      "learning_rate": 4.8236623018031215e-06,
+      "loss": 1.0455,
+      "step": 18139
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.003234386444092,
+      "learning_rate": 4.822825586746434e-06,
+      "loss": 0.396,
+      "step": 18140
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 7.79044246673584,
+      "learning_rate": 4.821988871689747e-06,
+      "loss": 0.9295,
+      "step": 18141
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 26.89201545715332,
+      "learning_rate": 4.821152156633059e-06,
+      "loss": 1.0806,
+      "step": 18142
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 15.89769172668457,
+      "learning_rate": 4.820315441576371e-06,
+      "loss": 1.8888,
+      "step": 18143
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.993673324584961,
+      "learning_rate": 4.819478726519684e-06,
+      "loss": 2.3131,
+      "step": 18144
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 8.432538032531738,
+      "learning_rate": 4.818642011462997e-06,
+      "loss": 0.524,
+      "step": 18145
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.847290515899658,
+      "learning_rate": 4.8178052964063094e-06,
+      "loss": 0.535,
+      "step": 18146
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 15.385242462158203,
+      "learning_rate": 4.816968581349622e-06,
+      "loss": 1.5261,
+      "step": 18147
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 16.392929077148438,
+      "learning_rate": 4.816131866292935e-06,
+      "loss": 1.4277,
+      "step": 18148
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 10.277228355407715,
+      "learning_rate": 4.815295151236247e-06,
+      "loss": 0.6431,
+      "step": 18149
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 13.693373680114746,
+      "learning_rate": 4.814458436179559e-06,
+      "loss": 1.8388,
+      "step": 18150
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 9.805331230163574,
+      "learning_rate": 4.813621721122872e-06,
+      "loss": 0.3037,
+      "step": 18151
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 17.375608444213867,
+      "learning_rate": 4.8127850060661845e-06,
+      "loss": 0.5494,
+      "step": 18152
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.424117088317871,
+      "learning_rate": 4.811948291009497e-06,
+      "loss": 0.2041,
+      "step": 18153
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 14.018604278564453,
+      "learning_rate": 4.81111157595281e-06,
+      "loss": 1.1859,
+      "step": 18154
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 7.766754150390625,
+      "learning_rate": 4.810274860896122e-06,
+      "loss": 0.4826,
+      "step": 18155
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 6.565408229827881,
+      "learning_rate": 4.809438145839435e-06,
+      "loss": 0.5125,
+      "step": 18156
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 40.847049713134766,
+      "learning_rate": 4.808601430782747e-06,
+      "loss": 2.2631,
+      "step": 18157
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 148.59217834472656,
+      "learning_rate": 4.80776471572606e-06,
+      "loss": 2.0997,
+      "step": 18158
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 9.460793495178223,
+      "learning_rate": 4.8069280006693725e-06,
+      "loss": 0.4907,
+      "step": 18159
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.744081020355225,
+      "learning_rate": 4.806091285612685e-06,
+      "loss": 0.2731,
+      "step": 18160
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 27.180835723876953,
+      "learning_rate": 4.805254570555998e-06,
+      "loss": 1.5487,
+      "step": 18161
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 22.261077880859375,
+      "learning_rate": 4.80441785549931e-06,
+      "loss": 1.4461,
+      "step": 18162
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 13.491981506347656,
+      "learning_rate": 4.803581140442623e-06,
+      "loss": 1.6574,
+      "step": 18163
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 18.313154220581055,
+      "learning_rate": 4.802744425385935e-06,
+      "loss": 0.9745,
+      "step": 18164
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 15.0607328414917,
+      "learning_rate": 4.801907710329248e-06,
+      "loss": 0.5434,
+      "step": 18165
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 22.887287139892578,
+      "learning_rate": 4.80107099527256e-06,
+      "loss": 0.5663,
+      "step": 18166
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 19.480941772460938,
+      "learning_rate": 4.800234280215873e-06,
+      "loss": 0.8002,
+      "step": 18167
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 25.959819793701172,
+      "learning_rate": 4.799397565159186e-06,
+      "loss": 1.2098,
+      "step": 18168
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 17.326189041137695,
+      "learning_rate": 4.798560850102498e-06,
+      "loss": 1.2297,
+      "step": 18169
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 11.509465217590332,
+      "learning_rate": 4.797724135045811e-06,
+      "loss": 1.7532,
+      "step": 18170
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 6.875067234039307,
+      "learning_rate": 4.796887419989123e-06,
+      "loss": 1.0477,
+      "step": 18171
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 14.742862701416016,
+      "learning_rate": 4.7960507049324355e-06,
+      "loss": 0.5861,
+      "step": 18172
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 6.9289679527282715,
+      "learning_rate": 4.795213989875748e-06,
+      "loss": 0.2185,
+      "step": 18173
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 16.963138580322266,
+      "learning_rate": 4.794377274819061e-06,
+      "loss": 0.7712,
+      "step": 18174
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 13.129985809326172,
+      "learning_rate": 4.793540559762374e-06,
+      "loss": 0.7799,
+      "step": 18175
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 15.65756607055664,
+      "learning_rate": 4.792703844705686e-06,
+      "loss": 1.0235,
+      "step": 18176
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 23.486530303955078,
+      "learning_rate": 4.791867129648999e-06,
+      "loss": 0.6155,
+      "step": 18177
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 17.494380950927734,
+      "learning_rate": 4.791030414592311e-06,
+      "loss": 0.9875,
+      "step": 18178
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 12.82445240020752,
+      "learning_rate": 4.790193699535623e-06,
+      "loss": 0.5564,
+      "step": 18179
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 12.545409202575684,
+      "learning_rate": 4.789356984478936e-06,
+      "loss": 0.9441,
+      "step": 18180
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 10.12964916229248,
+      "learning_rate": 4.788520269422249e-06,
+      "loss": 0.5128,
+      "step": 18181
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 11.175071716308594,
+      "learning_rate": 4.787683554365562e-06,
+      "loss": 0.5869,
+      "step": 18182
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 10.429204940795898,
+      "learning_rate": 4.786846839308874e-06,
+      "loss": 0.9771,
+      "step": 18183
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 7.880436897277832,
+      "learning_rate": 4.786010124252186e-06,
+      "loss": 0.5035,
+      "step": 18184
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 6.829651355743408,
+      "learning_rate": 4.7851734091954985e-06,
+      "loss": 0.282,
+      "step": 18185
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 9.256832122802734,
+      "learning_rate": 4.784336694138811e-06,
+      "loss": 0.9679,
+      "step": 18186
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 82.81836700439453,
+      "learning_rate": 4.783499979082124e-06,
+      "loss": 2.0453,
+      "step": 18187
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 13.147477149963379,
+      "learning_rate": 4.782663264025437e-06,
+      "loss": 0.5534,
+      "step": 18188
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 11.433621406555176,
+      "learning_rate": 4.78182654896875e-06,
+      "loss": 0.9656,
+      "step": 18189
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 20.815425872802734,
+      "learning_rate": 4.780989833912062e-06,
+      "loss": 1.3373,
+      "step": 18190
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 29.052854537963867,
+      "learning_rate": 4.780153118855374e-06,
+      "loss": 0.7903,
+      "step": 18191
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 14.508295059204102,
+      "learning_rate": 4.7793164037986864e-06,
+      "loss": 0.673,
+      "step": 18192
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 17.395936965942383,
+      "learning_rate": 4.778479688741999e-06,
+      "loss": 0.8457,
+      "step": 18193
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 17.987634658813477,
+      "learning_rate": 4.777642973685312e-06,
+      "loss": 0.7368,
+      "step": 18194
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 20.08308982849121,
+      "learning_rate": 4.776806258628625e-06,
+      "loss": 2.3677,
+      "step": 18195
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 9.142012596130371,
+      "learning_rate": 4.775969543571937e-06,
+      "loss": 1.2506,
+      "step": 18196
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 21.252647399902344,
+      "learning_rate": 4.77513282851525e-06,
+      "loss": 1.0209,
+      "step": 18197
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 18.741111755371094,
+      "learning_rate": 4.7742961134585616e-06,
+      "loss": 0.4748,
+      "step": 18198
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.1713852882385254,
+      "learning_rate": 4.773459398401874e-06,
+      "loss": 0.1406,
+      "step": 18199
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 13.61368465423584,
+      "learning_rate": 4.772622683345187e-06,
+      "loss": 0.6536,
+      "step": 18200
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 13.201034545898438,
+      "learning_rate": 4.7717859682885e-06,
+      "loss": 0.8234,
+      "step": 18201
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 12.031088829040527,
+      "learning_rate": 4.770949253231813e-06,
+      "loss": 1.6413,
+      "step": 18202
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 31.76029396057129,
+      "learning_rate": 4.770112538175125e-06,
+      "loss": 3.4188,
+      "step": 18203
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 11.155171394348145,
+      "learning_rate": 4.7692758231184375e-06,
+      "loss": 1.4523,
+      "step": 18204
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 67.39051818847656,
+      "learning_rate": 4.7684391080617495e-06,
+      "loss": 1.882,
+      "step": 18205
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 19.147981643676758,
+      "learning_rate": 4.767602393005062e-06,
+      "loss": 1.1568,
+      "step": 18206
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 6.553123950958252,
+      "learning_rate": 4.766765677948375e-06,
+      "loss": 1.0164,
+      "step": 18207
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 12.664617538452148,
+      "learning_rate": 4.765928962891688e-06,
+      "loss": 0.9444,
+      "step": 18208
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 16.09284210205078,
+      "learning_rate": 4.765092247835001e-06,
+      "loss": 0.5439,
+      "step": 18209
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 8.732881546020508,
+      "learning_rate": 4.764255532778313e-06,
+      "loss": 0.7299,
+      "step": 18210
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 45.31431579589844,
+      "learning_rate": 4.7634188177216254e-06,
+      "loss": 1.6899,
+      "step": 18211
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 24.906145095825195,
+      "learning_rate": 4.762582102664937e-06,
+      "loss": 1.6587,
+      "step": 18212
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 16.11197280883789,
+      "learning_rate": 4.76174538760825e-06,
+      "loss": 1.0412,
+      "step": 18213
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 9.287578582763672,
+      "learning_rate": 4.760908672551563e-06,
+      "loss": 1.0026,
+      "step": 18214
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 16.759666442871094,
+      "learning_rate": 4.760071957494876e-06,
+      "loss": 0.8681,
+      "step": 18215
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 20.16606903076172,
+      "learning_rate": 4.759235242438189e-06,
+      "loss": 0.8982,
+      "step": 18216
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 9.273433685302734,
+      "learning_rate": 4.7583985273815006e-06,
+      "loss": 1.404,
+      "step": 18217
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 34.79890441894531,
+      "learning_rate": 4.757561812324813e-06,
+      "loss": 3.0047,
+      "step": 18218
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 31.379962921142578,
+      "learning_rate": 4.756725097268125e-06,
+      "loss": 0.9652,
+      "step": 18219
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 19.541475296020508,
+      "learning_rate": 4.755888382211438e-06,
+      "loss": 0.7403,
+      "step": 18220
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 15.139787673950195,
+      "learning_rate": 4.755051667154751e-06,
+      "loss": 0.4795,
+      "step": 18221
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 6.544916152954102,
+      "learning_rate": 4.754214952098064e-06,
+      "loss": 0.9741,
+      "step": 18222
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 30.706077575683594,
+      "learning_rate": 4.7533782370413765e-06,
+      "loss": 1.6226,
+      "step": 18223
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 7.352999210357666,
+      "learning_rate": 4.7525415219846885e-06,
+      "loss": 0.7358,
+      "step": 18224
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 10.688814163208008,
+      "learning_rate": 4.7517048069280004e-06,
+      "loss": 1.2113,
+      "step": 18225
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 6.308061599731445,
+      "learning_rate": 4.750868091871313e-06,
+      "loss": 0.3193,
+      "step": 18226
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 18.884119033813477,
+      "learning_rate": 4.750031376814626e-06,
+      "loss": 1.5468,
+      "step": 18227
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 18.070905685424805,
+      "learning_rate": 4.749194661757939e-06,
+      "loss": 1.1872,
+      "step": 18228
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 6.125032901763916,
+      "learning_rate": 4.748357946701252e-06,
+      "loss": 1.0369,
+      "step": 18229
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 8.317168235778809,
+      "learning_rate": 4.747521231644564e-06,
+      "loss": 0.38,
+      "step": 18230
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.549513816833496,
+      "learning_rate": 4.746684516587876e-06,
+      "loss": 0.5862,
+      "step": 18231
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 6.371419906616211,
+      "learning_rate": 4.745847801531188e-06,
+      "loss": 0.5998,
+      "step": 18232
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 15.010042190551758,
+      "learning_rate": 4.745011086474501e-06,
+      "loss": 0.6881,
+      "step": 18233
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 32.16387939453125,
+      "learning_rate": 4.744174371417814e-06,
+      "loss": 1.2754,
+      "step": 18234
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.978853225708008,
+      "learning_rate": 4.743337656361127e-06,
+      "loss": 0.3576,
+      "step": 18235
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 16.613771438598633,
+      "learning_rate": 4.7425009413044396e-06,
+      "loss": 1.196,
+      "step": 18236
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 13.595804214477539,
+      "learning_rate": 4.7416642262477515e-06,
+      "loss": 0.8008,
+      "step": 18237
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 92.11951446533203,
+      "learning_rate": 4.740827511191064e-06,
+      "loss": 1.0349,
+      "step": 18238
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 6.98962926864624,
+      "learning_rate": 4.739990796134376e-06,
+      "loss": 0.2019,
+      "step": 18239
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 6.2486162185668945,
+      "learning_rate": 4.739154081077689e-06,
+      "loss": 0.2373,
+      "step": 18240
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 10.029749870300293,
+      "learning_rate": 4.738317366021002e-06,
+      "loss": 1.55,
+      "step": 18241
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 18.082401275634766,
+      "learning_rate": 4.737480650964315e-06,
+      "loss": 0.5872,
+      "step": 18242
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 18.084856033325195,
+      "learning_rate": 4.7366439359076275e-06,
+      "loss": 1.0304,
+      "step": 18243
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 11.300666809082031,
+      "learning_rate": 4.7358072208509394e-06,
+      "loss": 0.6484,
+      "step": 18244
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 8.962103843688965,
+      "learning_rate": 4.734970505794252e-06,
+      "loss": 1.3226,
+      "step": 18245
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 21.546951293945312,
+      "learning_rate": 4.734133790737564e-06,
+      "loss": 0.226,
+      "step": 18246
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 9.672987937927246,
+      "learning_rate": 4.733297075680877e-06,
+      "loss": 0.9477,
+      "step": 18247
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 19.209959030151367,
+      "learning_rate": 4.73246036062419e-06,
+      "loss": 4.4495,
+      "step": 18248
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 6.786340236663818,
+      "learning_rate": 4.731623645567503e-06,
+      "loss": 1.3741,
+      "step": 18249
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 6.299871921539307,
+      "learning_rate": 4.730786930510815e-06,
+      "loss": 0.6948,
+      "step": 18250
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 43.83751678466797,
+      "learning_rate": 4.729950215454127e-06,
+      "loss": 3.2393,
+      "step": 18251
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 6.27683162689209,
+      "learning_rate": 4.72911350039744e-06,
+      "loss": 0.1366,
+      "step": 18252
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 9.838420867919922,
+      "learning_rate": 4.728276785340752e-06,
+      "loss": 0.7473,
+      "step": 18253
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 16.24065399169922,
+      "learning_rate": 4.727440070284065e-06,
+      "loss": 0.6801,
+      "step": 18254
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 14.735170364379883,
+      "learning_rate": 4.726603355227378e-06,
+      "loss": 1.2568,
+      "step": 18255
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 10.91494083404541,
+      "learning_rate": 4.7257666401706905e-06,
+      "loss": 0.3934,
+      "step": 18256
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 25.261390686035156,
+      "learning_rate": 4.724929925114003e-06,
+      "loss": 2.6683,
+      "step": 18257
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 18.75349235534668,
+      "learning_rate": 4.724093210057315e-06,
+      "loss": 1.5721,
+      "step": 18258
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 9.904398918151855,
+      "learning_rate": 4.723256495000628e-06,
+      "loss": 0.417,
+      "step": 18259
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 34.27606964111328,
+      "learning_rate": 4.72241977994394e-06,
+      "loss": 0.6526,
+      "step": 18260
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 51.746421813964844,
+      "learning_rate": 4.721583064887253e-06,
+      "loss": 2.392,
+      "step": 18261
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 13.128945350646973,
+      "learning_rate": 4.720746349830566e-06,
+      "loss": 0.4282,
+      "step": 18262
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 11.959078788757324,
+      "learning_rate": 4.7199096347738784e-06,
+      "loss": 0.9903,
+      "step": 18263
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 13.84642219543457,
+      "learning_rate": 4.719072919717191e-06,
+      "loss": 1.7438,
+      "step": 18264
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 17.713260650634766,
+      "learning_rate": 4.718236204660503e-06,
+      "loss": 1.294,
+      "step": 18265
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 12.827953338623047,
+      "learning_rate": 4.717399489603815e-06,
+      "loss": 0.7201,
+      "step": 18266
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 6.821693420410156,
+      "learning_rate": 4.716562774547128e-06,
+      "loss": 0.5782,
+      "step": 18267
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 7.889687538146973,
+      "learning_rate": 4.715726059490441e-06,
+      "loss": 0.2661,
+      "step": 18268
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 16.12978172302246,
+      "learning_rate": 4.7148893444337536e-06,
+      "loss": 1.2605,
+      "step": 18269
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 7.606167316436768,
+      "learning_rate": 4.714052629377066e-06,
+      "loss": 0.5176,
+      "step": 18270
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 12.436556816101074,
+      "learning_rate": 4.713215914320378e-06,
+      "loss": 1.8853,
+      "step": 18271
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 150.93994140625,
+      "learning_rate": 4.712379199263691e-06,
+      "loss": 1.0326,
+      "step": 18272
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 16.166446685791016,
+      "learning_rate": 4.711542484207003e-06,
+      "loss": 0.3729,
+      "step": 18273
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 16.094757080078125,
+      "learning_rate": 4.710705769150316e-06,
+      "loss": 0.5804,
+      "step": 18274
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 18.42324447631836,
+      "learning_rate": 4.709869054093629e-06,
+      "loss": 0.9788,
+      "step": 18275
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 7.328489303588867,
+      "learning_rate": 4.7090323390369415e-06,
+      "loss": 0.2255,
+      "step": 18276
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 30.318105697631836,
+      "learning_rate": 4.708195623980254e-06,
+      "loss": 1.2918,
+      "step": 18277
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 8.805867195129395,
+      "learning_rate": 4.707358908923566e-06,
+      "loss": 2.0253,
+      "step": 18278
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 27.83954620361328,
+      "learning_rate": 4.706522193866879e-06,
+      "loss": 0.7115,
+      "step": 18279
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 17.944978713989258,
+      "learning_rate": 4.705685478810191e-06,
+      "loss": 2.4743,
+      "step": 18280
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 11.114306449890137,
+      "learning_rate": 4.704848763753504e-06,
+      "loss": 0.944,
+      "step": 18281
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 14.47452449798584,
+      "learning_rate": 4.704012048696817e-06,
+      "loss": 0.3426,
+      "step": 18282
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 9.617793083190918,
+      "learning_rate": 4.703175333640129e-06,
+      "loss": 0.9436,
+      "step": 18283
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 9.082965850830078,
+      "learning_rate": 4.702338618583442e-06,
+      "loss": 0.39,
+      "step": 18284
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 11.9785795211792,
+      "learning_rate": 4.701501903526754e-06,
+      "loss": 0.3713,
+      "step": 18285
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 6.59641170501709,
+      "learning_rate": 4.700665188470067e-06,
+      "loss": 0.614,
+      "step": 18286
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 12.665901184082031,
+      "learning_rate": 4.699828473413379e-06,
+      "loss": 0.4018,
+      "step": 18287
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.5972509384155273,
+      "learning_rate": 4.698991758356692e-06,
+      "loss": 0.087,
+      "step": 18288
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 22.941980361938477,
+      "learning_rate": 4.6981550433000045e-06,
+      "loss": 2.0695,
+      "step": 18289
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.258419036865234,
+      "learning_rate": 4.697318328243317e-06,
+      "loss": 0.2443,
+      "step": 18290
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 11.865102767944336,
+      "learning_rate": 4.69648161318663e-06,
+      "loss": 0.6209,
+      "step": 18291
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 21.162181854248047,
+      "learning_rate": 4.695644898129942e-06,
+      "loss": 2.312,
+      "step": 18292
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.515084743499756,
+      "learning_rate": 4.694808183073255e-06,
+      "loss": 0.4826,
+      "step": 18293
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 11.08837890625,
+      "learning_rate": 4.693971468016567e-06,
+      "loss": 1.0752,
+      "step": 18294
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 95.66429138183594,
+      "learning_rate": 4.69313475295988e-06,
+      "loss": 1.3982,
+      "step": 18295
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 19.24129867553711,
+      "learning_rate": 4.6922980379031924e-06,
+      "loss": 0.3854,
+      "step": 18296
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 22.71967887878418,
+      "learning_rate": 4.691461322846505e-06,
+      "loss": 1.989,
+      "step": 18297
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 8.208455085754395,
+      "learning_rate": 4.690624607789818e-06,
+      "loss": 0.253,
+      "step": 18298
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 22.519729614257812,
+      "learning_rate": 4.68978789273313e-06,
+      "loss": 1.0901,
+      "step": 18299
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 11.487154960632324,
+      "learning_rate": 4.688951177676443e-06,
+      "loss": 2.3736,
+      "step": 18300
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 17.971080780029297,
+      "learning_rate": 4.688114462619755e-06,
+      "loss": 2.133,
+      "step": 18301
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 19.557613372802734,
+      "learning_rate": 4.6872777475630675e-06,
+      "loss": 0.9644,
+      "step": 18302
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 9.254117012023926,
+      "learning_rate": 4.68644103250638e-06,
+      "loss": 0.6961,
+      "step": 18303
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 15.289875030517578,
+      "learning_rate": 4.685604317449693e-06,
+      "loss": 0.338,
+      "step": 18304
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 163.48834228515625,
+      "learning_rate": 4.684767602393006e-06,
+      "loss": 1.5944,
+      "step": 18305
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 7.755233287811279,
+      "learning_rate": 4.683930887336318e-06,
+      "loss": 1.7118,
+      "step": 18306
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 7.73464298248291,
+      "learning_rate": 4.683094172279631e-06,
+      "loss": 0.4349,
+      "step": 18307
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 22.640239715576172,
+      "learning_rate": 4.682257457222943e-06,
+      "loss": 1.8137,
+      "step": 18308
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 8.948088645935059,
+      "learning_rate": 4.6814207421662555e-06,
+      "loss": 0.481,
+      "step": 18309
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 14.048690795898438,
+      "learning_rate": 4.680584027109568e-06,
+      "loss": 1.3658,
+      "step": 18310
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 15.745485305786133,
+      "learning_rate": 4.679747312052881e-06,
+      "loss": 0.4563,
+      "step": 18311
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 11.852276802062988,
+      "learning_rate": 4.678910596996193e-06,
+      "loss": 0.5245,
+      "step": 18312
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 41.801578521728516,
+      "learning_rate": 4.678073881939506e-06,
+      "loss": 0.6852,
+      "step": 18313
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.1565005779266357,
+      "learning_rate": 4.677237166882819e-06,
+      "loss": 0.1697,
+      "step": 18314
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 19.444913864135742,
+      "learning_rate": 4.676400451826131e-06,
+      "loss": 0.5341,
+      "step": 18315
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 6.4638800621032715,
+      "learning_rate": 4.675563736769443e-06,
+      "loss": 0.3213,
+      "step": 18316
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 26.27005386352539,
+      "learning_rate": 4.674727021712756e-06,
+      "loss": 1.0434,
+      "step": 18317
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 15.692587852478027,
+      "learning_rate": 4.673890306656069e-06,
+      "loss": 0.9039,
+      "step": 18318
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 13.831748008728027,
+      "learning_rate": 4.673053591599381e-06,
+      "loss": 1.1145,
+      "step": 18319
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 41.15101623535156,
+      "learning_rate": 4.672216876542694e-06,
+      "loss": 1.4555,
+      "step": 18320
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 8.727827072143555,
+      "learning_rate": 4.6713801614860065e-06,
+      "loss": 0.6902,
+      "step": 18321
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 26.913434982299805,
+      "learning_rate": 4.6705434464293185e-06,
+      "loss": 1.9168,
+      "step": 18322
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 19.45763397216797,
+      "learning_rate": 4.669706731372631e-06,
+      "loss": 0.8972,
+      "step": 18323
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 16.61663246154785,
+      "learning_rate": 4.668870016315944e-06,
+      "loss": 0.5028,
+      "step": 18324
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 36.733219146728516,
+      "learning_rate": 4.668033301259257e-06,
+      "loss": 1.1269,
+      "step": 18325
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 16.540794372558594,
+      "learning_rate": 4.667196586202569e-06,
+      "loss": 0.585,
+      "step": 18326
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 9.48679256439209,
+      "learning_rate": 4.666359871145882e-06,
+      "loss": 0.4882,
+      "step": 18327
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 7.973689079284668,
+      "learning_rate": 4.6655231560891945e-06,
+      "loss": 0.63,
+      "step": 18328
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 10.795661926269531,
+      "learning_rate": 4.664686441032506e-06,
+      "loss": 0.9274,
+      "step": 18329
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 11.213825225830078,
+      "learning_rate": 4.663849725975819e-06,
+      "loss": 2.3832,
+      "step": 18330
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 10.533309936523438,
+      "learning_rate": 4.663013010919132e-06,
+      "loss": 1.4456,
+      "step": 18331
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 17.412017822265625,
+      "learning_rate": 4.662176295862445e-06,
+      "loss": 1.477,
+      "step": 18332
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 22.808351516723633,
+      "learning_rate": 4.661339580805757e-06,
+      "loss": 1.0247,
+      "step": 18333
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 17.05181884765625,
+      "learning_rate": 4.6605028657490696e-06,
+      "loss": 0.617,
+      "step": 18334
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 14.080843925476074,
+      "learning_rate": 4.6596661506923815e-06,
+      "loss": 0.6881,
+      "step": 18335
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 12.616730690002441,
+      "learning_rate": 4.658829435635694e-06,
+      "loss": 0.4547,
+      "step": 18336
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 13.244577407836914,
+      "learning_rate": 4.657992720579007e-06,
+      "loss": 0.5683,
+      "step": 18337
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 10.909563064575195,
+      "learning_rate": 4.65715600552232e-06,
+      "loss": 0.9421,
+      "step": 18338
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 99.05635070800781,
+      "learning_rate": 4.656319290465633e-06,
+      "loss": 2.5841,
+      "step": 18339
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 20.45166778564453,
+      "learning_rate": 4.655482575408945e-06,
+      "loss": 1.4672,
+      "step": 18340
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 33.56233596801758,
+      "learning_rate": 4.6546458603522575e-06,
+      "loss": 1.5021,
+      "step": 18341
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 8.5343656539917,
+      "learning_rate": 4.6538091452955694e-06,
+      "loss": 0.3094,
+      "step": 18342
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 12.456859588623047,
+      "learning_rate": 4.652972430238882e-06,
+      "loss": 0.4925,
+      "step": 18343
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 14.237940788269043,
+      "learning_rate": 4.652135715182195e-06,
+      "loss": 0.6235,
+      "step": 18344
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 22.024959564208984,
+      "learning_rate": 4.651299000125508e-06,
+      "loss": 0.6861,
+      "step": 18345
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 9.857805252075195,
+      "learning_rate": 4.650462285068821e-06,
+      "loss": 1.6765,
+      "step": 18346
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 7.209195137023926,
+      "learning_rate": 4.649625570012133e-06,
+      "loss": 0.4586,
+      "step": 18347
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 15.141064643859863,
+      "learning_rate": 4.648788854955445e-06,
+      "loss": 1.2863,
+      "step": 18348
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 20.099699020385742,
+      "learning_rate": 4.647952139898757e-06,
+      "loss": 1.2703,
+      "step": 18349
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 18.63115882873535,
+      "learning_rate": 4.64711542484207e-06,
+      "loss": 1.5081,
+      "step": 18350
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 13.551779747009277,
+      "learning_rate": 4.646278709785383e-06,
+      "loss": 0.6705,
+      "step": 18351
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 20.256752014160156,
+      "learning_rate": 4.645441994728696e-06,
+      "loss": 2.3797,
+      "step": 18352
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.774133682250977,
+      "learning_rate": 4.644605279672008e-06,
+      "loss": 0.1291,
+      "step": 18353
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 10.925572395324707,
+      "learning_rate": 4.6437685646153205e-06,
+      "loss": 0.9005,
+      "step": 18354
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 12.335229873657227,
+      "learning_rate": 4.642931849558633e-06,
+      "loss": 0.8607,
+      "step": 18355
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 10.979912757873535,
+      "learning_rate": 4.642095134501945e-06,
+      "loss": 1.9296,
+      "step": 18356
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 23.525379180908203,
+      "learning_rate": 4.641258419445258e-06,
+      "loss": 1.5312,
+      "step": 18357
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 18.172266006469727,
+      "learning_rate": 4.640421704388571e-06,
+      "loss": 1.4989,
+      "step": 18358
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 12.313436508178711,
+      "learning_rate": 4.639584989331884e-06,
+      "loss": 0.8825,
+      "step": 18359
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 103.46768188476562,
+      "learning_rate": 4.638748274275196e-06,
+      "loss": 1.6756,
+      "step": 18360
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 10.648122787475586,
+      "learning_rate": 4.6379115592185084e-06,
+      "loss": 0.7992,
+      "step": 18361
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 11.252740859985352,
+      "learning_rate": 4.637074844161821e-06,
+      "loss": 0.7059,
+      "step": 18362
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 97.60618591308594,
+      "learning_rate": 4.636238129105133e-06,
+      "loss": 1.5978,
+      "step": 18363
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 10.907837867736816,
+      "learning_rate": 4.635401414048446e-06,
+      "loss": 0.5228,
+      "step": 18364
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 14.384221076965332,
+      "learning_rate": 4.634564698991759e-06,
+      "loss": 1.6483,
+      "step": 18365
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 41.83057403564453,
+      "learning_rate": 4.633727983935072e-06,
+      "loss": 1.7768,
+      "step": 18366
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 24.17619514465332,
+      "learning_rate": 4.6328912688783836e-06,
+      "loss": 0.3494,
+      "step": 18367
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 6.7195563316345215,
+      "learning_rate": 4.632054553821696e-06,
+      "loss": 1.5434,
+      "step": 18368
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 30.15358543395996,
+      "learning_rate": 4.631217838765009e-06,
+      "loss": 2.3482,
+      "step": 18369
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 6.931991100311279,
+      "learning_rate": 4.630381123708321e-06,
+      "loss": 0.6997,
+      "step": 18370
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 11.640711784362793,
+      "learning_rate": 4.629544408651634e-06,
+      "loss": 1.0447,
+      "step": 18371
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 13.369149208068848,
+      "learning_rate": 4.628707693594947e-06,
+      "loss": 0.4712,
+      "step": 18372
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 14.961252212524414,
+      "learning_rate": 4.6278709785382595e-06,
+      "loss": 1.9146,
+      "step": 18373
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 9.926864624023438,
+      "learning_rate": 4.6270342634815715e-06,
+      "loss": 0.2889,
+      "step": 18374
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.883518695831299,
+      "learning_rate": 4.626197548424884e-06,
+      "loss": 0.2071,
+      "step": 18375
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 14.534547805786133,
+      "learning_rate": 4.625360833368197e-06,
+      "loss": 1.4073,
+      "step": 18376
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 28.588939666748047,
+      "learning_rate": 4.624524118311509e-06,
+      "loss": 0.9256,
+      "step": 18377
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 17.355749130249023,
+      "learning_rate": 4.623687403254822e-06,
+      "loss": 0.3963,
+      "step": 18378
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 11.834649085998535,
+      "learning_rate": 4.622850688198135e-06,
+      "loss": 0.6306,
+      "step": 18379
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.803377151489258,
+      "learning_rate": 4.6220139731414474e-06,
+      "loss": 0.364,
+      "step": 18380
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 22.250408172607422,
+      "learning_rate": 4.621177258084759e-06,
+      "loss": 1.0079,
+      "step": 18381
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 116.30833435058594,
+      "learning_rate": 4.620340543028072e-06,
+      "loss": 2.0143,
+      "step": 18382
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 16.099084854125977,
+      "learning_rate": 4.619503827971385e-06,
+      "loss": 1.2797,
+      "step": 18383
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 8.6182222366333,
+      "learning_rate": 4.618667112914697e-06,
+      "loss": 1.729,
+      "step": 18384
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 12.962079048156738,
+      "learning_rate": 4.61783039785801e-06,
+      "loss": 0.3484,
+      "step": 18385
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 22.328760147094727,
+      "learning_rate": 4.6169936828013226e-06,
+      "loss": 1.6485,
+      "step": 18386
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 20.145177841186523,
+      "learning_rate": 4.616156967744635e-06,
+      "loss": 1.6653,
+      "step": 18387
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 24.25278663635254,
+      "learning_rate": 4.615320252687947e-06,
+      "loss": 0.9849,
+      "step": 18388
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 15.110502243041992,
+      "learning_rate": 4.61448353763126e-06,
+      "loss": 1.1584,
+      "step": 18389
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 6.980915546417236,
+      "learning_rate": 4.613646822574573e-06,
+      "loss": 0.6934,
+      "step": 18390
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 103.16082763671875,
+      "learning_rate": 4.612810107517885e-06,
+      "loss": 1.3066,
+      "step": 18391
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 9.676926612854004,
+      "learning_rate": 4.611973392461198e-06,
+      "loss": 0.4227,
+      "step": 18392
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 41.39181900024414,
+      "learning_rate": 4.6111366774045105e-06,
+      "loss": 1.914,
+      "step": 18393
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 51.430519104003906,
+      "learning_rate": 4.6102999623478224e-06,
+      "loss": 1.3135,
+      "step": 18394
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 12.843317031860352,
+      "learning_rate": 4.609463247291135e-06,
+      "loss": 2.5327,
+      "step": 18395
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 60.07124710083008,
+      "learning_rate": 4.608626532234448e-06,
+      "loss": 1.1639,
+      "step": 18396
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 18.975269317626953,
+      "learning_rate": 4.607789817177761e-06,
+      "loss": 0.9971,
+      "step": 18397
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 17.61123275756836,
+      "learning_rate": 4.606953102121073e-06,
+      "loss": 0.5639,
+      "step": 18398
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 17.708599090576172,
+      "learning_rate": 4.606116387064386e-06,
+      "loss": 1.7949,
+      "step": 18399
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 12.286175727844238,
+      "learning_rate": 4.605279672007698e-06,
+      "loss": 0.6957,
+      "step": 18400
+    },
+    {
+      "epoch": 2.31,
+      "eval_loss": 0.0826125219464302,
+      "eval_runtime": 95.608,
+      "eval_samples_per_second": 37.047,
+      "eval_steps_per_second": 37.047,
+      "step": 18400
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 18.678001403808594,
+      "learning_rate": 4.60444295695101e-06,
+      "loss": 1.502,
+      "step": 18401
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 18.006704330444336,
+      "learning_rate": 4.603606241894323e-06,
+      "loss": 1.5553,
+      "step": 18402
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 13.556264877319336,
+      "learning_rate": 4.602769526837636e-06,
+      "loss": 1.2797,
+      "step": 18403
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 85.38465118408203,
+      "learning_rate": 4.601932811780949e-06,
+      "loss": 3.1793,
+      "step": 18404
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 48.9052848815918,
+      "learning_rate": 4.601096096724261e-06,
+      "loss": 1.9324,
+      "step": 18405
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 15.101606369018555,
+      "learning_rate": 4.6002593816675735e-06,
+      "loss": 0.5255,
+      "step": 18406
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 15.13419246673584,
+      "learning_rate": 4.599422666610886e-06,
+      "loss": 0.7577,
+      "step": 18407
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 14.737383842468262,
+      "learning_rate": 4.598585951554198e-06,
+      "loss": 1.2584,
+      "step": 18408
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 7.4333295822143555,
+      "learning_rate": 4.597749236497511e-06,
+      "loss": 0.402,
+      "step": 18409
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 61.6894645690918,
+      "learning_rate": 4.596912521440824e-06,
+      "loss": 1.2618,
+      "step": 18410
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 8.353595733642578,
+      "learning_rate": 4.596075806384137e-06,
+      "loss": 0.5188,
+      "step": 18411
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 12.857970237731934,
+      "learning_rate": 4.595239091327449e-06,
+      "loss": 1.075,
+      "step": 18412
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 10.973597526550293,
+      "learning_rate": 4.5944023762707614e-06,
+      "loss": 1.0918,
+      "step": 18413
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 19.118207931518555,
+      "learning_rate": 4.593565661214074e-06,
+      "loss": 1.0175,
+      "step": 18414
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 23.525999069213867,
+      "learning_rate": 4.592728946157386e-06,
+      "loss": 1.0776,
+      "step": 18415
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.870167255401611,
+      "learning_rate": 4.591892231100699e-06,
+      "loss": 1.8001,
+      "step": 18416
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 7.887999057769775,
+      "learning_rate": 4.591055516044012e-06,
+      "loss": 1.8062,
+      "step": 18417
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 10.264204025268555,
+      "learning_rate": 4.590218800987325e-06,
+      "loss": 0.6321,
+      "step": 18418
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 17.61655616760254,
+      "learning_rate": 4.5893820859306366e-06,
+      "loss": 1.7462,
+      "step": 18419
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 39.52225875854492,
+      "learning_rate": 4.588545370873949e-06,
+      "loss": 0.6963,
+      "step": 18420
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 14.193148612976074,
+      "learning_rate": 4.587708655817262e-06,
+      "loss": 1.3271,
+      "step": 18421
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 14.687078475952148,
+      "learning_rate": 4.586871940760574e-06,
+      "loss": 0.9129,
+      "step": 18422
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 42.63068771362305,
+      "learning_rate": 4.586035225703887e-06,
+      "loss": 0.7915,
+      "step": 18423
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 92.52960968017578,
+      "learning_rate": 4.5851985106472e-06,
+      "loss": 1.8236,
+      "step": 18424
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 11.0745267868042,
+      "learning_rate": 4.584361795590512e-06,
+      "loss": 0.7362,
+      "step": 18425
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 7.666187763214111,
+      "learning_rate": 4.5835250805338245e-06,
+      "loss": 0.2181,
+      "step": 18426
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 41.562591552734375,
+      "learning_rate": 4.582688365477137e-06,
+      "loss": 1.4681,
+      "step": 18427
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 11.841303825378418,
+      "learning_rate": 4.58185165042045e-06,
+      "loss": 1.4834,
+      "step": 18428
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 19.387245178222656,
+      "learning_rate": 4.581014935363762e-06,
+      "loss": 1.9407,
+      "step": 18429
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 7.08765172958374,
+      "learning_rate": 4.580178220307075e-06,
+      "loss": 0.6848,
+      "step": 18430
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 16.13463592529297,
+      "learning_rate": 4.579341505250388e-06,
+      "loss": 0.6158,
+      "step": 18431
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 16.1623477935791,
+      "learning_rate": 4.5785047901937e-06,
+      "loss": 1.6799,
+      "step": 18432
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 16.288482666015625,
+      "learning_rate": 4.577668075137012e-06,
+      "loss": 0.726,
+      "step": 18433
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 34.15134811401367,
+      "learning_rate": 4.576831360080325e-06,
+      "loss": 0.9051,
+      "step": 18434
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 48.93670654296875,
+      "learning_rate": 4.575994645023637e-06,
+      "loss": 1.1806,
+      "step": 18435
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 67.82980346679688,
+      "learning_rate": 4.57515792996695e-06,
+      "loss": 1.4423,
+      "step": 18436
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 7.230655670166016,
+      "learning_rate": 4.574321214910263e-06,
+      "loss": 0.3903,
+      "step": 18437
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 18.597454071044922,
+      "learning_rate": 4.5734844998535756e-06,
+      "loss": 1.1491,
+      "step": 18438
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 6.0403242111206055,
+      "learning_rate": 4.5726477847968875e-06,
+      "loss": 0.9506,
+      "step": 18439
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 18.730148315429688,
+      "learning_rate": 4.5718110697402e-06,
+      "loss": 1.3363,
+      "step": 18440
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 30.04512596130371,
+      "learning_rate": 4.570974354683513e-06,
+      "loss": 1.4447,
+      "step": 18441
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 9.06648063659668,
+      "learning_rate": 4.570137639626825e-06,
+      "loss": 0.5994,
+      "step": 18442
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 194.48912048339844,
+      "learning_rate": 4.569300924570138e-06,
+      "loss": 1.0488,
+      "step": 18443
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 12.337285995483398,
+      "learning_rate": 4.568464209513451e-06,
+      "loss": 0.8462,
+      "step": 18444
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 15.318734169006348,
+      "learning_rate": 4.5676274944567635e-06,
+      "loss": 0.4775,
+      "step": 18445
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 15.811105728149414,
+      "learning_rate": 4.5667907794000754e-06,
+      "loss": 0.4122,
+      "step": 18446
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 10.545247077941895,
+      "learning_rate": 4.565954064343388e-06,
+      "loss": 1.5772,
+      "step": 18447
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 9.892328262329102,
+      "learning_rate": 4.565117349286701e-06,
+      "loss": 0.3988,
+      "step": 18448
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.395293235778809,
+      "learning_rate": 4.564280634230013e-06,
+      "loss": 1.4455,
+      "step": 18449
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.257930278778076,
+      "learning_rate": 4.563443919173326e-06,
+      "loss": 0.6387,
+      "step": 18450
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 23.21833038330078,
+      "learning_rate": 4.562607204116639e-06,
+      "loss": 1.8555,
+      "step": 18451
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 39.41880416870117,
+      "learning_rate": 4.561770489059951e-06,
+      "loss": 0.9496,
+      "step": 18452
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 30.933761596679688,
+      "learning_rate": 4.560933774003263e-06,
+      "loss": 0.8729,
+      "step": 18453
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 70.10780334472656,
+      "learning_rate": 4.560097058946576e-06,
+      "loss": 1.2494,
+      "step": 18454
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 14.925946235656738,
+      "learning_rate": 4.559260343889889e-06,
+      "loss": 0.3035,
+      "step": 18455
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 6.578594207763672,
+      "learning_rate": 4.558423628833201e-06,
+      "loss": 0.5906,
+      "step": 18456
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 17.706449508666992,
+      "learning_rate": 4.557586913776514e-06,
+      "loss": 0.6118,
+      "step": 18457
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 10.585360527038574,
+      "learning_rate": 4.5567501987198265e-06,
+      "loss": 0.2812,
+      "step": 18458
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 13.294159889221191,
+      "learning_rate": 4.555913483663139e-06,
+      "loss": 0.7182,
+      "step": 18459
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 6.902012348175049,
+      "learning_rate": 4.555076768606451e-06,
+      "loss": 0.3839,
+      "step": 18460
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 12.276602745056152,
+      "learning_rate": 4.554240053549764e-06,
+      "loss": 0.9393,
+      "step": 18461
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 7.4398651123046875,
+      "learning_rate": 4.553403338493077e-06,
+      "loss": 0.6542,
+      "step": 18462
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 35.33179473876953,
+      "learning_rate": 4.552566623436389e-06,
+      "loss": 0.4444,
+      "step": 18463
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 24.448781967163086,
+      "learning_rate": 4.551729908379702e-06,
+      "loss": 0.9474,
+      "step": 18464
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.1484103202819824,
+      "learning_rate": 4.5508931933230144e-06,
+      "loss": 0.0949,
+      "step": 18465
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 26.908098220825195,
+      "learning_rate": 4.550056478266327e-06,
+      "loss": 1.1604,
+      "step": 18466
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 31.584924697875977,
+      "learning_rate": 4.549219763209639e-06,
+      "loss": 1.0898,
+      "step": 18467
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 8.06916332244873,
+      "learning_rate": 4.548383048152952e-06,
+      "loss": 0.2024,
+      "step": 18468
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 53.92453384399414,
+      "learning_rate": 4.547546333096265e-06,
+      "loss": 0.8322,
+      "step": 18469
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 10.829703330993652,
+      "learning_rate": 4.546709618039577e-06,
+      "loss": 0.546,
+      "step": 18470
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 15.220983505249023,
+      "learning_rate": 4.5458729029828895e-06,
+      "loss": 1.5201,
+      "step": 18471
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 48.43095779418945,
+      "learning_rate": 4.545036187926202e-06,
+      "loss": 1.083,
+      "step": 18472
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 34.10857391357422,
+      "learning_rate": 4.544199472869515e-06,
+      "loss": 1.2666,
+      "step": 18473
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 38.154232025146484,
+      "learning_rate": 4.543362757812827e-06,
+      "loss": 2.316,
+      "step": 18474
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 19.14337921142578,
+      "learning_rate": 4.54252604275614e-06,
+      "loss": 1.2413,
+      "step": 18475
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 60.2432975769043,
+      "learning_rate": 4.541689327699452e-06,
+      "loss": 1.3024,
+      "step": 18476
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 12.001728057861328,
+      "learning_rate": 4.540852612642765e-06,
+      "loss": 1.1734,
+      "step": 18477
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 6.172059535980225,
+      "learning_rate": 4.5400158975860775e-06,
+      "loss": 0.5006,
+      "step": 18478
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 17.862943649291992,
+      "learning_rate": 4.53917918252939e-06,
+      "loss": 0.7339,
+      "step": 18479
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 6.126031398773193,
+      "learning_rate": 4.538342467472703e-06,
+      "loss": 0.3133,
+      "step": 18480
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.250203609466553,
+      "learning_rate": 4.537505752416015e-06,
+      "loss": 0.1674,
+      "step": 18481
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 9.197067260742188,
+      "learning_rate": 4.536669037359328e-06,
+      "loss": 0.5205,
+      "step": 18482
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 9.302022933959961,
+      "learning_rate": 4.53583232230264e-06,
+      "loss": 0.3086,
+      "step": 18483
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 24.13341522216797,
+      "learning_rate": 4.5349956072459526e-06,
+      "loss": 0.71,
+      "step": 18484
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.410804748535156,
+      "learning_rate": 4.534158892189265e-06,
+      "loss": 0.2536,
+      "step": 18485
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 16.58905792236328,
+      "learning_rate": 4.533322177132578e-06,
+      "loss": 0.5471,
+      "step": 18486
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 9.57160472869873,
+      "learning_rate": 4.532485462075891e-06,
+      "loss": 0.5688,
+      "step": 18487
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 86.57952117919922,
+      "learning_rate": 4.531648747019203e-06,
+      "loss": 1.9876,
+      "step": 18488
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 29.423912048339844,
+      "learning_rate": 4.530812031962516e-06,
+      "loss": 0.0888,
+      "step": 18489
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.2133326530456543,
+      "learning_rate": 4.529975316905828e-06,
+      "loss": 0.2878,
+      "step": 18490
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 44.775272369384766,
+      "learning_rate": 4.5291386018491405e-06,
+      "loss": 0.8185,
+      "step": 18491
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 22.77064323425293,
+      "learning_rate": 4.528301886792453e-06,
+      "loss": 1.3059,
+      "step": 18492
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 18.844072341918945,
+      "learning_rate": 4.527465171735766e-06,
+      "loss": 1.5743,
+      "step": 18493
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.893050670623779,
+      "learning_rate": 4.526628456679079e-06,
+      "loss": 0.1682,
+      "step": 18494
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 9.104093551635742,
+      "learning_rate": 4.525791741622391e-06,
+      "loss": 0.1361,
+      "step": 18495
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 14.211118698120117,
+      "learning_rate": 4.524955026565704e-06,
+      "loss": 0.8007,
+      "step": 18496
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 19.555896759033203,
+      "learning_rate": 4.524118311509016e-06,
+      "loss": 2.5899,
+      "step": 18497
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 24.445524215698242,
+      "learning_rate": 4.523281596452328e-06,
+      "loss": 1.299,
+      "step": 18498
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 9.455183982849121,
+      "learning_rate": 4.522444881395641e-06,
+      "loss": 0.2394,
+      "step": 18499
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 15.775891304016113,
+      "learning_rate": 4.521608166338954e-06,
+      "loss": 1.1262,
+      "step": 18500
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 8.089310646057129,
+      "learning_rate": 4.520771451282267e-06,
+      "loss": 0.1442,
+      "step": 18501
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 16.34964942932129,
+      "learning_rate": 4.519934736225579e-06,
+      "loss": 1.0115,
+      "step": 18502
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 28.648630142211914,
+      "learning_rate": 4.5190980211688916e-06,
+      "loss": 0.2435,
+      "step": 18503
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 7.589015483856201,
+      "learning_rate": 4.5182613061122035e-06,
+      "loss": 1.3313,
+      "step": 18504
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 16.20108413696289,
+      "learning_rate": 4.517424591055516e-06,
+      "loss": 0.7038,
+      "step": 18505
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 9.961381912231445,
+      "learning_rate": 4.516587875998829e-06,
+      "loss": 0.3543,
+      "step": 18506
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 15.873948097229004,
+      "learning_rate": 4.515751160942142e-06,
+      "loss": 2.2536,
+      "step": 18507
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 15.497950553894043,
+      "learning_rate": 4.514914445885454e-06,
+      "loss": 0.6082,
+      "step": 18508
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.966700553894043,
+      "learning_rate": 4.514077730828767e-06,
+      "loss": 0.3721,
+      "step": 18509
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 56.19346237182617,
+      "learning_rate": 4.5132410157720795e-06,
+      "loss": 1.6361,
+      "step": 18510
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 76.994873046875,
+      "learning_rate": 4.5124043007153914e-06,
+      "loss": 1.6389,
+      "step": 18511
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 18.834623336791992,
+      "learning_rate": 4.511567585658704e-06,
+      "loss": 0.6549,
+      "step": 18512
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 37.12870788574219,
+      "learning_rate": 4.510730870602017e-06,
+      "loss": 1.274,
+      "step": 18513
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 18.4162654876709,
+      "learning_rate": 4.50989415554533e-06,
+      "loss": 0.626,
+      "step": 18514
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 183.23519897460938,
+      "learning_rate": 4.509057440488642e-06,
+      "loss": 0.9285,
+      "step": 18515
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 31.22499656677246,
+      "learning_rate": 4.508220725431955e-06,
+      "loss": 0.8322,
+      "step": 18516
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 43.19390869140625,
+      "learning_rate": 4.5073840103752666e-06,
+      "loss": 1.4242,
+      "step": 18517
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 10.707412719726562,
+      "learning_rate": 4.506547295318579e-06,
+      "loss": 1.2822,
+      "step": 18518
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 18.027921676635742,
+      "learning_rate": 4.505710580261892e-06,
+      "loss": 0.9674,
+      "step": 18519
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 24.536752700805664,
+      "learning_rate": 4.504873865205205e-06,
+      "loss": 1.9647,
+      "step": 18520
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 36.74241256713867,
+      "learning_rate": 4.504037150148518e-06,
+      "loss": 1.7273,
+      "step": 18521
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 21.1619873046875,
+      "learning_rate": 4.50320043509183e-06,
+      "loss": 0.9094,
+      "step": 18522
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 92.76055145263672,
+      "learning_rate": 4.5023637200351425e-06,
+      "loss": 2.894,
+      "step": 18523
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 13.162256240844727,
+      "learning_rate": 4.5015270049784545e-06,
+      "loss": 0.9359,
+      "step": 18524
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 19.254579544067383,
+      "learning_rate": 4.500690289921767e-06,
+      "loss": 3.4928,
+      "step": 18525
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 33.88831329345703,
+      "learning_rate": 4.49985357486508e-06,
+      "loss": 3.0827,
+      "step": 18526
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 8.00139045715332,
+      "learning_rate": 4.499016859808393e-06,
+      "loss": 0.3537,
+      "step": 18527
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 18.81523895263672,
+      "learning_rate": 4.498180144751706e-06,
+      "loss": 1.4291,
+      "step": 18528
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 28.315595626831055,
+      "learning_rate": 4.497343429695018e-06,
+      "loss": 1.0837,
+      "step": 18529
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 11.951425552368164,
+      "learning_rate": 4.4965067146383304e-06,
+      "loss": 1.8781,
+      "step": 18530
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 27.449447631835938,
+      "learning_rate": 4.495669999581642e-06,
+      "loss": 1.3868,
+      "step": 18531
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 16.96086883544922,
+      "learning_rate": 4.494833284524955e-06,
+      "loss": 0.7793,
+      "step": 18532
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 19.07918930053711,
+      "learning_rate": 4.493996569468268e-06,
+      "loss": 2.4194,
+      "step": 18533
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 18.13289451599121,
+      "learning_rate": 4.493159854411581e-06,
+      "loss": 1.1208,
+      "step": 18534
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 7.961818218231201,
+      "learning_rate": 4.492323139354894e-06,
+      "loss": 0.3971,
+      "step": 18535
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 9.359652519226074,
+      "learning_rate": 4.4914864242982056e-06,
+      "loss": 1.1236,
+      "step": 18536
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 10.581709861755371,
+      "learning_rate": 4.490649709241518e-06,
+      "loss": 0.3395,
+      "step": 18537
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 10.72530746459961,
+      "learning_rate": 4.48981299418483e-06,
+      "loss": 0.5704,
+      "step": 18538
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 19.760509490966797,
+      "learning_rate": 4.488976279128143e-06,
+      "loss": 0.24,
+      "step": 18539
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 63.88747024536133,
+      "learning_rate": 4.488139564071456e-06,
+      "loss": 1.9412,
+      "step": 18540
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 8.674443244934082,
+      "learning_rate": 4.487302849014769e-06,
+      "loss": 0.7178,
+      "step": 18541
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 12.277080535888672,
+      "learning_rate": 4.4864661339580815e-06,
+      "loss": 0.3274,
+      "step": 18542
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 19.937883377075195,
+      "learning_rate": 4.4856294189013935e-06,
+      "loss": 1.2808,
+      "step": 18543
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 12.608200073242188,
+      "learning_rate": 4.484792703844706e-06,
+      "loss": 1.8176,
+      "step": 18544
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 11.698739051818848,
+      "learning_rate": 4.483955988788018e-06,
+      "loss": 0.6196,
+      "step": 18545
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 9.430562019348145,
+      "learning_rate": 4.483119273731331e-06,
+      "loss": 0.9752,
+      "step": 18546
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 6.099495887756348,
+      "learning_rate": 4.482282558674644e-06,
+      "loss": 0.3059,
+      "step": 18547
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 11.467565536499023,
+      "learning_rate": 4.481445843617957e-06,
+      "loss": 1.7283,
+      "step": 18548
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 177.7603302001953,
+      "learning_rate": 4.4806091285612694e-06,
+      "loss": 2.3344,
+      "step": 18549
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 19.29451560974121,
+      "learning_rate": 4.479772413504581e-06,
+      "loss": 0.4299,
+      "step": 18550
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 20.672033309936523,
+      "learning_rate": 4.478935698447893e-06,
+      "loss": 0.8978,
+      "step": 18551
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 9.668931007385254,
+      "learning_rate": 4.478098983391206e-06,
+      "loss": 0.6274,
+      "step": 18552
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 8.82039737701416,
+      "learning_rate": 4.477262268334519e-06,
+      "loss": 1.276,
+      "step": 18553
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 24.4532527923584,
+      "learning_rate": 4.476425553277832e-06,
+      "loss": 1.1815,
+      "step": 18554
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 27.42000389099121,
+      "learning_rate": 4.4755888382211446e-06,
+      "loss": 1.7023,
+      "step": 18555
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 19.05655860900879,
+      "learning_rate": 4.474752123164457e-06,
+      "loss": 1.3701,
+      "step": 18556
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 31.000120162963867,
+      "learning_rate": 4.473915408107769e-06,
+      "loss": 0.8982,
+      "step": 18557
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 8.06107234954834,
+      "learning_rate": 4.473078693051081e-06,
+      "loss": 1.0744,
+      "step": 18558
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.505249977111816,
+      "learning_rate": 4.472241977994394e-06,
+      "loss": 0.4442,
+      "step": 18559
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 31.695676803588867,
+      "learning_rate": 4.471405262937707e-06,
+      "loss": 0.9121,
+      "step": 18560
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 130.29441833496094,
+      "learning_rate": 4.47056854788102e-06,
+      "loss": 0.6397,
+      "step": 18561
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 25.62512969970703,
+      "learning_rate": 4.4697318328243325e-06,
+      "loss": 2.3557,
+      "step": 18562
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.5183045864105225,
+      "learning_rate": 4.4688951177676444e-06,
+      "loss": 0.2687,
+      "step": 18563
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.5712084770202637,
+      "learning_rate": 4.468058402710957e-06,
+      "loss": 0.0653,
+      "step": 18564
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.8861128091812134,
+      "learning_rate": 4.467221687654269e-06,
+      "loss": 0.0338,
+      "step": 18565
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 11.420167922973633,
+      "learning_rate": 4.466384972597582e-06,
+      "loss": 1.7012,
+      "step": 18566
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 13.980409622192383,
+      "learning_rate": 4.465548257540895e-06,
+      "loss": 0.7014,
+      "step": 18567
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 23.168249130249023,
+      "learning_rate": 4.464711542484208e-06,
+      "loss": 0.7185,
+      "step": 18568
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 8.879084587097168,
+      "learning_rate": 4.46387482742752e-06,
+      "loss": 0.6503,
+      "step": 18569
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 8.347853660583496,
+      "learning_rate": 4.463038112370832e-06,
+      "loss": 0.7873,
+      "step": 18570
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 27.89021110534668,
+      "learning_rate": 4.462201397314145e-06,
+      "loss": 3.0137,
+      "step": 18571
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 24.764833450317383,
+      "learning_rate": 4.461364682257457e-06,
+      "loss": 0.7871,
+      "step": 18572
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 6.496964454650879,
+      "learning_rate": 4.46052796720077e-06,
+      "loss": 0.9638,
+      "step": 18573
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 7.120611667633057,
+      "learning_rate": 4.459691252144083e-06,
+      "loss": 0.2862,
+      "step": 18574
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 24.407320022583008,
+      "learning_rate": 4.4588545370873955e-06,
+      "loss": 0.5231,
+      "step": 18575
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 7.571482181549072,
+      "learning_rate": 4.458017822030708e-06,
+      "loss": 0.6017,
+      "step": 18576
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 20.153623580932617,
+      "learning_rate": 4.45718110697402e-06,
+      "loss": 0.7772,
+      "step": 18577
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 10.780214309692383,
+      "learning_rate": 4.456344391917333e-06,
+      "loss": 0.4093,
+      "step": 18578
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 134.1143035888672,
+      "learning_rate": 4.455507676860645e-06,
+      "loss": 2.7139,
+      "step": 18579
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 22.914356231689453,
+      "learning_rate": 4.454670961803958e-06,
+      "loss": 1.1819,
+      "step": 18580
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 16.470748901367188,
+      "learning_rate": 4.453834246747271e-06,
+      "loss": 1.6799,
+      "step": 18581
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 28.637752532958984,
+      "learning_rate": 4.4529975316905834e-06,
+      "loss": 0.5893,
+      "step": 18582
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 17.95598793029785,
+      "learning_rate": 4.452160816633896e-06,
+      "loss": 1.992,
+      "step": 18583
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 16.640546798706055,
+      "learning_rate": 4.451324101577208e-06,
+      "loss": 1.4794,
+      "step": 18584
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 12.653728485107422,
+      "learning_rate": 4.450487386520521e-06,
+      "loss": 0.5668,
+      "step": 18585
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 21.595863342285156,
+      "learning_rate": 4.449650671463833e-06,
+      "loss": 0.7837,
+      "step": 18586
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.787606239318848,
+      "learning_rate": 4.448813956407146e-06,
+      "loss": 1.5406,
+      "step": 18587
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 21.696575164794922,
+      "learning_rate": 4.4479772413504586e-06,
+      "loss": 0.848,
+      "step": 18588
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 11.923542022705078,
+      "learning_rate": 4.447140526293771e-06,
+      "loss": 1.5221,
+      "step": 18589
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 8.494603157043457,
+      "learning_rate": 4.446303811237084e-06,
+      "loss": 0.9219,
+      "step": 18590
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 15.257559776306152,
+      "learning_rate": 4.445467096180396e-06,
+      "loss": 1.4104,
+      "step": 18591
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 8.365575790405273,
+      "learning_rate": 4.444630381123708e-06,
+      "loss": 1.4944,
+      "step": 18592
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 10.471525192260742,
+      "learning_rate": 4.443793666067021e-06,
+      "loss": 1.1119,
+      "step": 18593
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 16.419118881225586,
+      "learning_rate": 4.442956951010334e-06,
+      "loss": 0.9984,
+      "step": 18594
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 29.719350814819336,
+      "learning_rate": 4.4421202359536465e-06,
+      "loss": 3.2967,
+      "step": 18595
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 12.81863021850586,
+      "learning_rate": 4.441283520896959e-06,
+      "loss": 0.3091,
+      "step": 18596
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 24.091392517089844,
+      "learning_rate": 4.440446805840272e-06,
+      "loss": 0.2168,
+      "step": 18597
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 14.132325172424316,
+      "learning_rate": 4.439610090783584e-06,
+      "loss": 0.8024,
+      "step": 18598
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 26.080467224121094,
+      "learning_rate": 4.438773375726896e-06,
+      "loss": 2.0559,
+      "step": 18599
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 14.235190391540527,
+      "learning_rate": 4.437936660670209e-06,
+      "loss": 1.5803,
+      "step": 18600
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 10.642189979553223,
+      "learning_rate": 4.437099945613522e-06,
+      "loss": 1.5307,
+      "step": 18601
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 10.55996036529541,
+      "learning_rate": 4.436263230556834e-06,
+      "loss": 0.5519,
+      "step": 18602
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 17.189367294311523,
+      "learning_rate": 4.435426515500147e-06,
+      "loss": 1.8954,
+      "step": 18603
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 17.33152961730957,
+      "learning_rate": 4.434589800443459e-06,
+      "loss": 1.3626,
+      "step": 18604
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 16.388660430908203,
+      "learning_rate": 4.433753085386772e-06,
+      "loss": 0.9414,
+      "step": 18605
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 13.897096633911133,
+      "learning_rate": 4.432916370330084e-06,
+      "loss": 1.7795,
+      "step": 18606
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 74.13624572753906,
+      "learning_rate": 4.432079655273397e-06,
+      "loss": 0.8079,
+      "step": 18607
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 23.349058151245117,
+      "learning_rate": 4.4312429402167095e-06,
+      "loss": 1.7441,
+      "step": 18608
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 8.103655815124512,
+      "learning_rate": 4.430406225160022e-06,
+      "loss": 1.6228,
+      "step": 18609
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 67.40815734863281,
+      "learning_rate": 4.429569510103335e-06,
+      "loss": 2.4471,
+      "step": 18610
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 12.260972023010254,
+      "learning_rate": 4.428732795046647e-06,
+      "loss": 0.5967,
+      "step": 18611
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 10.175232887268066,
+      "learning_rate": 4.42789607998996e-06,
+      "loss": 0.8402,
+      "step": 18612
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 90.39164733886719,
+      "learning_rate": 4.427059364933272e-06,
+      "loss": 1.1611,
+      "step": 18613
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 47.40072250366211,
+      "learning_rate": 4.426222649876585e-06,
+      "loss": 1.7785,
+      "step": 18614
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 7.856509685516357,
+      "learning_rate": 4.4253859348198974e-06,
+      "loss": 0.5466,
+      "step": 18615
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 10.867964744567871,
+      "learning_rate": 4.42454921976321e-06,
+      "loss": 0.3609,
+      "step": 18616
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 10.657840728759766,
+      "learning_rate": 4.423712504706523e-06,
+      "loss": 0.6803,
+      "step": 18617
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.646101474761963,
+      "learning_rate": 4.422875789649835e-06,
+      "loss": 0.2756,
+      "step": 18618
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 14.785065650939941,
+      "learning_rate": 4.422039074593148e-06,
+      "loss": 0.9255,
+      "step": 18619
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 17.859464645385742,
+      "learning_rate": 4.42120235953646e-06,
+      "loss": 0.3477,
+      "step": 18620
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 6.709313869476318,
+      "learning_rate": 4.4203656444797725e-06,
+      "loss": 0.2985,
+      "step": 18621
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.3436636924743652,
+      "learning_rate": 4.419528929423085e-06,
+      "loss": 0.1535,
+      "step": 18622
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 8.863362312316895,
+      "learning_rate": 4.418692214366398e-06,
+      "loss": 0.7028,
+      "step": 18623
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 8.461028099060059,
+      "learning_rate": 4.417855499309711e-06,
+      "loss": 1.6624,
+      "step": 18624
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.642300605773926,
+      "learning_rate": 4.417018784253023e-06,
+      "loss": 0.1512,
+      "step": 18625
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 6.611247539520264,
+      "learning_rate": 4.416182069196336e-06,
+      "loss": 1.3047,
+      "step": 18626
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 23.731626510620117,
+      "learning_rate": 4.415345354139648e-06,
+      "loss": 0.9144,
+      "step": 18627
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 10.380583763122559,
+      "learning_rate": 4.4145086390829605e-06,
+      "loss": 0.186,
+      "step": 18628
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 13.79790210723877,
+      "learning_rate": 4.413671924026273e-06,
+      "loss": 0.5185,
+      "step": 18629
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 22.302764892578125,
+      "learning_rate": 4.412835208969586e-06,
+      "loss": 1.3934,
+      "step": 18630
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 259.0495300292969,
+      "learning_rate": 4.411998493912899e-06,
+      "loss": 1.516,
+      "step": 18631
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 11.745481491088867,
+      "learning_rate": 4.411161778856211e-06,
+      "loss": 2.1137,
+      "step": 18632
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 72.0488510131836,
+      "learning_rate": 4.410325063799523e-06,
+      "loss": 1.3036,
+      "step": 18633
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 8.15693473815918,
+      "learning_rate": 4.4094883487428356e-06,
+      "loss": 0.3817,
+      "step": 18634
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 9.523024559020996,
+      "learning_rate": 4.408651633686148e-06,
+      "loss": 1.111,
+      "step": 18635
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 10.763395309448242,
+      "learning_rate": 4.407814918629461e-06,
+      "loss": 1.9269,
+      "step": 18636
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 80.72051239013672,
+      "learning_rate": 4.406978203572774e-06,
+      "loss": 1.6392,
+      "step": 18637
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 16.553476333618164,
+      "learning_rate": 4.406141488516087e-06,
+      "loss": 0.6152,
+      "step": 18638
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.874630928039551,
+      "learning_rate": 4.405304773459399e-06,
+      "loss": 0.456,
+      "step": 18639
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 47.115177154541016,
+      "learning_rate": 4.404468058402711e-06,
+      "loss": 0.8628,
+      "step": 18640
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 8.854296684265137,
+      "learning_rate": 4.4036313433460235e-06,
+      "loss": 0.135,
+      "step": 18641
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 13.961197853088379,
+      "learning_rate": 4.402794628289336e-06,
+      "loss": 1.3367,
+      "step": 18642
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.6081061363220215,
+      "learning_rate": 4.401957913232649e-06,
+      "loss": 0.7036,
+      "step": 18643
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 13.064496994018555,
+      "learning_rate": 4.401121198175962e-06,
+      "loss": 0.8997,
+      "step": 18644
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 15.707348823547363,
+      "learning_rate": 4.400284483119274e-06,
+      "loss": 0.9013,
+      "step": 18645
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 11.664067268371582,
+      "learning_rate": 4.399447768062587e-06,
+      "loss": 0.6092,
+      "step": 18646
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 25.071107864379883,
+      "learning_rate": 4.398611053005899e-06,
+      "loss": 0.6235,
+      "step": 18647
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 9.626585006713867,
+      "learning_rate": 4.397774337949211e-06,
+      "loss": 0.6089,
+      "step": 18648
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.4906086921691895,
+      "learning_rate": 4.396937622892524e-06,
+      "loss": 0.3915,
+      "step": 18649
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 21.848379135131836,
+      "learning_rate": 4.396100907835837e-06,
+      "loss": 2.2813,
+      "step": 18650
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 100.59906768798828,
+      "learning_rate": 4.39526419277915e-06,
+      "loss": 1.7135,
+      "step": 18651
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.7957444190979,
+      "learning_rate": 4.394427477722462e-06,
+      "loss": 0.1458,
+      "step": 18652
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 6.3409037590026855,
+      "learning_rate": 4.3935907626657746e-06,
+      "loss": 1.2853,
+      "step": 18653
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 21.08176040649414,
+      "learning_rate": 4.3927540476090865e-06,
+      "loss": 2.1313,
+      "step": 18654
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 15.40987777709961,
+      "learning_rate": 4.391917332552399e-06,
+      "loss": 0.9029,
+      "step": 18655
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 11.423988342285156,
+      "learning_rate": 4.391080617495712e-06,
+      "loss": 0.675,
+      "step": 18656
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 12.644664764404297,
+      "learning_rate": 4.390243902439025e-06,
+      "loss": 0.3895,
+      "step": 18657
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 8.19571304321289,
+      "learning_rate": 4.389407187382338e-06,
+      "loss": 0.1933,
+      "step": 18658
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 16.45088005065918,
+      "learning_rate": 4.38857047232565e-06,
+      "loss": 0.5284,
+      "step": 18659
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 43.56208419799805,
+      "learning_rate": 4.3877337572689625e-06,
+      "loss": 0.7396,
+      "step": 18660
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 6.200404644012451,
+      "learning_rate": 4.3868970422122744e-06,
+      "loss": 0.8841,
+      "step": 18661
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 18.50584602355957,
+      "learning_rate": 4.386060327155587e-06,
+      "loss": 2.0774,
+      "step": 18662
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 122.0450668334961,
+      "learning_rate": 4.3852236120989e-06,
+      "loss": 0.9103,
+      "step": 18663
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 15.36992073059082,
+      "learning_rate": 4.384386897042213e-06,
+      "loss": 0.8376,
+      "step": 18664
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 15.341522216796875,
+      "learning_rate": 4.383550181985526e-06,
+      "loss": 1.083,
+      "step": 18665
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.4258575439453125,
+      "learning_rate": 4.382713466928838e-06,
+      "loss": 0.4676,
+      "step": 18666
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.513173818588257,
+      "learning_rate": 4.38187675187215e-06,
+      "loss": 0.0561,
+      "step": 18667
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 16.44916534423828,
+      "learning_rate": 4.381040036815462e-06,
+      "loss": 1.1331,
+      "step": 18668
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 27.556068420410156,
+      "learning_rate": 4.380203321758775e-06,
+      "loss": 2.3297,
+      "step": 18669
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 19.5424861907959,
+      "learning_rate": 4.379366606702088e-06,
+      "loss": 1.7086,
+      "step": 18670
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 17.760272979736328,
+      "learning_rate": 4.378529891645401e-06,
+      "loss": 0.8763,
+      "step": 18671
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 8.427785873413086,
+      "learning_rate": 4.3776931765887136e-06,
+      "loss": 2.0768,
+      "step": 18672
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 21.770397186279297,
+      "learning_rate": 4.3768564615320255e-06,
+      "loss": 0.9538,
+      "step": 18673
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 14.882251739501953,
+      "learning_rate": 4.376019746475338e-06,
+      "loss": 0.4401,
+      "step": 18674
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 24.103191375732422,
+      "learning_rate": 4.37518303141865e-06,
+      "loss": 3.058,
+      "step": 18675
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 6.864442825317383,
+      "learning_rate": 4.374346316361963e-06,
+      "loss": 0.479,
+      "step": 18676
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 45.01991271972656,
+      "learning_rate": 4.373509601305276e-06,
+      "loss": 1.2495,
+      "step": 18677
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 33.04206085205078,
+      "learning_rate": 4.372672886248589e-06,
+      "loss": 0.9721,
+      "step": 18678
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 83.6139144897461,
+      "learning_rate": 4.371836171191901e-06,
+      "loss": 2.5946,
+      "step": 18679
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 26.812896728515625,
+      "learning_rate": 4.3709994561352134e-06,
+      "loss": 0.6091,
+      "step": 18680
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 13.539735794067383,
+      "learning_rate": 4.370162741078526e-06,
+      "loss": 0.6702,
+      "step": 18681
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 7.730309963226318,
+      "learning_rate": 4.369326026021838e-06,
+      "loss": 0.4636,
+      "step": 18682
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 13.025385856628418,
+      "learning_rate": 4.368489310965151e-06,
+      "loss": 1.4847,
+      "step": 18683
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 6.699886798858643,
+      "learning_rate": 4.367652595908464e-06,
+      "loss": 0.309,
+      "step": 18684
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 10.381994247436523,
+      "learning_rate": 4.366815880851777e-06,
+      "loss": 0.4492,
+      "step": 18685
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 13.109643936157227,
+      "learning_rate": 4.3659791657950886e-06,
+      "loss": 1.0229,
+      "step": 18686
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 8.719829559326172,
+      "learning_rate": 4.365142450738401e-06,
+      "loss": 0.4225,
+      "step": 18687
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 6.568873405456543,
+      "learning_rate": 4.364305735681713e-06,
+      "loss": 0.3474,
+      "step": 18688
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.817454814910889,
+      "learning_rate": 4.363469020625026e-06,
+      "loss": 1.064,
+      "step": 18689
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 43.017364501953125,
+      "learning_rate": 4.362632305568339e-06,
+      "loss": 1.0785,
+      "step": 18690
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 13.391928672790527,
+      "learning_rate": 4.361795590511652e-06,
+      "loss": 0.4956,
+      "step": 18691
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 19.47562599182129,
+      "learning_rate": 4.3609588754549645e-06,
+      "loss": 0.5848,
+      "step": 18692
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 34.49109649658203,
+      "learning_rate": 4.3601221603982765e-06,
+      "loss": 2.0613,
+      "step": 18693
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 9.013361930847168,
+      "learning_rate": 4.359285445341589e-06,
+      "loss": 0.4564,
+      "step": 18694
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 9.572412490844727,
+      "learning_rate": 4.358448730284901e-06,
+      "loss": 0.7554,
+      "step": 18695
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 11.869797706604004,
+      "learning_rate": 4.357612015228214e-06,
+      "loss": 0.4906,
+      "step": 18696
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 30.388763427734375,
+      "learning_rate": 4.356775300171527e-06,
+      "loss": 1.9539,
+      "step": 18697
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 13.3001070022583,
+      "learning_rate": 4.35593858511484e-06,
+      "loss": 0.5377,
+      "step": 18698
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 8.962652206420898,
+      "learning_rate": 4.3551018700581524e-06,
+      "loss": 0.4298,
+      "step": 18699
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 14.28447437286377,
+      "learning_rate": 4.354265155001464e-06,
+      "loss": 1.4159,
+      "step": 18700
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 18.35455894470215,
+      "learning_rate": 4.353428439944777e-06,
+      "loss": 0.6595,
+      "step": 18701
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 18.493106842041016,
+      "learning_rate": 4.352591724888089e-06,
+      "loss": 0.7556,
+      "step": 18702
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 18.51642417907715,
+      "learning_rate": 4.351755009831402e-06,
+      "loss": 0.7741,
+      "step": 18703
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 20.658401489257812,
+      "learning_rate": 4.350918294774715e-06,
+      "loss": 2.6664,
+      "step": 18704
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 22.133594512939453,
+      "learning_rate": 4.3500815797180276e-06,
+      "loss": 2.4023,
+      "step": 18705
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 8.598179817199707,
+      "learning_rate": 4.34924486466134e-06,
+      "loss": 0.282,
+      "step": 18706
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 14.833199501037598,
+      "learning_rate": 4.348408149604652e-06,
+      "loss": 0.4488,
+      "step": 18707
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 16.867813110351562,
+      "learning_rate": 4.347571434547965e-06,
+      "loss": 1.0466,
+      "step": 18708
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 18.45764923095703,
+      "learning_rate": 4.346734719491277e-06,
+      "loss": 1.0179,
+      "step": 18709
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.755275726318359,
+      "learning_rate": 4.34589800443459e-06,
+      "loss": 0.7205,
+      "step": 18710
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 17.404754638671875,
+      "learning_rate": 4.345061289377903e-06,
+      "loss": 0.7211,
+      "step": 18711
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 19.29561996459961,
+      "learning_rate": 4.3442245743212155e-06,
+      "loss": 3.2392,
+      "step": 18712
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 27.662883758544922,
+      "learning_rate": 4.343387859264528e-06,
+      "loss": 1.2601,
+      "step": 18713
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 14.027308464050293,
+      "learning_rate": 4.34255114420784e-06,
+      "loss": 0.3639,
+      "step": 18714
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 10.645654678344727,
+      "learning_rate": 4.341714429151153e-06,
+      "loss": 1.0314,
+      "step": 18715
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 23.480266571044922,
+      "learning_rate": 4.340877714094465e-06,
+      "loss": 0.5757,
+      "step": 18716
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.5198709964752197,
+      "learning_rate": 4.340040999037778e-06,
+      "loss": 0.3255,
+      "step": 18717
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 21.50994300842285,
+      "learning_rate": 4.339204283981091e-06,
+      "loss": 1.4896,
+      "step": 18718
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 13.328341484069824,
+      "learning_rate": 4.338367568924403e-06,
+      "loss": 0.7514,
+      "step": 18719
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 19.248104095458984,
+      "learning_rate": 4.337530853867715e-06,
+      "loss": 0.8781,
+      "step": 18720
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 11.229989051818848,
+      "learning_rate": 4.336694138811028e-06,
+      "loss": 0.5211,
+      "step": 18721
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 12.407174110412598,
+      "learning_rate": 4.335857423754341e-06,
+      "loss": 0.6381,
+      "step": 18722
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 10.882194519042969,
+      "learning_rate": 4.335020708697653e-06,
+      "loss": 0.6095,
+      "step": 18723
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 18.214351654052734,
+      "learning_rate": 4.334183993640966e-06,
+      "loss": 0.5243,
+      "step": 18724
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.372857570648193,
+      "learning_rate": 4.3333472785842785e-06,
+      "loss": 0.3892,
+      "step": 18725
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.732825756072998,
+      "learning_rate": 4.332510563527591e-06,
+      "loss": 0.1582,
+      "step": 18726
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 14.107687950134277,
+      "learning_rate": 4.331673848470903e-06,
+      "loss": 0.8475,
+      "step": 18727
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 16.790781021118164,
+      "learning_rate": 4.330837133414216e-06,
+      "loss": 1.5795,
+      "step": 18728
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 13.135433197021484,
+      "learning_rate": 4.330000418357529e-06,
+      "loss": 1.7949,
+      "step": 18729
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 17.481163024902344,
+      "learning_rate": 4.329163703300841e-06,
+      "loss": 0.461,
+      "step": 18730
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 10.736077308654785,
+      "learning_rate": 4.328326988244154e-06,
+      "loss": 0.5273,
+      "step": 18731
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 16.15587043762207,
+      "learning_rate": 4.3274902731874664e-06,
+      "loss": 0.436,
+      "step": 18732
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 7.515250205993652,
+      "learning_rate": 4.326653558130779e-06,
+      "loss": 0.9062,
+      "step": 18733
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 28.70311164855957,
+      "learning_rate": 4.325816843074091e-06,
+      "loss": 1.467,
+      "step": 18734
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 505.750732421875,
+      "learning_rate": 4.324980128017404e-06,
+      "loss": 1.8738,
+      "step": 18735
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 14.68479061126709,
+      "learning_rate": 4.324143412960717e-06,
+      "loss": 0.5343,
+      "step": 18736
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 15.57192325592041,
+      "learning_rate": 4.323306697904029e-06,
+      "loss": 0.7605,
+      "step": 18737
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 6.905477046966553,
+      "learning_rate": 4.3224699828473416e-06,
+      "loss": 0.3782,
+      "step": 18738
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 6.218075275421143,
+      "learning_rate": 4.321633267790654e-06,
+      "loss": 0.601,
+      "step": 18739
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 64.63713073730469,
+      "learning_rate": 4.320796552733967e-06,
+      "loss": 1.1643,
+      "step": 18740
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 13.204005241394043,
+      "learning_rate": 4.319959837677279e-06,
+      "loss": 0.649,
+      "step": 18741
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 36.709774017333984,
+      "learning_rate": 4.319123122620592e-06,
+      "loss": 0.6826,
+      "step": 18742
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 16.327075958251953,
+      "learning_rate": 4.318286407563905e-06,
+      "loss": 0.9353,
+      "step": 18743
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 13.860923767089844,
+      "learning_rate": 4.317449692507217e-06,
+      "loss": 1.3019,
+      "step": 18744
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 25.233509063720703,
+      "learning_rate": 4.3166129774505295e-06,
+      "loss": 0.9742,
+      "step": 18745
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 41.156368255615234,
+      "learning_rate": 4.315776262393842e-06,
+      "loss": 0.8902,
+      "step": 18746
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 22.470033645629883,
+      "learning_rate": 4.314939547337155e-06,
+      "loss": 0.7067,
+      "step": 18747
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 14.529524803161621,
+      "learning_rate": 4.314102832280467e-06,
+      "loss": 0.7574,
+      "step": 18748
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 10.950799942016602,
+      "learning_rate": 4.31326611722378e-06,
+      "loss": 0.5338,
+      "step": 18749
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 16.41820526123047,
+      "learning_rate": 4.312429402167093e-06,
+      "loss": 0.3557,
+      "step": 18750
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 14.858510971069336,
+      "learning_rate": 4.311592687110405e-06,
+      "loss": 1.075,
+      "step": 18751
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 9.458222389221191,
+      "learning_rate": 4.310755972053717e-06,
+      "loss": 1.0552,
+      "step": 18752
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 23.35594367980957,
+      "learning_rate": 4.30991925699703e-06,
+      "loss": 0.7038,
+      "step": 18753
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 40.7728157043457,
+      "learning_rate": 4.309082541940343e-06,
+      "loss": 0.3855,
+      "step": 18754
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 20.873706817626953,
+      "learning_rate": 4.308245826883655e-06,
+      "loss": 1.2008,
+      "step": 18755
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 90.79370880126953,
+      "learning_rate": 4.307409111826968e-06,
+      "loss": 2.0384,
+      "step": 18756
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 11.139209747314453,
+      "learning_rate": 4.3065723967702806e-06,
+      "loss": 0.581,
+      "step": 18757
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.312123775482178,
+      "learning_rate": 4.3057356817135925e-06,
+      "loss": 0.1415,
+      "step": 18758
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 22.708269119262695,
+      "learning_rate": 4.304898966656905e-06,
+      "loss": 1.34,
+      "step": 18759
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 28.71159553527832,
+      "learning_rate": 4.304062251600218e-06,
+      "loss": 1.1707,
+      "step": 18760
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 8.738513946533203,
+      "learning_rate": 4.30322553654353e-06,
+      "loss": 0.621,
+      "step": 18761
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 15.694025993347168,
+      "learning_rate": 4.302388821486843e-06,
+      "loss": 0.8531,
+      "step": 18762
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 13.099310874938965,
+      "learning_rate": 4.301552106430156e-06,
+      "loss": 1.1724,
+      "step": 18763
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 12.900650978088379,
+      "learning_rate": 4.3007153913734685e-06,
+      "loss": 0.8502,
+      "step": 18764
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 10.844757080078125,
+      "learning_rate": 4.2998786763167804e-06,
+      "loss": 0.7511,
+      "step": 18765
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 10.86279010772705,
+      "learning_rate": 4.299041961260093e-06,
+      "loss": 0.4016,
+      "step": 18766
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 30.131702423095703,
+      "learning_rate": 4.298205246203406e-06,
+      "loss": 1.2946,
+      "step": 18767
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 11.3428316116333,
+      "learning_rate": 4.297368531146718e-06,
+      "loss": 0.44,
+      "step": 18768
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 17.472829818725586,
+      "learning_rate": 4.296531816090031e-06,
+      "loss": 0.6456,
+      "step": 18769
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 28.170934677124023,
+      "learning_rate": 4.295695101033344e-06,
+      "loss": 2.0749,
+      "step": 18770
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 11.849146842956543,
+      "learning_rate": 4.294858385976656e-06,
+      "loss": 1.1018,
+      "step": 18771
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 26.870027542114258,
+      "learning_rate": 4.294021670919968e-06,
+      "loss": 1.7141,
+      "step": 18772
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 32.702880859375,
+      "learning_rate": 4.293184955863281e-06,
+      "loss": 1.5337,
+      "step": 18773
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 22.25931167602539,
+      "learning_rate": 4.292348240806594e-06,
+      "loss": 2.4034,
+      "step": 18774
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 566.87744140625,
+      "learning_rate": 4.291511525749906e-06,
+      "loss": 2.0262,
+      "step": 18775
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.847529411315918,
+      "learning_rate": 4.290674810693219e-06,
+      "loss": 0.593,
+      "step": 18776
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 8.884088516235352,
+      "learning_rate": 4.2898380956365315e-06,
+      "loss": 0.5754,
+      "step": 18777
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 9.90809154510498,
+      "learning_rate": 4.2890013805798435e-06,
+      "loss": 0.5576,
+      "step": 18778
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.6491923332214355,
+      "learning_rate": 4.288164665523156e-06,
+      "loss": 0.1102,
+      "step": 18779
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 57.226985931396484,
+      "learning_rate": 4.287327950466469e-06,
+      "loss": 1.3927,
+      "step": 18780
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 17.85209083557129,
+      "learning_rate": 4.286491235409782e-06,
+      "loss": 1.3865,
+      "step": 18781
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 14.462690353393555,
+      "learning_rate": 4.285654520353094e-06,
+      "loss": 1.7501,
+      "step": 18782
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 17.78639030456543,
+      "learning_rate": 4.284817805296407e-06,
+      "loss": 0.9139,
+      "step": 18783
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 11.665761947631836,
+      "learning_rate": 4.283981090239719e-06,
+      "loss": 0.6823,
+      "step": 18784
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 11.269731521606445,
+      "learning_rate": 4.283144375183031e-06,
+      "loss": 0.2417,
+      "step": 18785
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 9.31551742553711,
+      "learning_rate": 4.282307660126344e-06,
+      "loss": 0.2929,
+      "step": 18786
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 15.734121322631836,
+      "learning_rate": 4.281470945069657e-06,
+      "loss": 1.9836,
+      "step": 18787
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 12.540379524230957,
+      "learning_rate": 4.28063423001297e-06,
+      "loss": 0.3442,
+      "step": 18788
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 21.346942901611328,
+      "learning_rate": 4.279797514956282e-06,
+      "loss": 0.7449,
+      "step": 18789
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 12.596646308898926,
+      "learning_rate": 4.2789607998995945e-06,
+      "loss": 1.1474,
+      "step": 18790
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 18.674962997436523,
+      "learning_rate": 4.278124084842907e-06,
+      "loss": 0.7168,
+      "step": 18791
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 21.470760345458984,
+      "learning_rate": 4.277287369786219e-06,
+      "loss": 1.6533,
+      "step": 18792
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 16.76324462890625,
+      "learning_rate": 4.276450654729532e-06,
+      "loss": 0.7499,
+      "step": 18793
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 8.946023941040039,
+      "learning_rate": 4.275613939672845e-06,
+      "loss": 0.4861,
+      "step": 18794
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 11.926936149597168,
+      "learning_rate": 4.274777224616158e-06,
+      "loss": 1.6696,
+      "step": 18795
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 8.531956672668457,
+      "learning_rate": 4.27394050955947e-06,
+      "loss": 0.7272,
+      "step": 18796
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 45.68627166748047,
+      "learning_rate": 4.2731037945027825e-06,
+      "loss": 1.5998,
+      "step": 18797
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 49.42822265625,
+      "learning_rate": 4.272267079446095e-06,
+      "loss": 0.6409,
+      "step": 18798
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.116422653198242,
+      "learning_rate": 4.271430364389407e-06,
+      "loss": 0.1939,
+      "step": 18799
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 16.58919906616211,
+      "learning_rate": 4.27059364933272e-06,
+      "loss": 0.7827,
+      "step": 18800
+    },
+    {
+      "epoch": 2.36,
+      "eval_loss": 0.08024092018604279,
+      "eval_runtime": 94.8209,
+      "eval_samples_per_second": 37.355,
+      "eval_steps_per_second": 37.355,
+      "step": 18800
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 16.0667724609375,
+      "learning_rate": 4.269756934276033e-06,
+      "loss": 1.9254,
+      "step": 18801
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 193.6348876953125,
+      "learning_rate": 4.268920219219345e-06,
+      "loss": 1.5117,
+      "step": 18802
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 12.501949310302734,
+      "learning_rate": 4.2680835041626576e-06,
+      "loss": 1.2545,
+      "step": 18803
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 7.939141750335693,
+      "learning_rate": 4.26724678910597e-06,
+      "loss": 0.6291,
+      "step": 18804
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 10.727766036987305,
+      "learning_rate": 4.266410074049283e-06,
+      "loss": 0.5534,
+      "step": 18805
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 20.324636459350586,
+      "learning_rate": 4.265573358992595e-06,
+      "loss": 0.537,
+      "step": 18806
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.044099807739258,
+      "learning_rate": 4.264736643935908e-06,
+      "loss": 0.1942,
+      "step": 18807
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 26.941146850585938,
+      "learning_rate": 4.263899928879221e-06,
+      "loss": 1.7258,
+      "step": 18808
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 58.14222717285156,
+      "learning_rate": 4.263063213822533e-06,
+      "loss": 1.8982,
+      "step": 18809
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 9.319318771362305,
+      "learning_rate": 4.2622264987658455e-06,
+      "loss": 0.6623,
+      "step": 18810
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 21.797510147094727,
+      "learning_rate": 4.261389783709158e-06,
+      "loss": 1.3951,
+      "step": 18811
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 32.75272750854492,
+      "learning_rate": 4.260553068652471e-06,
+      "loss": 1.5829,
+      "step": 18812
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.681191921234131,
+      "learning_rate": 4.259716353595783e-06,
+      "loss": 0.5909,
+      "step": 18813
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 10.198139190673828,
+      "learning_rate": 4.258879638539096e-06,
+      "loss": 0.2836,
+      "step": 18814
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 18.730148315429688,
+      "learning_rate": 4.258042923482409e-06,
+      "loss": 2.1234,
+      "step": 18815
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 12.210789680480957,
+      "learning_rate": 4.257206208425721e-06,
+      "loss": 0.394,
+      "step": 18816
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 8.160884857177734,
+      "learning_rate": 4.256369493369033e-06,
+      "loss": 0.9246,
+      "step": 18817
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 12.022786140441895,
+      "learning_rate": 4.255532778312346e-06,
+      "loss": 0.6435,
+      "step": 18818
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 17.289894104003906,
+      "learning_rate": 4.254696063255659e-06,
+      "loss": 1.4093,
+      "step": 18819
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 16.079078674316406,
+      "learning_rate": 4.253859348198971e-06,
+      "loss": 1.842,
+      "step": 18820
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 13.1908540725708,
+      "learning_rate": 4.253022633142284e-06,
+      "loss": 0.8832,
+      "step": 18821
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 25.93243408203125,
+      "learning_rate": 4.2521859180855966e-06,
+      "loss": 1.1607,
+      "step": 18822
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 10.018243789672852,
+      "learning_rate": 4.2513492030289085e-06,
+      "loss": 0.3035,
+      "step": 18823
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 9.124788284301758,
+      "learning_rate": 4.250512487972221e-06,
+      "loss": 0.9549,
+      "step": 18824
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.4602513313293457,
+      "learning_rate": 4.249675772915534e-06,
+      "loss": 0.0339,
+      "step": 18825
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 18.3439998626709,
+      "learning_rate": 4.248839057858847e-06,
+      "loss": 0.6064,
+      "step": 18826
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 16.524682998657227,
+      "learning_rate": 4.248002342802159e-06,
+      "loss": 0.5423,
+      "step": 18827
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 15.11779499053955,
+      "learning_rate": 4.247165627745472e-06,
+      "loss": 0.4983,
+      "step": 18828
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 9.199140548706055,
+      "learning_rate": 4.2463289126887845e-06,
+      "loss": 0.3444,
+      "step": 18829
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 12.126413345336914,
+      "learning_rate": 4.2454921976320964e-06,
+      "loss": 0.4077,
+      "step": 18830
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 8.376360893249512,
+      "learning_rate": 4.244655482575409e-06,
+      "loss": 1.4381,
+      "step": 18831
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 14.740618705749512,
+      "learning_rate": 4.243818767518722e-06,
+      "loss": 1.1537,
+      "step": 18832
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 21.037717819213867,
+      "learning_rate": 4.242982052462035e-06,
+      "loss": 1.0074,
+      "step": 18833
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 14.283814430236816,
+      "learning_rate": 4.242145337405347e-06,
+      "loss": 0.6166,
+      "step": 18834
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.688212871551514,
+      "learning_rate": 4.24130862234866e-06,
+      "loss": 0.4822,
+      "step": 18835
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 23.023244857788086,
+      "learning_rate": 4.240471907291972e-06,
+      "loss": 1.8606,
+      "step": 18836
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 6.0786662101745605,
+      "learning_rate": 4.239635192235284e-06,
+      "loss": 0.2515,
+      "step": 18837
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 24.129573822021484,
+      "learning_rate": 4.238798477178597e-06,
+      "loss": 0.8353,
+      "step": 18838
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 113.80068969726562,
+      "learning_rate": 4.23796176212191e-06,
+      "loss": 0.8471,
+      "step": 18839
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 25.95708656311035,
+      "learning_rate": 4.237125047065223e-06,
+      "loss": 1.8908,
+      "step": 18840
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 9.736194610595703,
+      "learning_rate": 4.236288332008535e-06,
+      "loss": 1.5688,
+      "step": 18841
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 25.874277114868164,
+      "learning_rate": 4.2354516169518475e-06,
+      "loss": 2.0212,
+      "step": 18842
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 8.549764633178711,
+      "learning_rate": 4.2346149018951595e-06,
+      "loss": 1.0517,
+      "step": 18843
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 31.702428817749023,
+      "learning_rate": 4.233778186838472e-06,
+      "loss": 1.0722,
+      "step": 18844
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.378050804138184,
+      "learning_rate": 4.232941471781785e-06,
+      "loss": 0.5427,
+      "step": 18845
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 14.59392261505127,
+      "learning_rate": 4.232104756725098e-06,
+      "loss": 0.6592,
+      "step": 18846
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 11.767300605773926,
+      "learning_rate": 4.231268041668411e-06,
+      "loss": 0.2533,
+      "step": 18847
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 17.252201080322266,
+      "learning_rate": 4.230431326611723e-06,
+      "loss": 0.4975,
+      "step": 18848
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.5346503257751465,
+      "learning_rate": 4.2295946115550354e-06,
+      "loss": 0.1133,
+      "step": 18849
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 44.0077018737793,
+      "learning_rate": 4.228757896498347e-06,
+      "loss": 1.6376,
+      "step": 18850
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 25.099565505981445,
+      "learning_rate": 4.22792118144166e-06,
+      "loss": 1.5878,
+      "step": 18851
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 7.18458890914917,
+      "learning_rate": 4.227084466384973e-06,
+      "loss": 1.1702,
+      "step": 18852
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 9.28117561340332,
+      "learning_rate": 4.226247751328286e-06,
+      "loss": 0.7456,
+      "step": 18853
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 16.449676513671875,
+      "learning_rate": 4.225411036271599e-06,
+      "loss": 1.1489,
+      "step": 18854
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.339559316635132,
+      "learning_rate": 4.2245743212149106e-06,
+      "loss": 0.0947,
+      "step": 18855
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 26.09206771850586,
+      "learning_rate": 4.223737606158223e-06,
+      "loss": 2.6489,
+      "step": 18856
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 15.931502342224121,
+      "learning_rate": 4.222900891101535e-06,
+      "loss": 0.9297,
+      "step": 18857
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 19.132822036743164,
+      "learning_rate": 4.222064176044848e-06,
+      "loss": 0.3773,
+      "step": 18858
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 27.903467178344727,
+      "learning_rate": 4.221227460988161e-06,
+      "loss": 0.9293,
+      "step": 18859
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 15.282549858093262,
+      "learning_rate": 4.220390745931474e-06,
+      "loss": 0.7332,
+      "step": 18860
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 9.827350616455078,
+      "learning_rate": 4.2195540308747865e-06,
+      "loss": 1.2795,
+      "step": 18861
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 11.952923774719238,
+      "learning_rate": 4.2187173158180985e-06,
+      "loss": 1.1583,
+      "step": 18862
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 8.408980369567871,
+      "learning_rate": 4.217880600761411e-06,
+      "loss": 0.4653,
+      "step": 18863
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 51.14714431762695,
+      "learning_rate": 4.217043885704723e-06,
+      "loss": 0.923,
+      "step": 18864
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 10.286130905151367,
+      "learning_rate": 4.216207170648036e-06,
+      "loss": 0.9121,
+      "step": 18865
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 23.672351837158203,
+      "learning_rate": 4.215370455591349e-06,
+      "loss": 0.8421,
+      "step": 18866
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 17.699554443359375,
+      "learning_rate": 4.214533740534662e-06,
+      "loss": 2.0303,
+      "step": 18867
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 15.324867248535156,
+      "learning_rate": 4.213697025477974e-06,
+      "loss": 2.3224,
+      "step": 18868
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 6.052308559417725,
+      "learning_rate": 4.212860310421286e-06,
+      "loss": 0.9046,
+      "step": 18869
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 18.945478439331055,
+      "learning_rate": 4.212023595364599e-06,
+      "loss": 1.6304,
+      "step": 18870
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 10.311890602111816,
+      "learning_rate": 4.211186880307911e-06,
+      "loss": 0.23,
+      "step": 18871
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.769211769104004,
+      "learning_rate": 4.210350165251224e-06,
+      "loss": 0.1743,
+      "step": 18872
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 14.367249488830566,
+      "learning_rate": 4.209513450194537e-06,
+      "loss": 0.8107,
+      "step": 18873
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 20.644311904907227,
+      "learning_rate": 4.2086767351378496e-06,
+      "loss": 2.0886,
+      "step": 18874
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 68.03089904785156,
+      "learning_rate": 4.2078400200811615e-06,
+      "loss": 3.0096,
+      "step": 18875
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 29.12495994567871,
+      "learning_rate": 4.207003305024474e-06,
+      "loss": 0.7108,
+      "step": 18876
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 19.217121124267578,
+      "learning_rate": 4.206166589967787e-06,
+      "loss": 0.7231,
+      "step": 18877
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 115.30744171142578,
+      "learning_rate": 4.205329874911099e-06,
+      "loss": 1.8465,
+      "step": 18878
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 23.02408218383789,
+      "learning_rate": 4.204493159854412e-06,
+      "loss": 1.339,
+      "step": 18879
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 87.8171615600586,
+      "learning_rate": 4.203656444797725e-06,
+      "loss": 0.7673,
+      "step": 18880
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 2.9575674533843994,
+      "learning_rate": 4.2028197297410375e-06,
+      "loss": 0.1289,
+      "step": 18881
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 10.533477783203125,
+      "learning_rate": 4.2019830146843494e-06,
+      "loss": 0.6978,
+      "step": 18882
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 12.552096366882324,
+      "learning_rate": 4.201146299627662e-06,
+      "loss": 0.348,
+      "step": 18883
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 13.85686206817627,
+      "learning_rate": 4.200309584570974e-06,
+      "loss": 1.8697,
+      "step": 18884
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 17.216442108154297,
+      "learning_rate": 4.199472869514287e-06,
+      "loss": 0.3324,
+      "step": 18885
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 13.309085845947266,
+      "learning_rate": 4.1986361544576e-06,
+      "loss": 0.4239,
+      "step": 18886
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 20.115856170654297,
+      "learning_rate": 4.197799439400913e-06,
+      "loss": 0.7594,
+      "step": 18887
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 17.214750289916992,
+      "learning_rate": 4.196962724344225e-06,
+      "loss": 0.2819,
+      "step": 18888
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 6.4202775955200195,
+      "learning_rate": 4.196126009287537e-06,
+      "loss": 1.2622,
+      "step": 18889
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 8.798897743225098,
+      "learning_rate": 4.19528929423085e-06,
+      "loss": 0.3203,
+      "step": 18890
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 14.010072708129883,
+      "learning_rate": 4.194452579174162e-06,
+      "loss": 0.4699,
+      "step": 18891
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 2.101485013961792,
+      "learning_rate": 4.193615864117475e-06,
+      "loss": 0.097,
+      "step": 18892
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 71.57234191894531,
+      "learning_rate": 4.192779149060788e-06,
+      "loss": 2.0226,
+      "step": 18893
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.3469460010528564,
+      "learning_rate": 4.1919424340041005e-06,
+      "loss": 0.2548,
+      "step": 18894
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 6.6450934410095215,
+      "learning_rate": 4.191105718947413e-06,
+      "loss": 0.385,
+      "step": 18895
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 16.732669830322266,
+      "learning_rate": 4.190269003890725e-06,
+      "loss": 1.6495,
+      "step": 18896
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 9.60965633392334,
+      "learning_rate": 4.189432288834038e-06,
+      "loss": 1.7285,
+      "step": 18897
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 8.528308868408203,
+      "learning_rate": 4.18859557377735e-06,
+      "loss": 0.4976,
+      "step": 18898
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 46.66935729980469,
+      "learning_rate": 4.187758858720663e-06,
+      "loss": 1.3655,
+      "step": 18899
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 87.4046859741211,
+      "learning_rate": 4.186922143663976e-06,
+      "loss": 2.0976,
+      "step": 18900
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 15.820965766906738,
+      "learning_rate": 4.1860854286072884e-06,
+      "loss": 1.2496,
+      "step": 18901
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 17.30780601501465,
+      "learning_rate": 4.185248713550601e-06,
+      "loss": 1.9525,
+      "step": 18902
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 10.334717750549316,
+      "learning_rate": 4.184411998493913e-06,
+      "loss": 0.5088,
+      "step": 18903
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 18.360502243041992,
+      "learning_rate": 4.183575283437226e-06,
+      "loss": 0.9067,
+      "step": 18904
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 13.89871883392334,
+      "learning_rate": 4.182738568380538e-06,
+      "loss": 2.7615,
+      "step": 18905
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 16.200584411621094,
+      "learning_rate": 4.181901853323851e-06,
+      "loss": 0.6003,
+      "step": 18906
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 7.320597171783447,
+      "learning_rate": 4.1810651382671636e-06,
+      "loss": 0.1048,
+      "step": 18907
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 45.176151275634766,
+      "learning_rate": 4.180228423210476e-06,
+      "loss": 0.6831,
+      "step": 18908
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 17.75602149963379,
+      "learning_rate": 4.179391708153789e-06,
+      "loss": 0.4882,
+      "step": 18909
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 28.60691261291504,
+      "learning_rate": 4.178554993097101e-06,
+      "loss": 0.5252,
+      "step": 18910
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 22.014205932617188,
+      "learning_rate": 4.177718278040414e-06,
+      "loss": 0.673,
+      "step": 18911
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 16.342103958129883,
+      "learning_rate": 4.176881562983726e-06,
+      "loss": 1.3377,
+      "step": 18912
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 23.191085815429688,
+      "learning_rate": 4.176044847927039e-06,
+      "loss": 1.2594,
+      "step": 18913
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.5212202072143555,
+      "learning_rate": 4.1752081328703515e-06,
+      "loss": 0.0523,
+      "step": 18914
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 20.338064193725586,
+      "learning_rate": 4.174371417813664e-06,
+      "loss": 0.7773,
+      "step": 18915
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 10.315778732299805,
+      "learning_rate": 4.173534702756977e-06,
+      "loss": 0.6839,
+      "step": 18916
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 18.50375747680664,
+      "learning_rate": 4.172697987700289e-06,
+      "loss": 1.3235,
+      "step": 18917
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.669938564300537,
+      "learning_rate": 4.171861272643602e-06,
+      "loss": 0.2752,
+      "step": 18918
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.4706034660339355,
+      "learning_rate": 4.171024557586914e-06,
+      "loss": 0.4694,
+      "step": 18919
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 13.295501708984375,
+      "learning_rate": 4.170187842530227e-06,
+      "loss": 0.321,
+      "step": 18920
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.5812788009643555,
+      "learning_rate": 4.169351127473539e-06,
+      "loss": 0.0755,
+      "step": 18921
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 12.092533111572266,
+      "learning_rate": 4.168514412416852e-06,
+      "loss": 0.3939,
+      "step": 18922
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 6.5334391593933105,
+      "learning_rate": 4.167677697360165e-06,
+      "loss": 0.4008,
+      "step": 18923
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 6.879368782043457,
+      "learning_rate": 4.166840982303477e-06,
+      "loss": 0.4942,
+      "step": 18924
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 17.075260162353516,
+      "learning_rate": 4.166004267246789e-06,
+      "loss": 0.6626,
+      "step": 18925
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 12.275047302246094,
+      "learning_rate": 4.165167552190102e-06,
+      "loss": 0.9725,
+      "step": 18926
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 22.844661712646484,
+      "learning_rate": 4.1643308371334145e-06,
+      "loss": 0.3083,
+      "step": 18927
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 9.049379348754883,
+      "learning_rate": 4.163494122076727e-06,
+      "loss": 1.3238,
+      "step": 18928
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 17.66362190246582,
+      "learning_rate": 4.16265740702004e-06,
+      "loss": 1.2612,
+      "step": 18929
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.301072597503662,
+      "learning_rate": 4.161820691963352e-06,
+      "loss": 1.4726,
+      "step": 18930
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 14.616764068603516,
+      "learning_rate": 4.160983976906665e-06,
+      "loss": 0.5514,
+      "step": 18931
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 17.993520736694336,
+      "learning_rate": 4.160147261849977e-06,
+      "loss": 0.8817,
+      "step": 18932
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 11.278755187988281,
+      "learning_rate": 4.15931054679329e-06,
+      "loss": 1.2055,
+      "step": 18933
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 59.47795104980469,
+      "learning_rate": 4.158473831736602e-06,
+      "loss": 1.5626,
+      "step": 18934
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 8.581803321838379,
+      "learning_rate": 4.157637116679915e-06,
+      "loss": 0.274,
+      "step": 18935
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.019199371337891,
+      "learning_rate": 4.156800401623228e-06,
+      "loss": 0.4222,
+      "step": 18936
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 14.822456359863281,
+      "learning_rate": 4.15596368656654e-06,
+      "loss": 0.8128,
+      "step": 18937
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 17.66838836669922,
+      "learning_rate": 4.155126971509853e-06,
+      "loss": 1.3863,
+      "step": 18938
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 24.690292358398438,
+      "learning_rate": 4.154290256453165e-06,
+      "loss": 2.2689,
+      "step": 18939
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 9.440573692321777,
+      "learning_rate": 4.1534535413964775e-06,
+      "loss": 0.6847,
+      "step": 18940
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 13.989275932312012,
+      "learning_rate": 4.15261682633979e-06,
+      "loss": 0.7083,
+      "step": 18941
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.123847007751465,
+      "learning_rate": 4.151780111283103e-06,
+      "loss": 0.1782,
+      "step": 18942
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 14.532522201538086,
+      "learning_rate": 4.150943396226416e-06,
+      "loss": 0.697,
+      "step": 18943
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 19.20111846923828,
+      "learning_rate": 4.150106681169728e-06,
+      "loss": 0.7649,
+      "step": 18944
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 37.36458969116211,
+      "learning_rate": 4.149269966113041e-06,
+      "loss": 3.0226,
+      "step": 18945
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 7.685119152069092,
+      "learning_rate": 4.148433251056353e-06,
+      "loss": 0.3855,
+      "step": 18946
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 14.21370792388916,
+      "learning_rate": 4.1475965359996655e-06,
+      "loss": 0.8942,
+      "step": 18947
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 27.41675567626953,
+      "learning_rate": 4.146759820942978e-06,
+      "loss": 3.6937,
+      "step": 18948
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 8.781967163085938,
+      "learning_rate": 4.145923105886291e-06,
+      "loss": 0.6549,
+      "step": 18949
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.892064571380615,
+      "learning_rate": 4.145086390829604e-06,
+      "loss": 0.2349,
+      "step": 18950
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 11.855024337768555,
+      "learning_rate": 4.144249675772916e-06,
+      "loss": 0.5825,
+      "step": 18951
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 75.8827896118164,
+      "learning_rate": 4.143412960716229e-06,
+      "loss": 1.6483,
+      "step": 18952
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 35.53962707519531,
+      "learning_rate": 4.1425762456595406e-06,
+      "loss": 1.5198,
+      "step": 18953
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.657505989074707,
+      "learning_rate": 4.141739530602853e-06,
+      "loss": 0.3553,
+      "step": 18954
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 12.064531326293945,
+      "learning_rate": 4.140902815546166e-06,
+      "loss": 0.2139,
+      "step": 18955
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.048888206481934,
+      "learning_rate": 4.140066100489479e-06,
+      "loss": 0.3971,
+      "step": 18956
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 9.145115852355957,
+      "learning_rate": 4.139229385432792e-06,
+      "loss": 0.4083,
+      "step": 18957
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.697978973388672,
+      "learning_rate": 4.138392670376104e-06,
+      "loss": 0.4068,
+      "step": 18958
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 10.511747360229492,
+      "learning_rate": 4.1375559553194165e-06,
+      "loss": 1.393,
+      "step": 18959
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.4989471435546875,
+      "learning_rate": 4.1367192402627285e-06,
+      "loss": 0.0813,
+      "step": 18960
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 14.240618705749512,
+      "learning_rate": 4.135882525206041e-06,
+      "loss": 0.3611,
+      "step": 18961
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 12.395482063293457,
+      "learning_rate": 4.135045810149354e-06,
+      "loss": 1.7359,
+      "step": 18962
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 92.8197250366211,
+      "learning_rate": 4.134209095092667e-06,
+      "loss": 2.6389,
+      "step": 18963
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 19.586116790771484,
+      "learning_rate": 4.13337238003598e-06,
+      "loss": 0.3887,
+      "step": 18964
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 12.205307960510254,
+      "learning_rate": 4.132535664979292e-06,
+      "loss": 1.1692,
+      "step": 18965
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 59.4495964050293,
+      "learning_rate": 4.131698949922604e-06,
+      "loss": 2.6039,
+      "step": 18966
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.818330764770508,
+      "learning_rate": 4.130862234865916e-06,
+      "loss": 0.5976,
+      "step": 18967
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 12.355748176574707,
+      "learning_rate": 4.130025519809229e-06,
+      "loss": 1.4166,
+      "step": 18968
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 21.618093490600586,
+      "learning_rate": 4.129188804752542e-06,
+      "loss": 1.5143,
+      "step": 18969
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 21.394245147705078,
+      "learning_rate": 4.128352089695855e-06,
+      "loss": 0.7113,
+      "step": 18970
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 15.48207950592041,
+      "learning_rate": 4.127515374639167e-06,
+      "loss": 0.1245,
+      "step": 18971
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 11.195927619934082,
+      "learning_rate": 4.1266786595824796e-06,
+      "loss": 0.7492,
+      "step": 18972
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 18.20484733581543,
+      "learning_rate": 4.1258419445257915e-06,
+      "loss": 0.6182,
+      "step": 18973
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 13.016801834106445,
+      "learning_rate": 4.125005229469104e-06,
+      "loss": 0.7796,
+      "step": 18974
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 12.392509460449219,
+      "learning_rate": 4.124168514412417e-06,
+      "loss": 1.2688,
+      "step": 18975
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 27.529605865478516,
+      "learning_rate": 4.12333179935573e-06,
+      "loss": 1.6192,
+      "step": 18976
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 14.227264404296875,
+      "learning_rate": 4.122495084299043e-06,
+      "loss": 0.5007,
+      "step": 18977
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.663363456726074,
+      "learning_rate": 4.121658369242355e-06,
+      "loss": 0.3257,
+      "step": 18978
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 16.237409591674805,
+      "learning_rate": 4.1208216541856675e-06,
+      "loss": 0.8368,
+      "step": 18979
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.636362552642822,
+      "learning_rate": 4.1199849391289794e-06,
+      "loss": 0.9216,
+      "step": 18980
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 15.949992179870605,
+      "learning_rate": 4.119148224072292e-06,
+      "loss": 0.783,
+      "step": 18981
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 8.62476921081543,
+      "learning_rate": 4.118311509015605e-06,
+      "loss": 0.4269,
+      "step": 18982
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 13.7176513671875,
+      "learning_rate": 4.117474793958918e-06,
+      "loss": 0.3969,
+      "step": 18983
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.941378116607666,
+      "learning_rate": 4.116638078902231e-06,
+      "loss": 0.2093,
+      "step": 18984
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 21.555898666381836,
+      "learning_rate": 4.115801363845543e-06,
+      "loss": 0.7242,
+      "step": 18985
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 19.81019401550293,
+      "learning_rate": 4.114964648788855e-06,
+      "loss": 0.4602,
+      "step": 18986
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 11.031840324401855,
+      "learning_rate": 4.114127933732167e-06,
+      "loss": 0.4644,
+      "step": 18987
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 18.960458755493164,
+      "learning_rate": 4.11329121867548e-06,
+      "loss": 0.6422,
+      "step": 18988
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 10.122111320495605,
+      "learning_rate": 4.112454503618793e-06,
+      "loss": 1.5367,
+      "step": 18989
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 21.96918487548828,
+      "learning_rate": 4.111617788562106e-06,
+      "loss": 0.5202,
+      "step": 18990
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 39.133480072021484,
+      "learning_rate": 4.1107810735054186e-06,
+      "loss": 1.3954,
+      "step": 18991
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 14.091297149658203,
+      "learning_rate": 4.1099443584487305e-06,
+      "loss": 0.4381,
+      "step": 18992
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 23.905752182006836,
+      "learning_rate": 4.109107643392043e-06,
+      "loss": 1.4625,
+      "step": 18993
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 8.82052993774414,
+      "learning_rate": 4.108270928335355e-06,
+      "loss": 1.88,
+      "step": 18994
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 31.28426742553711,
+      "learning_rate": 4.107434213278668e-06,
+      "loss": 0.8463,
+      "step": 18995
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 25.97709083557129,
+      "learning_rate": 4.106597498221981e-06,
+      "loss": 1.4488,
+      "step": 18996
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.084327220916748,
+      "learning_rate": 4.105760783165294e-06,
+      "loss": 0.4668,
+      "step": 18997
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.0948405265808105,
+      "learning_rate": 4.1049240681086065e-06,
+      "loss": 0.7042,
+      "step": 18998
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 458.3492431640625,
+      "learning_rate": 4.1040873530519184e-06,
+      "loss": 2.2982,
+      "step": 18999
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 16.414735794067383,
+      "learning_rate": 4.10325063799523e-06,
+      "loss": 0.6207,
+      "step": 19000
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 13.103593826293945,
+      "learning_rate": 4.102413922938543e-06,
+      "loss": 0.9018,
+      "step": 19001
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.032532691955566,
+      "learning_rate": 4.101577207881856e-06,
+      "loss": 0.1367,
+      "step": 19002
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 30.239131927490234,
+      "learning_rate": 4.100740492825169e-06,
+      "loss": 1.9775,
+      "step": 19003
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 16.82728385925293,
+      "learning_rate": 4.099903777768482e-06,
+      "loss": 1.0877,
+      "step": 19004
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 56.356014251708984,
+      "learning_rate": 4.099067062711794e-06,
+      "loss": 2.3591,
+      "step": 19005
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 36.4920539855957,
+      "learning_rate": 4.098230347655106e-06,
+      "loss": 1.2582,
+      "step": 19006
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 10.160242080688477,
+      "learning_rate": 4.097393632598418e-06,
+      "loss": 0.9966,
+      "step": 19007
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.926279544830322,
+      "learning_rate": 4.096556917541731e-06,
+      "loss": 0.0569,
+      "step": 19008
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 13.520187377929688,
+      "learning_rate": 4.095720202485044e-06,
+      "loss": 0.714,
+      "step": 19009
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 6.462107181549072,
+      "learning_rate": 4.094883487428357e-06,
+      "loss": 0.1518,
+      "step": 19010
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 124.68399810791016,
+      "learning_rate": 4.0940467723716695e-06,
+      "loss": 2.5772,
+      "step": 19011
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 53.15134048461914,
+      "learning_rate": 4.0932100573149815e-06,
+      "loss": 0.9663,
+      "step": 19012
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 38.00040817260742,
+      "learning_rate": 4.092373342258294e-06,
+      "loss": 1.5956,
+      "step": 19013
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 9.098489761352539,
+      "learning_rate": 4.091536627201606e-06,
+      "loss": 0.3268,
+      "step": 19014
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 46.67536544799805,
+      "learning_rate": 4.090699912144919e-06,
+      "loss": 3.1959,
+      "step": 19015
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 50.20859146118164,
+      "learning_rate": 4.089863197088232e-06,
+      "loss": 0.9767,
+      "step": 19016
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 11.621726036071777,
+      "learning_rate": 4.089026482031545e-06,
+      "loss": 0.9935,
+      "step": 19017
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 6.445420742034912,
+      "learning_rate": 4.0881897669748574e-06,
+      "loss": 0.3812,
+      "step": 19018
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 25.69889259338379,
+      "learning_rate": 4.087353051918169e-06,
+      "loss": 0.7589,
+      "step": 19019
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.7841275930404663,
+      "learning_rate": 4.086516336861482e-06,
+      "loss": 0.1167,
+      "step": 19020
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 6.596001625061035,
+      "learning_rate": 4.085679621804794e-06,
+      "loss": 0.5149,
+      "step": 19021
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 17.258195877075195,
+      "learning_rate": 4.084842906748107e-06,
+      "loss": 0.4251,
+      "step": 19022
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 18.278339385986328,
+      "learning_rate": 4.08400619169142e-06,
+      "loss": 0.9196,
+      "step": 19023
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 15.891921043395996,
+      "learning_rate": 4.0831694766347326e-06,
+      "loss": 1.2198,
+      "step": 19024
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 12.87957763671875,
+      "learning_rate": 4.082332761578045e-06,
+      "loss": 1.3263,
+      "step": 19025
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 8.608296394348145,
+      "learning_rate": 4.081496046521357e-06,
+      "loss": 0.342,
+      "step": 19026
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 17.928401947021484,
+      "learning_rate": 4.08065933146467e-06,
+      "loss": 1.6522,
+      "step": 19027
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 9.473708152770996,
+      "learning_rate": 4.079822616407982e-06,
+      "loss": 0.8328,
+      "step": 19028
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 10.671655654907227,
+      "learning_rate": 4.078985901351295e-06,
+      "loss": 1.4546,
+      "step": 19029
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 69.41703033447266,
+      "learning_rate": 4.078149186294608e-06,
+      "loss": 1.2106,
+      "step": 19030
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 28.889257431030273,
+      "learning_rate": 4.0773124712379205e-06,
+      "loss": 1.7755,
+      "step": 19031
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 9.402454376220703,
+      "learning_rate": 4.076475756181233e-06,
+      "loss": 0.3341,
+      "step": 19032
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 7.450305938720703,
+      "learning_rate": 4.075639041124545e-06,
+      "loss": 0.3173,
+      "step": 19033
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 8.831427574157715,
+      "learning_rate": 4.074802326067858e-06,
+      "loss": 1.5428,
+      "step": 19034
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 18.86667823791504,
+      "learning_rate": 4.07396561101117e-06,
+      "loss": 0.18,
+      "step": 19035
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 16.75472068786621,
+      "learning_rate": 4.073128895954483e-06,
+      "loss": 0.6608,
+      "step": 19036
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 23.9716854095459,
+      "learning_rate": 4.072292180897796e-06,
+      "loss": 1.7005,
+      "step": 19037
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 10.454980850219727,
+      "learning_rate": 4.071455465841108e-06,
+      "loss": 1.2324,
+      "step": 19038
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 33.47016143798828,
+      "learning_rate": 4.070618750784421e-06,
+      "loss": 0.7504,
+      "step": 19039
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 18.456640243530273,
+      "learning_rate": 4.069782035727733e-06,
+      "loss": 1.3785,
+      "step": 19040
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 19.197938919067383,
+      "learning_rate": 4.068945320671046e-06,
+      "loss": 1.41,
+      "step": 19041
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 18.04644012451172,
+      "learning_rate": 4.068108605614358e-06,
+      "loss": 0.7376,
+      "step": 19042
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 22.988388061523438,
+      "learning_rate": 4.067271890557671e-06,
+      "loss": 1.5383,
+      "step": 19043
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.129838943481445,
+      "learning_rate": 4.0664351755009835e-06,
+      "loss": 0.1589,
+      "step": 19044
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 15.090490341186523,
+      "learning_rate": 4.065598460444296e-06,
+      "loss": 0.4905,
+      "step": 19045
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 21.809528350830078,
+      "learning_rate": 4.064761745387609e-06,
+      "loss": 2.1405,
+      "step": 19046
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 40.71038055419922,
+      "learning_rate": 4.063925030330921e-06,
+      "loss": 0.5077,
+      "step": 19047
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 15.136262893676758,
+      "learning_rate": 4.063088315274233e-06,
+      "loss": 1.4303,
+      "step": 19048
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 27.37680435180664,
+      "learning_rate": 4.062251600217546e-06,
+      "loss": 2.7442,
+      "step": 19049
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 7.498646259307861,
+      "learning_rate": 4.061414885160859e-06,
+      "loss": 1.2976,
+      "step": 19050
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 12.314183235168457,
+      "learning_rate": 4.0605781701041714e-06,
+      "loss": 0.6545,
+      "step": 19051
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 13.800204277038574,
+      "learning_rate": 4.059741455047484e-06,
+      "loss": 1.7657,
+      "step": 19052
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 13.44913101196289,
+      "learning_rate": 4.058904739990796e-06,
+      "loss": 0.4265,
+      "step": 19053
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 19.050992965698242,
+      "learning_rate": 4.058068024934109e-06,
+      "loss": 0.9761,
+      "step": 19054
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 30.88131332397461,
+      "learning_rate": 4.057231309877421e-06,
+      "loss": 1.2247,
+      "step": 19055
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 21.770322799682617,
+      "learning_rate": 4.056394594820734e-06,
+      "loss": 0.6648,
+      "step": 19056
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 33.90979766845703,
+      "learning_rate": 4.0555578797640466e-06,
+      "loss": 1.006,
+      "step": 19057
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 26.196422576904297,
+      "learning_rate": 4.054721164707359e-06,
+      "loss": 0.5683,
+      "step": 19058
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.3883891105651855,
+      "learning_rate": 4.053884449650672e-06,
+      "loss": 0.5584,
+      "step": 19059
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 22.187639236450195,
+      "learning_rate": 4.053047734593984e-06,
+      "loss": 2.7119,
+      "step": 19060
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 40.63218688964844,
+      "learning_rate": 4.052211019537297e-06,
+      "loss": 1.6813,
+      "step": 19061
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 9.716521263122559,
+      "learning_rate": 4.051374304480609e-06,
+      "loss": 0.2526,
+      "step": 19062
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 9.759636878967285,
+      "learning_rate": 4.050537589423922e-06,
+      "loss": 0.4243,
+      "step": 19063
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 27.638835906982422,
+      "learning_rate": 4.0497008743672345e-06,
+      "loss": 0.8442,
+      "step": 19064
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 15.038436889648438,
+      "learning_rate": 4.048864159310547e-06,
+      "loss": 0.7551,
+      "step": 19065
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 369.1249084472656,
+      "learning_rate": 4.04802744425386e-06,
+      "loss": 2.6777,
+      "step": 19066
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 12.90273666381836,
+      "learning_rate": 4.047190729197172e-06,
+      "loss": 1.0085,
+      "step": 19067
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 16.095760345458984,
+      "learning_rate": 4.046354014140485e-06,
+      "loss": 0.7467,
+      "step": 19068
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 6.69637393951416,
+      "learning_rate": 4.045517299083797e-06,
+      "loss": 0.5625,
+      "step": 19069
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 19.717710494995117,
+      "learning_rate": 4.04468058402711e-06,
+      "loss": 0.9955,
+      "step": 19070
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 15.726923942565918,
+      "learning_rate": 4.043843868970422e-06,
+      "loss": 2.3147,
+      "step": 19071
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 6.696619033813477,
+      "learning_rate": 4.043007153913735e-06,
+      "loss": 0.4699,
+      "step": 19072
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 14.654072761535645,
+      "learning_rate": 4.042170438857048e-06,
+      "loss": 0.4272,
+      "step": 19073
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 20.2203311920166,
+      "learning_rate": 4.04133372380036e-06,
+      "loss": 1.061,
+      "step": 19074
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 21.050973892211914,
+      "learning_rate": 4.040497008743673e-06,
+      "loss": 0.7387,
+      "step": 19075
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 13.224677085876465,
+      "learning_rate": 4.039660293686985e-06,
+      "loss": 0.5062,
+      "step": 19076
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 10.98463249206543,
+      "learning_rate": 4.0388235786302975e-06,
+      "loss": 1.1087,
+      "step": 19077
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 130.39227294921875,
+      "learning_rate": 4.03798686357361e-06,
+      "loss": 1.1727,
+      "step": 19078
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 20.21027374267578,
+      "learning_rate": 4.037150148516923e-06,
+      "loss": 1.056,
+      "step": 19079
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 9.529265403747559,
+      "learning_rate": 4.036313433460236e-06,
+      "loss": 0.851,
+      "step": 19080
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 18.91885757446289,
+      "learning_rate": 4.035476718403548e-06,
+      "loss": 1.0852,
+      "step": 19081
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 19.511430740356445,
+      "learning_rate": 4.034640003346861e-06,
+      "loss": 1.0616,
+      "step": 19082
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 26.91684341430664,
+      "learning_rate": 4.033803288290173e-06,
+      "loss": 0.1865,
+      "step": 19083
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 11.549873352050781,
+      "learning_rate": 4.032966573233485e-06,
+      "loss": 1.4849,
+      "step": 19084
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 21.561246871948242,
+      "learning_rate": 4.032129858176798e-06,
+      "loss": 1.0529,
+      "step": 19085
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 6.6802754402160645,
+      "learning_rate": 4.031293143120111e-06,
+      "loss": 0.7517,
+      "step": 19086
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 30.852449417114258,
+      "learning_rate": 4.030456428063424e-06,
+      "loss": 0.6937,
+      "step": 19087
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 65.0311050415039,
+      "learning_rate": 4.029619713006736e-06,
+      "loss": 1.7491,
+      "step": 19088
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 21.973583221435547,
+      "learning_rate": 4.028782997950049e-06,
+      "loss": 1.6574,
+      "step": 19089
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 10.833308219909668,
+      "learning_rate": 4.0279462828933605e-06,
+      "loss": 1.639,
+      "step": 19090
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 15.514629364013672,
+      "learning_rate": 4.027109567836673e-06,
+      "loss": 0.9653,
+      "step": 19091
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 30.378694534301758,
+      "learning_rate": 4.026272852779986e-06,
+      "loss": 1.764,
+      "step": 19092
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 22.07337760925293,
+      "learning_rate": 4.025436137723299e-06,
+      "loss": 1.5077,
+      "step": 19093
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.136129856109619,
+      "learning_rate": 4.024599422666611e-06,
+      "loss": 0.1062,
+      "step": 19094
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 6.182592868804932,
+      "learning_rate": 4.023762707609924e-06,
+      "loss": 0.6325,
+      "step": 19095
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 14.115240097045898,
+      "learning_rate": 4.0229259925532365e-06,
+      "loss": 1.8024,
+      "step": 19096
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 44.72590255737305,
+      "learning_rate": 4.0220892774965485e-06,
+      "loss": 3.7641,
+      "step": 19097
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 37.94907760620117,
+      "learning_rate": 4.021252562439861e-06,
+      "loss": 1.1817,
+      "step": 19098
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 26.96363639831543,
+      "learning_rate": 4.020415847383174e-06,
+      "loss": 0.6482,
+      "step": 19099
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 17.43169593811035,
+      "learning_rate": 4.019579132326487e-06,
+      "loss": 0.6672,
+      "step": 19100
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 8.113611221313477,
+      "learning_rate": 4.018742417269799e-06,
+      "loss": 0.5586,
+      "step": 19101
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 33.545021057128906,
+      "learning_rate": 4.017905702213112e-06,
+      "loss": 2.3937,
+      "step": 19102
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 7.373061656951904,
+      "learning_rate": 4.017068987156424e-06,
+      "loss": 1.0229,
+      "step": 19103
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 24.821794509887695,
+      "learning_rate": 4.016232272099736e-06,
+      "loss": 0.8089,
+      "step": 19104
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 11.429865837097168,
+      "learning_rate": 4.015395557043049e-06,
+      "loss": 1.1606,
+      "step": 19105
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 86.04908752441406,
+      "learning_rate": 4.014558841986362e-06,
+      "loss": 1.2803,
+      "step": 19106
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 8.572955131530762,
+      "learning_rate": 4.013722126929675e-06,
+      "loss": 1.4158,
+      "step": 19107
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 19.158706665039062,
+      "learning_rate": 4.012885411872987e-06,
+      "loss": 0.7274,
+      "step": 19108
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 7.97062349319458,
+      "learning_rate": 4.0120486968162995e-06,
+      "loss": 2.0569,
+      "step": 19109
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 22.070228576660156,
+      "learning_rate": 4.011211981759612e-06,
+      "loss": 0.5531,
+      "step": 19110
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 11.128243446350098,
+      "learning_rate": 4.010375266702924e-06,
+      "loss": 0.5593,
+      "step": 19111
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 9.367835998535156,
+      "learning_rate": 4.009538551646237e-06,
+      "loss": 0.3618,
+      "step": 19112
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 23.092803955078125,
+      "learning_rate": 4.00870183658955e-06,
+      "loss": 0.8735,
+      "step": 19113
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 7.3652849197387695,
+      "learning_rate": 4.007865121532863e-06,
+      "loss": 0.729,
+      "step": 19114
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.7102227210998535,
+      "learning_rate": 4.007028406476175e-06,
+      "loss": 0.2235,
+      "step": 19115
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 13.90251636505127,
+      "learning_rate": 4.0061916914194875e-06,
+      "loss": 1.2297,
+      "step": 19116
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 36.02486801147461,
+      "learning_rate": 4.0053549763628e-06,
+      "loss": 1.7103,
+      "step": 19117
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 44.666255950927734,
+      "learning_rate": 4.004518261306112e-06,
+      "loss": 1.6022,
+      "step": 19118
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 29.21457862854004,
+      "learning_rate": 4.003681546249425e-06,
+      "loss": 1.1878,
+      "step": 19119
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 13.307560920715332,
+      "learning_rate": 4.002844831192738e-06,
+      "loss": 0.7035,
+      "step": 19120
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 17.413768768310547,
+      "learning_rate": 4.002008116136051e-06,
+      "loss": 0.7771,
+      "step": 19121
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.2511372566223145,
+      "learning_rate": 4.0011714010793626e-06,
+      "loss": 0.1663,
+      "step": 19122
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 10.610640525817871,
+      "learning_rate": 4.000334686022675e-06,
+      "loss": 0.7032,
+      "step": 19123
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 8.91777515411377,
+      "learning_rate": 3.999497970965988e-06,
+      "loss": 2.1738,
+      "step": 19124
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 17.62346076965332,
+      "learning_rate": 3.9986612559093e-06,
+      "loss": 0.7938,
+      "step": 19125
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 23.286380767822266,
+      "learning_rate": 3.997824540852613e-06,
+      "loss": 1.5049,
+      "step": 19126
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 21.39017105102539,
+      "learning_rate": 3.996987825795926e-06,
+      "loss": 1.7285,
+      "step": 19127
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 21.869333267211914,
+      "learning_rate": 3.9961511107392385e-06,
+      "loss": 1.0293,
+      "step": 19128
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 19.568134307861328,
+      "learning_rate": 3.9953143956825505e-06,
+      "loss": 1.342,
+      "step": 19129
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 13.146456718444824,
+      "learning_rate": 3.994477680625863e-06,
+      "loss": 1.8469,
+      "step": 19130
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 17.104389190673828,
+      "learning_rate": 3.993640965569176e-06,
+      "loss": 0.6416,
+      "step": 19131
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 14.854141235351562,
+      "learning_rate": 3.992804250512488e-06,
+      "loss": 0.7681,
+      "step": 19132
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 31.403024673461914,
+      "learning_rate": 3.991967535455801e-06,
+      "loss": 2.6604,
+      "step": 19133
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 58.17258071899414,
+      "learning_rate": 3.991130820399114e-06,
+      "loss": 1.5447,
+      "step": 19134
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 20.589649200439453,
+      "learning_rate": 3.990294105342426e-06,
+      "loss": 3.3073,
+      "step": 19135
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 11.538389205932617,
+      "learning_rate": 3.989457390285738e-06,
+      "loss": 0.4872,
+      "step": 19136
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 13.821202278137207,
+      "learning_rate": 3.988620675229051e-06,
+      "loss": 1.3287,
+      "step": 19137
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 38.13163375854492,
+      "learning_rate": 3.987783960172363e-06,
+      "loss": 1.3356,
+      "step": 19138
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 69.0101089477539,
+      "learning_rate": 3.986947245115676e-06,
+      "loss": 1.0536,
+      "step": 19139
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 38.037227630615234,
+      "learning_rate": 3.986110530058989e-06,
+      "loss": 0.4617,
+      "step": 19140
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.5604493618011475,
+      "learning_rate": 3.9852738150023016e-06,
+      "loss": 0.1397,
+      "step": 19141
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 11.344226837158203,
+      "learning_rate": 3.9844370999456135e-06,
+      "loss": 0.3115,
+      "step": 19142
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 22.58791732788086,
+      "learning_rate": 3.983600384888926e-06,
+      "loss": 1.9093,
+      "step": 19143
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 14.093499183654785,
+      "learning_rate": 3.982763669832239e-06,
+      "loss": 0.4885,
+      "step": 19144
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 12.28605842590332,
+      "learning_rate": 3.981926954775551e-06,
+      "loss": 1.5395,
+      "step": 19145
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 200.0440673828125,
+      "learning_rate": 3.981090239718864e-06,
+      "loss": 1.5086,
+      "step": 19146
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 15.651595115661621,
+      "learning_rate": 3.980253524662177e-06,
+      "loss": 0.5725,
+      "step": 19147
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 99.0028076171875,
+      "learning_rate": 3.9794168096054895e-06,
+      "loss": 1.0011,
+      "step": 19148
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 31.75554847717285,
+      "learning_rate": 3.9785800945488014e-06,
+      "loss": 0.994,
+      "step": 19149
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 35.45192337036133,
+      "learning_rate": 3.977743379492114e-06,
+      "loss": 1.4886,
+      "step": 19150
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 14.628107070922852,
+      "learning_rate": 3.976906664435427e-06,
+      "loss": 0.4394,
+      "step": 19151
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 14.06891918182373,
+      "learning_rate": 3.976069949378739e-06,
+      "loss": 1.1679,
+      "step": 19152
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 12.731145858764648,
+      "learning_rate": 3.975233234322052e-06,
+      "loss": 1.7926,
+      "step": 19153
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 11.016291618347168,
+      "learning_rate": 3.974396519265365e-06,
+      "loss": 0.4143,
+      "step": 19154
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 26.453344345092773,
+      "learning_rate": 3.973559804208677e-06,
+      "loss": 1.2922,
+      "step": 19155
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 22.435813903808594,
+      "learning_rate": 3.972723089151989e-06,
+      "loss": 0.9505,
+      "step": 19156
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 14.01541805267334,
+      "learning_rate": 3.971886374095302e-06,
+      "loss": 0.505,
+      "step": 19157
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 18.743432998657227,
+      "learning_rate": 3.971049659038615e-06,
+      "loss": 1.2548,
+      "step": 19158
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 23.11290740966797,
+      "learning_rate": 3.970212943981927e-06,
+      "loss": 1.0154,
+      "step": 19159
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 24.722064971923828,
+      "learning_rate": 3.96937622892524e-06,
+      "loss": 0.6556,
+      "step": 19160
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 10.541865348815918,
+      "learning_rate": 3.9685395138685525e-06,
+      "loss": 0.4639,
+      "step": 19161
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 31.57651710510254,
+      "learning_rate": 3.967702798811865e-06,
+      "loss": 1.409,
+      "step": 19162
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 6.214509963989258,
+      "learning_rate": 3.966866083755177e-06,
+      "loss": 0.5104,
+      "step": 19163
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 11.101883888244629,
+      "learning_rate": 3.96602936869849e-06,
+      "loss": 0.7904,
+      "step": 19164
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 22.653268814086914,
+      "learning_rate": 3.965192653641803e-06,
+      "loss": 2.0983,
+      "step": 19165
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 20.678075790405273,
+      "learning_rate": 3.964355938585115e-06,
+      "loss": 1.9968,
+      "step": 19166
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 80.93486785888672,
+      "learning_rate": 3.963519223528428e-06,
+      "loss": 0.8472,
+      "step": 19167
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 9.310606956481934,
+      "learning_rate": 3.9626825084717404e-06,
+      "loss": 0.5267,
+      "step": 19168
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 17.505313873291016,
+      "learning_rate": 3.961845793415052e-06,
+      "loss": 2.1095,
+      "step": 19169
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 6.054166793823242,
+      "learning_rate": 3.961009078358365e-06,
+      "loss": 0.2008,
+      "step": 19170
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 28.922883987426758,
+      "learning_rate": 3.960172363301678e-06,
+      "loss": 1.1252,
+      "step": 19171
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 7.792020320892334,
+      "learning_rate": 3.959335648244991e-06,
+      "loss": 0.5446,
+      "step": 19172
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 55.393497467041016,
+      "learning_rate": 3.958498933188303e-06,
+      "loss": 2.2708,
+      "step": 19173
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 10.111420631408691,
+      "learning_rate": 3.9576622181316156e-06,
+      "loss": 1.2057,
+      "step": 19174
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 17.73046112060547,
+      "learning_rate": 3.956825503074928e-06,
+      "loss": 1.1219,
+      "step": 19175
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 8.470479011535645,
+      "learning_rate": 3.95598878801824e-06,
+      "loss": 0.6359,
+      "step": 19176
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.160224199295044,
+      "learning_rate": 3.955152072961553e-06,
+      "loss": 0.1297,
+      "step": 19177
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 7.111017227172852,
+      "learning_rate": 3.954315357904866e-06,
+      "loss": 0.3905,
+      "step": 19178
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 15.66545295715332,
+      "learning_rate": 3.953478642848179e-06,
+      "loss": 0.825,
+      "step": 19179
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 27.49559211730957,
+      "learning_rate": 3.952641927791491e-06,
+      "loss": 0.533,
+      "step": 19180
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 22.51540756225586,
+      "learning_rate": 3.9518052127348035e-06,
+      "loss": 1.4591,
+      "step": 19181
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 23.426937103271484,
+      "learning_rate": 3.950968497678116e-06,
+      "loss": 1.9108,
+      "step": 19182
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 27.181591033935547,
+      "learning_rate": 3.950131782621428e-06,
+      "loss": 1.0922,
+      "step": 19183
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 7.732165813446045,
+      "learning_rate": 3.949295067564741e-06,
+      "loss": 0.2493,
+      "step": 19184
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 10.207642555236816,
+      "learning_rate": 3.948458352508054e-06,
+      "loss": 1.5212,
+      "step": 19185
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 9.41203498840332,
+      "learning_rate": 3.947621637451367e-06,
+      "loss": 1.6123,
+      "step": 19186
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 9.990777015686035,
+      "learning_rate": 3.946784922394679e-06,
+      "loss": 0.4817,
+      "step": 19187
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.777212142944336,
+      "learning_rate": 3.945948207337991e-06,
+      "loss": 0.3308,
+      "step": 19188
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 16.02861976623535,
+      "learning_rate": 3.945111492281304e-06,
+      "loss": 0.4803,
+      "step": 19189
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 9.46855354309082,
+      "learning_rate": 3.944274777224616e-06,
+      "loss": 0.8299,
+      "step": 19190
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.9646720886230469,
+      "learning_rate": 3.943438062167929e-06,
+      "loss": 0.073,
+      "step": 19191
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 10.14421272277832,
+      "learning_rate": 3.942601347111242e-06,
+      "loss": 0.4138,
+      "step": 19192
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 8.660815238952637,
+      "learning_rate": 3.9417646320545546e-06,
+      "loss": 0.7271,
+      "step": 19193
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 18.654617309570312,
+      "learning_rate": 3.9409279169978665e-06,
+      "loss": 1.6773,
+      "step": 19194
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 120.55852508544922,
+      "learning_rate": 3.940091201941179e-06,
+      "loss": 2.4435,
+      "step": 19195
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 6.149465084075928,
+      "learning_rate": 3.939254486884492e-06,
+      "loss": 0.2222,
+      "step": 19196
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 23.639225006103516,
+      "learning_rate": 3.938417771827804e-06,
+      "loss": 0.5304,
+      "step": 19197
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 13.503562927246094,
+      "learning_rate": 3.937581056771117e-06,
+      "loss": 0.5163,
+      "step": 19198
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 39.53706741333008,
+      "learning_rate": 3.93674434171443e-06,
+      "loss": 1.2687,
+      "step": 19199
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 6.831623554229736,
+      "learning_rate": 3.9359076266577425e-06,
+      "loss": 1.3298,
+      "step": 19200
+    },
+    {
+      "epoch": 2.41,
+      "eval_loss": 0.07908567786216736,
+      "eval_runtime": 96.7748,
+      "eval_samples_per_second": 36.6,
+      "eval_steps_per_second": 36.6,
+      "step": 19200
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 117.7660140991211,
+      "learning_rate": 3.9350709116010544e-06,
+      "loss": 1.8972,
+      "step": 19201
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 17.90907859802246,
+      "learning_rate": 3.934234196544367e-06,
+      "loss": 0.7705,
+      "step": 19202
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 11.033977508544922,
+      "learning_rate": 3.93339748148768e-06,
+      "loss": 0.4761,
+      "step": 19203
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 6.191403865814209,
+      "learning_rate": 3.932560766430992e-06,
+      "loss": 0.3172,
+      "step": 19204
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 18.340211868286133,
+      "learning_rate": 3.931724051374305e-06,
+      "loss": 0.963,
+      "step": 19205
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 16.469741821289062,
+      "learning_rate": 3.930887336317618e-06,
+      "loss": 1.8976,
+      "step": 19206
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 10.327061653137207,
+      "learning_rate": 3.93005062126093e-06,
+      "loss": 0.8535,
+      "step": 19207
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 20.400331497192383,
+      "learning_rate": 3.929213906204242e-06,
+      "loss": 0.5511,
+      "step": 19208
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 8.918400764465332,
+      "learning_rate": 3.928377191147555e-06,
+      "loss": 0.5542,
+      "step": 19209
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 16.06703758239746,
+      "learning_rate": 3.927540476090867e-06,
+      "loss": 1.9891,
+      "step": 19210
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 42.195884704589844,
+      "learning_rate": 3.92670376103418e-06,
+      "loss": 1.3129,
+      "step": 19211
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 30.93027114868164,
+      "learning_rate": 3.925867045977493e-06,
+      "loss": 1.1214,
+      "step": 19212
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.7529571056365967,
+      "learning_rate": 3.9250303309208055e-06,
+      "loss": 0.1369,
+      "step": 19213
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 6.6867828369140625,
+      "learning_rate": 3.924193615864118e-06,
+      "loss": 0.6473,
+      "step": 19214
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 7.075080394744873,
+      "learning_rate": 3.92335690080743e-06,
+      "loss": 0.2434,
+      "step": 19215
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 18.304428100585938,
+      "learning_rate": 3.922520185750743e-06,
+      "loss": 1.6788,
+      "step": 19216
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 7.171009540557861,
+      "learning_rate": 3.921683470694055e-06,
+      "loss": 0.171,
+      "step": 19217
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 13.496152877807617,
+      "learning_rate": 3.920846755637368e-06,
+      "loss": 1.196,
+      "step": 19218
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 14.460394859313965,
+      "learning_rate": 3.920010040580681e-06,
+      "loss": 1.0907,
+      "step": 19219
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 16.681568145751953,
+      "learning_rate": 3.9191733255239934e-06,
+      "loss": 0.7428,
+      "step": 19220
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 14.806492805480957,
+      "learning_rate": 3.918336610467306e-06,
+      "loss": 1.4748,
+      "step": 19221
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.13444709777832,
+      "learning_rate": 3.917499895410618e-06,
+      "loss": 1.0966,
+      "step": 19222
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 10.014902114868164,
+      "learning_rate": 3.916663180353931e-06,
+      "loss": 0.2904,
+      "step": 19223
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 19.748510360717773,
+      "learning_rate": 3.915826465297243e-06,
+      "loss": 2.6818,
+      "step": 19224
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 15.84622859954834,
+      "learning_rate": 3.914989750240556e-06,
+      "loss": 1.2931,
+      "step": 19225
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 12.150506019592285,
+      "learning_rate": 3.9141530351838685e-06,
+      "loss": 0.8015,
+      "step": 19226
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 11.955732345581055,
+      "learning_rate": 3.913316320127181e-06,
+      "loss": 0.9369,
+      "step": 19227
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 12.21967887878418,
+      "learning_rate": 3.912479605070493e-06,
+      "loss": 0.6531,
+      "step": 19228
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 24.59291648864746,
+      "learning_rate": 3.911642890013806e-06,
+      "loss": 1.1242,
+      "step": 19229
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 33.7893180847168,
+      "learning_rate": 3.910806174957119e-06,
+      "loss": 1.9844,
+      "step": 19230
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 32.203125,
+      "learning_rate": 3.909969459900431e-06,
+      "loss": 1.6182,
+      "step": 19231
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 18.622575759887695,
+      "learning_rate": 3.909132744843744e-06,
+      "loss": 0.869,
+      "step": 19232
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 14.749338150024414,
+      "learning_rate": 3.9082960297870565e-06,
+      "loss": 0.7788,
+      "step": 19233
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 14.998071670532227,
+      "learning_rate": 3.907459314730369e-06,
+      "loss": 1.0434,
+      "step": 19234
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 148.5703887939453,
+      "learning_rate": 3.906622599673681e-06,
+      "loss": 1.4984,
+      "step": 19235
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 7.36978006362915,
+      "learning_rate": 3.905785884616994e-06,
+      "loss": 0.4504,
+      "step": 19236
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 8.224552154541016,
+      "learning_rate": 3.904949169560307e-06,
+      "loss": 0.4879,
+      "step": 19237
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 32.19786834716797,
+      "learning_rate": 3.904112454503619e-06,
+      "loss": 0.5529,
+      "step": 19238
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 40.72687530517578,
+      "learning_rate": 3.903275739446932e-06,
+      "loss": 1.2672,
+      "step": 19239
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 22.027299880981445,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 1.2204,
+      "step": 19240
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 60.05638885498047,
+      "learning_rate": 3.901602309333557e-06,
+      "loss": 1.807,
+      "step": 19241
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 18.618318557739258,
+      "learning_rate": 3.900765594276869e-06,
+      "loss": 1.3002,
+      "step": 19242
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 11.675912857055664,
+      "learning_rate": 3.899928879220182e-06,
+      "loss": 0.534,
+      "step": 19243
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 21.279949188232422,
+      "learning_rate": 3.899092164163495e-06,
+      "loss": 0.5592,
+      "step": 19244
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 105.58536529541016,
+      "learning_rate": 3.898255449106807e-06,
+      "loss": 1.5623,
+      "step": 19245
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 30.80086326599121,
+      "learning_rate": 3.8974187340501195e-06,
+      "loss": 1.1525,
+      "step": 19246
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 6.376068115234375,
+      "learning_rate": 3.896582018993432e-06,
+      "loss": 1.0781,
+      "step": 19247
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 11.469795227050781,
+      "learning_rate": 3.895745303936745e-06,
+      "loss": 0.9443,
+      "step": 19248
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 11.311722755432129,
+      "learning_rate": 3.894908588880057e-06,
+      "loss": 1.2709,
+      "step": 19249
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 8.450668334960938,
+      "learning_rate": 3.89407187382337e-06,
+      "loss": 1.0794,
+      "step": 19250
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 33.71548080444336,
+      "learning_rate": 3.893235158766682e-06,
+      "loss": 1.5056,
+      "step": 19251
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 115.70248413085938,
+      "learning_rate": 3.892398443709995e-06,
+      "loss": 0.8059,
+      "step": 19252
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 13.588934898376465,
+      "learning_rate": 3.891561728653307e-06,
+      "loss": 2.1443,
+      "step": 19253
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 20.4144344329834,
+      "learning_rate": 3.89072501359662e-06,
+      "loss": 2.0834,
+      "step": 19254
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 13.07769775390625,
+      "learning_rate": 3.889888298539933e-06,
+      "loss": 0.9905,
+      "step": 19255
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 59.85422134399414,
+      "learning_rate": 3.889051583483245e-06,
+      "loss": 1.6588,
+      "step": 19256
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 17.695573806762695,
+      "learning_rate": 3.888214868426558e-06,
+      "loss": 1.8516,
+      "step": 19257
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.559680461883545,
+      "learning_rate": 3.88737815336987e-06,
+      "loss": 1.1526,
+      "step": 19258
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 11.505620956420898,
+      "learning_rate": 3.8865414383131825e-06,
+      "loss": 0.9919,
+      "step": 19259
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 9.828024864196777,
+      "learning_rate": 3.885704723256495e-06,
+      "loss": 0.8659,
+      "step": 19260
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 10.574488639831543,
+      "learning_rate": 3.884868008199808e-06,
+      "loss": 1.7149,
+      "step": 19261
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 8.784329414367676,
+      "learning_rate": 3.884031293143121e-06,
+      "loss": 0.2339,
+      "step": 19262
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 31.123836517333984,
+      "learning_rate": 3.883194578086433e-06,
+      "loss": 1.7855,
+      "step": 19263
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 26.213031768798828,
+      "learning_rate": 3.882357863029746e-06,
+      "loss": 1.5882,
+      "step": 19264
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 10.312332153320312,
+      "learning_rate": 3.881521147973058e-06,
+      "loss": 0.5542,
+      "step": 19265
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 10.933004379272461,
+      "learning_rate": 3.8806844329163705e-06,
+      "loss": 0.526,
+      "step": 19266
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 9.574731826782227,
+      "learning_rate": 3.879847717859683e-06,
+      "loss": 0.3757,
+      "step": 19267
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 10.87021255493164,
+      "learning_rate": 3.879011002802996e-06,
+      "loss": 2.3607,
+      "step": 19268
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 6.682435989379883,
+      "learning_rate": 3.878174287746309e-06,
+      "loss": 1.4867,
+      "step": 19269
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 380.7222900390625,
+      "learning_rate": 3.877337572689621e-06,
+      "loss": 1.6735,
+      "step": 19270
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 17.722379684448242,
+      "learning_rate": 3.876500857632934e-06,
+      "loss": 1.0548,
+      "step": 19271
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 8.213561058044434,
+      "learning_rate": 3.8756641425762456e-06,
+      "loss": 0.6286,
+      "step": 19272
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 10.587211608886719,
+      "learning_rate": 3.874827427519558e-06,
+      "loss": 0.4255,
+      "step": 19273
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 21.14800262451172,
+      "learning_rate": 3.873990712462871e-06,
+      "loss": 0.9089,
+      "step": 19274
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 26.88003921508789,
+      "learning_rate": 3.873153997406184e-06,
+      "loss": 2.877,
+      "step": 19275
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.936546325683594,
+      "learning_rate": 3.872317282349497e-06,
+      "loss": 0.4939,
+      "step": 19276
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 18.24165153503418,
+      "learning_rate": 3.871480567292809e-06,
+      "loss": 0.6357,
+      "step": 19277
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 10.405789375305176,
+      "learning_rate": 3.8706438522361215e-06,
+      "loss": 1.3881,
+      "step": 19278
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 8.698683738708496,
+      "learning_rate": 3.8698071371794335e-06,
+      "loss": 0.9599,
+      "step": 19279
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.385855674743652,
+      "learning_rate": 3.868970422122746e-06,
+      "loss": 0.2265,
+      "step": 19280
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 10.598146438598633,
+      "learning_rate": 3.868133707066059e-06,
+      "loss": 1.0209,
+      "step": 19281
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 6.441844463348389,
+      "learning_rate": 3.867296992009372e-06,
+      "loss": 0.7301,
+      "step": 19282
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 15.066563606262207,
+      "learning_rate": 3.866460276952685e-06,
+      "loss": 0.4305,
+      "step": 19283
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 26.297943115234375,
+      "learning_rate": 3.865623561895997e-06,
+      "loss": 1.183,
+      "step": 19284
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 2.300495147705078,
+      "learning_rate": 3.8647868468393095e-06,
+      "loss": 0.1023,
+      "step": 19285
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 17.563278198242188,
+      "learning_rate": 3.863950131782621e-06,
+      "loss": 1.1672,
+      "step": 19286
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 29.381669998168945,
+      "learning_rate": 3.863113416725934e-06,
+      "loss": 0.734,
+      "step": 19287
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 42.8314094543457,
+      "learning_rate": 3.862276701669247e-06,
+      "loss": 1.2531,
+      "step": 19288
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 11.0601167678833,
+      "learning_rate": 3.86143998661256e-06,
+      "loss": 0.5414,
+      "step": 19289
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 39.98393249511719,
+      "learning_rate": 3.860603271555873e-06,
+      "loss": 1.9857,
+      "step": 19290
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 157.44747924804688,
+      "learning_rate": 3.8597665564991846e-06,
+      "loss": 1.1885,
+      "step": 19291
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 11.619778633117676,
+      "learning_rate": 3.8589298414424965e-06,
+      "loss": 1.5894,
+      "step": 19292
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 14.429756164550781,
+      "learning_rate": 3.858093126385809e-06,
+      "loss": 0.9863,
+      "step": 19293
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 22.345094680786133,
+      "learning_rate": 3.857256411329122e-06,
+      "loss": 1.3794,
+      "step": 19294
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.763068199157715,
+      "learning_rate": 3.856419696272435e-06,
+      "loss": 1.5051,
+      "step": 19295
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 29.449905395507812,
+      "learning_rate": 3.855582981215748e-06,
+      "loss": 1.2129,
+      "step": 19296
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 16.893463134765625,
+      "learning_rate": 3.85474626615906e-06,
+      "loss": 0.7417,
+      "step": 19297
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.6840009689331055,
+      "learning_rate": 3.8539095511023725e-06,
+      "loss": 0.2945,
+      "step": 19298
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 55.94939041137695,
+      "learning_rate": 3.8530728360456844e-06,
+      "loss": 0.2093,
+      "step": 19299
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 188.28958129882812,
+      "learning_rate": 3.852236120988997e-06,
+      "loss": 1.5405,
+      "step": 19300
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 6.573770046234131,
+      "learning_rate": 3.85139940593231e-06,
+      "loss": 0.0694,
+      "step": 19301
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 17.996570587158203,
+      "learning_rate": 3.850562690875623e-06,
+      "loss": 1.7903,
+      "step": 19302
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 6.312846660614014,
+      "learning_rate": 3.849725975818936e-06,
+      "loss": 0.1564,
+      "step": 19303
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 16.640140533447266,
+      "learning_rate": 3.848889260762248e-06,
+      "loss": 1.5704,
+      "step": 19304
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 18.343538284301758,
+      "learning_rate": 3.84805254570556e-06,
+      "loss": 1.6799,
+      "step": 19305
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 44.953529357910156,
+      "learning_rate": 3.847215830648872e-06,
+      "loss": 0.5269,
+      "step": 19306
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 18.762069702148438,
+      "learning_rate": 3.846379115592185e-06,
+      "loss": 0.8328,
+      "step": 19307
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 13.904192924499512,
+      "learning_rate": 3.845542400535498e-06,
+      "loss": 0.6432,
+      "step": 19308
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 20.194324493408203,
+      "learning_rate": 3.844705685478811e-06,
+      "loss": 0.8736,
+      "step": 19309
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 23.708715438842773,
+      "learning_rate": 3.8438689704221236e-06,
+      "loss": 0.9779,
+      "step": 19310
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 10.490565299987793,
+      "learning_rate": 3.8430322553654355e-06,
+      "loss": 0.8575,
+      "step": 19311
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 9.342899322509766,
+      "learning_rate": 3.842195540308748e-06,
+      "loss": 0.2516,
+      "step": 19312
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 6.65676212310791,
+      "learning_rate": 3.84135882525206e-06,
+      "loss": 0.4363,
+      "step": 19313
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 19.738954544067383,
+      "learning_rate": 3.840522110195373e-06,
+      "loss": 0.7261,
+      "step": 19314
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 9.836162567138672,
+      "learning_rate": 3.839685395138686e-06,
+      "loss": 0.5298,
+      "step": 19315
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.317717552185059,
+      "learning_rate": 3.838848680081999e-06,
+      "loss": 0.46,
+      "step": 19316
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 12.451818466186523,
+      "learning_rate": 3.8380119650253115e-06,
+      "loss": 0.8643,
+      "step": 19317
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 11.131498336791992,
+      "learning_rate": 3.8371752499686234e-06,
+      "loss": 0.6231,
+      "step": 19318
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 6.2766523361206055,
+      "learning_rate": 3.836338534911936e-06,
+      "loss": 1.1419,
+      "step": 19319
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 12.388383865356445,
+      "learning_rate": 3.835501819855248e-06,
+      "loss": 0.3583,
+      "step": 19320
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 21.208301544189453,
+      "learning_rate": 3.834665104798561e-06,
+      "loss": 0.4861,
+      "step": 19321
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 16.611286163330078,
+      "learning_rate": 3.833828389741874e-06,
+      "loss": 0.612,
+      "step": 19322
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 11.50299072265625,
+      "learning_rate": 3.832991674685187e-06,
+      "loss": 1.377,
+      "step": 19323
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 12.140826225280762,
+      "learning_rate": 3.832154959628499e-06,
+      "loss": 0.5706,
+      "step": 19324
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 20.794418334960938,
+      "learning_rate": 3.831318244571811e-06,
+      "loss": 1.0795,
+      "step": 19325
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 27.155277252197266,
+      "learning_rate": 3.830481529515124e-06,
+      "loss": 1.8459,
+      "step": 19326
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 25.26287269592285,
+      "learning_rate": 3.829644814458436e-06,
+      "loss": 1.5525,
+      "step": 19327
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 6.756760597229004,
+      "learning_rate": 3.828808099401749e-06,
+      "loss": 0.44,
+      "step": 19328
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 10.38017463684082,
+      "learning_rate": 3.827971384345062e-06,
+      "loss": 0.3319,
+      "step": 19329
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 16.2598934173584,
+      "learning_rate": 3.8271346692883745e-06,
+      "loss": 1.7218,
+      "step": 19330
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 12.016488075256348,
+      "learning_rate": 3.826297954231687e-06,
+      "loss": 0.7483,
+      "step": 19331
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 8.901429176330566,
+      "learning_rate": 3.825461239174999e-06,
+      "loss": 1.7516,
+      "step": 19332
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 25.796611785888672,
+      "learning_rate": 3.824624524118311e-06,
+      "loss": 0.5186,
+      "step": 19333
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 112.78844451904297,
+      "learning_rate": 3.823787809061624e-06,
+      "loss": 0.7428,
+      "step": 19334
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 48.96705627441406,
+      "learning_rate": 3.822951094004937e-06,
+      "loss": 1.0306,
+      "step": 19335
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 12.96463394165039,
+      "learning_rate": 3.82211437894825e-06,
+      "loss": 2.505,
+      "step": 19336
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 25.676706314086914,
+      "learning_rate": 3.8212776638915624e-06,
+      "loss": 0.8836,
+      "step": 19337
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 6.562462329864502,
+      "learning_rate": 3.820440948834874e-06,
+      "loss": 0.2133,
+      "step": 19338
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 39.64939880371094,
+      "learning_rate": 3.819604233778187e-06,
+      "loss": 0.7905,
+      "step": 19339
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 18.159381866455078,
+      "learning_rate": 3.818767518721499e-06,
+      "loss": 0.7596,
+      "step": 19340
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 6.936798572540283,
+      "learning_rate": 3.817930803664812e-06,
+      "loss": 0.2851,
+      "step": 19341
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 14.394283294677734,
+      "learning_rate": 3.817094088608125e-06,
+      "loss": 1.9691,
+      "step": 19342
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 11.488725662231445,
+      "learning_rate": 3.8162573735514376e-06,
+      "loss": 1.0727,
+      "step": 19343
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 6.755542755126953,
+      "learning_rate": 3.81542065849475e-06,
+      "loss": 0.238,
+      "step": 19344
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 15.270042419433594,
+      "learning_rate": 3.8145839434380627e-06,
+      "loss": 0.4638,
+      "step": 19345
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 22.63208770751953,
+      "learning_rate": 3.8137472283813747e-06,
+      "loss": 1.8979,
+      "step": 19346
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 45.051815032958984,
+      "learning_rate": 3.8129105133246875e-06,
+      "loss": 1.3845,
+      "step": 19347
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.2802832126617432,
+      "learning_rate": 3.812073798268e-06,
+      "loss": 0.0313,
+      "step": 19348
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 17.48478126525879,
+      "learning_rate": 3.8112370832113127e-06,
+      "loss": 0.9503,
+      "step": 19349
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 10.073248863220215,
+      "learning_rate": 3.8104003681546255e-06,
+      "loss": 0.4555,
+      "step": 19350
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 30.915050506591797,
+      "learning_rate": 3.809563653097938e-06,
+      "loss": 1.0442,
+      "step": 19351
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 10.166717529296875,
+      "learning_rate": 3.8087269380412507e-06,
+      "loss": 1.2188,
+      "step": 19352
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 11.059699058532715,
+      "learning_rate": 3.8078902229845626e-06,
+      "loss": 0.0837,
+      "step": 19353
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 21.718486785888672,
+      "learning_rate": 3.8070535079278754e-06,
+      "loss": 2.2156,
+      "step": 19354
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 16.999832153320312,
+      "learning_rate": 3.806216792871188e-06,
+      "loss": 0.7584,
+      "step": 19355
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 11.521601676940918,
+      "learning_rate": 3.8053800778145006e-06,
+      "loss": 1.0624,
+      "step": 19356
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 16.258691787719727,
+      "learning_rate": 3.8045433627578134e-06,
+      "loss": 0.4942,
+      "step": 19357
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 8.650497436523438,
+      "learning_rate": 3.8037066477011258e-06,
+      "loss": 1.1155,
+      "step": 19358
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 10.786202430725098,
+      "learning_rate": 3.8028699326444386e-06,
+      "loss": 0.5909,
+      "step": 19359
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 33.60354232788086,
+      "learning_rate": 3.8020332175877505e-06,
+      "loss": 0.3585,
+      "step": 19360
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 6.411097526550293,
+      "learning_rate": 3.8011965025310633e-06,
+      "loss": 0.22,
+      "step": 19361
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 9.783259391784668,
+      "learning_rate": 3.8003597874743757e-06,
+      "loss": 0.9092,
+      "step": 19362
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 8.718094825744629,
+      "learning_rate": 3.7995230724176885e-06,
+      "loss": 0.5791,
+      "step": 19363
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 24.100753784179688,
+      "learning_rate": 3.7986863573610013e-06,
+      "loss": 0.7816,
+      "step": 19364
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 25.01043701171875,
+      "learning_rate": 3.7978496423043137e-06,
+      "loss": 2.6373,
+      "step": 19365
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 14.57491683959961,
+      "learning_rate": 3.7970129272476265e-06,
+      "loss": 0.3222,
+      "step": 19366
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 6.600212097167969,
+      "learning_rate": 3.7961762121909385e-06,
+      "loss": 0.1352,
+      "step": 19367
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 14.061747550964355,
+      "learning_rate": 3.7953394971342513e-06,
+      "loss": 0.5259,
+      "step": 19368
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 35.42488098144531,
+      "learning_rate": 3.7945027820775636e-06,
+      "loss": 1.9063,
+      "step": 19369
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 17.88988494873047,
+      "learning_rate": 3.7936660670208764e-06,
+      "loss": 1.4857,
+      "step": 19370
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 12.470402717590332,
+      "learning_rate": 3.792829351964189e-06,
+      "loss": 0.6173,
+      "step": 19371
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.856550931930542,
+      "learning_rate": 3.7919926369075016e-06,
+      "loss": 0.0938,
+      "step": 19372
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 12.61774730682373,
+      "learning_rate": 3.7911559218508144e-06,
+      "loss": 0.1798,
+      "step": 19373
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 181.10760498046875,
+      "learning_rate": 3.7903192067941264e-06,
+      "loss": 2.1395,
+      "step": 19374
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 44.0191650390625,
+      "learning_rate": 3.789482491737439e-06,
+      "loss": 1.591,
+      "step": 19375
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 87.66629028320312,
+      "learning_rate": 3.7886457766807515e-06,
+      "loss": 0.7762,
+      "step": 19376
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 11.412515640258789,
+      "learning_rate": 3.7878090616240644e-06,
+      "loss": 0.5234,
+      "step": 19377
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 43.73072814941406,
+      "learning_rate": 3.7869723465673767e-06,
+      "loss": 2.9941,
+      "step": 19378
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 12.963578224182129,
+      "learning_rate": 3.7861356315106895e-06,
+      "loss": 0.8023,
+      "step": 19379
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 10.066496849060059,
+      "learning_rate": 3.7852989164540023e-06,
+      "loss": 1.0838,
+      "step": 19380
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 2.953787088394165,
+      "learning_rate": 3.7844622013973143e-06,
+      "loss": 0.2104,
+      "step": 19381
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 20.860986709594727,
+      "learning_rate": 3.7836254863406267e-06,
+      "loss": 1.8658,
+      "step": 19382
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 17.35547637939453,
+      "learning_rate": 3.7827887712839395e-06,
+      "loss": 0.6201,
+      "step": 19383
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 9.459033012390137,
+      "learning_rate": 3.7819520562272523e-06,
+      "loss": 0.2003,
+      "step": 19384
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 15.11513614654541,
+      "learning_rate": 3.7811153411705646e-06,
+      "loss": 0.4576,
+      "step": 19385
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.2749271392822266,
+      "learning_rate": 3.7802786261138775e-06,
+      "loss": 0.0915,
+      "step": 19386
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 8.152831077575684,
+      "learning_rate": 3.7794419110571903e-06,
+      "loss": 0.7088,
+      "step": 19387
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 7.579389572143555,
+      "learning_rate": 3.778605196000502e-06,
+      "loss": 0.7091,
+      "step": 19388
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 10.391526222229004,
+      "learning_rate": 3.7777684809438146e-06,
+      "loss": 0.7642,
+      "step": 19389
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 86.48945617675781,
+      "learning_rate": 3.7769317658871274e-06,
+      "loss": 1.9806,
+      "step": 19390
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 10.321609497070312,
+      "learning_rate": 3.77609505083044e-06,
+      "loss": 0.9803,
+      "step": 19391
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 8.198219299316406,
+      "learning_rate": 3.7752583357737526e-06,
+      "loss": 0.5157,
+      "step": 19392
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 12.072665214538574,
+      "learning_rate": 3.7744216207170654e-06,
+      "loss": 1.122,
+      "step": 19393
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 16.971315383911133,
+      "learning_rate": 3.7735849056603777e-06,
+      "loss": 0.9175,
+      "step": 19394
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 6.218404769897461,
+      "learning_rate": 3.77274819060369e-06,
+      "loss": 0.4567,
+      "step": 19395
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 11.903154373168945,
+      "learning_rate": 3.7719114755470025e-06,
+      "loss": 1.9367,
+      "step": 19396
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 35.32255172729492,
+      "learning_rate": 3.7710747604903153e-06,
+      "loss": 0.7885,
+      "step": 19397
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 10.045111656188965,
+      "learning_rate": 3.770238045433628e-06,
+      "loss": 0.3523,
+      "step": 19398
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 6.714310169219971,
+      "learning_rate": 3.7694013303769405e-06,
+      "loss": 0.8446,
+      "step": 19399
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 214.91531372070312,
+      "learning_rate": 3.7685646153202533e-06,
+      "loss": 1.7278,
+      "step": 19400
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 12.572393417358398,
+      "learning_rate": 3.7677279002635652e-06,
+      "loss": 0.8203,
+      "step": 19401
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 42.073081970214844,
+      "learning_rate": 3.766891185206878e-06,
+      "loss": 2.2245,
+      "step": 19402
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.576661586761475,
+      "learning_rate": 3.7660544701501904e-06,
+      "loss": 0.2068,
+      "step": 19403
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 9.408468246459961,
+      "learning_rate": 3.7652177550935032e-06,
+      "loss": 0.6677,
+      "step": 19404
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 17.082490921020508,
+      "learning_rate": 3.764381040036816e-06,
+      "loss": 1.1447,
+      "step": 19405
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 8.84753131866455,
+      "learning_rate": 3.7635443249801284e-06,
+      "loss": 0.7181,
+      "step": 19406
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 12.232440948486328,
+      "learning_rate": 3.762707609923441e-06,
+      "loss": 0.8382,
+      "step": 19407
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 29.958683013916016,
+      "learning_rate": 3.761870894866753e-06,
+      "loss": 1.5797,
+      "step": 19408
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 26.029563903808594,
+      "learning_rate": 3.761034179810066e-06,
+      "loss": 2.1999,
+      "step": 19409
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 14.655035972595215,
+      "learning_rate": 3.7601974647533783e-06,
+      "loss": 0.9913,
+      "step": 19410
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 7.407000541687012,
+      "learning_rate": 3.759360749696691e-06,
+      "loss": 0.831,
+      "step": 19411
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 11.39689826965332,
+      "learning_rate": 3.7585240346400035e-06,
+      "loss": 0.4357,
+      "step": 19412
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 21.788785934448242,
+      "learning_rate": 3.7576873195833163e-06,
+      "loss": 0.5291,
+      "step": 19413
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 8.13565731048584,
+      "learning_rate": 3.756850604526629e-06,
+      "loss": 0.3877,
+      "step": 19414
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 6.812713623046875,
+      "learning_rate": 3.756013889469941e-06,
+      "loss": 1.2595,
+      "step": 19415
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 6.118829727172852,
+      "learning_rate": 3.755177174413254e-06,
+      "loss": 0.4958,
+      "step": 19416
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.148907661437988,
+      "learning_rate": 3.7543404593565663e-06,
+      "loss": 0.1913,
+      "step": 19417
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 21.72593116760254,
+      "learning_rate": 3.753503744299879e-06,
+      "loss": 0.746,
+      "step": 19418
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 14.517721176147461,
+      "learning_rate": 3.7526670292431914e-06,
+      "loss": 0.9438,
+      "step": 19419
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 14.233525276184082,
+      "learning_rate": 3.7518303141865042e-06,
+      "loss": 0.743,
+      "step": 19420
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 18.300806045532227,
+      "learning_rate": 3.750993599129817e-06,
+      "loss": 0.945,
+      "step": 19421
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 11.133574485778809,
+      "learning_rate": 3.750156884073129e-06,
+      "loss": 1.0218,
+      "step": 19422
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 11.053126335144043,
+      "learning_rate": 3.7493201690164414e-06,
+      "loss": 2.1753,
+      "step": 19423
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 18.2052001953125,
+      "learning_rate": 3.748483453959754e-06,
+      "loss": 2.6961,
+      "step": 19424
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 38.304786682128906,
+      "learning_rate": 3.747646738903067e-06,
+      "loss": 0.769,
+      "step": 19425
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 11.873785972595215,
+      "learning_rate": 3.7468100238463794e-06,
+      "loss": 0.4924,
+      "step": 19426
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 13.371991157531738,
+      "learning_rate": 3.745973308789692e-06,
+      "loss": 0.8206,
+      "step": 19427
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 27.54091453552246,
+      "learning_rate": 3.745136593733005e-06,
+      "loss": 1.5238,
+      "step": 19428
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 20.18380355834961,
+      "learning_rate": 3.744299878676317e-06,
+      "loss": 1.0788,
+      "step": 19429
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 9.077404022216797,
+      "learning_rate": 3.7434631636196293e-06,
+      "loss": 0.922,
+      "step": 19430
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 15.031689643859863,
+      "learning_rate": 3.742626448562942e-06,
+      "loss": 0.9573,
+      "step": 19431
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 23.64993667602539,
+      "learning_rate": 3.741789733506255e-06,
+      "loss": 1.1587,
+      "step": 19432
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 15.048856735229492,
+      "learning_rate": 3.7409530184495673e-06,
+      "loss": 0.7992,
+      "step": 19433
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 50.3357048034668,
+      "learning_rate": 3.74011630339288e-06,
+      "loss": 2.5979,
+      "step": 19434
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.9730583429336548,
+      "learning_rate": 3.7392795883361925e-06,
+      "loss": 0.0196,
+      "step": 19435
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 11.490036010742188,
+      "learning_rate": 3.738442873279505e-06,
+      "loss": 1.5979,
+      "step": 19436
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 69.22692108154297,
+      "learning_rate": 3.7376061582228172e-06,
+      "loss": 4.0118,
+      "step": 19437
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 14.819534301757812,
+      "learning_rate": 3.73676944316613e-06,
+      "loss": 1.1017,
+      "step": 19438
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 35.04603958129883,
+      "learning_rate": 3.735932728109443e-06,
+      "loss": 1.308,
+      "step": 19439
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 11.587259292602539,
+      "learning_rate": 3.735096013052755e-06,
+      "loss": 0.4893,
+      "step": 19440
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 21.826257705688477,
+      "learning_rate": 3.734259297996068e-06,
+      "loss": 2.334,
+      "step": 19441
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 24.667236328125,
+      "learning_rate": 3.7334225829393804e-06,
+      "loss": 0.9229,
+      "step": 19442
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 23.616796493530273,
+      "learning_rate": 3.7325858678826928e-06,
+      "loss": 1.2666,
+      "step": 19443
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 36.222328186035156,
+      "learning_rate": 3.731749152826005e-06,
+      "loss": 1.8234,
+      "step": 19444
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 12.682394027709961,
+      "learning_rate": 3.730912437769318e-06,
+      "loss": 0.3684,
+      "step": 19445
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 115.44532012939453,
+      "learning_rate": 3.7300757227126303e-06,
+      "loss": 2.5413,
+      "step": 19446
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.9555535316467285,
+      "learning_rate": 3.729239007655943e-06,
+      "loss": 0.2132,
+      "step": 19447
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.5936670303344727,
+      "learning_rate": 3.728402292599256e-06,
+      "loss": 0.3806,
+      "step": 19448
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 40.56935501098633,
+      "learning_rate": 3.7275655775425683e-06,
+      "loss": 0.8785,
+      "step": 19449
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 23.914955139160156,
+      "learning_rate": 3.7267288624858807e-06,
+      "loss": 1.6017,
+      "step": 19450
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.980844020843506,
+      "learning_rate": 3.725892147429193e-06,
+      "loss": 0.2545,
+      "step": 19451
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 12.170584678649902,
+      "learning_rate": 3.725055432372506e-06,
+      "loss": 0.9213,
+      "step": 19452
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 18.933670043945312,
+      "learning_rate": 3.7242187173158182e-06,
+      "loss": 1.3134,
+      "step": 19453
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 32.968238830566406,
+      "learning_rate": 3.723382002259131e-06,
+      "loss": 0.8626,
+      "step": 19454
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 6.560389995574951,
+      "learning_rate": 3.722545287202444e-06,
+      "loss": 0.5033,
+      "step": 19455
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 9.095024108886719,
+      "learning_rate": 3.7217085721457562e-06,
+      "loss": 0.5097,
+      "step": 19456
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 7.687363147735596,
+      "learning_rate": 3.7208718570890686e-06,
+      "loss": 0.3373,
+      "step": 19457
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 22.17128562927246,
+      "learning_rate": 3.720035142032381e-06,
+      "loss": 0.6416,
+      "step": 19458
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 10.849040985107422,
+      "learning_rate": 3.7191984269756938e-06,
+      "loss": 1.394,
+      "step": 19459
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.1686325073242188,
+      "learning_rate": 3.718361711919006e-06,
+      "loss": 0.252,
+      "step": 19460
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 94.78205108642578,
+      "learning_rate": 3.717524996862319e-06,
+      "loss": 2.0621,
+      "step": 19461
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 13.970884323120117,
+      "learning_rate": 3.7166882818056318e-06,
+      "loss": 0.522,
+      "step": 19462
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 10.70833683013916,
+      "learning_rate": 3.715851566748944e-06,
+      "loss": 1.0751,
+      "step": 19463
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 113.82579040527344,
+      "learning_rate": 3.715014851692256e-06,
+      "loss": 1.6744,
+      "step": 19464
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 11.373026847839355,
+      "learning_rate": 3.714178136635569e-06,
+      "loss": 0.5601,
+      "step": 19465
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 17.753814697265625,
+      "learning_rate": 3.7133414215788817e-06,
+      "loss": 1.9597,
+      "step": 19466
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 15.960187911987305,
+      "learning_rate": 3.712504706522194e-06,
+      "loss": 2.7553,
+      "step": 19467
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 6.061026096343994,
+      "learning_rate": 3.711667991465507e-06,
+      "loss": 0.3813,
+      "step": 19468
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.9911651611328125,
+      "learning_rate": 3.7108312764088197e-06,
+      "loss": 0.0639,
+      "step": 19469
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 14.849716186523438,
+      "learning_rate": 3.709994561352132e-06,
+      "loss": 0.3714,
+      "step": 19470
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 6.802692413330078,
+      "learning_rate": 3.709157846295444e-06,
+      "loss": 0.4296,
+      "step": 19471
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 57.396915435791016,
+      "learning_rate": 3.708321131238757e-06,
+      "loss": 2.195,
+      "step": 19472
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 12.485960006713867,
+      "learning_rate": 3.7074844161820696e-06,
+      "loss": 1.3668,
+      "step": 19473
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 6.738912105560303,
+      "learning_rate": 3.706647701125382e-06,
+      "loss": 0.2507,
+      "step": 19474
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.0292773246765137,
+      "learning_rate": 3.7058109860686948e-06,
+      "loss": 0.0491,
+      "step": 19475
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 7.762231349945068,
+      "learning_rate": 3.704974271012007e-06,
+      "loss": 1.7823,
+      "step": 19476
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 8.054952621459961,
+      "learning_rate": 3.70413755595532e-06,
+      "loss": 0.5199,
+      "step": 19477
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 27.710302352905273,
+      "learning_rate": 3.703300840898632e-06,
+      "loss": 1.8586,
+      "step": 19478
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 21.074888229370117,
+      "learning_rate": 3.7024641258419447e-06,
+      "loss": 0.7223,
+      "step": 19479
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 48.69511032104492,
+      "learning_rate": 3.7016274107852575e-06,
+      "loss": 0.6589,
+      "step": 19480
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 18.42178726196289,
+      "learning_rate": 3.70079069572857e-06,
+      "loss": 0.9482,
+      "step": 19481
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 12.237410545349121,
+      "learning_rate": 3.6999539806718827e-06,
+      "loss": 0.443,
+      "step": 19482
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 87.9030532836914,
+      "learning_rate": 3.699117265615195e-06,
+      "loss": 1.164,
+      "step": 19483
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 10.423582077026367,
+      "learning_rate": 3.698280550558508e-06,
+      "loss": 1.4942,
+      "step": 19484
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 15.563819885253906,
+      "learning_rate": 3.69744383550182e-06,
+      "loss": 0.3124,
+      "step": 19485
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 26.41593360900879,
+      "learning_rate": 3.6966071204451326e-06,
+      "loss": 1.2273,
+      "step": 19486
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 8.330730438232422,
+      "learning_rate": 3.695770405388445e-06,
+      "loss": 0.3219,
+      "step": 19487
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 46.10406494140625,
+      "learning_rate": 3.694933690331758e-06,
+      "loss": 1.6226,
+      "step": 19488
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 9.911351203918457,
+      "learning_rate": 3.6940969752750706e-06,
+      "loss": 1.0932,
+      "step": 19489
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 7.939240455627441,
+      "learning_rate": 3.693260260218383e-06,
+      "loss": 0.6091,
+      "step": 19490
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 14.360014915466309,
+      "learning_rate": 3.6924235451616954e-06,
+      "loss": 0.5849,
+      "step": 19491
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 17.98020362854004,
+      "learning_rate": 3.6915868301050078e-06,
+      "loss": 0.8839,
+      "step": 19492
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 20.01367950439453,
+      "learning_rate": 3.6907501150483206e-06,
+      "loss": 1.1177,
+      "step": 19493
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 8.152234077453613,
+      "learning_rate": 3.689913399991633e-06,
+      "loss": 1.4944,
+      "step": 19494
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 15.775879859924316,
+      "learning_rate": 3.6890766849349457e-06,
+      "loss": 0.8828,
+      "step": 19495
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 18.101858139038086,
+      "learning_rate": 3.6882399698782585e-06,
+      "loss": 0.6848,
+      "step": 19496
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 14.405749320983887,
+      "learning_rate": 3.687403254821571e-06,
+      "loss": 2.2475,
+      "step": 19497
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 29.071508407592773,
+      "learning_rate": 3.6865665397648833e-06,
+      "loss": 1.3265,
+      "step": 19498
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 15.056915283203125,
+      "learning_rate": 3.6857298247081957e-06,
+      "loss": 1.1738,
+      "step": 19499
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 12.642854690551758,
+      "learning_rate": 3.6848931096515085e-06,
+      "loss": 1.3507,
+      "step": 19500
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 9.297966003417969,
+      "learning_rate": 3.684056394594821e-06,
+      "loss": 0.3365,
+      "step": 19501
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.268497467041016,
+      "learning_rate": 3.6832196795381337e-06,
+      "loss": 0.2009,
+      "step": 19502
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 7.838582992553711,
+      "learning_rate": 3.6823829644814465e-06,
+      "loss": 0.1295,
+      "step": 19503
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 9.778390884399414,
+      "learning_rate": 3.681546249424759e-06,
+      "loss": 1.5067,
+      "step": 19504
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.048859119415283,
+      "learning_rate": 3.680709534368071e-06,
+      "loss": 0.1969,
+      "step": 19505
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 7.797203063964844,
+      "learning_rate": 3.6798728193113836e-06,
+      "loss": 1.3604,
+      "step": 19506
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 17.60069465637207,
+      "learning_rate": 3.6790361042546964e-06,
+      "loss": 1.4674,
+      "step": 19507
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 15.349143981933594,
+      "learning_rate": 3.6781993891980088e-06,
+      "loss": 1.5963,
+      "step": 19508
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 20.265392303466797,
+      "learning_rate": 3.6773626741413216e-06,
+      "loss": 1.4714,
+      "step": 19509
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 37.8333740234375,
+      "learning_rate": 3.676525959084634e-06,
+      "loss": 1.4036,
+      "step": 19510
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 17.156679153442383,
+      "learning_rate": 3.6756892440279468e-06,
+      "loss": 1.0847,
+      "step": 19511
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 18.195537567138672,
+      "learning_rate": 3.6748525289712587e-06,
+      "loss": 0.7521,
+      "step": 19512
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 16.99808120727539,
+      "learning_rate": 3.6740158139145715e-06,
+      "loss": 0.2767,
+      "step": 19513
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 20.308456420898438,
+      "learning_rate": 3.6731790988578843e-06,
+      "loss": 1.6816,
+      "step": 19514
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 15.886495590209961,
+      "learning_rate": 3.6723423838011967e-06,
+      "loss": 2.2394,
+      "step": 19515
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 110.13824462890625,
+      "learning_rate": 3.6715056687445095e-06,
+      "loss": 2.045,
+      "step": 19516
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 14.06793212890625,
+      "learning_rate": 3.670668953687822e-06,
+      "loss": 0.3872,
+      "step": 19517
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 6.715120315551758,
+      "learning_rate": 3.6698322386311347e-06,
+      "loss": 1.452,
+      "step": 19518
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 37.93868637084961,
+      "learning_rate": 3.6689955235744466e-06,
+      "loss": 2.0243,
+      "step": 19519
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 21.736034393310547,
+      "learning_rate": 3.6681588085177594e-06,
+      "loss": 0.7348,
+      "step": 19520
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 6.163998603820801,
+      "learning_rate": 3.6673220934610722e-06,
+      "loss": 0.1308,
+      "step": 19521
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 12.94540023803711,
+      "learning_rate": 3.6664853784043846e-06,
+      "loss": 0.5496,
+      "step": 19522
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 11.611268997192383,
+      "learning_rate": 3.6656486633476974e-06,
+      "loss": 1.0797,
+      "step": 19523
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 14.129745483398438,
+      "learning_rate": 3.66481194829101e-06,
+      "loss": 1.3862,
+      "step": 19524
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 46.87518310546875,
+      "learning_rate": 3.6639752332343226e-06,
+      "loss": 1.7547,
+      "step": 19525
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 8.3310546875,
+      "learning_rate": 3.6631385181776345e-06,
+      "loss": 0.8219,
+      "step": 19526
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 11.203372955322266,
+      "learning_rate": 3.6623018031209474e-06,
+      "loss": 0.7145,
+      "step": 19527
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 27.463226318359375,
+      "learning_rate": 3.6614650880642597e-06,
+      "loss": 0.6897,
+      "step": 19528
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 16.030445098876953,
+      "learning_rate": 3.6606283730075725e-06,
+      "loss": 1.0668,
+      "step": 19529
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 14.853161811828613,
+      "learning_rate": 3.6597916579508853e-06,
+      "loss": 1.0309,
+      "step": 19530
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 19.87004852294922,
+      "learning_rate": 3.6589549428941977e-06,
+      "loss": 0.8711,
+      "step": 19531
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 15.398309707641602,
+      "learning_rate": 3.6581182278375105e-06,
+      "loss": 1.8404,
+      "step": 19532
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.537481307983398,
+      "learning_rate": 3.6572815127808225e-06,
+      "loss": 0.5392,
+      "step": 19533
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 6.660141468048096,
+      "learning_rate": 3.6564447977241353e-06,
+      "loss": 0.3912,
+      "step": 19534
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 10.438698768615723,
+      "learning_rate": 3.6556080826674476e-06,
+      "loss": 0.4849,
+      "step": 19535
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.368213653564453,
+      "learning_rate": 3.6547713676107605e-06,
+      "loss": 1.1956,
+      "step": 19536
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 7.393680572509766,
+      "learning_rate": 3.6539346525540733e-06,
+      "loss": 1.0433,
+      "step": 19537
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 11.357941627502441,
+      "learning_rate": 3.6530979374973856e-06,
+      "loss": 0.5289,
+      "step": 19538
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 9.764481544494629,
+      "learning_rate": 3.6522612224406984e-06,
+      "loss": 0.3442,
+      "step": 19539
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 9.39493465423584,
+      "learning_rate": 3.6514245073840104e-06,
+      "loss": 1.3427,
+      "step": 19540
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 11.347414016723633,
+      "learning_rate": 3.650587792327323e-06,
+      "loss": 1.4416,
+      "step": 19541
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 11.80206298828125,
+      "learning_rate": 3.6497510772706356e-06,
+      "loss": 0.6181,
+      "step": 19542
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 23.547332763671875,
+      "learning_rate": 3.6489143622139484e-06,
+      "loss": 1.0659,
+      "step": 19543
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 15.252655982971191,
+      "learning_rate": 3.648077647157261e-06,
+      "loss": 0.5836,
+      "step": 19544
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 17.29207992553711,
+      "learning_rate": 3.6472409321005735e-06,
+      "loss": 0.8593,
+      "step": 19545
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 25.271240234375,
+      "learning_rate": 3.6464042170438864e-06,
+      "loss": 2.1257,
+      "step": 19546
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 6.511344909667969,
+      "learning_rate": 3.6455675019871983e-06,
+      "loss": 0.4393,
+      "step": 19547
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 2.7553915977478027,
+      "learning_rate": 3.644730786930511e-06,
+      "loss": 1.0807,
+      "step": 19548
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 66.39590454101562,
+      "learning_rate": 3.6438940718738235e-06,
+      "loss": 1.712,
+      "step": 19549
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 25.083534240722656,
+      "learning_rate": 3.6430573568171363e-06,
+      "loss": 1.3258,
+      "step": 19550
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 16.88987159729004,
+      "learning_rate": 3.6422206417604487e-06,
+      "loss": 1.274,
+      "step": 19551
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 28.262577056884766,
+      "learning_rate": 3.6413839267037615e-06,
+      "loss": 1.6496,
+      "step": 19552
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 69.0420150756836,
+      "learning_rate": 3.6405472116470743e-06,
+      "loss": 1.1903,
+      "step": 19553
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 10.95007610321045,
+      "learning_rate": 3.6397104965903862e-06,
+      "loss": 0.1559,
+      "step": 19554
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 12.338822364807129,
+      "learning_rate": 3.638873781533699e-06,
+      "loss": 0.8436,
+      "step": 19555
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 9.753349304199219,
+      "learning_rate": 3.6380370664770114e-06,
+      "loss": 0.4897,
+      "step": 19556
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 12.568263053894043,
+      "learning_rate": 3.637200351420324e-06,
+      "loss": 0.6605,
+      "step": 19557
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 60.659671783447266,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 2.3283,
+      "step": 19558
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 16.86094093322754,
+      "learning_rate": 3.6355269213069494e-06,
+      "loss": 0.4849,
+      "step": 19559
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 17.64162254333496,
+      "learning_rate": 3.634690206250262e-06,
+      "loss": 0.8689,
+      "step": 19560
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 13.738029479980469,
+      "learning_rate": 3.633853491193574e-06,
+      "loss": 0.8648,
+      "step": 19561
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 14.854199409484863,
+      "learning_rate": 3.633016776136887e-06,
+      "loss": 1.7406,
+      "step": 19562
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 21.5056095123291,
+      "learning_rate": 3.6321800610801993e-06,
+      "loss": 0.4041,
+      "step": 19563
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 19.813793182373047,
+      "learning_rate": 3.631343346023512e-06,
+      "loss": 1.316,
+      "step": 19564
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 17.993206024169922,
+      "learning_rate": 3.6305066309668245e-06,
+      "loss": 0.6492,
+      "step": 19565
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 8.353463172912598,
+      "learning_rate": 3.6296699159101373e-06,
+      "loss": 1.2699,
+      "step": 19566
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 49.2107048034668,
+      "learning_rate": 3.62883320085345e-06,
+      "loss": 0.7511,
+      "step": 19567
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 14.674556732177734,
+      "learning_rate": 3.627996485796762e-06,
+      "loss": 0.4794,
+      "step": 19568
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 10.151453971862793,
+      "learning_rate": 3.6271597707400744e-06,
+      "loss": 0.3333,
+      "step": 19569
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 11.01027774810791,
+      "learning_rate": 3.6263230556833872e-06,
+      "loss": 1.0966,
+      "step": 19570
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 15.187713623046875,
+      "learning_rate": 3.6254863406267e-06,
+      "loss": 1.5509,
+      "step": 19571
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 52.74294662475586,
+      "learning_rate": 3.6246496255700124e-06,
+      "loss": 1.6114,
+      "step": 19572
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2.404120683670044,
+      "learning_rate": 3.6238129105133252e-06,
+      "loss": 0.1087,
+      "step": 19573
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 13.5791654586792,
+      "learning_rate": 3.6229761954566376e-06,
+      "loss": 0.5679,
+      "step": 19574
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 9.572688102722168,
+      "learning_rate": 3.62213948039995e-06,
+      "loss": 0.4548,
+      "step": 19575
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 26.158750534057617,
+      "learning_rate": 3.6213027653432624e-06,
+      "loss": 2.7927,
+      "step": 19576
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 11.314776420593262,
+      "learning_rate": 3.620466050286575e-06,
+      "loss": 0.4775,
+      "step": 19577
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 5.215376853942871,
+      "learning_rate": 3.619629335229888e-06,
+      "loss": 0.3213,
+      "step": 19578
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 22.112407684326172,
+      "learning_rate": 3.6187926201732003e-06,
+      "loss": 1.6694,
+      "step": 19579
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 18.88576889038086,
+      "learning_rate": 3.617955905116513e-06,
+      "loss": 0.802,
+      "step": 19580
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 14.532073974609375,
+      "learning_rate": 3.617119190059825e-06,
+      "loss": 0.8773,
+      "step": 19581
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 21.64769744873047,
+      "learning_rate": 3.616282475003138e-06,
+      "loss": 0.8309,
+      "step": 19582
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 7.648311138153076,
+      "learning_rate": 3.6154457599464503e-06,
+      "loss": 0.7514,
+      "step": 19583
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 18.131874084472656,
+      "learning_rate": 3.614609044889763e-06,
+      "loss": 0.5386,
+      "step": 19584
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 78.26471710205078,
+      "learning_rate": 3.613772329833076e-06,
+      "loss": 1.9049,
+      "step": 19585
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 19.853479385375977,
+      "learning_rate": 3.6129356147763883e-06,
+      "loss": 0.8589,
+      "step": 19586
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 11.111654281616211,
+      "learning_rate": 3.612098899719701e-06,
+      "loss": 0.9744,
+      "step": 19587
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 31.242660522460938,
+      "learning_rate": 3.611262184663013e-06,
+      "loss": 1.9299,
+      "step": 19588
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 33.60970687866211,
+      "learning_rate": 3.610425469606326e-06,
+      "loss": 1.4131,
+      "step": 19589
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 5.124098300933838,
+      "learning_rate": 3.609588754549638e-06,
+      "loss": 0.4458,
+      "step": 19590
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.3736352920532227,
+      "learning_rate": 3.608752039492951e-06,
+      "loss": 0.1404,
+      "step": 19591
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 114.87059783935547,
+      "learning_rate": 3.6079153244362634e-06,
+      "loss": 2.0552,
+      "step": 19592
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 6.528724193572998,
+      "learning_rate": 3.607078609379576e-06,
+      "loss": 0.4047,
+      "step": 19593
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 14.813525199890137,
+      "learning_rate": 3.606241894322889e-06,
+      "loss": 0.3514,
+      "step": 19594
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 6.653222560882568,
+      "learning_rate": 3.605405179266201e-06,
+      "loss": 0.4972,
+      "step": 19595
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 29.276601791381836,
+      "learning_rate": 3.6045684642095137e-06,
+      "loss": 1.7808,
+      "step": 19596
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 13.689377784729004,
+      "learning_rate": 3.603731749152826e-06,
+      "loss": 0.602,
+      "step": 19597
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 7.918407917022705,
+      "learning_rate": 3.602895034096139e-06,
+      "loss": 0.3228,
+      "step": 19598
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 40.56988525390625,
+      "learning_rate": 3.6020583190394513e-06,
+      "loss": 1.5681,
+      "step": 19599
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 9.635811805725098,
+      "learning_rate": 3.601221603982764e-06,
+      "loss": 0.2254,
+      "step": 19600
+    },
+    {
+      "epoch": 2.46,
+      "eval_loss": 0.08711016178131104,
+      "eval_runtime": 95.4209,
+      "eval_samples_per_second": 37.12,
+      "eval_steps_per_second": 37.12,
+      "step": 19600
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 17.06913185119629,
+      "learning_rate": 3.600384888926077e-06,
+      "loss": 1.5398,
+      "step": 19601
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 15.327908515930176,
+      "learning_rate": 3.599548173869389e-06,
+      "loss": 1.126,
+      "step": 19602
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 7.690364837646484,
+      "learning_rate": 3.5987114588127017e-06,
+      "loss": 0.8513,
+      "step": 19603
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 14.208024024963379,
+      "learning_rate": 3.597874743756014e-06,
+      "loss": 0.637,
+      "step": 19604
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 27.198848724365234,
+      "learning_rate": 3.597038028699327e-06,
+      "loss": 0.8151,
+      "step": 19605
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 8.483991622924805,
+      "learning_rate": 3.5962013136426392e-06,
+      "loss": 0.2585,
+      "step": 19606
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 10.895722389221191,
+      "learning_rate": 3.595364598585952e-06,
+      "loss": 1.4322,
+      "step": 19607
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.930285930633545,
+      "learning_rate": 3.594527883529265e-06,
+      "loss": 0.2244,
+      "step": 19608
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 20.113201141357422,
+      "learning_rate": 3.5936911684725768e-06,
+      "loss": 3.2422,
+      "step": 19609
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.7983405590057373,
+      "learning_rate": 3.592854453415889e-06,
+      "loss": 0.0766,
+      "step": 19610
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 10.632842063903809,
+      "learning_rate": 3.592017738359202e-06,
+      "loss": 0.5352,
+      "step": 19611
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 40.46181869506836,
+      "learning_rate": 3.5911810233025148e-06,
+      "loss": 0.6148,
+      "step": 19612
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 14.521509170532227,
+      "learning_rate": 3.590344308245827e-06,
+      "loss": 0.8839,
+      "step": 19613
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 11.604077339172363,
+      "learning_rate": 3.58950759318914e-06,
+      "loss": 0.8917,
+      "step": 19614
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 14.674836158752441,
+      "learning_rate": 3.5886708781324523e-06,
+      "loss": 1.3426,
+      "step": 19615
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 16.176982879638672,
+      "learning_rate": 3.5878341630757647e-06,
+      "loss": 0.6628,
+      "step": 19616
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 14.715084075927734,
+      "learning_rate": 3.586997448019077e-06,
+      "loss": 1.1123,
+      "step": 19617
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 31.576904296875,
+      "learning_rate": 3.58616073296239e-06,
+      "loss": 1.2923,
+      "step": 19618
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 10.267995834350586,
+      "learning_rate": 3.5853240179057027e-06,
+      "loss": 0.6754,
+      "step": 19619
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 15.240357398986816,
+      "learning_rate": 3.584487302849015e-06,
+      "loss": 1.5771,
+      "step": 19620
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 111.63504028320312,
+      "learning_rate": 3.583650587792328e-06,
+      "loss": 2.1596,
+      "step": 19621
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 11.189507484436035,
+      "learning_rate": 3.5828138727356402e-06,
+      "loss": 0.9857,
+      "step": 19622
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 84.83074951171875,
+      "learning_rate": 3.5819771576789526e-06,
+      "loss": 1.9947,
+      "step": 19623
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.97153377532959,
+      "learning_rate": 3.581140442622265e-06,
+      "loss": 0.1164,
+      "step": 19624
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 8.341716766357422,
+      "learning_rate": 3.5803037275655778e-06,
+      "loss": 1.4288,
+      "step": 19625
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 31.253576278686523,
+      "learning_rate": 3.5794670125088906e-06,
+      "loss": 2.1505,
+      "step": 19626
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 26.299104690551758,
+      "learning_rate": 3.578630297452203e-06,
+      "loss": 1.4682,
+      "step": 19627
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 23.523075103759766,
+      "learning_rate": 3.5777935823955158e-06,
+      "loss": 1.395,
+      "step": 19628
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 14.520183563232422,
+      "learning_rate": 3.576956867338828e-06,
+      "loss": 0.8351,
+      "step": 19629
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 11.813876152038574,
+      "learning_rate": 3.5761201522821405e-06,
+      "loss": 2.7818,
+      "step": 19630
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 16.819231033325195,
+      "learning_rate": 3.575283437225453e-06,
+      "loss": 1.6929,
+      "step": 19631
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 12.025693893432617,
+      "learning_rate": 3.5744467221687657e-06,
+      "loss": 0.6277,
+      "step": 19632
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 119.55182647705078,
+      "learning_rate": 3.573610007112078e-06,
+      "loss": 0.5821,
+      "step": 19633
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 43.22489547729492,
+      "learning_rate": 3.572773292055391e-06,
+      "loss": 0.9943,
+      "step": 19634
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 170.36593627929688,
+      "learning_rate": 3.5719365769987037e-06,
+      "loss": 1.7828,
+      "step": 19635
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 8.834037780761719,
+      "learning_rate": 3.571099861942016e-06,
+      "loss": 0.8836,
+      "step": 19636
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 13.344812393188477,
+      "learning_rate": 3.5702631468853284e-06,
+      "loss": 2.1031,
+      "step": 19637
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 8.85362434387207,
+      "learning_rate": 3.569426431828641e-06,
+      "loss": 0.9738,
+      "step": 19638
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 15.561305046081543,
+      "learning_rate": 3.5685897167719536e-06,
+      "loss": 1.2766,
+      "step": 19639
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 15.3659086227417,
+      "learning_rate": 3.567753001715266e-06,
+      "loss": 0.8837,
+      "step": 19640
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 24.762359619140625,
+      "learning_rate": 3.566916286658579e-06,
+      "loss": 0.8039,
+      "step": 19641
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 20.780590057373047,
+      "learning_rate": 3.5660795716018916e-06,
+      "loss": 0.9226,
+      "step": 19642
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 13.365804672241211,
+      "learning_rate": 3.565242856545204e-06,
+      "loss": 0.4648,
+      "step": 19643
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 9.629612922668457,
+      "learning_rate": 3.5644061414885164e-06,
+      "loss": 1.1486,
+      "step": 19644
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.953568458557129,
+      "learning_rate": 3.5635694264318287e-06,
+      "loss": 0.3317,
+      "step": 19645
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.891057968139648,
+      "learning_rate": 3.5627327113751415e-06,
+      "loss": 0.3262,
+      "step": 19646
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 13.948858261108398,
+      "learning_rate": 3.561895996318454e-06,
+      "loss": 0.934,
+      "step": 19647
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 14.469639778137207,
+      "learning_rate": 3.5610592812617667e-06,
+      "loss": 0.7603,
+      "step": 19648
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 32.848106384277344,
+      "learning_rate": 3.5602225662050795e-06,
+      "loss": 0.9696,
+      "step": 19649
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.0937618017196655,
+      "learning_rate": 3.559385851148392e-06,
+      "loss": 0.0276,
+      "step": 19650
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 10.17780876159668,
+      "learning_rate": 3.558549136091704e-06,
+      "loss": 0.3886,
+      "step": 19651
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 17.023239135742188,
+      "learning_rate": 3.5577124210350167e-06,
+      "loss": 0.7694,
+      "step": 19652
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 14.045829772949219,
+      "learning_rate": 3.5568757059783295e-06,
+      "loss": 0.4005,
+      "step": 19653
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 17.317874908447266,
+      "learning_rate": 3.556038990921642e-06,
+      "loss": 0.3614,
+      "step": 19654
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 6.471225261688232,
+      "learning_rate": 3.5552022758649546e-06,
+      "loss": 0.5899,
+      "step": 19655
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 18.64377784729004,
+      "learning_rate": 3.554365560808267e-06,
+      "loss": 0.8984,
+      "step": 19656
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 55.732696533203125,
+      "learning_rate": 3.55352884575158e-06,
+      "loss": 1.6754,
+      "step": 19657
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 15.808688163757324,
+      "learning_rate": 3.5526921306948918e-06,
+      "loss": 0.6719,
+      "step": 19658
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 76.85340118408203,
+      "learning_rate": 3.5518554156382046e-06,
+      "loss": 2.71,
+      "step": 19659
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 11.649335861206055,
+      "learning_rate": 3.5510187005815174e-06,
+      "loss": 0.6056,
+      "step": 19660
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 14.526211738586426,
+      "learning_rate": 3.5501819855248298e-06,
+      "loss": 0.4754,
+      "step": 19661
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 111.2694091796875,
+      "learning_rate": 3.5493452704681426e-06,
+      "loss": 1.8594,
+      "step": 19662
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 23.605730056762695,
+      "learning_rate": 3.548508555411455e-06,
+      "loss": 0.8543,
+      "step": 19663
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 45.71969223022461,
+      "learning_rate": 3.5476718403547677e-06,
+      "loss": 1.9528,
+      "step": 19664
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.812906742095947,
+      "learning_rate": 3.5468351252980797e-06,
+      "loss": 0.0803,
+      "step": 19665
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 24.693025588989258,
+      "learning_rate": 3.5459984102413925e-06,
+      "loss": 2.4052,
+      "step": 19666
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 23.21428871154785,
+      "learning_rate": 3.5451616951847053e-06,
+      "loss": 0.9809,
+      "step": 19667
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 14.185833930969238,
+      "learning_rate": 3.5443249801280177e-06,
+      "loss": 0.5601,
+      "step": 19668
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 16.369680404663086,
+      "learning_rate": 3.5434882650713305e-06,
+      "loss": 2.8719,
+      "step": 19669
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 10.543824195861816,
+      "learning_rate": 3.542651550014643e-06,
+      "loss": 0.1977,
+      "step": 19670
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 15.344318389892578,
+      "learning_rate": 3.5418148349579552e-06,
+      "loss": 2.0361,
+      "step": 19671
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 23.386581420898438,
+      "learning_rate": 3.5409781199012676e-06,
+      "loss": 1.4351,
+      "step": 19672
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 14.04124927520752,
+      "learning_rate": 3.5401414048445804e-06,
+      "loss": 0.4966,
+      "step": 19673
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 23.20058822631836,
+      "learning_rate": 3.539304689787893e-06,
+      "loss": 1.6201,
+      "step": 19674
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 14.27526569366455,
+      "learning_rate": 3.5384679747312056e-06,
+      "loss": 0.8549,
+      "step": 19675
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 17.937129974365234,
+      "learning_rate": 3.5376312596745184e-06,
+      "loss": 0.8855,
+      "step": 19676
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 45.540687561035156,
+      "learning_rate": 3.5367945446178308e-06,
+      "loss": 2.1641,
+      "step": 19677
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.103177070617676,
+      "learning_rate": 3.535957829561143e-06,
+      "loss": 0.3367,
+      "step": 19678
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 22.65744972229004,
+      "learning_rate": 3.5351211145044555e-06,
+      "loss": 1.3482,
+      "step": 19679
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 26.80577278137207,
+      "learning_rate": 3.5342843994477683e-06,
+      "loss": 0.8469,
+      "step": 19680
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.383388519287109,
+      "learning_rate": 3.5334476843910807e-06,
+      "loss": 0.1368,
+      "step": 19681
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 42.43025588989258,
+      "learning_rate": 3.5326109693343935e-06,
+      "loss": 1.5467,
+      "step": 19682
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 106.4570083618164,
+      "learning_rate": 3.5317742542777063e-06,
+      "loss": 1.3021,
+      "step": 19683
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 66.1055679321289,
+      "learning_rate": 3.5309375392210187e-06,
+      "loss": 1.7628,
+      "step": 19684
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 17.448001861572266,
+      "learning_rate": 3.530100824164331e-06,
+      "loss": 1.1025,
+      "step": 19685
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 11.699751853942871,
+      "learning_rate": 3.5292641091076435e-06,
+      "loss": 0.3397,
+      "step": 19686
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 11.920134544372559,
+      "learning_rate": 3.5284273940509563e-06,
+      "loss": 0.5954,
+      "step": 19687
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 39.143470764160156,
+      "learning_rate": 3.5275906789942686e-06,
+      "loss": 1.0038,
+      "step": 19688
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 18.545639038085938,
+      "learning_rate": 3.5267539639375814e-06,
+      "loss": 2.4994,
+      "step": 19689
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 64.48916625976562,
+      "learning_rate": 3.5259172488808942e-06,
+      "loss": 1.3953,
+      "step": 19690
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 13.650774002075195,
+      "learning_rate": 3.5250805338242066e-06,
+      "loss": 0.3642,
+      "step": 19691
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 48.111083984375,
+      "learning_rate": 3.5242438187675186e-06,
+      "loss": 1.8149,
+      "step": 19692
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.1658313274383545,
+      "learning_rate": 3.5234071037108314e-06,
+      "loss": 0.075,
+      "step": 19693
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 9.8385648727417,
+      "learning_rate": 3.522570388654144e-06,
+      "loss": 0.8082,
+      "step": 19694
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 18.068117141723633,
+      "learning_rate": 3.5217336735974565e-06,
+      "loss": 2.5033,
+      "step": 19695
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 23.014169692993164,
+      "learning_rate": 3.5208969585407694e-06,
+      "loss": 0.8867,
+      "step": 19696
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 233.1300811767578,
+      "learning_rate": 3.5200602434840817e-06,
+      "loss": 1.2144,
+      "step": 19697
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 17.034318923950195,
+      "learning_rate": 3.5192235284273945e-06,
+      "loss": 1.1768,
+      "step": 19698
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 10.715003967285156,
+      "learning_rate": 3.5183868133707065e-06,
+      "loss": 0.2994,
+      "step": 19699
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 18.686586380004883,
+      "learning_rate": 3.5175500983140193e-06,
+      "loss": 0.476,
+      "step": 19700
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 17.066011428833008,
+      "learning_rate": 3.516713383257332e-06,
+      "loss": 1.4482,
+      "step": 19701
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 14.509745597839355,
+      "learning_rate": 3.5158766682006445e-06,
+      "loss": 1.2477,
+      "step": 19702
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.128937244415283,
+      "learning_rate": 3.5150399531439573e-06,
+      "loss": 0.0904,
+      "step": 19703
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 10.236748695373535,
+      "learning_rate": 3.5142032380872696e-06,
+      "loss": 1.3713,
+      "step": 19704
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 15.119993209838867,
+      "learning_rate": 3.5133665230305824e-06,
+      "loss": 1.3188,
+      "step": 19705
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 12.767293930053711,
+      "learning_rate": 3.5125298079738944e-06,
+      "loss": 0.7497,
+      "step": 19706
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 11.747312545776367,
+      "learning_rate": 3.511693092917207e-06,
+      "loss": 0.6895,
+      "step": 19707
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 11.619932174682617,
+      "learning_rate": 3.51085637786052e-06,
+      "loss": 0.7908,
+      "step": 19708
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 17.017072677612305,
+      "learning_rate": 3.5100196628038324e-06,
+      "loss": 1.3293,
+      "step": 19709
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 13.582147598266602,
+      "learning_rate": 3.509182947747145e-06,
+      "loss": 1.2597,
+      "step": 19710
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 6.918244361877441,
+      "learning_rate": 3.5083462326904576e-06,
+      "loss": 0.5073,
+      "step": 19711
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.3982391357421875,
+      "learning_rate": 3.5075095176337704e-06,
+      "loss": 2.3984,
+      "step": 19712
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 8.981369018554688,
+      "learning_rate": 3.5066728025770823e-06,
+      "loss": 0.6294,
+      "step": 19713
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.048067569732666,
+      "learning_rate": 3.505836087520395e-06,
+      "loss": 0.1552,
+      "step": 19714
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 16.41480255126953,
+      "learning_rate": 3.5049993724637075e-06,
+      "loss": 0.676,
+      "step": 19715
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 10.085628509521484,
+      "learning_rate": 3.5041626574070203e-06,
+      "loss": 0.1062,
+      "step": 19716
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 10.479913711547852,
+      "learning_rate": 3.503325942350333e-06,
+      "loss": 0.6488,
+      "step": 19717
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 8.183189392089844,
+      "learning_rate": 3.5024892272936455e-06,
+      "loss": 0.3081,
+      "step": 19718
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 26.590784072875977,
+      "learning_rate": 3.5016525122369583e-06,
+      "loss": 1.0394,
+      "step": 19719
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.6834235191345215,
+      "learning_rate": 3.5008157971802702e-06,
+      "loss": 0.5966,
+      "step": 19720
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 10.283235549926758,
+      "learning_rate": 3.499979082123583e-06,
+      "loss": 1.4802,
+      "step": 19721
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 13.874239921569824,
+      "learning_rate": 3.4991423670668954e-06,
+      "loss": 1.2384,
+      "step": 19722
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 14.061607360839844,
+      "learning_rate": 3.4983056520102082e-06,
+      "loss": 0.635,
+      "step": 19723
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.137641906738281,
+      "learning_rate": 3.497468936953521e-06,
+      "loss": 0.8886,
+      "step": 19724
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 16.545169830322266,
+      "learning_rate": 3.4966322218968334e-06,
+      "loss": 0.7775,
+      "step": 19725
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.7449951171875,
+      "learning_rate": 3.495795506840146e-06,
+      "loss": 0.9171,
+      "step": 19726
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 35.38679885864258,
+      "learning_rate": 3.494958791783458e-06,
+      "loss": 0.6957,
+      "step": 19727
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 163.88958740234375,
+      "learning_rate": 3.494122076726771e-06,
+      "loss": 2.2629,
+      "step": 19728
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 33.55429458618164,
+      "learning_rate": 3.4932853616700833e-06,
+      "loss": 1.0029,
+      "step": 19729
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 17.60959815979004,
+      "learning_rate": 3.492448646613396e-06,
+      "loss": 0.6596,
+      "step": 19730
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 172.82119750976562,
+      "learning_rate": 3.491611931556709e-06,
+      "loss": 2.0909,
+      "step": 19731
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 71.0232162475586,
+      "learning_rate": 3.4907752165000213e-06,
+      "loss": 2.8295,
+      "step": 19732
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 16.076486587524414,
+      "learning_rate": 3.489938501443334e-06,
+      "loss": 1.1559,
+      "step": 19733
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 11.610815048217773,
+      "learning_rate": 3.489101786386646e-06,
+      "loss": 1.2087,
+      "step": 19734
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 14.32686710357666,
+      "learning_rate": 3.488265071329959e-06,
+      "loss": 0.5971,
+      "step": 19735
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 13.450050354003906,
+      "learning_rate": 3.4874283562732713e-06,
+      "loss": 0.6311,
+      "step": 19736
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 21.91973304748535,
+      "learning_rate": 3.486591641216584e-06,
+      "loss": 0.7215,
+      "step": 19737
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.609929084777832,
+      "learning_rate": 3.4857549261598964e-06,
+      "loss": 0.3477,
+      "step": 19738
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 11.93908977508545,
+      "learning_rate": 3.4849182111032092e-06,
+      "loss": 0.4525,
+      "step": 19739
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 8.637784004211426,
+      "learning_rate": 3.484081496046522e-06,
+      "loss": 0.5528,
+      "step": 19740
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 9.774800300598145,
+      "learning_rate": 3.483244780989834e-06,
+      "loss": 0.6823,
+      "step": 19741
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 16.305648803710938,
+      "learning_rate": 3.482408065933147e-06,
+      "loss": 2.189,
+      "step": 19742
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 7.879077434539795,
+      "learning_rate": 3.481571350876459e-06,
+      "loss": 0.5587,
+      "step": 19743
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 10.909440040588379,
+      "learning_rate": 3.480734635819772e-06,
+      "loss": 0.664,
+      "step": 19744
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 10.326915740966797,
+      "learning_rate": 3.4798979207630844e-06,
+      "loss": 1.8753,
+      "step": 19745
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 12.269172668457031,
+      "learning_rate": 3.479061205706397e-06,
+      "loss": 0.7914,
+      "step": 19746
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 25.903446197509766,
+      "learning_rate": 3.47822449064971e-06,
+      "loss": 0.7501,
+      "step": 19747
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 7.591154098510742,
+      "learning_rate": 3.477387775593022e-06,
+      "loss": 0.4238,
+      "step": 19748
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 13.29786491394043,
+      "learning_rate": 3.4765510605363347e-06,
+      "loss": 0.5984,
+      "step": 19749
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.9905242919921875,
+      "learning_rate": 3.475714345479647e-06,
+      "loss": 0.1857,
+      "step": 19750
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 18.47494888305664,
+      "learning_rate": 3.47487763042296e-06,
+      "loss": 1.0523,
+      "step": 19751
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 80.27205657958984,
+      "learning_rate": 3.4740409153662723e-06,
+      "loss": 2.1472,
+      "step": 19752
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 19.939136505126953,
+      "learning_rate": 3.473204200309585e-06,
+      "loss": 0.6476,
+      "step": 19753
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 15.560843467712402,
+      "learning_rate": 3.472367485252898e-06,
+      "loss": 0.425,
+      "step": 19754
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 14.479058265686035,
+      "learning_rate": 3.47153077019621e-06,
+      "loss": 1.2733,
+      "step": 19755
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 11.08471965789795,
+      "learning_rate": 3.4706940551395222e-06,
+      "loss": 0.4975,
+      "step": 19756
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 18.777667999267578,
+      "learning_rate": 3.469857340082835e-06,
+      "loss": 0.1296,
+      "step": 19757
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 44.677101135253906,
+      "learning_rate": 3.469020625026148e-06,
+      "loss": 0.9832,
+      "step": 19758
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.612020492553711,
+      "learning_rate": 3.46818390996946e-06,
+      "loss": 0.4032,
+      "step": 19759
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 24.597963333129883,
+      "learning_rate": 3.467347194912773e-06,
+      "loss": 2.2152,
+      "step": 19760
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 11.08047866821289,
+      "learning_rate": 3.466510479856085e-06,
+      "loss": 1.2333,
+      "step": 19761
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 6.552147388458252,
+      "learning_rate": 3.4656737647993978e-06,
+      "loss": 0.975,
+      "step": 19762
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 10.885284423828125,
+      "learning_rate": 3.46483704974271e-06,
+      "loss": 0.6009,
+      "step": 19763
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 8.18996810913086,
+      "learning_rate": 3.464000334686023e-06,
+      "loss": 0.4458,
+      "step": 19764
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 15.223823547363281,
+      "learning_rate": 3.4631636196293357e-06,
+      "loss": 0.5961,
+      "step": 19765
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.9410605430603027,
+      "learning_rate": 3.462326904572648e-06,
+      "loss": 0.2058,
+      "step": 19766
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 62.11387252807617,
+      "learning_rate": 3.461490189515961e-06,
+      "loss": 1.2756,
+      "step": 19767
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 23.682811737060547,
+      "learning_rate": 3.460653474459273e-06,
+      "loss": 1.4904,
+      "step": 19768
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 19.57076072692871,
+      "learning_rate": 3.4598167594025857e-06,
+      "loss": 0.6619,
+      "step": 19769
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 15.864212989807129,
+      "learning_rate": 3.458980044345898e-06,
+      "loss": 0.6604,
+      "step": 19770
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 91.69075775146484,
+      "learning_rate": 3.458143329289211e-06,
+      "loss": 1.7616,
+      "step": 19771
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 9.820259094238281,
+      "learning_rate": 3.4573066142325237e-06,
+      "loss": 0.5641,
+      "step": 19772
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 31.864337921142578,
+      "learning_rate": 3.456469899175836e-06,
+      "loss": 0.957,
+      "step": 19773
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 7.831910610198975,
+      "learning_rate": 3.455633184119149e-06,
+      "loss": 1.7723,
+      "step": 19774
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 17.594562530517578,
+      "learning_rate": 3.4547964690624608e-06,
+      "loss": 0.9591,
+      "step": 19775
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 8.951953887939453,
+      "learning_rate": 3.4539597540057736e-06,
+      "loss": 1.1267,
+      "step": 19776
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 21.579511642456055,
+      "learning_rate": 3.453123038949086e-06,
+      "loss": 0.7413,
+      "step": 19777
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 15.402226448059082,
+      "learning_rate": 3.4522863238923988e-06,
+      "loss": 0.555,
+      "step": 19778
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 21.79979133605957,
+      "learning_rate": 3.451449608835711e-06,
+      "loss": 2.6856,
+      "step": 19779
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 15.194982528686523,
+      "learning_rate": 3.450612893779024e-06,
+      "loss": 1.3308,
+      "step": 19780
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.8116750717163086,
+      "learning_rate": 3.4497761787223368e-06,
+      "loss": 0.1193,
+      "step": 19781
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 11.502127647399902,
+      "learning_rate": 3.4489394636656487e-06,
+      "loss": 0.5772,
+      "step": 19782
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 27.45766258239746,
+      "learning_rate": 3.4481027486089615e-06,
+      "loss": 1.7305,
+      "step": 19783
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 16.907424926757812,
+      "learning_rate": 3.447266033552274e-06,
+      "loss": 2.2742,
+      "step": 19784
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.822242259979248,
+      "learning_rate": 3.4464293184955867e-06,
+      "loss": 0.5652,
+      "step": 19785
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 20.32942771911621,
+      "learning_rate": 3.445592603438899e-06,
+      "loss": 0.4808,
+      "step": 19786
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 8.529967308044434,
+      "learning_rate": 3.444755888382212e-06,
+      "loss": 0.1695,
+      "step": 19787
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 16.412992477416992,
+      "learning_rate": 3.4439191733255247e-06,
+      "loss": 0.4608,
+      "step": 19788
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 18.60361671447754,
+      "learning_rate": 3.4430824582688366e-06,
+      "loss": 1.1064,
+      "step": 19789
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 18.82691192626953,
+      "learning_rate": 3.4422457432121494e-06,
+      "loss": 1.9421,
+      "step": 19790
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 43.89506149291992,
+      "learning_rate": 3.441409028155462e-06,
+      "loss": 1.5197,
+      "step": 19791
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 7.936665058135986,
+      "learning_rate": 3.4405723130987746e-06,
+      "loss": 0.7861,
+      "step": 19792
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 18.61930274963379,
+      "learning_rate": 3.439735598042087e-06,
+      "loss": 1.1221,
+      "step": 19793
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.85394811630249,
+      "learning_rate": 3.4388988829853998e-06,
+      "loss": 1.7637,
+      "step": 19794
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 12.541993141174316,
+      "learning_rate": 3.4380621679287126e-06,
+      "loss": 0.9118,
+      "step": 19795
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 27.030841827392578,
+      "learning_rate": 3.4372254528720245e-06,
+      "loss": 1.0295,
+      "step": 19796
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 36.323028564453125,
+      "learning_rate": 3.436388737815337e-06,
+      "loss": 0.4287,
+      "step": 19797
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 13.478815078735352,
+      "learning_rate": 3.4355520227586497e-06,
+      "loss": 0.7857,
+      "step": 19798
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.058920860290527,
+      "learning_rate": 3.4347153077019625e-06,
+      "loss": 0.4471,
+      "step": 19799
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 6.7582597732543945,
+      "learning_rate": 3.433878592645275e-06,
+      "loss": 0.1954,
+      "step": 19800
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 37.303916931152344,
+      "learning_rate": 3.4330418775885877e-06,
+      "loss": 1.8653,
+      "step": 19801
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 15.346481323242188,
+      "learning_rate": 3.4322051625319e-06,
+      "loss": 0.3085,
+      "step": 19802
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 7.7009429931640625,
+      "learning_rate": 3.4313684474752125e-06,
+      "loss": 0.4781,
+      "step": 19803
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 7.781106948852539,
+      "learning_rate": 3.430531732418525e-06,
+      "loss": 0.936,
+      "step": 19804
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 21.661930084228516,
+      "learning_rate": 3.4296950173618376e-06,
+      "loss": 0.8793,
+      "step": 19805
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.812716960906982,
+      "learning_rate": 3.4288583023051504e-06,
+      "loss": 0.518,
+      "step": 19806
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 15.22900676727295,
+      "learning_rate": 3.428021587248463e-06,
+      "loss": 0.683,
+      "step": 19807
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 61.01286697387695,
+      "learning_rate": 3.4271848721917756e-06,
+      "loss": 1.0509,
+      "step": 19808
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 11.959772109985352,
+      "learning_rate": 3.426348157135088e-06,
+      "loss": 0.9398,
+      "step": 19809
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 27.148923873901367,
+      "learning_rate": 3.4255114420784004e-06,
+      "loss": 1.8368,
+      "step": 19810
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 9.010526657104492,
+      "learning_rate": 3.4246747270217128e-06,
+      "loss": 0.8156,
+      "step": 19811
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 30.53544807434082,
+      "learning_rate": 3.4238380119650256e-06,
+      "loss": 0.9557,
+      "step": 19812
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 24.31468391418457,
+      "learning_rate": 3.4230012969083384e-06,
+      "loss": 3.0352,
+      "step": 19813
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 11.948583602905273,
+      "learning_rate": 3.4221645818516507e-06,
+      "loss": 0.784,
+      "step": 19814
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 14.328639030456543,
+      "learning_rate": 3.4213278667949635e-06,
+      "loss": 0.307,
+      "step": 19815
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 20.781282424926758,
+      "learning_rate": 3.420491151738276e-06,
+      "loss": 0.6423,
+      "step": 19816
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 34.052772521972656,
+      "learning_rate": 3.4196544366815883e-06,
+      "loss": 2.9036,
+      "step": 19817
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 10.930405616760254,
+      "learning_rate": 3.4188177216249007e-06,
+      "loss": 1.4673,
+      "step": 19818
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 27.7197322845459,
+      "learning_rate": 3.4179810065682135e-06,
+      "loss": 0.7177,
+      "step": 19819
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.597465753555298,
+      "learning_rate": 3.417144291511526e-06,
+      "loss": 0.1199,
+      "step": 19820
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 18.610553741455078,
+      "learning_rate": 3.4163075764548387e-06,
+      "loss": 1.3404,
+      "step": 19821
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 24.10991096496582,
+      "learning_rate": 3.4154708613981515e-06,
+      "loss": 1.0394,
+      "step": 19822
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 7.0872931480407715,
+      "learning_rate": 3.414634146341464e-06,
+      "loss": 0.6687,
+      "step": 19823
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 65.09789276123047,
+      "learning_rate": 3.4137974312847762e-06,
+      "loss": 3.3621,
+      "step": 19824
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 28.820966720581055,
+      "learning_rate": 3.4129607162280886e-06,
+      "loss": 1.3284,
+      "step": 19825
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 27.640705108642578,
+      "learning_rate": 3.4121240011714014e-06,
+      "loss": 1.0373,
+      "step": 19826
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 21.718908309936523,
+      "learning_rate": 3.4112872861147138e-06,
+      "loss": 1.6755,
+      "step": 19827
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 16.77260398864746,
+      "learning_rate": 3.4104505710580266e-06,
+      "loss": 0.7976,
+      "step": 19828
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 13.465864181518555,
+      "learning_rate": 3.4096138560013394e-06,
+      "loss": 1.2967,
+      "step": 19829
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 31.5263729095459,
+      "learning_rate": 3.4087771409446518e-06,
+      "loss": 1.9992,
+      "step": 19830
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 10.719361305236816,
+      "learning_rate": 3.4079404258879637e-06,
+      "loss": 0.7871,
+      "step": 19831
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 339.76422119140625,
+      "learning_rate": 3.4071037108312765e-06,
+      "loss": 0.8432,
+      "step": 19832
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 15.445401191711426,
+      "learning_rate": 3.4062669957745893e-06,
+      "loss": 0.5135,
+      "step": 19833
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 23.20089340209961,
+      "learning_rate": 3.4054302807179017e-06,
+      "loss": 1.6641,
+      "step": 19834
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 10.486080169677734,
+      "learning_rate": 3.4045935656612145e-06,
+      "loss": 0.7401,
+      "step": 19835
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 34.3616828918457,
+      "learning_rate": 3.4037568506045273e-06,
+      "loss": 1.6526,
+      "step": 19836
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 10.02161693572998,
+      "learning_rate": 3.4029201355478397e-06,
+      "loss": 0.9273,
+      "step": 19837
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 10.030625343322754,
+      "learning_rate": 3.4020834204911516e-06,
+      "loss": 0.452,
+      "step": 19838
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 21.02669906616211,
+      "learning_rate": 3.4012467054344644e-06,
+      "loss": 0.3191,
+      "step": 19839
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 6.957141876220703,
+      "learning_rate": 3.4004099903777772e-06,
+      "loss": 0.3309,
+      "step": 19840
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 10.954840660095215,
+      "learning_rate": 3.3995732753210896e-06,
+      "loss": 0.6704,
+      "step": 19841
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 9.489827156066895,
+      "learning_rate": 3.3987365602644024e-06,
+      "loss": 0.3872,
+      "step": 19842
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 12.518203735351562,
+      "learning_rate": 3.397899845207715e-06,
+      "loss": 1.1024,
+      "step": 19843
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 13.665872573852539,
+      "learning_rate": 3.3970631301510276e-06,
+      "loss": 0.9706,
+      "step": 19844
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 10.394923210144043,
+      "learning_rate": 3.3962264150943395e-06,
+      "loss": 1.0683,
+      "step": 19845
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 37.72017288208008,
+      "learning_rate": 3.3953897000376524e-06,
+      "loss": 2.3307,
+      "step": 19846
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 27.23324966430664,
+      "learning_rate": 3.394552984980965e-06,
+      "loss": 0.8453,
+      "step": 19847
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 9.878998756408691,
+      "learning_rate": 3.3937162699242775e-06,
+      "loss": 0.452,
+      "step": 19848
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 37.411582946777344,
+      "learning_rate": 3.3928795548675903e-06,
+      "loss": 1.9002,
+      "step": 19849
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 8.64536190032959,
+      "learning_rate": 3.3920428398109027e-06,
+      "loss": 0.7126,
+      "step": 19850
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 32.167545318603516,
+      "learning_rate": 3.391206124754215e-06,
+      "loss": 2.47,
+      "step": 19851
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 8.786133766174316,
+      "learning_rate": 3.3903694096975275e-06,
+      "loss": 0.5359,
+      "step": 19852
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 11.467194557189941,
+      "learning_rate": 3.3895326946408403e-06,
+      "loss": 0.3455,
+      "step": 19853
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 7.817058563232422,
+      "learning_rate": 3.388695979584153e-06,
+      "loss": 0.3058,
+      "step": 19854
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 17.731460571289062,
+      "learning_rate": 3.3878592645274654e-06,
+      "loss": 0.9054,
+      "step": 19855
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 26.260496139526367,
+      "learning_rate": 3.3870225494707783e-06,
+      "loss": 1.6367,
+      "step": 19856
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 18.765504837036133,
+      "learning_rate": 3.3861858344140906e-06,
+      "loss": 0.3956,
+      "step": 19857
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 8.115496635437012,
+      "learning_rate": 3.385349119357403e-06,
+      "loss": 0.9168,
+      "step": 19858
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 6.180880546569824,
+      "learning_rate": 3.3845124043007154e-06,
+      "loss": 0.5112,
+      "step": 19859
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 11.935192108154297,
+      "learning_rate": 3.383675689244028e-06,
+      "loss": 0.4514,
+      "step": 19860
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 21.07114028930664,
+      "learning_rate": 3.3828389741873406e-06,
+      "loss": 1.3732,
+      "step": 19861
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 33.54058837890625,
+      "learning_rate": 3.3820022591306534e-06,
+      "loss": 1.4688,
+      "step": 19862
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 13.222314834594727,
+      "learning_rate": 3.381165544073966e-06,
+      "loss": 0.6227,
+      "step": 19863
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 36.876625061035156,
+      "learning_rate": 3.3803288290172785e-06,
+      "loss": 2.2184,
+      "step": 19864
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 7.127270221710205,
+      "learning_rate": 3.379492113960591e-06,
+      "loss": 0.4504,
+      "step": 19865
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 17.18406867980957,
+      "learning_rate": 3.3786553989039033e-06,
+      "loss": 1.1996,
+      "step": 19866
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 17.484281539916992,
+      "learning_rate": 3.377818683847216e-06,
+      "loss": 0.9784,
+      "step": 19867
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 19.296897888183594,
+      "learning_rate": 3.3769819687905285e-06,
+      "loss": 1.0986,
+      "step": 19868
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 341.8594970703125,
+      "learning_rate": 3.3761452537338413e-06,
+      "loss": 1.4915,
+      "step": 19869
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 6.280468463897705,
+      "learning_rate": 3.375308538677154e-06,
+      "loss": 0.8342,
+      "step": 19870
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 155.24281311035156,
+      "learning_rate": 3.3744718236204665e-06,
+      "loss": 0.5996,
+      "step": 19871
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 20.076494216918945,
+      "learning_rate": 3.3736351085637784e-06,
+      "loss": 1.6459,
+      "step": 19872
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 6.135019779205322,
+      "learning_rate": 3.3727983935070912e-06,
+      "loss": 0.4506,
+      "step": 19873
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 12.224918365478516,
+      "learning_rate": 3.371961678450404e-06,
+      "loss": 0.916,
+      "step": 19874
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 27.757692337036133,
+      "learning_rate": 3.3711249633937164e-06,
+      "loss": 1.8756,
+      "step": 19875
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 10.30232048034668,
+      "learning_rate": 3.370288248337029e-06,
+      "loss": 0.4556,
+      "step": 19876
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 11.233203887939453,
+      "learning_rate": 3.369451533280342e-06,
+      "loss": 1.4636,
+      "step": 19877
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.675137519836426,
+      "learning_rate": 3.3686148182236544e-06,
+      "loss": 1.0059,
+      "step": 19878
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 47.12982177734375,
+      "learning_rate": 3.3677781031669663e-06,
+      "loss": 1.6327,
+      "step": 19879
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 197.06178283691406,
+      "learning_rate": 3.366941388110279e-06,
+      "loss": 1.3262,
+      "step": 19880
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 20.33978271484375,
+      "learning_rate": 3.366104673053592e-06,
+      "loss": 1.2696,
+      "step": 19881
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 15.062004089355469,
+      "learning_rate": 3.3652679579969043e-06,
+      "loss": 1.9937,
+      "step": 19882
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 9.023582458496094,
+      "learning_rate": 3.364431242940217e-06,
+      "loss": 0.3869,
+      "step": 19883
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 27.49932098388672,
+      "learning_rate": 3.3635945278835295e-06,
+      "loss": 1.3255,
+      "step": 19884
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 6.58065128326416,
+      "learning_rate": 3.3627578128268423e-06,
+      "loss": 0.5256,
+      "step": 19885
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 10.925614356994629,
+      "learning_rate": 3.3619210977701543e-06,
+      "loss": 0.7766,
+      "step": 19886
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 7.92704963684082,
+      "learning_rate": 3.361084382713467e-06,
+      "loss": 0.3684,
+      "step": 19887
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 7.970696926116943,
+      "learning_rate": 3.36024766765678e-06,
+      "loss": 1.202,
+      "step": 19888
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 73.66069793701172,
+      "learning_rate": 3.3594109526000922e-06,
+      "loss": 1.3298,
+      "step": 19889
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 19.252840042114258,
+      "learning_rate": 3.358574237543405e-06,
+      "loss": 1.8018,
+      "step": 19890
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 12.6245756149292,
+      "learning_rate": 3.3577375224867174e-06,
+      "loss": 1.8099,
+      "step": 19891
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 19.752887725830078,
+      "learning_rate": 3.3569008074300302e-06,
+      "loss": 0.9063,
+      "step": 19892
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 47.04121398925781,
+      "learning_rate": 3.356064092373342e-06,
+      "loss": 1.9994,
+      "step": 19893
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 20.247621536254883,
+      "learning_rate": 3.355227377316655e-06,
+      "loss": 2.1466,
+      "step": 19894
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.817601203918457,
+      "learning_rate": 3.3543906622599674e-06,
+      "loss": 0.2317,
+      "step": 19895
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 12.793375968933105,
+      "learning_rate": 3.35355394720328e-06,
+      "loss": 1.6354,
+      "step": 19896
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 25.820037841796875,
+      "learning_rate": 3.352717232146593e-06,
+      "loss": 1.1934,
+      "step": 19897
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 25.317333221435547,
+      "learning_rate": 3.3518805170899053e-06,
+      "loss": 1.4453,
+      "step": 19898
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 12.54918384552002,
+      "learning_rate": 3.351043802033218e-06,
+      "loss": 1.9853,
+      "step": 19899
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 10.937554359436035,
+      "learning_rate": 3.35020708697653e-06,
+      "loss": 0.5546,
+      "step": 19900
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 11.002120018005371,
+      "learning_rate": 3.349370371919843e-06,
+      "loss": 0.548,
+      "step": 19901
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 28.960004806518555,
+      "learning_rate": 3.3485336568631553e-06,
+      "loss": 1.6595,
+      "step": 19902
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 8.909430503845215,
+      "learning_rate": 3.347696941806468e-06,
+      "loss": 0.4689,
+      "step": 19903
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 7.743149757385254,
+      "learning_rate": 3.346860226749781e-06,
+      "loss": 0.4208,
+      "step": 19904
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 48.501033782958984,
+      "learning_rate": 3.3460235116930933e-06,
+      "loss": 1.1391,
+      "step": 19905
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 7.790011405944824,
+      "learning_rate": 3.345186796636406e-06,
+      "loss": 1.8365,
+      "step": 19906
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 10.871684074401855,
+      "learning_rate": 3.344350081579718e-06,
+      "loss": 0.418,
+      "step": 19907
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 42.4021110534668,
+      "learning_rate": 3.343513366523031e-06,
+      "loss": 0.9568,
+      "step": 19908
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 8.166342735290527,
+      "learning_rate": 3.342676651466343e-06,
+      "loss": 0.4243,
+      "step": 19909
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 27.085187911987305,
+      "learning_rate": 3.341839936409656e-06,
+      "loss": 1.7439,
+      "step": 19910
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 11.258872032165527,
+      "learning_rate": 3.341003221352969e-06,
+      "loss": 1.4339,
+      "step": 19911
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 13.332852363586426,
+      "learning_rate": 3.340166506296281e-06,
+      "loss": 0.8385,
+      "step": 19912
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.6175317764282227,
+      "learning_rate": 3.339329791239594e-06,
+      "loss": 0.3455,
+      "step": 19913
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 8.716652870178223,
+      "learning_rate": 3.338493076182906e-06,
+      "loss": 0.758,
+      "step": 19914
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 40.15550231933594,
+      "learning_rate": 3.3376563611262187e-06,
+      "loss": 2.2447,
+      "step": 19915
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 7.4539794921875,
+      "learning_rate": 3.336819646069531e-06,
+      "loss": 1.1303,
+      "step": 19916
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 15.161958694458008,
+      "learning_rate": 3.335982931012844e-06,
+      "loss": 1.203,
+      "step": 19917
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.728916645050049,
+      "learning_rate": 3.3351462159561567e-06,
+      "loss": 0.0629,
+      "step": 19918
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 12.251533508300781,
+      "learning_rate": 3.334309500899469e-06,
+      "loss": 0.5393,
+      "step": 19919
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 17.577667236328125,
+      "learning_rate": 3.333472785842782e-06,
+      "loss": 0.5843,
+      "step": 19920
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 30.012460708618164,
+      "learning_rate": 3.332636070786094e-06,
+      "loss": 1.8977,
+      "step": 19921
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 6.592525005340576,
+      "learning_rate": 3.3317993557294067e-06,
+      "loss": 0.1557,
+      "step": 19922
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 9.875282287597656,
+      "learning_rate": 3.330962640672719e-06,
+      "loss": 0.7951,
+      "step": 19923
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 17.681758880615234,
+      "learning_rate": 3.330125925616032e-06,
+      "loss": 0.4883,
+      "step": 19924
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.01639723777771,
+      "learning_rate": 3.329289210559344e-06,
+      "loss": 0.111,
+      "step": 19925
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 21.93477439880371,
+      "learning_rate": 3.328452495502657e-06,
+      "loss": 0.7063,
+      "step": 19926
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 8.804841995239258,
+      "learning_rate": 3.32761578044597e-06,
+      "loss": 0.7265,
+      "step": 19927
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 19.41964340209961,
+      "learning_rate": 3.3267790653892818e-06,
+      "loss": 1.9393,
+      "step": 19928
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 7.241828918457031,
+      "learning_rate": 3.3259423503325946e-06,
+      "loss": 0.6794,
+      "step": 19929
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.395716190338135,
+      "learning_rate": 3.325105635275907e-06,
+      "loss": 0.0924,
+      "step": 19930
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 13.952709197998047,
+      "learning_rate": 3.3242689202192198e-06,
+      "loss": 1.1084,
+      "step": 19931
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 9.51970100402832,
+      "learning_rate": 3.323432205162532e-06,
+      "loss": 0.6472,
+      "step": 19932
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 26.60560417175293,
+      "learning_rate": 3.322595490105845e-06,
+      "loss": 1.0385,
+      "step": 19933
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 22.788707733154297,
+      "learning_rate": 3.3217587750491577e-06,
+      "loss": 0.7261,
+      "step": 19934
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 17.438146591186523,
+      "learning_rate": 3.3209220599924697e-06,
+      "loss": 1.7893,
+      "step": 19935
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.608642578125,
+      "learning_rate": 3.320085344935782e-06,
+      "loss": 0.5657,
+      "step": 19936
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 26.571809768676758,
+      "learning_rate": 3.319248629879095e-06,
+      "loss": 1.0257,
+      "step": 19937
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 14.450227737426758,
+      "learning_rate": 3.3184119148224077e-06,
+      "loss": 1.7862,
+      "step": 19938
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 17.081966400146484,
+      "learning_rate": 3.31757519976572e-06,
+      "loss": 1.4273,
+      "step": 19939
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 9.602093696594238,
+      "learning_rate": 3.316738484709033e-06,
+      "loss": 0.7716,
+      "step": 19940
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 15.299277305603027,
+      "learning_rate": 3.315901769652345e-06,
+      "loss": 0.7906,
+      "step": 19941
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.3168649673461914,
+      "learning_rate": 3.3150650545956576e-06,
+      "loss": 0.1147,
+      "step": 19942
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 10.020484924316406,
+      "learning_rate": 3.31422833953897e-06,
+      "loss": 0.5444,
+      "step": 19943
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 8.818445205688477,
+      "learning_rate": 3.3133916244822828e-06,
+      "loss": 0.7655,
+      "step": 19944
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 49.77157974243164,
+      "learning_rate": 3.3125549094255956e-06,
+      "loss": 2.2813,
+      "step": 19945
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 18.44257164001465,
+      "learning_rate": 3.311718194368908e-06,
+      "loss": 0.9881,
+      "step": 19946
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 75.80842590332031,
+      "learning_rate": 3.3108814793122208e-06,
+      "loss": 5.0154,
+      "step": 19947
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 60.95927429199219,
+      "learning_rate": 3.3100447642555327e-06,
+      "loss": 0.6309,
+      "step": 19948
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 17.889598846435547,
+      "learning_rate": 3.3092080491988455e-06,
+      "loss": 1.1538,
+      "step": 19949
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.751458168029785,
+      "learning_rate": 3.308371334142158e-06,
+      "loss": 0.3345,
+      "step": 19950
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 26.583431243896484,
+      "learning_rate": 3.3075346190854707e-06,
+      "loss": 0.6315,
+      "step": 19951
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 20.479509353637695,
+      "learning_rate": 3.3066979040287835e-06,
+      "loss": 3.3615,
+      "step": 19952
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.811371326446533,
+      "learning_rate": 3.305861188972096e-06,
+      "loss": 0.4642,
+      "step": 19953
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 8.042828559875488,
+      "learning_rate": 3.3050244739154087e-06,
+      "loss": 0.9753,
+      "step": 19954
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 14.338342666625977,
+      "learning_rate": 3.3041877588587206e-06,
+      "loss": 1.716,
+      "step": 19955
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 9.702662467956543,
+      "learning_rate": 3.3033510438020334e-06,
+      "loss": 0.5687,
+      "step": 19956
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 6.313868522644043,
+      "learning_rate": 3.302514328745346e-06,
+      "loss": 0.5918,
+      "step": 19957
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.752679824829102,
+      "learning_rate": 3.3016776136886586e-06,
+      "loss": 0.6654,
+      "step": 19958
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 68.26758575439453,
+      "learning_rate": 3.300840898631971e-06,
+      "loss": 1.3103,
+      "step": 19959
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 37.147972106933594,
+      "learning_rate": 3.300004183575284e-06,
+      "loss": 1.3215,
+      "step": 19960
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.803019046783447,
+      "learning_rate": 3.2991674685185966e-06,
+      "loss": 0.1563,
+      "step": 19961
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 7.469912052154541,
+      "learning_rate": 3.2983307534619086e-06,
+      "loss": 0.104,
+      "step": 19962
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 18.475568771362305,
+      "learning_rate": 3.2974940384052214e-06,
+      "loss": 0.9115,
+      "step": 19963
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 7.950811386108398,
+      "learning_rate": 3.2966573233485337e-06,
+      "loss": 0.4655,
+      "step": 19964
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.4812703132629395,
+      "learning_rate": 3.2958206082918465e-06,
+      "loss": 0.1384,
+      "step": 19965
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.8312828540802,
+      "learning_rate": 3.294983893235159e-06,
+      "loss": 0.4218,
+      "step": 19966
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 19.539031982421875,
+      "learning_rate": 3.2941471781784717e-06,
+      "loss": 0.4395,
+      "step": 19967
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 12.390055656433105,
+      "learning_rate": 3.2933104631217845e-06,
+      "loss": 1.0971,
+      "step": 19968
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 11.946459770202637,
+      "learning_rate": 3.2924737480650965e-06,
+      "loss": 0.2893,
+      "step": 19969
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.550558567047119,
+      "learning_rate": 3.2916370330084093e-06,
+      "loss": 0.2716,
+      "step": 19970
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 22.58316993713379,
+      "learning_rate": 3.2908003179517217e-06,
+      "loss": 1.5036,
+      "step": 19971
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 39.19320297241211,
+      "learning_rate": 3.2899636028950345e-06,
+      "loss": 1.392,
+      "step": 19972
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 11.933780670166016,
+      "learning_rate": 3.289126887838347e-06,
+      "loss": 0.567,
+      "step": 19973
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 2.2914276123046875,
+      "learning_rate": 3.2882901727816596e-06,
+      "loss": 0.1324,
+      "step": 19974
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 77.90569305419922,
+      "learning_rate": 3.2874534577249724e-06,
+      "loss": 1.653,
+      "step": 19975
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 9.612706184387207,
+      "learning_rate": 3.2866167426682844e-06,
+      "loss": 0.4752,
+      "step": 19976
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 16.720169067382812,
+      "learning_rate": 3.2857800276115968e-06,
+      "loss": 2.4077,
+      "step": 19977
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 18.49120330810547,
+      "learning_rate": 3.2849433125549096e-06,
+      "loss": 0.6799,
+      "step": 19978
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 22.447124481201172,
+      "learning_rate": 3.2841065974982224e-06,
+      "loss": 1.8789,
+      "step": 19979
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 6.493262767791748,
+      "learning_rate": 3.2832698824415348e-06,
+      "loss": 0.4613,
+      "step": 19980
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 57.3089599609375,
+      "learning_rate": 3.2824331673848476e-06,
+      "loss": 1.4242,
+      "step": 19981
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 2.4070045948028564,
+      "learning_rate": 3.2815964523281604e-06,
+      "loss": 0.17,
+      "step": 19982
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 10.130541801452637,
+      "learning_rate": 3.2807597372714723e-06,
+      "loss": 0.6856,
+      "step": 19983
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 21.3435115814209,
+      "learning_rate": 3.2799230222147847e-06,
+      "loss": 1.7294,
+      "step": 19984
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 16.479320526123047,
+      "learning_rate": 3.2790863071580975e-06,
+      "loss": 1.3781,
+      "step": 19985
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 17.575529098510742,
+      "learning_rate": 3.2782495921014103e-06,
+      "loss": 1.0384,
+      "step": 19986
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 48.52682113647461,
+      "learning_rate": 3.2774128770447227e-06,
+      "loss": 2.9254,
+      "step": 19987
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 15.873946189880371,
+      "learning_rate": 3.2765761619880355e-06,
+      "loss": 1.6611,
+      "step": 19988
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 15.20587158203125,
+      "learning_rate": 3.275739446931348e-06,
+      "loss": 0.3941,
+      "step": 19989
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 7.7490363121032715,
+      "learning_rate": 3.2749027318746602e-06,
+      "loss": 0.4194,
+      "step": 19990
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 18.263975143432617,
+      "learning_rate": 3.2740660168179726e-06,
+      "loss": 0.7866,
+      "step": 19991
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 17.40163230895996,
+      "learning_rate": 3.2732293017612854e-06,
+      "loss": 1.5032,
+      "step": 19992
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 21.275117874145508,
+      "learning_rate": 3.2723925867045982e-06,
+      "loss": 1.906,
+      "step": 19993
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 12.57906723022461,
+      "learning_rate": 3.2715558716479106e-06,
+      "loss": 0.7331,
+      "step": 19994
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 12.67908000946045,
+      "learning_rate": 3.2707191565912234e-06,
+      "loss": 0.4227,
+      "step": 19995
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 10.568775177001953,
+      "learning_rate": 3.2698824415345358e-06,
+      "loss": 0.8041,
+      "step": 19996
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 6.025681495666504,
+      "learning_rate": 3.269045726477848e-06,
+      "loss": 0.1539,
+      "step": 19997
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 13.698433876037598,
+      "learning_rate": 3.2682090114211605e-06,
+      "loss": 1.3535,
+      "step": 19998
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 2.734616279602051,
+      "learning_rate": 3.2673722963644733e-06,
+      "loss": 0.1233,
+      "step": 19999
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 25.240158081054688,
+      "learning_rate": 3.2665355813077857e-06,
+      "loss": 1.041,
+      "step": 20000
+    },
+    {
+      "epoch": 2.51,
+      "eval_loss": 0.0808543786406517,
+      "eval_runtime": 95.0651,
+      "eval_samples_per_second": 37.259,
+      "eval_steps_per_second": 37.259,
+      "step": 20000
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 39.454795837402344,
+      "learning_rate": 3.2656988662510985e-06,
+      "loss": 1.9426,
+      "step": 20001
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 12.454354286193848,
+      "learning_rate": 3.2648621511944113e-06,
+      "loss": 0.2503,
+      "step": 20002
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 10.543244361877441,
+      "learning_rate": 3.2640254361377237e-06,
+      "loss": 0.2657,
+      "step": 20003
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 75.90203094482422,
+      "learning_rate": 3.263188721081036e-06,
+      "loss": 0.8773,
+      "step": 20004
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 7.747319221496582,
+      "learning_rate": 3.2623520060243484e-06,
+      "loss": 0.5037,
+      "step": 20005
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 8.718236923217773,
+      "learning_rate": 3.2615152909676613e-06,
+      "loss": 0.3986,
+      "step": 20006
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 8.183634757995605,
+      "learning_rate": 3.2606785759109736e-06,
+      "loss": 1.8096,
+      "step": 20007
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 27.24098014831543,
+      "learning_rate": 3.2598418608542864e-06,
+      "loss": 0.724,
+      "step": 20008
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 10.727471351623535,
+      "learning_rate": 3.2590051457975992e-06,
+      "loss": 0.8958,
+      "step": 20009
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 21.750059127807617,
+      "learning_rate": 3.2581684307409116e-06,
+      "loss": 2.5499,
+      "step": 20010
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.738742351531982,
+      "learning_rate": 3.257331715684224e-06,
+      "loss": 1.2078,
+      "step": 20011
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 30.94146728515625,
+      "learning_rate": 3.2564950006275364e-06,
+      "loss": 0.4637,
+      "step": 20012
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 10.234822273254395,
+      "learning_rate": 3.255658285570849e-06,
+      "loss": 0.6016,
+      "step": 20013
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 20.73149871826172,
+      "learning_rate": 3.2548215705141615e-06,
+      "loss": 1.7852,
+      "step": 20014
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 2.1227056980133057,
+      "learning_rate": 3.2539848554574743e-06,
+      "loss": 0.1172,
+      "step": 20015
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 26.35969352722168,
+      "learning_rate": 3.253148140400787e-06,
+      "loss": 0.9049,
+      "step": 20016
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 10.881176948547363,
+      "learning_rate": 3.2523114253440995e-06,
+      "loss": 0.6887,
+      "step": 20017
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 15.723998069763184,
+      "learning_rate": 3.2514747102874115e-06,
+      "loss": 0.8719,
+      "step": 20018
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 14.749188423156738,
+      "learning_rate": 3.2506379952307243e-06,
+      "loss": 0.9361,
+      "step": 20019
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 20.851654052734375,
+      "learning_rate": 3.249801280174037e-06,
+      "loss": 2.1951,
+      "step": 20020
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 14.810416221618652,
+      "learning_rate": 3.2489645651173495e-06,
+      "loss": 0.7405,
+      "step": 20021
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 11.574252128601074,
+      "learning_rate": 3.2481278500606623e-06,
+      "loss": 0.6396,
+      "step": 20022
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 36.250186920166016,
+      "learning_rate": 3.247291135003975e-06,
+      "loss": 0.5705,
+      "step": 20023
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 12.263786315917969,
+      "learning_rate": 3.2464544199472874e-06,
+      "loss": 0.3928,
+      "step": 20024
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 25.44635009765625,
+      "learning_rate": 3.2456177048905994e-06,
+      "loss": 0.8051,
+      "step": 20025
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 7.821044921875,
+      "learning_rate": 3.244780989833912e-06,
+      "loss": 0.604,
+      "step": 20026
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 19.18406105041504,
+      "learning_rate": 3.243944274777225e-06,
+      "loss": 0.2634,
+      "step": 20027
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 26.439741134643555,
+      "learning_rate": 3.2431075597205374e-06,
+      "loss": 1.8381,
+      "step": 20028
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 11.98216724395752,
+      "learning_rate": 3.24227084466385e-06,
+      "loss": 1.3557,
+      "step": 20029
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 11.501047134399414,
+      "learning_rate": 3.2414341296071626e-06,
+      "loss": 0.9794,
+      "step": 20030
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 11.660356521606445,
+      "learning_rate": 3.240597414550475e-06,
+      "loss": 0.6197,
+      "step": 20031
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.886253833770752,
+      "learning_rate": 3.2397606994937873e-06,
+      "loss": 0.3043,
+      "step": 20032
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 40.572940826416016,
+      "learning_rate": 3.2389239844371e-06,
+      "loss": 0.3478,
+      "step": 20033
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 10.604790687561035,
+      "learning_rate": 3.238087269380413e-06,
+      "loss": 0.3787,
+      "step": 20034
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 18.762178421020508,
+      "learning_rate": 3.2372505543237253e-06,
+      "loss": 0.559,
+      "step": 20035
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.1380579471588135,
+      "learning_rate": 3.236413839267038e-06,
+      "loss": 0.2146,
+      "step": 20036
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 12.402771949768066,
+      "learning_rate": 3.2355771242103505e-06,
+      "loss": 1.3115,
+      "step": 20037
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 16.927108764648438,
+      "learning_rate": 3.234740409153663e-06,
+      "loss": 0.6548,
+      "step": 20038
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 24.82040786743164,
+      "learning_rate": 3.2339036940969752e-06,
+      "loss": 0.9741,
+      "step": 20039
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 18.516328811645508,
+      "learning_rate": 3.233066979040288e-06,
+      "loss": 1.3821,
+      "step": 20040
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 17.94833755493164,
+      "learning_rate": 3.2322302639836004e-06,
+      "loss": 0.525,
+      "step": 20041
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 10.300509452819824,
+      "learning_rate": 3.2313935489269132e-06,
+      "loss": 1.9777,
+      "step": 20042
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 25.85494613647461,
+      "learning_rate": 3.230556833870226e-06,
+      "loss": 1.0825,
+      "step": 20043
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 16.04183578491211,
+      "learning_rate": 3.2297201188135384e-06,
+      "loss": 1.3378,
+      "step": 20044
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.046987771987915,
+      "learning_rate": 3.2288834037568508e-06,
+      "loss": 0.0908,
+      "step": 20045
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 17.787328720092773,
+      "learning_rate": 3.228046688700163e-06,
+      "loss": 1.0892,
+      "step": 20046
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 114.71671295166016,
+      "learning_rate": 3.227209973643476e-06,
+      "loss": 1.1975,
+      "step": 20047
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 48.55499267578125,
+      "learning_rate": 3.2263732585867883e-06,
+      "loss": 1.4782,
+      "step": 20048
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.1245615482330322,
+      "learning_rate": 3.225536543530101e-06,
+      "loss": 0.0313,
+      "step": 20049
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 20.053237915039062,
+      "learning_rate": 3.224699828473414e-06,
+      "loss": 0.3424,
+      "step": 20050
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 9.567726135253906,
+      "learning_rate": 3.2238631134167263e-06,
+      "loss": 1.8609,
+      "step": 20051
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 17.328189849853516,
+      "learning_rate": 3.2230263983600387e-06,
+      "loss": 0.6531,
+      "step": 20052
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 13.401254653930664,
+      "learning_rate": 3.222189683303351e-06,
+      "loss": 1.1483,
+      "step": 20053
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 16.87504005432129,
+      "learning_rate": 3.221352968246664e-06,
+      "loss": 0.5526,
+      "step": 20054
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 15.760570526123047,
+      "learning_rate": 3.2205162531899763e-06,
+      "loss": 1.4285,
+      "step": 20055
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 13.635117530822754,
+      "learning_rate": 3.219679538133289e-06,
+      "loss": 0.4915,
+      "step": 20056
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 33.641998291015625,
+      "learning_rate": 3.218842823076602e-06,
+      "loss": 0.8777,
+      "step": 20057
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 16.73700523376465,
+      "learning_rate": 3.2180061080199142e-06,
+      "loss": 0.5806,
+      "step": 20058
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 11.086918830871582,
+      "learning_rate": 3.217169392963226e-06,
+      "loss": 0.8107,
+      "step": 20059
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 11.779153823852539,
+      "learning_rate": 3.216332677906539e-06,
+      "loss": 1.6249,
+      "step": 20060
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 12.519206047058105,
+      "learning_rate": 3.215495962849852e-06,
+      "loss": 1.0151,
+      "step": 20061
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 9.511662483215332,
+      "learning_rate": 3.214659247793164e-06,
+      "loss": 1.6965,
+      "step": 20062
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 9.904976844787598,
+      "learning_rate": 3.213822532736477e-06,
+      "loss": 0.5901,
+      "step": 20063
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 8.143558502197266,
+      "learning_rate": 3.2129858176797894e-06,
+      "loss": 0.58,
+      "step": 20064
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 2.6461517810821533,
+      "learning_rate": 3.212149102623102e-06,
+      "loss": 0.0457,
+      "step": 20065
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 13.534331321716309,
+      "learning_rate": 3.211312387566414e-06,
+      "loss": 1.4003,
+      "step": 20066
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 11.298065185546875,
+      "learning_rate": 3.210475672509727e-06,
+      "loss": 0.6139,
+      "step": 20067
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 16.307167053222656,
+      "learning_rate": 3.2096389574530397e-06,
+      "loss": 1.1053,
+      "step": 20068
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 15.244307518005371,
+      "learning_rate": 3.208802242396352e-06,
+      "loss": 0.5344,
+      "step": 20069
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 26.728551864624023,
+      "learning_rate": 3.207965527339665e-06,
+      "loss": 1.2105,
+      "step": 20070
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 22.271337509155273,
+      "learning_rate": 3.2071288122829773e-06,
+      "loss": 1.1998,
+      "step": 20071
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 11.722222328186035,
+      "learning_rate": 3.20629209722629e-06,
+      "loss": 0.5946,
+      "step": 20072
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 29.611454010009766,
+      "learning_rate": 3.205455382169602e-06,
+      "loss": 0.3169,
+      "step": 20073
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1529.944091796875,
+      "learning_rate": 3.204618667112915e-06,
+      "loss": 0.6748,
+      "step": 20074
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 10.68509292602539,
+      "learning_rate": 3.2037819520562276e-06,
+      "loss": 0.7615,
+      "step": 20075
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 18.093368530273438,
+      "learning_rate": 3.20294523699954e-06,
+      "loss": 2.1586,
+      "step": 20076
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.211366176605225,
+      "learning_rate": 3.202108521942853e-06,
+      "loss": 0.2439,
+      "step": 20077
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 25.321666717529297,
+      "learning_rate": 3.201271806886165e-06,
+      "loss": 1.5209,
+      "step": 20078
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 71.18042755126953,
+      "learning_rate": 3.200435091829478e-06,
+      "loss": 0.9226,
+      "step": 20079
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 8.386185646057129,
+      "learning_rate": 3.19959837677279e-06,
+      "loss": 0.2261,
+      "step": 20080
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 17.505979537963867,
+      "learning_rate": 3.1987616617161028e-06,
+      "loss": 1.5554,
+      "step": 20081
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 143.35447692871094,
+      "learning_rate": 3.197924946659415e-06,
+      "loss": 0.3874,
+      "step": 20082
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 9.683552742004395,
+      "learning_rate": 3.197088231602728e-06,
+      "loss": 1.7246,
+      "step": 20083
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 58.537803649902344,
+      "learning_rate": 3.1962515165460407e-06,
+      "loss": 0.4977,
+      "step": 20084
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 14.881050109863281,
+      "learning_rate": 3.195414801489353e-06,
+      "loss": 0.4981,
+      "step": 20085
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 37.98305130004883,
+      "learning_rate": 3.194578086432666e-06,
+      "loss": 2.2355,
+      "step": 20086
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 12.213129043579102,
+      "learning_rate": 3.193741371375978e-06,
+      "loss": 0.4452,
+      "step": 20087
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 11.402565002441406,
+      "learning_rate": 3.1929046563192907e-06,
+      "loss": 0.6451,
+      "step": 20088
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 29.905174255371094,
+      "learning_rate": 3.192067941262603e-06,
+      "loss": 1.5578,
+      "step": 20089
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 11.379555702209473,
+      "learning_rate": 3.191231226205916e-06,
+      "loss": 0.7022,
+      "step": 20090
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 15.473186492919922,
+      "learning_rate": 3.1903945111492287e-06,
+      "loss": 1.0074,
+      "step": 20091
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 9.624855041503906,
+      "learning_rate": 3.189557796092541e-06,
+      "loss": 0.9283,
+      "step": 20092
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 24.56440544128418,
+      "learning_rate": 3.188721081035854e-06,
+      "loss": 1.3809,
+      "step": 20093
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 135.54013061523438,
+      "learning_rate": 3.1878843659791658e-06,
+      "loss": 0.7548,
+      "step": 20094
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 15.073548316955566,
+      "learning_rate": 3.1870476509224786e-06,
+      "loss": 0.3575,
+      "step": 20095
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 15.966330528259277,
+      "learning_rate": 3.186210935865791e-06,
+      "loss": 0.6908,
+      "step": 20096
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 10.339390754699707,
+      "learning_rate": 3.1853742208091038e-06,
+      "loss": 0.6126,
+      "step": 20097
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 6.085317611694336,
+      "learning_rate": 3.1845375057524166e-06,
+      "loss": 1.5773,
+      "step": 20098
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 27.03826141357422,
+      "learning_rate": 3.183700790695729e-06,
+      "loss": 1.9262,
+      "step": 20099
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 10.994552612304688,
+      "learning_rate": 3.1828640756390417e-06,
+      "loss": 1.7028,
+      "step": 20100
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.914076805114746,
+      "learning_rate": 3.1820273605823537e-06,
+      "loss": 1.0003,
+      "step": 20101
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.8910298347473145,
+      "learning_rate": 3.1811906455256665e-06,
+      "loss": 1.4367,
+      "step": 20102
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 14.262541770935059,
+      "learning_rate": 3.180353930468979e-06,
+      "loss": 0.1744,
+      "step": 20103
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 18.70867919921875,
+      "learning_rate": 3.1795172154122917e-06,
+      "loss": 1.6976,
+      "step": 20104
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 11.734964370727539,
+      "learning_rate": 3.178680500355604e-06,
+      "loss": 0.6124,
+      "step": 20105
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.6546733379364014,
+      "learning_rate": 3.177843785298917e-06,
+      "loss": 1.3167,
+      "step": 20106
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 21.559032440185547,
+      "learning_rate": 3.1770070702422297e-06,
+      "loss": 0.9629,
+      "step": 20107
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 22.326847076416016,
+      "learning_rate": 3.1761703551855416e-06,
+      "loss": 1.1904,
+      "step": 20108
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 26.877647399902344,
+      "learning_rate": 3.1753336401288544e-06,
+      "loss": 1.3934,
+      "step": 20109
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 13.50160026550293,
+      "learning_rate": 3.174496925072167e-06,
+      "loss": 1.0344,
+      "step": 20110
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 63.82146072387695,
+      "learning_rate": 3.1736602100154796e-06,
+      "loss": 1.1811,
+      "step": 20111
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 10.990103721618652,
+      "learning_rate": 3.172823494958792e-06,
+      "loss": 0.4297,
+      "step": 20112
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 10.466145515441895,
+      "learning_rate": 3.1719867799021048e-06,
+      "loss": 0.2588,
+      "step": 20113
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 100.87720489501953,
+      "learning_rate": 3.1711500648454167e-06,
+      "loss": 1.0303,
+      "step": 20114
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 10.929286003112793,
+      "learning_rate": 3.1703133497887295e-06,
+      "loss": 0.6933,
+      "step": 20115
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 348.97802734375,
+      "learning_rate": 3.1694766347320423e-06,
+      "loss": 1.9972,
+      "step": 20116
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 15.033949851989746,
+      "learning_rate": 3.1686399196753547e-06,
+      "loss": 0.5968,
+      "step": 20117
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 13.286337852478027,
+      "learning_rate": 3.1678032046186675e-06,
+      "loss": 1.5688,
+      "step": 20118
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 83.24957275390625,
+      "learning_rate": 3.16696648956198e-06,
+      "loss": 2.0411,
+      "step": 20119
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 23.767091751098633,
+      "learning_rate": 3.1661297745052927e-06,
+      "loss": 1.3859,
+      "step": 20120
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 7.808661460876465,
+      "learning_rate": 3.1652930594486047e-06,
+      "loss": 0.5942,
+      "step": 20121
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 26.495922088623047,
+      "learning_rate": 3.1644563443919175e-06,
+      "loss": 1.1367,
+      "step": 20122
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 18.31427764892578,
+      "learning_rate": 3.16361962933523e-06,
+      "loss": 1.3548,
+      "step": 20123
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 15.174376487731934,
+      "learning_rate": 3.1627829142785426e-06,
+      "loss": 0.6136,
+      "step": 20124
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.846739768981934,
+      "learning_rate": 3.1619461992218554e-06,
+      "loss": 0.0403,
+      "step": 20125
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 21.373008728027344,
+      "learning_rate": 3.161109484165168e-06,
+      "loss": 1.0197,
+      "step": 20126
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 42.04866409301758,
+      "learning_rate": 3.1602727691084806e-06,
+      "loss": 0.5158,
+      "step": 20127
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 30.856718063354492,
+      "learning_rate": 3.1594360540517926e-06,
+      "loss": 3.1075,
+      "step": 20128
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 9.922351837158203,
+      "learning_rate": 3.1585993389951054e-06,
+      "loss": 1.4879,
+      "step": 20129
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 19.042163848876953,
+      "learning_rate": 3.1577626239384178e-06,
+      "loss": 1.5319,
+      "step": 20130
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 26.65334701538086,
+      "learning_rate": 3.1569259088817306e-06,
+      "loss": 0.5933,
+      "step": 20131
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 20.031673431396484,
+      "learning_rate": 3.1560891938250434e-06,
+      "loss": 0.939,
+      "step": 20132
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 9.657144546508789,
+      "learning_rate": 3.1552524787683557e-06,
+      "loss": 0.474,
+      "step": 20133
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 9.074506759643555,
+      "learning_rate": 3.1544157637116685e-06,
+      "loss": 1.8527,
+      "step": 20134
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 18.564651489257812,
+      "learning_rate": 3.1535790486549805e-06,
+      "loss": 0.4606,
+      "step": 20135
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 25.953868865966797,
+      "learning_rate": 3.1527423335982933e-06,
+      "loss": 0.716,
+      "step": 20136
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 18.119686126708984,
+      "learning_rate": 3.1519056185416057e-06,
+      "loss": 1.8624,
+      "step": 20137
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 13.626513481140137,
+      "learning_rate": 3.1510689034849185e-06,
+      "loss": 1.674,
+      "step": 20138
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 6.355057716369629,
+      "learning_rate": 3.1502321884282313e-06,
+      "loss": 0.4331,
+      "step": 20139
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 16.89761734008789,
+      "learning_rate": 3.1493954733715437e-06,
+      "loss": 1.8812,
+      "step": 20140
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 32.7131233215332,
+      "learning_rate": 3.1485587583148565e-06,
+      "loss": 1.5138,
+      "step": 20141
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 20.496959686279297,
+      "learning_rate": 3.1477220432581684e-06,
+      "loss": 0.4932,
+      "step": 20142
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 38.02460861206055,
+      "learning_rate": 3.1468853282014812e-06,
+      "loss": 0.3606,
+      "step": 20143
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 20.9803409576416,
+      "learning_rate": 3.1460486131447936e-06,
+      "loss": 0.8481,
+      "step": 20144
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 7.6487627029418945,
+      "learning_rate": 3.1452118980881064e-06,
+      "loss": 0.1384,
+      "step": 20145
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 25.933940887451172,
+      "learning_rate": 3.1443751830314188e-06,
+      "loss": 0.9629,
+      "step": 20146
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 138.8772735595703,
+      "learning_rate": 3.1435384679747316e-06,
+      "loss": 1.5815,
+      "step": 20147
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.3213183879852295,
+      "learning_rate": 3.1427017529180444e-06,
+      "loss": 0.0642,
+      "step": 20148
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 12.18281364440918,
+      "learning_rate": 3.1418650378613563e-06,
+      "loss": 1.4909,
+      "step": 20149
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 11.219976425170898,
+      "learning_rate": 3.141028322804669e-06,
+      "loss": 0.8455,
+      "step": 20150
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 19.861003875732422,
+      "learning_rate": 3.1401916077479815e-06,
+      "loss": 1.4017,
+      "step": 20151
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 24.716703414916992,
+      "learning_rate": 3.1393548926912943e-06,
+      "loss": 2.7867,
+      "step": 20152
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 74.31487274169922,
+      "learning_rate": 3.1385181776346067e-06,
+      "loss": 0.9074,
+      "step": 20153
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 15.867815971374512,
+      "learning_rate": 3.1376814625779195e-06,
+      "loss": 1.1159,
+      "step": 20154
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 18.009172439575195,
+      "learning_rate": 3.1368447475212323e-06,
+      "loss": 0.5687,
+      "step": 20155
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 15.600923538208008,
+      "learning_rate": 3.1360080324645443e-06,
+      "loss": 0.3674,
+      "step": 20156
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 559.9266967773438,
+      "learning_rate": 3.135171317407857e-06,
+      "loss": 2.0866,
+      "step": 20157
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 11.303703308105469,
+      "learning_rate": 3.1343346023511694e-06,
+      "loss": 0.85,
+      "step": 20158
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 17.95802879333496,
+      "learning_rate": 3.1334978872944822e-06,
+      "loss": 0.9572,
+      "step": 20159
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 14.622774124145508,
+      "learning_rate": 3.1326611722377946e-06,
+      "loss": 1.3898,
+      "step": 20160
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 9.754227638244629,
+      "learning_rate": 3.1318244571811074e-06,
+      "loss": 0.821,
+      "step": 20161
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 6.329394817352295,
+      "learning_rate": 3.1309877421244202e-06,
+      "loss": 0.5835,
+      "step": 20162
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 472.1472473144531,
+      "learning_rate": 3.130151027067732e-06,
+      "loss": 2.2185,
+      "step": 20163
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 8.972721099853516,
+      "learning_rate": 3.1293143120110445e-06,
+      "loss": 0.298,
+      "step": 20164
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 9.203935623168945,
+      "learning_rate": 3.1284775969543573e-06,
+      "loss": 1.1663,
+      "step": 20165
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 28.42084312438965,
+      "learning_rate": 3.12764088189767e-06,
+      "loss": 0.9144,
+      "step": 20166
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 239.5991668701172,
+      "learning_rate": 3.1268041668409825e-06,
+      "loss": 1.6324,
+      "step": 20167
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 41.38919448852539,
+      "learning_rate": 3.1259674517842953e-06,
+      "loss": 1.3735,
+      "step": 20168
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 7.04861307144165,
+      "learning_rate": 3.1251307367276077e-06,
+      "loss": 0.3261,
+      "step": 20169
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 7.5475077629089355,
+      "learning_rate": 3.12429402167092e-06,
+      "loss": 0.3679,
+      "step": 20170
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 25.39781951904297,
+      "learning_rate": 3.1234573066142325e-06,
+      "loss": 0.6875,
+      "step": 20171
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 21.164522171020508,
+      "learning_rate": 3.1226205915575453e-06,
+      "loss": 1.5491,
+      "step": 20172
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.918303966522217,
+      "learning_rate": 3.121783876500858e-06,
+      "loss": 0.3879,
+      "step": 20173
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 10.785502433776855,
+      "learning_rate": 3.1209471614441704e-06,
+      "loss": 1.1789,
+      "step": 20174
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.143613338470459,
+      "learning_rate": 3.1201104463874832e-06,
+      "loss": 0.6366,
+      "step": 20175
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 8.28675365447998,
+      "learning_rate": 3.1192737313307956e-06,
+      "loss": 0.4371,
+      "step": 20176
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 37.711341857910156,
+      "learning_rate": 3.118437016274108e-06,
+      "loss": 0.2428,
+      "step": 20177
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 38.26509475708008,
+      "learning_rate": 3.1176003012174204e-06,
+      "loss": 2.5128,
+      "step": 20178
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 39.81582260131836,
+      "learning_rate": 3.116763586160733e-06,
+      "loss": 1.7895,
+      "step": 20179
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 10.409308433532715,
+      "learning_rate": 3.115926871104046e-06,
+      "loss": 1.2166,
+      "step": 20180
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 11.536985397338867,
+      "learning_rate": 3.1150901560473584e-06,
+      "loss": 1.8168,
+      "step": 20181
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 17.858381271362305,
+      "learning_rate": 3.114253440990671e-06,
+      "loss": 0.383,
+      "step": 20182
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 36.268795013427734,
+      "learning_rate": 3.1134167259339835e-06,
+      "loss": 2.7144,
+      "step": 20183
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 6.797853946685791,
+      "learning_rate": 3.112580010877296e-06,
+      "loss": 0.3438,
+      "step": 20184
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.46155309677124,
+      "learning_rate": 3.1117432958206083e-06,
+      "loss": 0.1619,
+      "step": 20185
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 17.920578002929688,
+      "learning_rate": 3.110906580763921e-06,
+      "loss": 1.1489,
+      "step": 20186
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.990722179412842,
+      "learning_rate": 3.1100698657072335e-06,
+      "loss": 0.0987,
+      "step": 20187
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 29.773059844970703,
+      "learning_rate": 3.1092331506505463e-06,
+      "loss": 1.033,
+      "step": 20188
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 17.676555633544922,
+      "learning_rate": 3.108396435593859e-06,
+      "loss": 1.1681,
+      "step": 20189
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 16.256336212158203,
+      "learning_rate": 3.1075597205371715e-06,
+      "loss": 2.0528,
+      "step": 20190
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.727039337158203,
+      "learning_rate": 3.106723005480484e-06,
+      "loss": 0.435,
+      "step": 20191
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 7.141747951507568,
+      "learning_rate": 3.1058862904237962e-06,
+      "loss": 0.1768,
+      "step": 20192
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 18.85256576538086,
+      "learning_rate": 3.105049575367109e-06,
+      "loss": 0.7793,
+      "step": 20193
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 35.84891128540039,
+      "learning_rate": 3.1042128603104214e-06,
+      "loss": 1.0192,
+      "step": 20194
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 25.55327033996582,
+      "learning_rate": 3.103376145253734e-06,
+      "loss": 2.2958,
+      "step": 20195
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 9.817205429077148,
+      "learning_rate": 3.102539430197047e-06,
+      "loss": 0.2239,
+      "step": 20196
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 46.79560470581055,
+      "learning_rate": 3.1017027151403594e-06,
+      "loss": 0.8706,
+      "step": 20197
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 17.943552017211914,
+      "learning_rate": 3.1008660000836718e-06,
+      "loss": 1.6181,
+      "step": 20198
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 81.96236419677734,
+      "learning_rate": 3.100029285026984e-06,
+      "loss": 2.0724,
+      "step": 20199
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 16.38736915588379,
+      "learning_rate": 3.099192569970297e-06,
+      "loss": 1.3343,
+      "step": 20200
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 9.381766319274902,
+      "learning_rate": 3.0983558549136093e-06,
+      "loss": 0.3703,
+      "step": 20201
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 16.660558700561523,
+      "learning_rate": 3.097519139856922e-06,
+      "loss": 0.7836,
+      "step": 20202
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 18.434297561645508,
+      "learning_rate": 3.096682424800235e-06,
+      "loss": 1.3671,
+      "step": 20203
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 6.595559597015381,
+      "learning_rate": 3.095845709743547e-06,
+      "loss": 1.5472,
+      "step": 20204
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.3526664972305298,
+      "learning_rate": 3.0950089946868593e-06,
+      "loss": 0.0462,
+      "step": 20205
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 16.824804306030273,
+      "learning_rate": 3.094172279630172e-06,
+      "loss": 0.5207,
+      "step": 20206
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 18.006778717041016,
+      "learning_rate": 3.093335564573485e-06,
+      "loss": 0.8913,
+      "step": 20207
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 26.328611373901367,
+      "learning_rate": 3.0924988495167972e-06,
+      "loss": 1.2545,
+      "step": 20208
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 25.5814266204834,
+      "learning_rate": 3.09166213446011e-06,
+      "loss": 0.9104,
+      "step": 20209
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 61.892154693603516,
+      "learning_rate": 3.0908254194034224e-06,
+      "loss": 1.1347,
+      "step": 20210
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 15.218506813049316,
+      "learning_rate": 3.089988704346735e-06,
+      "loss": 1.982,
+      "step": 20211
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 16.15838623046875,
+      "learning_rate": 3.089151989290047e-06,
+      "loss": 1.277,
+      "step": 20212
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 25.595287322998047,
+      "learning_rate": 3.08831527423336e-06,
+      "loss": 0.8428,
+      "step": 20213
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.472421646118164,
+      "learning_rate": 3.0874785591766728e-06,
+      "loss": 0.4231,
+      "step": 20214
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.6701686382293701,
+      "learning_rate": 3.086641844119985e-06,
+      "loss": 0.0995,
+      "step": 20215
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 30.975154876708984,
+      "learning_rate": 3.085805129063298e-06,
+      "loss": 3.2722,
+      "step": 20216
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 18.317169189453125,
+      "learning_rate": 3.0849684140066103e-06,
+      "loss": 0.5636,
+      "step": 20217
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.2889814376831055,
+      "learning_rate": 3.0841316989499227e-06,
+      "loss": 0.2052,
+      "step": 20218
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 14.267701148986816,
+      "learning_rate": 3.083294983893235e-06,
+      "loss": 2.1028,
+      "step": 20219
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 9.833152770996094,
+      "learning_rate": 3.082458268836548e-06,
+      "loss": 0.8909,
+      "step": 20220
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 18.517662048339844,
+      "learning_rate": 3.0816215537798607e-06,
+      "loss": 1.3313,
+      "step": 20221
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 16.407983779907227,
+      "learning_rate": 3.080784838723173e-06,
+      "loss": 0.6524,
+      "step": 20222
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 16.671092987060547,
+      "learning_rate": 3.079948123666486e-06,
+      "loss": 1.1518,
+      "step": 20223
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 7.368433952331543,
+      "learning_rate": 3.0791114086097983e-06,
+      "loss": 0.5794,
+      "step": 20224
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 16.057140350341797,
+      "learning_rate": 3.0782746935531106e-06,
+      "loss": 1.0874,
+      "step": 20225
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 52.69207763671875,
+      "learning_rate": 3.077437978496423e-06,
+      "loss": 2.3154,
+      "step": 20226
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 15.292400360107422,
+      "learning_rate": 3.076601263439736e-06,
+      "loss": 1.1101,
+      "step": 20227
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 8.293540954589844,
+      "learning_rate": 3.075764548383048e-06,
+      "loss": 3.1448,
+      "step": 20228
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.3503389358520508,
+      "learning_rate": 3.074927833326361e-06,
+      "loss": 0.0068,
+      "step": 20229
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 13.330592155456543,
+      "learning_rate": 3.074091118269674e-06,
+      "loss": 0.7966,
+      "step": 20230
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 13.415767669677734,
+      "learning_rate": 3.073254403212986e-06,
+      "loss": 1.176,
+      "step": 20231
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.649020195007324,
+      "learning_rate": 3.0724176881562986e-06,
+      "loss": 1.315,
+      "step": 20232
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 10.661545753479004,
+      "learning_rate": 3.071580973099611e-06,
+      "loss": 0.5251,
+      "step": 20233
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 7.858932018280029,
+      "learning_rate": 3.0707442580429237e-06,
+      "loss": 0.8182,
+      "step": 20234
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 9.638903617858887,
+      "learning_rate": 3.069907542986236e-06,
+      "loss": 0.2427,
+      "step": 20235
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 12.211078643798828,
+      "learning_rate": 3.069070827929549e-06,
+      "loss": 0.9441,
+      "step": 20236
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 8.404287338256836,
+      "learning_rate": 3.0682341128728617e-06,
+      "loss": 0.3233,
+      "step": 20237
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 29.669578552246094,
+      "learning_rate": 3.067397397816174e-06,
+      "loss": 0.7231,
+      "step": 20238
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 19.04340171813965,
+      "learning_rate": 3.0665606827594865e-06,
+      "loss": 0.6144,
+      "step": 20239
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.203525066375732,
+      "learning_rate": 3.065723967702799e-06,
+      "loss": 1.0986,
+      "step": 20240
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 8.037447929382324,
+      "learning_rate": 3.0648872526461117e-06,
+      "loss": 0.6277,
+      "step": 20241
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 36.33158493041992,
+      "learning_rate": 3.064050537589424e-06,
+      "loss": 1.2463,
+      "step": 20242
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 8.591270446777344,
+      "learning_rate": 3.063213822532737e-06,
+      "loss": 0.8348,
+      "step": 20243
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 15.740802764892578,
+      "learning_rate": 3.0623771074760496e-06,
+      "loss": 0.4979,
+      "step": 20244
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 14.099591255187988,
+      "learning_rate": 3.061540392419362e-06,
+      "loss": 1.0288,
+      "step": 20245
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 16.36825942993164,
+      "learning_rate": 3.060703677362674e-06,
+      "loss": 0.52,
+      "step": 20246
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 9.811111450195312,
+      "learning_rate": 3.0598669623059868e-06,
+      "loss": 0.2784,
+      "step": 20247
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 17.12701988220215,
+      "learning_rate": 3.0590302472492996e-06,
+      "loss": 1.1396,
+      "step": 20248
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 11.216033935546875,
+      "learning_rate": 3.058193532192612e-06,
+      "loss": 0.8465,
+      "step": 20249
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.6195974349975586,
+      "learning_rate": 3.0573568171359247e-06,
+      "loss": 0.108,
+      "step": 20250
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.237502098083496,
+      "learning_rate": 3.056520102079237e-06,
+      "loss": 0.2449,
+      "step": 20251
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 27.812013626098633,
+      "learning_rate": 3.05568338702255e-06,
+      "loss": 0.866,
+      "step": 20252
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.755734920501709,
+      "learning_rate": 3.054846671965862e-06,
+      "loss": 0.3416,
+      "step": 20253
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 6.678841590881348,
+      "learning_rate": 3.0540099569091747e-06,
+      "loss": 0.5921,
+      "step": 20254
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 11.03495979309082,
+      "learning_rate": 3.0531732418524875e-06,
+      "loss": 1.387,
+      "step": 20255
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 24.642271041870117,
+      "learning_rate": 3.0523365267958e-06,
+      "loss": 1.3909,
+      "step": 20256
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 8.620196342468262,
+      "learning_rate": 3.0514998117391127e-06,
+      "loss": 0.7151,
+      "step": 20257
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 18.334409713745117,
+      "learning_rate": 3.050663096682425e-06,
+      "loss": 0.344,
+      "step": 20258
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.8685402870178223,
+      "learning_rate": 3.049826381625738e-06,
+      "loss": 0.0499,
+      "step": 20259
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 7.711960792541504,
+      "learning_rate": 3.04898966656905e-06,
+      "loss": 1.4663,
+      "step": 20260
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 127.82076263427734,
+      "learning_rate": 3.0481529515123626e-06,
+      "loss": 1.5027,
+      "step": 20261
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 16.69835662841797,
+      "learning_rate": 3.0473162364556754e-06,
+      "loss": 0.9731,
+      "step": 20262
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 46.56307601928711,
+      "learning_rate": 3.0464795213989878e-06,
+      "loss": 1.4718,
+      "step": 20263
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.6854890584945679,
+      "learning_rate": 3.0456428063423006e-06,
+      "loss": 0.0537,
+      "step": 20264
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 7.84578275680542,
+      "learning_rate": 3.044806091285613e-06,
+      "loss": 1.0442,
+      "step": 20265
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 64.02957153320312,
+      "learning_rate": 3.0439693762289258e-06,
+      "loss": 0.9431,
+      "step": 20266
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 6.88743257522583,
+      "learning_rate": 3.0431326611722377e-06,
+      "loss": 0.9714,
+      "step": 20267
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 8.80877685546875,
+      "learning_rate": 3.0422959461155505e-06,
+      "loss": 0.5565,
+      "step": 20268
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 105.63014221191406,
+      "learning_rate": 3.041459231058863e-06,
+      "loss": 0.6282,
+      "step": 20269
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 60.46650695800781,
+      "learning_rate": 3.0406225160021757e-06,
+      "loss": 2.6536,
+      "step": 20270
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 19.91202735900879,
+      "learning_rate": 3.0397858009454885e-06,
+      "loss": 1.1612,
+      "step": 20271
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 15.705401420593262,
+      "learning_rate": 3.038949085888801e-06,
+      "loss": 1.0115,
+      "step": 20272
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 12.167220115661621,
+      "learning_rate": 3.0381123708321137e-06,
+      "loss": 1.4634,
+      "step": 20273
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 85.41754150390625,
+      "learning_rate": 3.0372756557754256e-06,
+      "loss": 3.3383,
+      "step": 20274
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.647180557250977,
+      "learning_rate": 3.0364389407187384e-06,
+      "loss": 0.2814,
+      "step": 20275
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 8.535072326660156,
+      "learning_rate": 3.035602225662051e-06,
+      "loss": 1.3482,
+      "step": 20276
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.6375298500061035,
+      "learning_rate": 3.0347655106053636e-06,
+      "loss": 0.1065,
+      "step": 20277
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 44.29484558105469,
+      "learning_rate": 3.0339287955486764e-06,
+      "loss": 1.8078,
+      "step": 20278
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 15.028776168823242,
+      "learning_rate": 3.033092080491989e-06,
+      "loss": 0.9637,
+      "step": 20279
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 34.271339416503906,
+      "learning_rate": 3.0322553654353016e-06,
+      "loss": 1.7199,
+      "step": 20280
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 25.27798080444336,
+      "learning_rate": 3.0314186503786136e-06,
+      "loss": 0.869,
+      "step": 20281
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 429.1591796875,
+      "learning_rate": 3.0305819353219264e-06,
+      "loss": 2.6978,
+      "step": 20282
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 9.644042015075684,
+      "learning_rate": 3.0297452202652387e-06,
+      "loss": 0.7662,
+      "step": 20283
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 17.65026092529297,
+      "learning_rate": 3.0289085052085515e-06,
+      "loss": 1.4505,
+      "step": 20284
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 40.25577926635742,
+      "learning_rate": 3.0280717901518643e-06,
+      "loss": 1.4754,
+      "step": 20285
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 11.539678573608398,
+      "learning_rate": 3.0272350750951767e-06,
+      "loss": 0.1681,
+      "step": 20286
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.209381580352783,
+      "learning_rate": 3.0263983600384895e-06,
+      "loss": 0.2759,
+      "step": 20287
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.28467321395874,
+      "learning_rate": 3.0255616449818015e-06,
+      "loss": 0.4294,
+      "step": 20288
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 10.976964950561523,
+      "learning_rate": 3.0247249299251143e-06,
+      "loss": 0.4842,
+      "step": 20289
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 8.156913757324219,
+      "learning_rate": 3.0238882148684267e-06,
+      "loss": 1.174,
+      "step": 20290
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 18.176769256591797,
+      "learning_rate": 3.0230514998117395e-06,
+      "loss": 0.8032,
+      "step": 20291
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 13.518056869506836,
+      "learning_rate": 3.022214784755052e-06,
+      "loss": 0.6218,
+      "step": 20292
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 8.877962112426758,
+      "learning_rate": 3.0213780696983646e-06,
+      "loss": 0.9831,
+      "step": 20293
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 16.532529830932617,
+      "learning_rate": 3.0205413546416766e-06,
+      "loss": 1.6324,
+      "step": 20294
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 14.755925178527832,
+      "learning_rate": 3.0197046395849894e-06,
+      "loss": 1.2597,
+      "step": 20295
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 13.560070037841797,
+      "learning_rate": 3.018867924528302e-06,
+      "loss": 0.7244,
+      "step": 20296
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 15.837381362915039,
+      "learning_rate": 3.0180312094716146e-06,
+      "loss": 2.0255,
+      "step": 20297
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.487107515335083,
+      "learning_rate": 3.0171944944149274e-06,
+      "loss": 0.2261,
+      "step": 20298
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 11.393579483032227,
+      "learning_rate": 3.0163577793582398e-06,
+      "loss": 0.5304,
+      "step": 20299
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 19.697843551635742,
+      "learning_rate": 3.0155210643015526e-06,
+      "loss": 0.6604,
+      "step": 20300
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 434.9746398925781,
+      "learning_rate": 3.0146843492448645e-06,
+      "loss": 1.8302,
+      "step": 20301
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 29.384225845336914,
+      "learning_rate": 3.0138476341881773e-06,
+      "loss": 1.663,
+      "step": 20302
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 24.34198760986328,
+      "learning_rate": 3.01301091913149e-06,
+      "loss": 1.1187,
+      "step": 20303
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 9.198288917541504,
+      "learning_rate": 3.0121742040748025e-06,
+      "loss": 0.5688,
+      "step": 20304
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 6.318452835083008,
+      "learning_rate": 3.0113374890181153e-06,
+      "loss": 1.6171,
+      "step": 20305
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 32.8675651550293,
+      "learning_rate": 3.0105007739614277e-06,
+      "loss": 2.4725,
+      "step": 20306
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 223.1664276123047,
+      "learning_rate": 3.0096640589047405e-06,
+      "loss": 2.5462,
+      "step": 20307
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 10.688501358032227,
+      "learning_rate": 3.0088273438480524e-06,
+      "loss": 0.4247,
+      "step": 20308
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 23.226240158081055,
+      "learning_rate": 3.0079906287913652e-06,
+      "loss": 0.6099,
+      "step": 20309
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.763749599456787,
+      "learning_rate": 3.0071539137346776e-06,
+      "loss": 0.1512,
+      "step": 20310
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 14.509944915771484,
+      "learning_rate": 3.0063171986779904e-06,
+      "loss": 1.2319,
+      "step": 20311
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 12.403829574584961,
+      "learning_rate": 3.0054804836213032e-06,
+      "loss": 0.5058,
+      "step": 20312
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 25.466928482055664,
+      "learning_rate": 3.0046437685646156e-06,
+      "loss": 1.4126,
+      "step": 20313
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 17.499757766723633,
+      "learning_rate": 3.0038070535079284e-06,
+      "loss": 1.2541,
+      "step": 20314
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 10.804409980773926,
+      "learning_rate": 3.0029703384512403e-06,
+      "loss": 0.4218,
+      "step": 20315
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 13.253870964050293,
+      "learning_rate": 3.002133623394553e-06,
+      "loss": 0.518,
+      "step": 20316
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.310703754425049,
+      "learning_rate": 3.0012969083378655e-06,
+      "loss": 0.187,
+      "step": 20317
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 25.528017044067383,
+      "learning_rate": 3.0004601932811783e-06,
+      "loss": 0.6514,
+      "step": 20318
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 12.111616134643555,
+      "learning_rate": 2.999623478224491e-06,
+      "loss": 0.8347,
+      "step": 20319
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 20.67764663696289,
+      "learning_rate": 2.9987867631678035e-06,
+      "loss": 0.5942,
+      "step": 20320
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.750455379486084,
+      "learning_rate": 2.9979500481111163e-06,
+      "loss": 0.1571,
+      "step": 20321
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 8.119806289672852,
+      "learning_rate": 2.9971133330544283e-06,
+      "loss": 0.1028,
+      "step": 20322
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.8741098046302795,
+      "learning_rate": 2.996276617997741e-06,
+      "loss": 0.0391,
+      "step": 20323
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 15.446863174438477,
+      "learning_rate": 2.9954399029410534e-06,
+      "loss": 1.6925,
+      "step": 20324
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 12.320019721984863,
+      "learning_rate": 2.9946031878843662e-06,
+      "loss": 0.6637,
+      "step": 20325
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 19.610437393188477,
+      "learning_rate": 2.993766472827679e-06,
+      "loss": 1.6855,
+      "step": 20326
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 21.2055606842041,
+      "learning_rate": 2.9929297577709914e-06,
+      "loss": 1.7123,
+      "step": 20327
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 6.5165486335754395,
+      "learning_rate": 2.9920930427143042e-06,
+      "loss": 0.1991,
+      "step": 20328
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 29.296640396118164,
+      "learning_rate": 2.991256327657616e-06,
+      "loss": 0.3101,
+      "step": 20329
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 41.530357360839844,
+      "learning_rate": 2.990419612600929e-06,
+      "loss": 2.4167,
+      "step": 20330
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 6.716545104980469,
+      "learning_rate": 2.9895828975442414e-06,
+      "loss": 0.5641,
+      "step": 20331
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 57.292720794677734,
+      "learning_rate": 2.988746182487554e-06,
+      "loss": 0.7099,
+      "step": 20332
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 57.8223991394043,
+      "learning_rate": 2.9879094674308665e-06,
+      "loss": 2.4506,
+      "step": 20333
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 18.16729736328125,
+      "learning_rate": 2.9870727523741793e-06,
+      "loss": 0.7348,
+      "step": 20334
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 7.2328925132751465,
+      "learning_rate": 2.986236037317492e-06,
+      "loss": 2.323,
+      "step": 20335
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 9.983031272888184,
+      "learning_rate": 2.985399322260804e-06,
+      "loss": 0.6804,
+      "step": 20336
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 165.5161590576172,
+      "learning_rate": 2.984562607204117e-06,
+      "loss": 1.3491,
+      "step": 20337
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 10.617986679077148,
+      "learning_rate": 2.9837258921474293e-06,
+      "loss": 0.2747,
+      "step": 20338
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 11.300695419311523,
+      "learning_rate": 2.982889177090742e-06,
+      "loss": 0.497,
+      "step": 20339
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 9.961305618286133,
+      "learning_rate": 2.9820524620340545e-06,
+      "loss": 0.6737,
+      "step": 20340
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 21.09966278076172,
+      "learning_rate": 2.9812157469773673e-06,
+      "loss": 0.6899,
+      "step": 20341
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 12.62860107421875,
+      "learning_rate": 2.98037903192068e-06,
+      "loss": 0.5219,
+      "step": 20342
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 18.826616287231445,
+      "learning_rate": 2.979542316863992e-06,
+      "loss": 0.3157,
+      "step": 20343
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 26.843746185302734,
+      "learning_rate": 2.978705601807305e-06,
+      "loss": 0.947,
+      "step": 20344
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 12.240778923034668,
+      "learning_rate": 2.977868886750617e-06,
+      "loss": 0.5789,
+      "step": 20345
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 18.644920349121094,
+      "learning_rate": 2.97703217169393e-06,
+      "loss": 1.1404,
+      "step": 20346
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 20.328208923339844,
+      "learning_rate": 2.9761954566372424e-06,
+      "loss": 1.1604,
+      "step": 20347
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 31.434791564941406,
+      "learning_rate": 2.975358741580555e-06,
+      "loss": 1.3904,
+      "step": 20348
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.116194725036621,
+      "learning_rate": 2.974522026523868e-06,
+      "loss": 0.3696,
+      "step": 20349
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 14.46107292175293,
+      "learning_rate": 2.97368531146718e-06,
+      "loss": 2.9684,
+      "step": 20350
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 46.29600143432617,
+      "learning_rate": 2.9728485964104923e-06,
+      "loss": 1.2229,
+      "step": 20351
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 7.156410217285156,
+      "learning_rate": 2.972011881353805e-06,
+      "loss": 0.507,
+      "step": 20352
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 13.746549606323242,
+      "learning_rate": 2.971175166297118e-06,
+      "loss": 1.8664,
+      "step": 20353
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 71.00946807861328,
+      "learning_rate": 2.9703384512404303e-06,
+      "loss": 0.9309,
+      "step": 20354
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.014029502868652,
+      "learning_rate": 2.969501736183743e-06,
+      "loss": 0.4989,
+      "step": 20355
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 11.859457015991211,
+      "learning_rate": 2.9686650211270555e-06,
+      "loss": 0.4248,
+      "step": 20356
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.653107166290283,
+      "learning_rate": 2.967828306070368e-06,
+      "loss": 0.306,
+      "step": 20357
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 20.030885696411133,
+      "learning_rate": 2.9669915910136802e-06,
+      "loss": 1.5919,
+      "step": 20358
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 7.168249607086182,
+      "learning_rate": 2.966154875956993e-06,
+      "loss": 0.2919,
+      "step": 20359
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.518740653991699,
+      "learning_rate": 2.965318160900306e-06,
+      "loss": 1.2664,
+      "step": 20360
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 19.980792999267578,
+      "learning_rate": 2.9644814458436182e-06,
+      "loss": 1.2647,
+      "step": 20361
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 6.87418270111084,
+      "learning_rate": 2.963644730786931e-06,
+      "loss": 0.4924,
+      "step": 20362
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 13.117971420288086,
+      "learning_rate": 2.9628080157302434e-06,
+      "loss": 0.4402,
+      "step": 20363
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 155.3637237548828,
+      "learning_rate": 2.9619713006735558e-06,
+      "loss": 2.9843,
+      "step": 20364
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.862758159637451,
+      "learning_rate": 2.961134585616868e-06,
+      "loss": 0.2754,
+      "step": 20365
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 17.832685470581055,
+      "learning_rate": 2.960297870560181e-06,
+      "loss": 1.018,
+      "step": 20366
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 14.653788566589355,
+      "learning_rate": 2.9594611555034938e-06,
+      "loss": 0.7364,
+      "step": 20367
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 16.753034591674805,
+      "learning_rate": 2.958624440446806e-06,
+      "loss": 1.3807,
+      "step": 20368
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.317260265350342,
+      "learning_rate": 2.957787725390119e-06,
+      "loss": 0.3076,
+      "step": 20369
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 19.307132720947266,
+      "learning_rate": 2.9569510103334313e-06,
+      "loss": 0.7773,
+      "step": 20370
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 6.9656662940979,
+      "learning_rate": 2.9561142952767437e-06,
+      "loss": 0.4826,
+      "step": 20371
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 7.019276142120361,
+      "learning_rate": 2.955277580220056e-06,
+      "loss": 0.3119,
+      "step": 20372
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.526972770690918,
+      "learning_rate": 2.954440865163369e-06,
+      "loss": 0.3393,
+      "step": 20373
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 11.408474922180176,
+      "learning_rate": 2.9536041501066813e-06,
+      "loss": 2.8712,
+      "step": 20374
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 14.502580642700195,
+      "learning_rate": 2.952767435049994e-06,
+      "loss": 0.6733,
+      "step": 20375
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 20.133806228637695,
+      "learning_rate": 2.951930719993307e-06,
+      "loss": 1.2049,
+      "step": 20376
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 25.461997985839844,
+      "learning_rate": 2.9510940049366192e-06,
+      "loss": 0.4696,
+      "step": 20377
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 10.181278228759766,
+      "learning_rate": 2.9502572898799316e-06,
+      "loss": 1.8776,
+      "step": 20378
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 15.376132011413574,
+      "learning_rate": 2.949420574823244e-06,
+      "loss": 1.0615,
+      "step": 20379
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 25.411516189575195,
+      "learning_rate": 2.948583859766557e-06,
+      "loss": 1.2451,
+      "step": 20380
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 107.49919891357422,
+      "learning_rate": 2.947747144709869e-06,
+      "loss": 1.8063,
+      "step": 20381
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 12.071283340454102,
+      "learning_rate": 2.946910429653182e-06,
+      "loss": 0.4915,
+      "step": 20382
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 22.164794921875,
+      "learning_rate": 2.9460737145964948e-06,
+      "loss": 2.6652,
+      "step": 20383
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 26.348634719848633,
+      "learning_rate": 2.9452369995398067e-06,
+      "loss": 1.0793,
+      "step": 20384
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 9.477949142456055,
+      "learning_rate": 2.944400284483119e-06,
+      "loss": 0.4536,
+      "step": 20385
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 29.26583480834961,
+      "learning_rate": 2.943563569426432e-06,
+      "loss": 1.5501,
+      "step": 20386
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 25.907440185546875,
+      "learning_rate": 2.9427268543697447e-06,
+      "loss": 1.4007,
+      "step": 20387
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 30.64686393737793,
+      "learning_rate": 2.941890139313057e-06,
+      "loss": 1.2803,
+      "step": 20388
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 37.722618103027344,
+      "learning_rate": 2.94105342425637e-06,
+      "loss": 0.8912,
+      "step": 20389
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 6.256129264831543,
+      "learning_rate": 2.9402167091996827e-06,
+      "loss": 0.0959,
+      "step": 20390
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 15.23316764831543,
+      "learning_rate": 2.9393799941429947e-06,
+      "loss": 0.642,
+      "step": 20391
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 18.358585357666016,
+      "learning_rate": 2.938543279086307e-06,
+      "loss": 0.6939,
+      "step": 20392
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 20.84736442565918,
+      "learning_rate": 2.93770656402962e-06,
+      "loss": 1.0631,
+      "step": 20393
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.9894471168518066,
+      "learning_rate": 2.9368698489729326e-06,
+      "loss": 0.2495,
+      "step": 20394
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 20.384599685668945,
+      "learning_rate": 2.936033133916245e-06,
+      "loss": 2.1595,
+      "step": 20395
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.945254325866699,
+      "learning_rate": 2.935196418859558e-06,
+      "loss": 0.4665,
+      "step": 20396
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 6.275689601898193,
+      "learning_rate": 2.93435970380287e-06,
+      "loss": 0.6667,
+      "step": 20397
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 12.31562614440918,
+      "learning_rate": 2.9335229887461826e-06,
+      "loss": 0.3503,
+      "step": 20398
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.085023880004883,
+      "learning_rate": 2.932686273689495e-06,
+      "loss": 0.1873,
+      "step": 20399
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 47.73721694946289,
+      "learning_rate": 2.9318495586328077e-06,
+      "loss": 1.5451,
+      "step": 20400
+    },
+    {
+      "epoch": 2.56,
+      "eval_loss": 0.08377895504236221,
+      "eval_runtime": 95.4086,
+      "eval_samples_per_second": 37.125,
+      "eval_steps_per_second": 37.125,
+      "step": 20400
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 11.437736511230469,
+      "learning_rate": 2.9310128435761206e-06,
+      "loss": 0.221,
+      "step": 20401
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.5437731742858887,
+      "learning_rate": 2.930176128519433e-06,
+      "loss": 0.152,
+      "step": 20402
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 106.44786071777344,
+      "learning_rate": 2.9293394134627457e-06,
+      "loss": 0.878,
+      "step": 20403
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 8.05051326751709,
+      "learning_rate": 2.928502698406058e-06,
+      "loss": 1.2269,
+      "step": 20404
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 13.124874114990234,
+      "learning_rate": 2.9276659833493705e-06,
+      "loss": 0.6975,
+      "step": 20405
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 17.221309661865234,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 0.6167,
+      "step": 20406
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 13.0005464553833,
+      "learning_rate": 2.9259925532359957e-06,
+      "loss": 0.4355,
+      "step": 20407
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 16.78082275390625,
+      "learning_rate": 2.9251558381793085e-06,
+      "loss": 0.7117,
+      "step": 20408
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 26.075725555419922,
+      "learning_rate": 2.924319123122621e-06,
+      "loss": 1.4162,
+      "step": 20409
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 25.144702911376953,
+      "learning_rate": 2.9234824080659337e-06,
+      "loss": 1.8183,
+      "step": 20410
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 7.116442680358887,
+      "learning_rate": 2.922645693009246e-06,
+      "loss": 1.6661,
+      "step": 20411
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 16.31544303894043,
+      "learning_rate": 2.9218089779525584e-06,
+      "loss": 1.7183,
+      "step": 20412
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 14.379830360412598,
+      "learning_rate": 2.9209722628958708e-06,
+      "loss": 0.426,
+      "step": 20413
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 6.947141647338867,
+      "learning_rate": 2.9201355478391836e-06,
+      "loss": 0.7284,
+      "step": 20414
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 8.293611526489258,
+      "learning_rate": 2.919298832782496e-06,
+      "loss": 0.9619,
+      "step": 20415
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 8.01059627532959,
+      "learning_rate": 2.9184621177258088e-06,
+      "loss": 0.3516,
+      "step": 20416
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 22.62220573425293,
+      "learning_rate": 2.9176254026691216e-06,
+      "loss": 1.1697,
+      "step": 20417
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 20.267784118652344,
+      "learning_rate": 2.916788687612434e-06,
+      "loss": 1.1011,
+      "step": 20418
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.148131370544434,
+      "learning_rate": 2.9159519725557463e-06,
+      "loss": 0.3629,
+      "step": 20419
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 38.091129302978516,
+      "learning_rate": 2.9151152574990587e-06,
+      "loss": 0.6832,
+      "step": 20420
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 34.675601959228516,
+      "learning_rate": 2.9142785424423715e-06,
+      "loss": 1.2377,
+      "step": 20421
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 30.98348045349121,
+      "learning_rate": 2.913441827385684e-06,
+      "loss": 1.6618,
+      "step": 20422
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 13.842818260192871,
+      "learning_rate": 2.9126051123289967e-06,
+      "loss": 0.7118,
+      "step": 20423
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 9.606385231018066,
+      "learning_rate": 2.9117683972723095e-06,
+      "loss": 0.9371,
+      "step": 20424
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 29.835241317749023,
+      "learning_rate": 2.910931682215622e-06,
+      "loss": 0.8631,
+      "step": 20425
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 23.59840965270996,
+      "learning_rate": 2.910094967158934e-06,
+      "loss": 1.1475,
+      "step": 20426
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 22.209251403808594,
+      "learning_rate": 2.9092582521022466e-06,
+      "loss": 0.854,
+      "step": 20427
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 15.59139347076416,
+      "learning_rate": 2.9084215370455594e-06,
+      "loss": 0.8305,
+      "step": 20428
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 57.197265625,
+      "learning_rate": 2.907584821988872e-06,
+      "loss": 1.7459,
+      "step": 20429
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 35.6335563659668,
+      "learning_rate": 2.9067481069321846e-06,
+      "loss": 2.1248,
+      "step": 20430
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 19.992061614990234,
+      "learning_rate": 2.9059113918754974e-06,
+      "loss": 1.3376,
+      "step": 20431
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 17.135761260986328,
+      "learning_rate": 2.9050746768188098e-06,
+      "loss": 0.8492,
+      "step": 20432
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 14.199894905090332,
+      "learning_rate": 2.9042379617621217e-06,
+      "loss": 0.7577,
+      "step": 20433
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 9.9739351272583,
+      "learning_rate": 2.9034012467054345e-06,
+      "loss": 0.4578,
+      "step": 20434
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 19.584999084472656,
+      "learning_rate": 2.9025645316487473e-06,
+      "loss": 1.0605,
+      "step": 20435
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.4521515369415283,
+      "learning_rate": 2.9017278165920597e-06,
+      "loss": 0.1298,
+      "step": 20436
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.8404103517532349,
+      "learning_rate": 2.9008911015353725e-06,
+      "loss": 0.0307,
+      "step": 20437
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.019158363342285,
+      "learning_rate": 2.900054386478685e-06,
+      "loss": 0.3227,
+      "step": 20438
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 188.9952392578125,
+      "learning_rate": 2.8992176714219977e-06,
+      "loss": 1.2798,
+      "step": 20439
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 9.685225486755371,
+      "learning_rate": 2.8983809563653097e-06,
+      "loss": 2.7265,
+      "step": 20440
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 8.216483116149902,
+      "learning_rate": 2.8975442413086225e-06,
+      "loss": 0.3263,
+      "step": 20441
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 25.34651756286621,
+      "learning_rate": 2.8967075262519353e-06,
+      "loss": 0.7347,
+      "step": 20442
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 15.343186378479004,
+      "learning_rate": 2.8958708111952476e-06,
+      "loss": 0.7736,
+      "step": 20443
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 10.678083419799805,
+      "learning_rate": 2.8950340961385604e-06,
+      "loss": 0.7355,
+      "step": 20444
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 13.670865058898926,
+      "learning_rate": 2.894197381081873e-06,
+      "loss": 1.5327,
+      "step": 20445
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 13.043383598327637,
+      "learning_rate": 2.8933606660251856e-06,
+      "loss": 0.8371,
+      "step": 20446
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 15.65070629119873,
+      "learning_rate": 2.8925239509684976e-06,
+      "loss": 0.7348,
+      "step": 20447
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.623363971710205,
+      "learning_rate": 2.8916872359118104e-06,
+      "loss": 0.5187,
+      "step": 20448
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 13.084575653076172,
+      "learning_rate": 2.8908505208551228e-06,
+      "loss": 1.3617,
+      "step": 20449
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.675920486450195,
+      "learning_rate": 2.8900138057984356e-06,
+      "loss": 0.2809,
+      "step": 20450
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 15.445809364318848,
+      "learning_rate": 2.8891770907417484e-06,
+      "loss": 0.7837,
+      "step": 20451
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 9.706493377685547,
+      "learning_rate": 2.8883403756850607e-06,
+      "loss": 2.1096,
+      "step": 20452
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 12.382505416870117,
+      "learning_rate": 2.8875036606283735e-06,
+      "loss": 0.9842,
+      "step": 20453
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 17.63967514038086,
+      "learning_rate": 2.8866669455716855e-06,
+      "loss": 0.5586,
+      "step": 20454
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 6.769794464111328,
+      "learning_rate": 2.8858302305149983e-06,
+      "loss": 0.3738,
+      "step": 20455
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 156.72103881835938,
+      "learning_rate": 2.8849935154583107e-06,
+      "loss": 2.5701,
+      "step": 20456
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 7.69674015045166,
+      "learning_rate": 2.8841568004016235e-06,
+      "loss": 0.42,
+      "step": 20457
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 6.638584613800049,
+      "learning_rate": 2.8833200853449363e-06,
+      "loss": 0.1432,
+      "step": 20458
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 12.168359756469727,
+      "learning_rate": 2.8824833702882487e-06,
+      "loss": 0.616,
+      "step": 20459
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 11.028854370117188,
+      "learning_rate": 2.8816466552315615e-06,
+      "loss": 0.3877,
+      "step": 20460
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 16.257770538330078,
+      "learning_rate": 2.8808099401748734e-06,
+      "loss": 1.3306,
+      "step": 20461
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 13.83847427368164,
+      "learning_rate": 2.8799732251181862e-06,
+      "loss": 1.0759,
+      "step": 20462
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 7.097414970397949,
+      "learning_rate": 2.8791365100614986e-06,
+      "loss": 0.4918,
+      "step": 20463
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 15.4981107711792,
+      "learning_rate": 2.8782997950048114e-06,
+      "loss": 0.5027,
+      "step": 20464
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 8.238839149475098,
+      "learning_rate": 2.877463079948124e-06,
+      "loss": 0.2851,
+      "step": 20465
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 53.47518539428711,
+      "learning_rate": 2.8766263648914366e-06,
+      "loss": 0.889,
+      "step": 20466
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 8.977937698364258,
+      "learning_rate": 2.8757896498347494e-06,
+      "loss": 0.25,
+      "step": 20467
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 36.48308563232422,
+      "learning_rate": 2.8749529347780613e-06,
+      "loss": 1.6408,
+      "step": 20468
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 17.047109603881836,
+      "learning_rate": 2.874116219721374e-06,
+      "loss": 0.6356,
+      "step": 20469
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 15.693798065185547,
+      "learning_rate": 2.8732795046646865e-06,
+      "loss": 1.3917,
+      "step": 20470
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 71.5500259399414,
+      "learning_rate": 2.8724427896079993e-06,
+      "loss": 0.9439,
+      "step": 20471
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 21.63862419128418,
+      "learning_rate": 2.871606074551312e-06,
+      "loss": 0.925,
+      "step": 20472
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 84.76802825927734,
+      "learning_rate": 2.8707693594946245e-06,
+      "loss": 1.9239,
+      "step": 20473
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 18.43046760559082,
+      "learning_rate": 2.8699326444379364e-06,
+      "loss": 0.9415,
+      "step": 20474
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 31.983963012695312,
+      "learning_rate": 2.8690959293812492e-06,
+      "loss": 1.5056,
+      "step": 20475
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 6.9821014404296875,
+      "learning_rate": 2.868259214324562e-06,
+      "loss": 1.4984,
+      "step": 20476
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 7.553549766540527,
+      "learning_rate": 2.8674224992678744e-06,
+      "loss": 1.2257,
+      "step": 20477
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 13.389091491699219,
+      "learning_rate": 2.8665857842111872e-06,
+      "loss": 1.2428,
+      "step": 20478
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 67.30017852783203,
+      "learning_rate": 2.8657490691544996e-06,
+      "loss": 1.2801,
+      "step": 20479
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 12.074796676635742,
+      "learning_rate": 2.8649123540978124e-06,
+      "loss": 0.6248,
+      "step": 20480
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 8.72959041595459,
+      "learning_rate": 2.8640756390411244e-06,
+      "loss": 0.5387,
+      "step": 20481
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 10.535429000854492,
+      "learning_rate": 2.863238923984437e-06,
+      "loss": 0.5849,
+      "step": 20482
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.176032066345215,
+      "learning_rate": 2.86240220892775e-06,
+      "loss": 0.355,
+      "step": 20483
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.493794918060303,
+      "learning_rate": 2.8615654938710623e-06,
+      "loss": 0.1702,
+      "step": 20484
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 29.761999130249023,
+      "learning_rate": 2.860728778814375e-06,
+      "loss": 1.107,
+      "step": 20485
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 13.900596618652344,
+      "learning_rate": 2.8598920637576875e-06,
+      "loss": 2.0491,
+      "step": 20486
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 15.221857070922852,
+      "learning_rate": 2.8590553487010003e-06,
+      "loss": 0.3947,
+      "step": 20487
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 44.226226806640625,
+      "learning_rate": 2.8582186336443123e-06,
+      "loss": 2.401,
+      "step": 20488
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 16.93692970275879,
+      "learning_rate": 2.857381918587625e-06,
+      "loss": 1.8102,
+      "step": 20489
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 16.462223052978516,
+      "learning_rate": 2.8565452035309375e-06,
+      "loss": 0.8821,
+      "step": 20490
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 16.876440048217773,
+      "learning_rate": 2.8557084884742503e-06,
+      "loss": 0.4465,
+      "step": 20491
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 11.117237091064453,
+      "learning_rate": 2.854871773417563e-06,
+      "loss": 0.4618,
+      "step": 20492
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 22.775096893310547,
+      "learning_rate": 2.8540350583608754e-06,
+      "loss": 1.4381,
+      "step": 20493
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 21.94539451599121,
+      "learning_rate": 2.8531983433041882e-06,
+      "loss": 1.6329,
+      "step": 20494
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 15.146343231201172,
+      "learning_rate": 2.8523616282475e-06,
+      "loss": 0.6614,
+      "step": 20495
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 9.918079376220703,
+      "learning_rate": 2.851524913190813e-06,
+      "loss": 0.4859,
+      "step": 20496
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 18.03789520263672,
+      "learning_rate": 2.8506881981341254e-06,
+      "loss": 2.1463,
+      "step": 20497
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.780584335327148,
+      "learning_rate": 2.849851483077438e-06,
+      "loss": 1.5282,
+      "step": 20498
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 10.86877155303955,
+      "learning_rate": 2.849014768020751e-06,
+      "loss": 0.425,
+      "step": 20499
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 8.41763973236084,
+      "learning_rate": 2.8481780529640634e-06,
+      "loss": 0.9491,
+      "step": 20500
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 11.646855354309082,
+      "learning_rate": 2.847341337907376e-06,
+      "loss": 1.9672,
+      "step": 20501
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 19.35943031311035,
+      "learning_rate": 2.846504622850688e-06,
+      "loss": 1.1775,
+      "step": 20502
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 34.64932632446289,
+      "learning_rate": 2.845667907794001e-06,
+      "loss": 1.2362,
+      "step": 20503
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 10.847381591796875,
+      "learning_rate": 2.8448311927373133e-06,
+      "loss": 1.8057,
+      "step": 20504
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.648735523223877,
+      "learning_rate": 2.843994477680626e-06,
+      "loss": 0.1642,
+      "step": 20505
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 29.921693801879883,
+      "learning_rate": 2.843157762623939e-06,
+      "loss": 1.4823,
+      "step": 20506
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 21.329360961914062,
+      "learning_rate": 2.8423210475672513e-06,
+      "loss": 0.9152,
+      "step": 20507
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 9.264352798461914,
+      "learning_rate": 2.841484332510564e-06,
+      "loss": 0.5901,
+      "step": 20508
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 10.696232795715332,
+      "learning_rate": 2.840647617453876e-06,
+      "loss": 2.0568,
+      "step": 20509
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 104.81800079345703,
+      "learning_rate": 2.839810902397189e-06,
+      "loss": 1.7419,
+      "step": 20510
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 56.361446380615234,
+      "learning_rate": 2.8389741873405012e-06,
+      "loss": 2.2017,
+      "step": 20511
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 21.71269989013672,
+      "learning_rate": 2.838137472283814e-06,
+      "loss": 3.0182,
+      "step": 20512
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 17.92374038696289,
+      "learning_rate": 2.8373007572271264e-06,
+      "loss": 1.7016,
+      "step": 20513
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 7.789238929748535,
+      "learning_rate": 2.836464042170439e-06,
+      "loss": 1.2383,
+      "step": 20514
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 7.504380702972412,
+      "learning_rate": 2.835627327113752e-06,
+      "loss": 0.3392,
+      "step": 20515
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 7.736235618591309,
+      "learning_rate": 2.834790612057064e-06,
+      "loss": 1.0989,
+      "step": 20516
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 9.0857515335083,
+      "learning_rate": 2.8339538970003768e-06,
+      "loss": 1.3186,
+      "step": 20517
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 172.78436279296875,
+      "learning_rate": 2.833117181943689e-06,
+      "loss": 2.1974,
+      "step": 20518
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 25.050010681152344,
+      "learning_rate": 2.832280466887002e-06,
+      "loss": 0.5577,
+      "step": 20519
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.163219451904297,
+      "learning_rate": 2.8314437518303143e-06,
+      "loss": 0.0758,
+      "step": 20520
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 9.777112007141113,
+      "learning_rate": 2.830607036773627e-06,
+      "loss": 0.5397,
+      "step": 20521
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.4373345375061035,
+      "learning_rate": 2.82977032171694e-06,
+      "loss": 0.1589,
+      "step": 20522
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 12.718113899230957,
+      "learning_rate": 2.828933606660252e-06,
+      "loss": 0.3339,
+      "step": 20523
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 17.83120346069336,
+      "learning_rate": 2.8280968916035647e-06,
+      "loss": 1.3957,
+      "step": 20524
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 44.10799789428711,
+      "learning_rate": 2.827260176546877e-06,
+      "loss": 1.1544,
+      "step": 20525
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 35.75393295288086,
+      "learning_rate": 2.82642346149019e-06,
+      "loss": 3.0461,
+      "step": 20526
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 9.007635116577148,
+      "learning_rate": 2.8255867464335022e-06,
+      "loss": 0.9127,
+      "step": 20527
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 10.208179473876953,
+      "learning_rate": 2.824750031376815e-06,
+      "loss": 0.5074,
+      "step": 20528
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 33.458282470703125,
+      "learning_rate": 2.823913316320128e-06,
+      "loss": 1.4324,
+      "step": 20529
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.232382297515869,
+      "learning_rate": 2.82307660126344e-06,
+      "loss": 0.4225,
+      "step": 20530
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 19.59880828857422,
+      "learning_rate": 2.822239886206752e-06,
+      "loss": 1.464,
+      "step": 20531
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 8.07803726196289,
+      "learning_rate": 2.821403171150065e-06,
+      "loss": 1.4669,
+      "step": 20532
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 7.953825950622559,
+      "learning_rate": 2.8205664560933778e-06,
+      "loss": 0.4613,
+      "step": 20533
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 22.497541427612305,
+      "learning_rate": 2.81972974103669e-06,
+      "loss": 1.5066,
+      "step": 20534
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 107.39096069335938,
+      "learning_rate": 2.818893025980003e-06,
+      "loss": 2.8367,
+      "step": 20535
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 18.257076263427734,
+      "learning_rate": 2.8180563109233158e-06,
+      "loss": 1.0713,
+      "step": 20536
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 27.29205322265625,
+      "learning_rate": 2.8172195958666277e-06,
+      "loss": 1.4309,
+      "step": 20537
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 11.839692115783691,
+      "learning_rate": 2.81638288080994e-06,
+      "loss": 0.8319,
+      "step": 20538
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 8.173250198364258,
+      "learning_rate": 2.815546165753253e-06,
+      "loss": 0.2603,
+      "step": 20539
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 10.04826831817627,
+      "learning_rate": 2.8147094506965657e-06,
+      "loss": 0.5927,
+      "step": 20540
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 9.568615913391113,
+      "learning_rate": 2.813872735639878e-06,
+      "loss": 0.4649,
+      "step": 20541
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 2.9093575477600098,
+      "learning_rate": 2.813036020583191e-06,
+      "loss": 0.1411,
+      "step": 20542
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 514.5388793945312,
+      "learning_rate": 2.8121993055265033e-06,
+      "loss": 1.5976,
+      "step": 20543
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.592569351196289,
+      "learning_rate": 2.8113625904698156e-06,
+      "loss": 0.1667,
+      "step": 20544
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 14.131592750549316,
+      "learning_rate": 2.810525875413128e-06,
+      "loss": 0.8904,
+      "step": 20545
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 11.625960350036621,
+      "learning_rate": 2.809689160356441e-06,
+      "loss": 0.429,
+      "step": 20546
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 20.848854064941406,
+      "learning_rate": 2.8088524452997536e-06,
+      "loss": 1.1286,
+      "step": 20547
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 70.20594024658203,
+      "learning_rate": 2.808015730243066e-06,
+      "loss": 2.5154,
+      "step": 20548
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.269267559051514,
+      "learning_rate": 2.807179015186379e-06,
+      "loss": 0.3758,
+      "step": 20549
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 6.539968013763428,
+      "learning_rate": 2.806342300129691e-06,
+      "loss": 0.0825,
+      "step": 20550
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 17.589025497436523,
+      "learning_rate": 2.8055055850730036e-06,
+      "loss": 1.48,
+      "step": 20551
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 19.457406997680664,
+      "learning_rate": 2.804668870016316e-06,
+      "loss": 0.7478,
+      "step": 20552
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 22.284839630126953,
+      "learning_rate": 2.8038321549596287e-06,
+      "loss": 1.1945,
+      "step": 20553
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 13.882135391235352,
+      "learning_rate": 2.802995439902941e-06,
+      "loss": 0.636,
+      "step": 20554
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 60.36022186279297,
+      "learning_rate": 2.802158724846254e-06,
+      "loss": 1.6826,
+      "step": 20555
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 24.295879364013672,
+      "learning_rate": 2.8013220097895667e-06,
+      "loss": 0.9273,
+      "step": 20556
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.3162782192230225,
+      "learning_rate": 2.800485294732879e-06,
+      "loss": 0.1991,
+      "step": 20557
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.362887859344482,
+      "learning_rate": 2.7996485796761915e-06,
+      "loss": 0.3085,
+      "step": 20558
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 9.203067779541016,
+      "learning_rate": 2.798811864619504e-06,
+      "loss": 0.2045,
+      "step": 20559
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.6134934425354,
+      "learning_rate": 2.7979751495628167e-06,
+      "loss": 0.432,
+      "step": 20560
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 10.109713554382324,
+      "learning_rate": 2.797138434506129e-06,
+      "loss": 2.1859,
+      "step": 20561
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 13.00163459777832,
+      "learning_rate": 2.796301719449442e-06,
+      "loss": 0.5977,
+      "step": 20562
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 8.053282737731934,
+      "learning_rate": 2.7954650043927546e-06,
+      "loss": 1.171,
+      "step": 20563
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 6.237462997436523,
+      "learning_rate": 2.7946282893360666e-06,
+      "loss": 0.3323,
+      "step": 20564
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 15.753913879394531,
+      "learning_rate": 2.7937915742793794e-06,
+      "loss": 0.8277,
+      "step": 20565
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.748144149780273,
+      "learning_rate": 2.7929548592226918e-06,
+      "loss": 0.2545,
+      "step": 20566
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 7.599785327911377,
+      "learning_rate": 2.7921181441660046e-06,
+      "loss": 0.7825,
+      "step": 20567
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 17.133930206298828,
+      "learning_rate": 2.791281429109317e-06,
+      "loss": 1.7605,
+      "step": 20568
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 18.69016456604004,
+      "learning_rate": 2.7904447140526297e-06,
+      "loss": 1.2244,
+      "step": 20569
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 10.473901748657227,
+      "learning_rate": 2.7896079989959426e-06,
+      "loss": 0.9986,
+      "step": 20570
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.099677562713623,
+      "learning_rate": 2.7887712839392545e-06,
+      "loss": 0.3759,
+      "step": 20571
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 30.630308151245117,
+      "learning_rate": 2.787934568882567e-06,
+      "loss": 1.5521,
+      "step": 20572
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 28.276050567626953,
+      "learning_rate": 2.7870978538258797e-06,
+      "loss": 0.907,
+      "step": 20573
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 7.844485282897949,
+      "learning_rate": 2.7862611387691925e-06,
+      "loss": 1.8047,
+      "step": 20574
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 17.103439331054688,
+      "learning_rate": 2.785424423712505e-06,
+      "loss": 0.5604,
+      "step": 20575
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 11.931612968444824,
+      "learning_rate": 2.7845877086558177e-06,
+      "loss": 0.6014,
+      "step": 20576
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 14.460848808288574,
+      "learning_rate": 2.7837509935991305e-06,
+      "loss": 0.6772,
+      "step": 20577
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 7.24898624420166,
+      "learning_rate": 2.7829142785424424e-06,
+      "loss": 0.5393,
+      "step": 20578
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 8.05289363861084,
+      "learning_rate": 2.782077563485755e-06,
+      "loss": 0.3949,
+      "step": 20579
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 10.37386417388916,
+      "learning_rate": 2.7812408484290676e-06,
+      "loss": 0.6299,
+      "step": 20580
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 29.61414909362793,
+      "learning_rate": 2.7804041333723804e-06,
+      "loss": 1.3309,
+      "step": 20581
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 14.088113784790039,
+      "learning_rate": 2.7795674183156928e-06,
+      "loss": 0.4617,
+      "step": 20582
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 7.285881519317627,
+      "learning_rate": 2.7787307032590056e-06,
+      "loss": 0.5989,
+      "step": 20583
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 12.213241577148438,
+      "learning_rate": 2.777893988202318e-06,
+      "loss": 0.6455,
+      "step": 20584
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 8.219286918640137,
+      "learning_rate": 2.7770572731456303e-06,
+      "loss": 0.2536,
+      "step": 20585
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 139.68544006347656,
+      "learning_rate": 2.7762205580889427e-06,
+      "loss": 0.7538,
+      "step": 20586
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 16.023963928222656,
+      "learning_rate": 2.7753838430322555e-06,
+      "loss": 1.807,
+      "step": 20587
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.934849262237549,
+      "learning_rate": 2.7745471279755683e-06,
+      "loss": 0.2923,
+      "step": 20588
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 25.14888572692871,
+      "learning_rate": 2.7737104129188807e-06,
+      "loss": 1.7516,
+      "step": 20589
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 11.77205753326416,
+      "learning_rate": 2.7728736978621935e-06,
+      "loss": 0.9633,
+      "step": 20590
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 10.894165992736816,
+      "learning_rate": 2.772036982805506e-06,
+      "loss": 0.6362,
+      "step": 20591
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 16.57927894592285,
+      "learning_rate": 2.7712002677488183e-06,
+      "loss": 0.4187,
+      "step": 20592
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 15.53189754486084,
+      "learning_rate": 2.7703635526921306e-06,
+      "loss": 0.5672,
+      "step": 20593
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 18.74542808532715,
+      "learning_rate": 2.7695268376354434e-06,
+      "loss": 0.7687,
+      "step": 20594
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 29.633365631103516,
+      "learning_rate": 2.768690122578756e-06,
+      "loss": 0.8001,
+      "step": 20595
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 10.552553176879883,
+      "learning_rate": 2.7678534075220686e-06,
+      "loss": 1.1081,
+      "step": 20596
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 9.91150188446045,
+      "learning_rate": 2.7670166924653814e-06,
+      "loss": 0.1865,
+      "step": 20597
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 45.85779571533203,
+      "learning_rate": 2.766179977408694e-06,
+      "loss": 3.4106,
+      "step": 20598
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 20.061241149902344,
+      "learning_rate": 2.765343262352006e-06,
+      "loss": 0.3929,
+      "step": 20599
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 26.48193359375,
+      "learning_rate": 2.7645065472953186e-06,
+      "loss": 0.6826,
+      "step": 20600
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 10.189993858337402,
+      "learning_rate": 2.7636698322386314e-06,
+      "loss": 0.6115,
+      "step": 20601
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 14.606910705566406,
+      "learning_rate": 2.7628331171819437e-06,
+      "loss": 0.6867,
+      "step": 20602
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.479612112045288,
+      "learning_rate": 2.7619964021252565e-06,
+      "loss": 0.097,
+      "step": 20603
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 12.336150169372559,
+      "learning_rate": 2.7611596870685693e-06,
+      "loss": 1.465,
+      "step": 20604
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 6.281243801116943,
+      "learning_rate": 2.7603229720118817e-06,
+      "loss": 0.2678,
+      "step": 20605
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 12.324202537536621,
+      "learning_rate": 2.759486256955194e-06,
+      "loss": 1.1456,
+      "step": 20606
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 43.20732498168945,
+      "learning_rate": 2.7586495418985065e-06,
+      "loss": 2.068,
+      "step": 20607
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 15.503894805908203,
+      "learning_rate": 2.7578128268418193e-06,
+      "loss": 1.0782,
+      "step": 20608
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 106.66864013671875,
+      "learning_rate": 2.7569761117851317e-06,
+      "loss": 0.611,
+      "step": 20609
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 91.14187622070312,
+      "learning_rate": 2.7561393967284445e-06,
+      "loss": 2.117,
+      "step": 20610
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 14.441734313964844,
+      "learning_rate": 2.7553026816717573e-06,
+      "loss": 0.9908,
+      "step": 20611
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.973798990249634,
+      "learning_rate": 2.7544659666150696e-06,
+      "loss": 0.1908,
+      "step": 20612
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 15.52254581451416,
+      "learning_rate": 2.7536292515583816e-06,
+      "loss": 0.8581,
+      "step": 20613
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 14.244958877563477,
+      "learning_rate": 2.7527925365016944e-06,
+      "loss": 1.3271,
+      "step": 20614
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 14.75009822845459,
+      "learning_rate": 2.751955821445007e-06,
+      "loss": 1.0654,
+      "step": 20615
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.933854103088379,
+      "learning_rate": 2.7511191063883196e-06,
+      "loss": 0.3563,
+      "step": 20616
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 20.3519229888916,
+      "learning_rate": 2.7502823913316324e-06,
+      "loss": 1.71,
+      "step": 20617
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 15.010092735290527,
+      "learning_rate": 2.7494456762749448e-06,
+      "loss": 0.6177,
+      "step": 20618
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 6.581335544586182,
+      "learning_rate": 2.7486089612182576e-06,
+      "loss": 0.4071,
+      "step": 20619
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 36.17319869995117,
+      "learning_rate": 2.7477722461615695e-06,
+      "loss": 1.2728,
+      "step": 20620
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 28.553016662597656,
+      "learning_rate": 2.7469355311048823e-06,
+      "loss": 1.0801,
+      "step": 20621
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 7.8814311027526855,
+      "learning_rate": 2.746098816048195e-06,
+      "loss": 0.2771,
+      "step": 20622
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 31.427291870117188,
+      "learning_rate": 2.7452621009915075e-06,
+      "loss": 1.36,
+      "step": 20623
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 17.095361709594727,
+      "learning_rate": 2.7444253859348203e-06,
+      "loss": 1.497,
+      "step": 20624
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.985144138336182,
+      "learning_rate": 2.7435886708781327e-06,
+      "loss": 0.1673,
+      "step": 20625
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 20.383758544921875,
+      "learning_rate": 2.7427519558214455e-06,
+      "loss": 1.276,
+      "step": 20626
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 15.193656921386719,
+      "learning_rate": 2.7419152407647574e-06,
+      "loss": 1.4189,
+      "step": 20627
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.478861331939697,
+      "learning_rate": 2.7410785257080702e-06,
+      "loss": 0.1422,
+      "step": 20628
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.593821048736572,
+      "learning_rate": 2.740241810651383e-06,
+      "loss": 0.1697,
+      "step": 20629
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 17.445035934448242,
+      "learning_rate": 2.7394050955946954e-06,
+      "loss": 0.7018,
+      "step": 20630
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 10.315011978149414,
+      "learning_rate": 2.7385683805380082e-06,
+      "loss": 0.9595,
+      "step": 20631
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 7.475622653961182,
+      "learning_rate": 2.7377316654813206e-06,
+      "loss": 2.1201,
+      "step": 20632
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 15.363032341003418,
+      "learning_rate": 2.7368949504246334e-06,
+      "loss": 2.8288,
+      "step": 20633
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 10.100433349609375,
+      "learning_rate": 2.7360582353679453e-06,
+      "loss": 1.7027,
+      "step": 20634
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 7.44243049621582,
+      "learning_rate": 2.735221520311258e-06,
+      "loss": 0.7286,
+      "step": 20635
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 13.29871940612793,
+      "learning_rate": 2.7343848052545705e-06,
+      "loss": 1.0735,
+      "step": 20636
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 39.645599365234375,
+      "learning_rate": 2.7335480901978833e-06,
+      "loss": 0.4428,
+      "step": 20637
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 10.973965644836426,
+      "learning_rate": 2.732711375141196e-06,
+      "loss": 1.4176,
+      "step": 20638
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 9.956299781799316,
+      "learning_rate": 2.7318746600845085e-06,
+      "loss": 1.6256,
+      "step": 20639
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 10.297553062438965,
+      "learning_rate": 2.7310379450278213e-06,
+      "loss": 0.2793,
+      "step": 20640
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.664786338806152,
+      "learning_rate": 2.7302012299711333e-06,
+      "loss": 0.4077,
+      "step": 20641
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 19.875402450561523,
+      "learning_rate": 2.729364514914446e-06,
+      "loss": 1.0372,
+      "step": 20642
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 11.465612411499023,
+      "learning_rate": 2.7285277998577584e-06,
+      "loss": 0.5423,
+      "step": 20643
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 17.82269859313965,
+      "learning_rate": 2.7276910848010712e-06,
+      "loss": 0.5204,
+      "step": 20644
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 26.352035522460938,
+      "learning_rate": 2.726854369744384e-06,
+      "loss": 1.8282,
+      "step": 20645
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 31.096120834350586,
+      "learning_rate": 2.7260176546876964e-06,
+      "loss": 1.3168,
+      "step": 20646
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 12.857887268066406,
+      "learning_rate": 2.7251809396310092e-06,
+      "loss": 0.4651,
+      "step": 20647
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 15.641700744628906,
+      "learning_rate": 2.724344224574321e-06,
+      "loss": 2.7795,
+      "step": 20648
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.491549015045166,
+      "learning_rate": 2.723507509517634e-06,
+      "loss": 0.2071,
+      "step": 20649
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.6830925941467285,
+      "learning_rate": 2.7226707944609464e-06,
+      "loss": 0.5324,
+      "step": 20650
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 11.230381965637207,
+      "learning_rate": 2.721834079404259e-06,
+      "loss": 0.2106,
+      "step": 20651
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 7.093985080718994,
+      "learning_rate": 2.720997364347572e-06,
+      "loss": 0.7449,
+      "step": 20652
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 14.01636791229248,
+      "learning_rate": 2.7201606492908843e-06,
+      "loss": 0.2445,
+      "step": 20653
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 8.7313871383667,
+      "learning_rate": 2.7193239342341963e-06,
+      "loss": 1.6454,
+      "step": 20654
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 19.767362594604492,
+      "learning_rate": 2.718487219177509e-06,
+      "loss": 0.7345,
+      "step": 20655
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 92.36847686767578,
+      "learning_rate": 2.717650504120822e-06,
+      "loss": 3.7043,
+      "step": 20656
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 71.1680908203125,
+      "learning_rate": 2.7168137890641343e-06,
+      "loss": 1.4763,
+      "step": 20657
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.164688587188721,
+      "learning_rate": 2.715977074007447e-06,
+      "loss": 0.3721,
+      "step": 20658
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 31.245023727416992,
+      "learning_rate": 2.7151403589507595e-06,
+      "loss": 1.552,
+      "step": 20659
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 21.062349319458008,
+      "learning_rate": 2.7143036438940723e-06,
+      "loss": 0.4429,
+      "step": 20660
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 2.783233404159546,
+      "learning_rate": 2.7134669288373842e-06,
+      "loss": 0.1205,
+      "step": 20661
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 10.794570922851562,
+      "learning_rate": 2.712630213780697e-06,
+      "loss": 1.1679,
+      "step": 20662
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 9.779132843017578,
+      "learning_rate": 2.71179349872401e-06,
+      "loss": 0.6626,
+      "step": 20663
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 17.937089920043945,
+      "learning_rate": 2.710956783667322e-06,
+      "loss": 0.7875,
+      "step": 20664
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 7.253830909729004,
+      "learning_rate": 2.710120068610635e-06,
+      "loss": 0.4595,
+      "step": 20665
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 26.90309715270996,
+      "learning_rate": 2.7092833535539474e-06,
+      "loss": 1.2822,
+      "step": 20666
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 14.526180267333984,
+      "learning_rate": 2.70844663849726e-06,
+      "loss": 0.3683,
+      "step": 20667
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 28.100963592529297,
+      "learning_rate": 2.707609923440572e-06,
+      "loss": 1.9676,
+      "step": 20668
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 86.18717956542969,
+      "learning_rate": 2.706773208383885e-06,
+      "loss": 1.5185,
+      "step": 20669
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 45.607810974121094,
+      "learning_rate": 2.7059364933271977e-06,
+      "loss": 3.1953,
+      "step": 20670
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.311668872833252,
+      "learning_rate": 2.70509977827051e-06,
+      "loss": 0.4793,
+      "step": 20671
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.7336161136627197,
+      "learning_rate": 2.704263063213823e-06,
+      "loss": 0.1693,
+      "step": 20672
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 26.051963806152344,
+      "learning_rate": 2.7034263481571353e-06,
+      "loss": 1.4203,
+      "step": 20673
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 13.890101432800293,
+      "learning_rate": 2.702589633100448e-06,
+      "loss": 0.6348,
+      "step": 20674
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 14.046887397766113,
+      "learning_rate": 2.70175291804376e-06,
+      "loss": 0.6411,
+      "step": 20675
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 66.71613311767578,
+      "learning_rate": 2.700916202987073e-06,
+      "loss": 1.0706,
+      "step": 20676
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 9.172269821166992,
+      "learning_rate": 2.7000794879303852e-06,
+      "loss": 0.3427,
+      "step": 20677
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 58.57754898071289,
+      "learning_rate": 2.699242772873698e-06,
+      "loss": 1.3343,
+      "step": 20678
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 22.161231994628906,
+      "learning_rate": 2.698406057817011e-06,
+      "loss": 0.9301,
+      "step": 20679
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 20.561626434326172,
+      "learning_rate": 2.6975693427603232e-06,
+      "loss": 1.6658,
+      "step": 20680
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 14.982219696044922,
+      "learning_rate": 2.696732627703636e-06,
+      "loss": 1.2091,
+      "step": 20681
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 16.142711639404297,
+      "learning_rate": 2.695895912646948e-06,
+      "loss": 0.5801,
+      "step": 20682
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 11.208925247192383,
+      "learning_rate": 2.6950591975902608e-06,
+      "loss": 0.8101,
+      "step": 20683
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 23.70677375793457,
+      "learning_rate": 2.694222482533573e-06,
+      "loss": 2.8049,
+      "step": 20684
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 59.326507568359375,
+      "learning_rate": 2.693385767476886e-06,
+      "loss": 1.0946,
+      "step": 20685
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 50.2004280090332,
+      "learning_rate": 2.6925490524201988e-06,
+      "loss": 1.8323,
+      "step": 20686
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 19.800365447998047,
+      "learning_rate": 2.691712337363511e-06,
+      "loss": 1.8256,
+      "step": 20687
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 743.9962158203125,
+      "learning_rate": 2.690875622306824e-06,
+      "loss": 1.9372,
+      "step": 20688
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 20.274946212768555,
+      "learning_rate": 2.690038907250136e-06,
+      "loss": 1.0267,
+      "step": 20689
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 18.31287384033203,
+      "learning_rate": 2.6892021921934487e-06,
+      "loss": 0.5088,
+      "step": 20690
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 49.05681610107422,
+      "learning_rate": 2.688365477136761e-06,
+      "loss": 1.7442,
+      "step": 20691
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 10.268267631530762,
+      "learning_rate": 2.687528762080074e-06,
+      "loss": 0.5306,
+      "step": 20692
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 26.432613372802734,
+      "learning_rate": 2.6866920470233867e-06,
+      "loss": 0.9283,
+      "step": 20693
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 2.937655210494995,
+      "learning_rate": 2.685855331966699e-06,
+      "loss": 0.2135,
+      "step": 20694
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 8.145177841186523,
+      "learning_rate": 2.685018616910012e-06,
+      "loss": 0.2753,
+      "step": 20695
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 13.875649452209473,
+      "learning_rate": 2.684181901853324e-06,
+      "loss": 0.5858,
+      "step": 20696
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 41.01677703857422,
+      "learning_rate": 2.6833451867966366e-06,
+      "loss": 1.9673,
+      "step": 20697
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 13.268399238586426,
+      "learning_rate": 2.682508471739949e-06,
+      "loss": 1.0622,
+      "step": 20698
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 8.663495063781738,
+      "learning_rate": 2.681671756683262e-06,
+      "loss": 0.5457,
+      "step": 20699
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 14.659046173095703,
+      "learning_rate": 2.680835041626574e-06,
+      "loss": 0.5068,
+      "step": 20700
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 16.38235092163086,
+      "learning_rate": 2.679998326569887e-06,
+      "loss": 0.3942,
+      "step": 20701
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 68.92447662353516,
+      "learning_rate": 2.6791616115131998e-06,
+      "loss": 1.9818,
+      "step": 20702
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 2.925363779067993,
+      "learning_rate": 2.6783248964565117e-06,
+      "loss": 0.204,
+      "step": 20703
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 222.7730712890625,
+      "learning_rate": 2.6774881813998245e-06,
+      "loss": 1.5032,
+      "step": 20704
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 12.716299057006836,
+      "learning_rate": 2.676651466343137e-06,
+      "loss": 1.4752,
+      "step": 20705
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 9.49696159362793,
+      "learning_rate": 2.6758147512864497e-06,
+      "loss": 0.743,
+      "step": 20706
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 11.136838912963867,
+      "learning_rate": 2.674978036229762e-06,
+      "loss": 0.2524,
+      "step": 20707
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 20.2912540435791,
+      "learning_rate": 2.674141321173075e-06,
+      "loss": 0.4454,
+      "step": 20708
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 14.758130073547363,
+      "learning_rate": 2.6733046061163877e-06,
+      "loss": 1.7816,
+      "step": 20709
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 16.230806350708008,
+      "learning_rate": 2.6724678910596997e-06,
+      "loss": 1.0516,
+      "step": 20710
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.329649448394775,
+      "learning_rate": 2.6716311760030125e-06,
+      "loss": 0.1994,
+      "step": 20711
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 74.24472045898438,
+      "learning_rate": 2.670794460946325e-06,
+      "loss": 1.2867,
+      "step": 20712
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.160773515701294,
+      "learning_rate": 2.6699577458896376e-06,
+      "loss": 0.1505,
+      "step": 20713
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 52.667510986328125,
+      "learning_rate": 2.66912103083295e-06,
+      "loss": 3.4129,
+      "step": 20714
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 8.15816879272461,
+      "learning_rate": 2.668284315776263e-06,
+      "loss": 0.5077,
+      "step": 20715
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 18.959095001220703,
+      "learning_rate": 2.6674476007195756e-06,
+      "loss": 1.0521,
+      "step": 20716
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 14.158047676086426,
+      "learning_rate": 2.6666108856628876e-06,
+      "loss": 2.0452,
+      "step": 20717
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 10.375843048095703,
+      "learning_rate": 2.6657741706062e-06,
+      "loss": 0.461,
+      "step": 20718
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 14.602218627929688,
+      "learning_rate": 2.6649374555495127e-06,
+      "loss": 0.7571,
+      "step": 20719
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 12.750484466552734,
+      "learning_rate": 2.6641007404928256e-06,
+      "loss": 2.059,
+      "step": 20720
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 6.2839674949646,
+      "learning_rate": 2.663264025436138e-06,
+      "loss": 1.0687,
+      "step": 20721
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 7.21674919128418,
+      "learning_rate": 2.6624273103794507e-06,
+      "loss": 0.4122,
+      "step": 20722
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 12.052151679992676,
+      "learning_rate": 2.661590595322763e-06,
+      "loss": 0.6591,
+      "step": 20723
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 12.120726585388184,
+      "learning_rate": 2.6607538802660755e-06,
+      "loss": 0.1971,
+      "step": 20724
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 39.09234619140625,
+      "learning_rate": 2.659917165209388e-06,
+      "loss": 1.1153,
+      "step": 20725
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 17.80829429626465,
+      "learning_rate": 2.6590804501527007e-06,
+      "loss": 3.4628,
+      "step": 20726
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 27.43482208251953,
+      "learning_rate": 2.6582437350960135e-06,
+      "loss": 0.8496,
+      "step": 20727
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 24.271743774414062,
+      "learning_rate": 2.657407020039326e-06,
+      "loss": 1.3373,
+      "step": 20728
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 28.531177520751953,
+      "learning_rate": 2.6565703049826386e-06,
+      "loss": 1.9993,
+      "step": 20729
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 29.271854400634766,
+      "learning_rate": 2.655733589925951e-06,
+      "loss": 1.1161,
+      "step": 20730
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 9.77495288848877,
+      "learning_rate": 2.6548968748692634e-06,
+      "loss": 0.5064,
+      "step": 20731
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 8.433061599731445,
+      "learning_rate": 2.6540601598125758e-06,
+      "loss": 0.5588,
+      "step": 20732
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 16.231603622436523,
+      "learning_rate": 2.6532234447558886e-06,
+      "loss": 1.2543,
+      "step": 20733
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 16.82666778564453,
+      "learning_rate": 2.6523867296992014e-06,
+      "loss": 0.6834,
+      "step": 20734
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 34.19341278076172,
+      "learning_rate": 2.6515500146425138e-06,
+      "loss": 1.0255,
+      "step": 20735
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 29.689189910888672,
+      "learning_rate": 2.6507132995858266e-06,
+      "loss": 0.9617,
+      "step": 20736
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 17.329444885253906,
+      "learning_rate": 2.649876584529139e-06,
+      "loss": 0.4705,
+      "step": 20737
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 10.019248008728027,
+      "learning_rate": 2.6490398694724513e-06,
+      "loss": 2.0945,
+      "step": 20738
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 8.300518035888672,
+      "learning_rate": 2.6482031544157637e-06,
+      "loss": 1.34,
+      "step": 20739
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 15.227872848510742,
+      "learning_rate": 2.6473664393590765e-06,
+      "loss": 1.0722,
+      "step": 20740
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 22.152156829833984,
+      "learning_rate": 2.646529724302389e-06,
+      "loss": 0.3817,
+      "step": 20741
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 30.404407501220703,
+      "learning_rate": 2.6456930092457017e-06,
+      "loss": 0.8752,
+      "step": 20742
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 11.703365325927734,
+      "learning_rate": 2.6448562941890145e-06,
+      "loss": 1.8923,
+      "step": 20743
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 18.769855499267578,
+      "learning_rate": 2.6440195791323264e-06,
+      "loss": 0.882,
+      "step": 20744
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 8.443937301635742,
+      "learning_rate": 2.6431828640756392e-06,
+      "loss": 0.8397,
+      "step": 20745
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 15.188029289245605,
+      "learning_rate": 2.6423461490189516e-06,
+      "loss": 1.2173,
+      "step": 20746
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 17.374069213867188,
+      "learning_rate": 2.6415094339622644e-06,
+      "loss": 0.918,
+      "step": 20747
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 7.522876262664795,
+      "learning_rate": 2.640672718905577e-06,
+      "loss": 0.703,
+      "step": 20748
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 16.831457138061523,
+      "learning_rate": 2.6398360038488896e-06,
+      "loss": 1.0171,
+      "step": 20749
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 19.82137107849121,
+      "learning_rate": 2.6389992887922024e-06,
+      "loss": 0.7962,
+      "step": 20750
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 2.5512492656707764,
+      "learning_rate": 2.6381625737355144e-06,
+      "loss": 0.0613,
+      "step": 20751
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 17.773103713989258,
+      "learning_rate": 2.637325858678827e-06,
+      "loss": 0.988,
+      "step": 20752
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 20.804609298706055,
+      "learning_rate": 2.6364891436221395e-06,
+      "loss": 1.9218,
+      "step": 20753
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 41.56431198120117,
+      "learning_rate": 2.6356524285654523e-06,
+      "loss": 0.4315,
+      "step": 20754
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 22.21440887451172,
+      "learning_rate": 2.6348157135087647e-06,
+      "loss": 1.3783,
+      "step": 20755
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 7.141530990600586,
+      "learning_rate": 2.6339789984520775e-06,
+      "loss": 0.5846,
+      "step": 20756
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.123556613922119,
+      "learning_rate": 2.6331422833953903e-06,
+      "loss": 0.4207,
+      "step": 20757
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 18.96649169921875,
+      "learning_rate": 2.6323055683387023e-06,
+      "loss": 0.8664,
+      "step": 20758
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 11.451469421386719,
+      "learning_rate": 2.6314688532820147e-06,
+      "loss": 0.5508,
+      "step": 20759
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.617785930633545,
+      "learning_rate": 2.6306321382253275e-06,
+      "loss": 0.0696,
+      "step": 20760
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 2.501220226287842,
+      "learning_rate": 2.6297954231686403e-06,
+      "loss": 0.0922,
+      "step": 20761
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 8.61351490020752,
+      "learning_rate": 2.6289587081119526e-06,
+      "loss": 1.9235,
+      "step": 20762
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 33.558815002441406,
+      "learning_rate": 2.6281219930552654e-06,
+      "loss": 0.4547,
+      "step": 20763
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 58.36433029174805,
+      "learning_rate": 2.627285277998578e-06,
+      "loss": 0.8337,
+      "step": 20764
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 41.084014892578125,
+      "learning_rate": 2.62644856294189e-06,
+      "loss": 1.5015,
+      "step": 20765
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.41900110244751,
+      "learning_rate": 2.6256118478852026e-06,
+      "loss": 0.1704,
+      "step": 20766
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 14.344040870666504,
+      "learning_rate": 2.6247751328285154e-06,
+      "loss": 0.9146,
+      "step": 20767
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 11.356485366821289,
+      "learning_rate": 2.623938417771828e-06,
+      "loss": 0.6074,
+      "step": 20768
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 16.03408432006836,
+      "learning_rate": 2.6231017027151406e-06,
+      "loss": 1.0224,
+      "step": 20769
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 10.44325065612793,
+      "learning_rate": 2.6222649876584534e-06,
+      "loss": 1.5056,
+      "step": 20770
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 10.945624351501465,
+      "learning_rate": 2.6214282726017657e-06,
+      "loss": 0.9261,
+      "step": 20771
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 37.368927001953125,
+      "learning_rate": 2.620591557545078e-06,
+      "loss": 2.5891,
+      "step": 20772
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 21.220455169677734,
+      "learning_rate": 2.6197548424883905e-06,
+      "loss": 0.7615,
+      "step": 20773
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 16.384571075439453,
+      "learning_rate": 2.6189181274317033e-06,
+      "loss": 1.7023,
+      "step": 20774
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 20.78117561340332,
+      "learning_rate": 2.618081412375016e-06,
+      "loss": 1.7378,
+      "step": 20775
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 21.69114112854004,
+      "learning_rate": 2.6172446973183285e-06,
+      "loss": 0.93,
+      "step": 20776
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 17.99045753479004,
+      "learning_rate": 2.6164079822616413e-06,
+      "loss": 0.5936,
+      "step": 20777
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 53.71336364746094,
+      "learning_rate": 2.6155712672049537e-06,
+      "loss": 2.641,
+      "step": 20778
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 40.88047409057617,
+      "learning_rate": 2.614734552148266e-06,
+      "loss": 0.7631,
+      "step": 20779
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.7886650562286377,
+      "learning_rate": 2.6138978370915784e-06,
+      "loss": 0.3358,
+      "step": 20780
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 104.24272155761719,
+      "learning_rate": 2.6130611220348912e-06,
+      "loss": 1.4389,
+      "step": 20781
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 10.23959732055664,
+      "learning_rate": 2.6122244069782036e-06,
+      "loss": 1.752,
+      "step": 20782
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 18.32854461669922,
+      "learning_rate": 2.6113876919215164e-06,
+      "loss": 1.1634,
+      "step": 20783
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 12.131913185119629,
+      "learning_rate": 2.610550976864829e-06,
+      "loss": 0.943,
+      "step": 20784
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.605217456817627,
+      "learning_rate": 2.6097142618081416e-06,
+      "loss": 0.4482,
+      "step": 20785
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 17.47404670715332,
+      "learning_rate": 2.608877546751454e-06,
+      "loss": 1.6336,
+      "step": 20786
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 11.874710083007812,
+      "learning_rate": 2.6080408316947663e-06,
+      "loss": 0.8419,
+      "step": 20787
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 38.75798797607422,
+      "learning_rate": 2.607204116638079e-06,
+      "loss": 2.9181,
+      "step": 20788
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 37.21906661987305,
+      "learning_rate": 2.6063674015813915e-06,
+      "loss": 0.9856,
+      "step": 20789
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 12.168079376220703,
+      "learning_rate": 2.6055306865247043e-06,
+      "loss": 0.9537,
+      "step": 20790
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.899741172790527,
+      "learning_rate": 2.604693971468017e-06,
+      "loss": 0.594,
+      "step": 20791
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 14.10857105255127,
+      "learning_rate": 2.6038572564113295e-06,
+      "loss": 0.7821,
+      "step": 20792
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 42.20453643798828,
+      "learning_rate": 2.603020541354642e-06,
+      "loss": 1.8697,
+      "step": 20793
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 18.523605346679688,
+      "learning_rate": 2.6021838262979542e-06,
+      "loss": 0.7861,
+      "step": 20794
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 10.07702350616455,
+      "learning_rate": 2.601347111241267e-06,
+      "loss": 2.0771,
+      "step": 20795
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 26.76637077331543,
+      "learning_rate": 2.6005103961845794e-06,
+      "loss": 1.5734,
+      "step": 20796
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 12.20970630645752,
+      "learning_rate": 2.5996736811278922e-06,
+      "loss": 0.6719,
+      "step": 20797
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 18.1197452545166,
+      "learning_rate": 2.598836966071205e-06,
+      "loss": 1.5184,
+      "step": 20798
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 79.8078842163086,
+      "learning_rate": 2.5980002510145174e-06,
+      "loss": 1.9163,
+      "step": 20799
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 11.29383659362793,
+      "learning_rate": 2.5971635359578294e-06,
+      "loss": 1.6318,
+      "step": 20800
+    },
+    {
+      "epoch": 2.61,
+      "eval_loss": 0.08012223988771439,
+      "eval_runtime": 95.7084,
+      "eval_samples_per_second": 37.008,
+      "eval_steps_per_second": 37.008,
+      "step": 20800
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 27.131946563720703,
+      "learning_rate": 2.596326820901142e-06,
+      "loss": 0.7197,
+      "step": 20801
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 8.807607650756836,
+      "learning_rate": 2.595490105844455e-06,
+      "loss": 0.3941,
+      "step": 20802
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 35.305213928222656,
+      "learning_rate": 2.5946533907877673e-06,
+      "loss": 0.9097,
+      "step": 20803
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 26.552886962890625,
+      "learning_rate": 2.59381667573108e-06,
+      "loss": 3.5247,
+      "step": 20804
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 14.467986106872559,
+      "learning_rate": 2.5929799606743925e-06,
+      "loss": 1.6095,
+      "step": 20805
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 16.631698608398438,
+      "learning_rate": 2.5921432456177053e-06,
+      "loss": 0.4805,
+      "step": 20806
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 11.22146987915039,
+      "learning_rate": 2.5913065305610173e-06,
+      "loss": 0.6858,
+      "step": 20807
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.90317440032959,
+      "learning_rate": 2.59046981550433e-06,
+      "loss": 0.1063,
+      "step": 20808
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 12.939689636230469,
+      "learning_rate": 2.589633100447643e-06,
+      "loss": 0.3212,
+      "step": 20809
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 10.119101524353027,
+      "learning_rate": 2.5887963853909553e-06,
+      "loss": 0.0638,
+      "step": 20810
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.70673131942749,
+      "learning_rate": 2.587959670334268e-06,
+      "loss": 0.8487,
+      "step": 20811
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 14.181622505187988,
+      "learning_rate": 2.5871229552775804e-06,
+      "loss": 1.0025,
+      "step": 20812
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 291.0864562988281,
+      "learning_rate": 2.5862862402208932e-06,
+      "loss": 0.7056,
+      "step": 20813
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 16.884033203125,
+      "learning_rate": 2.585449525164205e-06,
+      "loss": 0.8283,
+      "step": 20814
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 13.720187187194824,
+      "learning_rate": 2.584612810107518e-06,
+      "loss": 0.6934,
+      "step": 20815
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 27.1307315826416,
+      "learning_rate": 2.583776095050831e-06,
+      "loss": 0.8664,
+      "step": 20816
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 15.082032203674316,
+      "learning_rate": 2.582939379994143e-06,
+      "loss": 0.5332,
+      "step": 20817
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 18.318191528320312,
+      "learning_rate": 2.582102664937456e-06,
+      "loss": 0.9966,
+      "step": 20818
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 6.35767126083374,
+      "learning_rate": 2.5812659498807684e-06,
+      "loss": 0.3516,
+      "step": 20819
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 7.867492198944092,
+      "learning_rate": 2.580429234824081e-06,
+      "loss": 0.3813,
+      "step": 20820
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.804530143737793,
+      "learning_rate": 2.579592519767393e-06,
+      "loss": 0.0702,
+      "step": 20821
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 13.48747444152832,
+      "learning_rate": 2.578755804710706e-06,
+      "loss": 1.7444,
+      "step": 20822
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 58.505714416503906,
+      "learning_rate": 2.5779190896540183e-06,
+      "loss": 0.99,
+      "step": 20823
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 10.6947603225708,
+      "learning_rate": 2.577082374597331e-06,
+      "loss": 0.7755,
+      "step": 20824
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 35.80773162841797,
+      "learning_rate": 2.576245659540644e-06,
+      "loss": 2.4354,
+      "step": 20825
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 13.465628623962402,
+      "learning_rate": 2.5754089444839563e-06,
+      "loss": 0.59,
+      "step": 20826
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 8.14944076538086,
+      "learning_rate": 2.5745722294272687e-06,
+      "loss": 2.6683,
+      "step": 20827
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 21.839962005615234,
+      "learning_rate": 2.573735514370581e-06,
+      "loss": 1.3925,
+      "step": 20828
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 20.817468643188477,
+      "learning_rate": 2.572898799313894e-06,
+      "loss": 2.7904,
+      "step": 20829
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 26.890777587890625,
+      "learning_rate": 2.5720620842572062e-06,
+      "loss": 1.6562,
+      "step": 20830
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 7.3743767738342285,
+      "learning_rate": 2.571225369200519e-06,
+      "loss": 0.5765,
+      "step": 20831
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 104.87767028808594,
+      "learning_rate": 2.570388654143832e-06,
+      "loss": 0.805,
+      "step": 20832
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 10.344313621520996,
+      "learning_rate": 2.569551939087144e-06,
+      "loss": 0.104,
+      "step": 20833
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 39.46104431152344,
+      "learning_rate": 2.568715224030456e-06,
+      "loss": 1.033,
+      "step": 20834
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 9.544102668762207,
+      "learning_rate": 2.567878508973769e-06,
+      "loss": 0.9579,
+      "step": 20835
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 11.002922058105469,
+      "learning_rate": 2.5670417939170818e-06,
+      "loss": 0.6541,
+      "step": 20836
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 105.73807525634766,
+      "learning_rate": 2.566205078860394e-06,
+      "loss": 1.4776,
+      "step": 20837
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 8.665227890014648,
+      "learning_rate": 2.565368363803707e-06,
+      "loss": 0.6,
+      "step": 20838
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 15.708761215209961,
+      "learning_rate": 2.5645316487470197e-06,
+      "loss": 1.0036,
+      "step": 20839
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 10.720295906066895,
+      "learning_rate": 2.563694933690332e-06,
+      "loss": 0.848,
+      "step": 20840
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 11.769108772277832,
+      "learning_rate": 2.562858218633644e-06,
+      "loss": 2.2958,
+      "step": 20841
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 75.2760238647461,
+      "learning_rate": 2.562021503576957e-06,
+      "loss": 1.1938,
+      "step": 20842
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 21.13252830505371,
+      "learning_rate": 2.5611847885202697e-06,
+      "loss": 1.2499,
+      "step": 20843
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 19.715547561645508,
+      "learning_rate": 2.560348073463582e-06,
+      "loss": 1.2686,
+      "step": 20844
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 17.419666290283203,
+      "learning_rate": 2.559511358406895e-06,
+      "loss": 0.7819,
+      "step": 20845
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 15.564249038696289,
+      "learning_rate": 2.5586746433502072e-06,
+      "loss": 0.6963,
+      "step": 20846
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 92.1093978881836,
+      "learning_rate": 2.55783792829352e-06,
+      "loss": 3.1407,
+      "step": 20847
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 42.56705856323242,
+      "learning_rate": 2.557001213236832e-06,
+      "loss": 1.3469,
+      "step": 20848
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 10.735899925231934,
+      "learning_rate": 2.556164498180145e-06,
+      "loss": 0.5852,
+      "step": 20849
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 14.951603889465332,
+      "learning_rate": 2.5553277831234576e-06,
+      "loss": 0.3663,
+      "step": 20850
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 13.693647384643555,
+      "learning_rate": 2.55449106806677e-06,
+      "loss": 0.6136,
+      "step": 20851
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 10.393128395080566,
+      "learning_rate": 2.5536543530100828e-06,
+      "loss": 0.2843,
+      "step": 20852
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 25.11324119567871,
+      "learning_rate": 2.552817637953395e-06,
+      "loss": 0.9215,
+      "step": 20853
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 21.964876174926758,
+      "learning_rate": 2.551980922896708e-06,
+      "loss": 1.7843,
+      "step": 20854
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 18.30719757080078,
+      "learning_rate": 2.55114420784002e-06,
+      "loss": 1.5675,
+      "step": 20855
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 6.828226089477539,
+      "learning_rate": 2.5503074927833327e-06,
+      "loss": 0.9239,
+      "step": 20856
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 30.28997802734375,
+      "learning_rate": 2.5494707777266455e-06,
+      "loss": 1.8447,
+      "step": 20857
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 7.1686320304870605,
+      "learning_rate": 2.548634062669958e-06,
+      "loss": 0.6699,
+      "step": 20858
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 10.375819206237793,
+      "learning_rate": 2.5477973476132707e-06,
+      "loss": 0.3918,
+      "step": 20859
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 116.3935546875,
+      "learning_rate": 2.546960632556583e-06,
+      "loss": 0.8598,
+      "step": 20860
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 43.87374496459961,
+      "learning_rate": 2.546123917499896e-06,
+      "loss": 1.1581,
+      "step": 20861
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 46.51468276977539,
+      "learning_rate": 2.545287202443208e-06,
+      "loss": 1.136,
+      "step": 20862
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 17.168806076049805,
+      "learning_rate": 2.5444504873865206e-06,
+      "loss": 1.3335,
+      "step": 20863
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 6.863992691040039,
+      "learning_rate": 2.543613772329833e-06,
+      "loss": 0.3917,
+      "step": 20864
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 36.62272644042969,
+      "learning_rate": 2.542777057273146e-06,
+      "loss": 1.0975,
+      "step": 20865
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 60.48590087890625,
+      "learning_rate": 2.5419403422164586e-06,
+      "loss": 1.7068,
+      "step": 20866
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.954254388809204,
+      "learning_rate": 2.541103627159771e-06,
+      "loss": 0.1199,
+      "step": 20867
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 232.37388610839844,
+      "learning_rate": 2.540266912103084e-06,
+      "loss": 1.3999,
+      "step": 20868
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 8.824796676635742,
+      "learning_rate": 2.5394301970463957e-06,
+      "loss": 0.4704,
+      "step": 20869
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 2.114812135696411,
+      "learning_rate": 2.5385934819897086e-06,
+      "loss": 0.0959,
+      "step": 20870
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.9958624839782715,
+      "learning_rate": 2.537756766933021e-06,
+      "loss": 0.1728,
+      "step": 20871
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 14.779882431030273,
+      "learning_rate": 2.5369200518763337e-06,
+      "loss": 0.7662,
+      "step": 20872
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 10.361798286437988,
+      "learning_rate": 2.5360833368196465e-06,
+      "loss": 1.5673,
+      "step": 20873
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 6.132584571838379,
+      "learning_rate": 2.535246621762959e-06,
+      "loss": 0.5689,
+      "step": 20874
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 13.274500846862793,
+      "learning_rate": 2.5344099067062717e-06,
+      "loss": 0.598,
+      "step": 20875
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 12.637325286865234,
+      "learning_rate": 2.5335731916495837e-06,
+      "loss": 0.4602,
+      "step": 20876
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 10.414531707763672,
+      "learning_rate": 2.5327364765928965e-06,
+      "loss": 0.3249,
+      "step": 20877
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 7.475411415100098,
+      "learning_rate": 2.531899761536209e-06,
+      "loss": 1.3807,
+      "step": 20878
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 34.25215148925781,
+      "learning_rate": 2.5310630464795216e-06,
+      "loss": 1.2075,
+      "step": 20879
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 20.318708419799805,
+      "learning_rate": 2.5302263314228345e-06,
+      "loss": 2.0952,
+      "step": 20880
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 16.875774383544922,
+      "learning_rate": 2.529389616366147e-06,
+      "loss": 2.1584,
+      "step": 20881
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 13.094490051269531,
+      "learning_rate": 2.5285529013094596e-06,
+      "loss": 1.6302,
+      "step": 20882
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 23.368240356445312,
+      "learning_rate": 2.5277161862527716e-06,
+      "loss": 0.7334,
+      "step": 20883
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 10.924906730651855,
+      "learning_rate": 2.5268794711960844e-06,
+      "loss": 1.2668,
+      "step": 20884
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 14.63503646850586,
+      "learning_rate": 2.5260427561393968e-06,
+      "loss": 0.242,
+      "step": 20885
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 13.004878044128418,
+      "learning_rate": 2.5252060410827096e-06,
+      "loss": 1.1775,
+      "step": 20886
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 8.729975700378418,
+      "learning_rate": 2.524369326026022e-06,
+      "loss": 1.2534,
+      "step": 20887
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 24.22922706604004,
+      "learning_rate": 2.5235326109693347e-06,
+      "loss": 0.6346,
+      "step": 20888
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 21.758121490478516,
+      "learning_rate": 2.5226958959126475e-06,
+      "loss": 0.6034,
+      "step": 20889
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 14.686166763305664,
+      "learning_rate": 2.5218591808559595e-06,
+      "loss": 0.9316,
+      "step": 20890
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 10.645651817321777,
+      "learning_rate": 2.5210224657992723e-06,
+      "loss": 0.7517,
+      "step": 20891
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.477910280227661,
+      "learning_rate": 2.5201857507425847e-06,
+      "loss": 0.2708,
+      "step": 20892
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 9.044672966003418,
+      "learning_rate": 2.5193490356858975e-06,
+      "loss": 1.2164,
+      "step": 20893
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 26.40950584411621,
+      "learning_rate": 2.51851232062921e-06,
+      "loss": 1.8651,
+      "step": 20894
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 11.842805862426758,
+      "learning_rate": 2.5176756055725227e-06,
+      "loss": 0.5844,
+      "step": 20895
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 15.054704666137695,
+      "learning_rate": 2.5168388905158355e-06,
+      "loss": 0.3755,
+      "step": 20896
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 12.256747245788574,
+      "learning_rate": 2.5160021754591474e-06,
+      "loss": 1.0146,
+      "step": 20897
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 11.8973970413208,
+      "learning_rate": 2.51516546040246e-06,
+      "loss": 0.7053,
+      "step": 20898
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 9.270255088806152,
+      "learning_rate": 2.5143287453457726e-06,
+      "loss": 0.736,
+      "step": 20899
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 15.950983047485352,
+      "learning_rate": 2.5134920302890854e-06,
+      "loss": 0.7235,
+      "step": 20900
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.982297897338867,
+      "learning_rate": 2.5126553152323978e-06,
+      "loss": 0.5387,
+      "step": 20901
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 8.136774063110352,
+      "learning_rate": 2.5118186001757106e-06,
+      "loss": 0.3629,
+      "step": 20902
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 26.423789978027344,
+      "learning_rate": 2.5109818851190234e-06,
+      "loss": 1.4861,
+      "step": 20903
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 26.003742218017578,
+      "learning_rate": 2.5101451700623353e-06,
+      "loss": 0.5039,
+      "step": 20904
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.174650192260742,
+      "learning_rate": 2.5093084550056477e-06,
+      "loss": 0.0985,
+      "step": 20905
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 23.96674346923828,
+      "learning_rate": 2.5084717399489605e-06,
+      "loss": 2.6189,
+      "step": 20906
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 20.15669059753418,
+      "learning_rate": 2.5076350248922733e-06,
+      "loss": 1.8705,
+      "step": 20907
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.903242588043213,
+      "learning_rate": 2.5067983098355857e-06,
+      "loss": 0.8006,
+      "step": 20908
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 36.92401885986328,
+      "learning_rate": 2.5059615947788985e-06,
+      "loss": 1.8831,
+      "step": 20909
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 11.049201965332031,
+      "learning_rate": 2.505124879722211e-06,
+      "loss": 1.0461,
+      "step": 20910
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 11.211837768554688,
+      "learning_rate": 2.5042881646655233e-06,
+      "loss": 0.9679,
+      "step": 20911
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 27.66070556640625,
+      "learning_rate": 2.5034514496088356e-06,
+      "loss": 1.9762,
+      "step": 20912
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 6.565810680389404,
+      "learning_rate": 2.5026147345521484e-06,
+      "loss": 0.1723,
+      "step": 20913
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 11.280132293701172,
+      "learning_rate": 2.5017780194954612e-06,
+      "loss": 1.4626,
+      "step": 20914
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 34.04676818847656,
+      "learning_rate": 2.5009413044387736e-06,
+      "loss": 0.9478,
+      "step": 20915
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 31.019742965698242,
+      "learning_rate": 2.5001045893820864e-06,
+      "loss": 0.7875,
+      "step": 20916
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 16.54743194580078,
+      "learning_rate": 2.499267874325399e-06,
+      "loss": 0.9548,
+      "step": 20917
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 22.695558547973633,
+      "learning_rate": 2.498431159268711e-06,
+      "loss": 3.3674,
+      "step": 20918
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 8.537609100341797,
+      "learning_rate": 2.4975944442120236e-06,
+      "loss": 0.4449,
+      "step": 20919
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.5578932762146,
+      "learning_rate": 2.4967577291553364e-06,
+      "loss": 0.6365,
+      "step": 20920
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 69.54683685302734,
+      "learning_rate": 2.495921014098649e-06,
+      "loss": 2.8953,
+      "step": 20921
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 19.73796844482422,
+      "learning_rate": 2.4950842990419615e-06,
+      "loss": 2.6961,
+      "step": 20922
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 20.43222999572754,
+      "learning_rate": 2.494247583985274e-06,
+      "loss": 1.7699,
+      "step": 20923
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 16.330984115600586,
+      "learning_rate": 2.4934108689285867e-06,
+      "loss": 0.7528,
+      "step": 20924
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 9.302873611450195,
+      "learning_rate": 2.492574153871899e-06,
+      "loss": 0.2775,
+      "step": 20925
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 17.74481201171875,
+      "learning_rate": 2.4917374388152115e-06,
+      "loss": 0.7376,
+      "step": 20926
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 21.127634048461914,
+      "learning_rate": 2.4909007237585243e-06,
+      "loss": 0.8145,
+      "step": 20927
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 15.73021125793457,
+      "learning_rate": 2.4900640087018367e-06,
+      "loss": 1.2041,
+      "step": 20928
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 11.025052070617676,
+      "learning_rate": 2.4892272936451495e-06,
+      "loss": 1.8824,
+      "step": 20929
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 17.70407485961914,
+      "learning_rate": 2.488390578588462e-06,
+      "loss": 0.5918,
+      "step": 20930
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 19.538415908813477,
+      "learning_rate": 2.4875538635317746e-06,
+      "loss": 0.3756,
+      "step": 20931
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 12.627802848815918,
+      "learning_rate": 2.486717148475087e-06,
+      "loss": 0.7633,
+      "step": 20932
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 23.011611938476562,
+      "learning_rate": 2.4858804334183994e-06,
+      "loss": 1.0695,
+      "step": 20933
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.7653117179870605,
+      "learning_rate": 2.485043718361712e-06,
+      "loss": 1.0073,
+      "step": 20934
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 12.809354782104492,
+      "learning_rate": 2.4842070033050246e-06,
+      "loss": 1.3837,
+      "step": 20935
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.444727659225464,
+      "learning_rate": 2.4833702882483374e-06,
+      "loss": 0.1017,
+      "step": 20936
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 8.391605377197266,
+      "learning_rate": 2.4825335731916498e-06,
+      "loss": 0.39,
+      "step": 20937
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 33.942481994628906,
+      "learning_rate": 2.4816968581349626e-06,
+      "loss": 1.4194,
+      "step": 20938
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.683876991271973,
+      "learning_rate": 2.480860143078275e-06,
+      "loss": 0.0859,
+      "step": 20939
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 6.307435035705566,
+      "learning_rate": 2.4800234280215873e-06,
+      "loss": 0.4947,
+      "step": 20940
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 10.36919116973877,
+      "learning_rate": 2.4791867129649e-06,
+      "loss": 1.7781,
+      "step": 20941
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.4842617511749268,
+      "learning_rate": 2.4783499979082125e-06,
+      "loss": 0.0286,
+      "step": 20942
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.505274295806885,
+      "learning_rate": 2.477513282851525e-06,
+      "loss": 0.136,
+      "step": 20943
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 19.212722778320312,
+      "learning_rate": 2.4766765677948377e-06,
+      "loss": 2.0218,
+      "step": 20944
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 88.36568450927734,
+      "learning_rate": 2.47583985273815e-06,
+      "loss": 0.3264,
+      "step": 20945
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 19.913543701171875,
+      "learning_rate": 2.475003137681463e-06,
+      "loss": 1.0815,
+      "step": 20946
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 24.3832950592041,
+      "learning_rate": 2.4741664226247752e-06,
+      "loss": 1.1981,
+      "step": 20947
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 21.547956466674805,
+      "learning_rate": 2.473329707568088e-06,
+      "loss": 1.0862,
+      "step": 20948
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 6.589284420013428,
+      "learning_rate": 2.4724929925114004e-06,
+      "loss": 0.2015,
+      "step": 20949
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.2801471948623657,
+      "learning_rate": 2.471656277454713e-06,
+      "loss": 0.0567,
+      "step": 20950
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 134.08868408203125,
+      "learning_rate": 2.4708195623980256e-06,
+      "loss": 1.8712,
+      "step": 20951
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 10.042970657348633,
+      "learning_rate": 2.469982847341338e-06,
+      "loss": 2.0202,
+      "step": 20952
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 21.065839767456055,
+      "learning_rate": 2.4691461322846508e-06,
+      "loss": 0.4961,
+      "step": 20953
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 14.301145553588867,
+      "learning_rate": 2.468309417227963e-06,
+      "loss": 0.8794,
+      "step": 20954
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 8.55249309539795,
+      "learning_rate": 2.467472702171276e-06,
+      "loss": 0.1541,
+      "step": 20955
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 11.381338119506836,
+      "learning_rate": 2.4666359871145883e-06,
+      "loss": 0.7351,
+      "step": 20956
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 60.15916442871094,
+      "learning_rate": 2.4657992720579007e-06,
+      "loss": 1.2717,
+      "step": 20957
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.352663993835449,
+      "learning_rate": 2.4649625570012135e-06,
+      "loss": 0.6574,
+      "step": 20958
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.918491363525391,
+      "learning_rate": 2.464125841944526e-06,
+      "loss": 0.7593,
+      "step": 20959
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 13.636199951171875,
+      "learning_rate": 2.4632891268878387e-06,
+      "loss": 0.521,
+      "step": 20960
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 9.359580039978027,
+      "learning_rate": 2.462452411831151e-06,
+      "loss": 1.1243,
+      "step": 20961
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 17.589763641357422,
+      "learning_rate": 2.461615696774464e-06,
+      "loss": 1.5709,
+      "step": 20962
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 10.143199920654297,
+      "learning_rate": 2.4607789817177762e-06,
+      "loss": 0.4982,
+      "step": 20963
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 11.696372032165527,
+      "learning_rate": 2.4599422666610886e-06,
+      "loss": 0.6827,
+      "step": 20964
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 26.767925262451172,
+      "learning_rate": 2.4591055516044014e-06,
+      "loss": 1.0194,
+      "step": 20965
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 14.609832763671875,
+      "learning_rate": 2.458268836547714e-06,
+      "loss": 0.858,
+      "step": 20966
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 10.931877136230469,
+      "learning_rate": 2.4574321214910266e-06,
+      "loss": 0.2819,
+      "step": 20967
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 16.908096313476562,
+      "learning_rate": 2.456595406434339e-06,
+      "loss": 0.8879,
+      "step": 20968
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 23.534486770629883,
+      "learning_rate": 2.4557586913776514e-06,
+      "loss": 2.2925,
+      "step": 20969
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 13.532764434814453,
+      "learning_rate": 2.454921976320964e-06,
+      "loss": 0.3926,
+      "step": 20970
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 8.230018615722656,
+      "learning_rate": 2.4540852612642765e-06,
+      "loss": 0.7927,
+      "step": 20971
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 15.438993453979492,
+      "learning_rate": 2.4532485462075893e-06,
+      "loss": 1.0728,
+      "step": 20972
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 34.309425354003906,
+      "learning_rate": 2.4524118311509017e-06,
+      "loss": 0.722,
+      "step": 20973
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.6777411103248596,
+      "learning_rate": 2.4515751160942145e-06,
+      "loss": 0.0491,
+      "step": 20974
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 7.284146785736084,
+      "learning_rate": 2.450738401037527e-06,
+      "loss": 0.1572,
+      "step": 20975
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.39168119430542,
+      "learning_rate": 2.4499016859808393e-06,
+      "loss": 0.2688,
+      "step": 20976
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 17.017398834228516,
+      "learning_rate": 2.449064970924152e-06,
+      "loss": 1.1687,
+      "step": 20977
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 125.64471435546875,
+      "learning_rate": 2.4482282558674645e-06,
+      "loss": 0.9681,
+      "step": 20978
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 8.142346382141113,
+      "learning_rate": 2.4473915408107773e-06,
+      "loss": 1.0965,
+      "step": 20979
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 10.373797416687012,
+      "learning_rate": 2.4465548257540896e-06,
+      "loss": 0.3774,
+      "step": 20980
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 15.762367248535156,
+      "learning_rate": 2.4457181106974024e-06,
+      "loss": 0.6187,
+      "step": 20981
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 16.356544494628906,
+      "learning_rate": 2.444881395640715e-06,
+      "loss": 2.3165,
+      "step": 20982
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 16.276365280151367,
+      "learning_rate": 2.444044680584027e-06,
+      "loss": 0.754,
+      "step": 20983
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 93.12158203125,
+      "learning_rate": 2.44320796552734e-06,
+      "loss": 1.9056,
+      "step": 20984
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 17.247848510742188,
+      "learning_rate": 2.4423712504706524e-06,
+      "loss": 0.5727,
+      "step": 20985
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 33.041507720947266,
+      "learning_rate": 2.4415345354139648e-06,
+      "loss": 0.3893,
+      "step": 20986
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 26.996477127075195,
+      "learning_rate": 2.4406978203572776e-06,
+      "loss": 1.1712,
+      "step": 20987
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 14.81056022644043,
+      "learning_rate": 2.43986110530059e-06,
+      "loss": 1.5086,
+      "step": 20988
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 35.87815856933594,
+      "learning_rate": 2.4390243902439027e-06,
+      "loss": 1.5195,
+      "step": 20989
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 13.832409858703613,
+      "learning_rate": 2.438187675187215e-06,
+      "loss": 1.307,
+      "step": 20990
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 22.23716163635254,
+      "learning_rate": 2.437350960130528e-06,
+      "loss": 2.4104,
+      "step": 20991
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 7.499285697937012,
+      "learning_rate": 2.4365142450738403e-06,
+      "loss": 0.1659,
+      "step": 20992
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 7.612830638885498,
+      "learning_rate": 2.4356775300171527e-06,
+      "loss": 0.8796,
+      "step": 20993
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 7.417469501495361,
+      "learning_rate": 2.4348408149604655e-06,
+      "loss": 0.7916,
+      "step": 20994
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 7.098451614379883,
+      "learning_rate": 2.434004099903778e-06,
+      "loss": 2.1647,
+      "step": 20995
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 8.97706127166748,
+      "learning_rate": 2.4331673848470907e-06,
+      "loss": 0.4958,
+      "step": 20996
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 33.11330032348633,
+      "learning_rate": 2.432330669790403e-06,
+      "loss": 1.3894,
+      "step": 20997
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 7.475759506225586,
+      "learning_rate": 2.431493954733716e-06,
+      "loss": 1.1272,
+      "step": 20998
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 6.357087135314941,
+      "learning_rate": 2.4306572396770282e-06,
+      "loss": 1.1372,
+      "step": 20999
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 15.309009552001953,
+      "learning_rate": 2.4298205246203406e-06,
+      "loss": 0.9697,
+      "step": 21000
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 14.061135292053223,
+      "learning_rate": 2.4289838095636534e-06,
+      "loss": 0.7943,
+      "step": 21001
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 6.136180400848389,
+      "learning_rate": 2.4281470945069658e-06,
+      "loss": 0.6436,
+      "step": 21002
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 8.432313919067383,
+      "learning_rate": 2.427310379450278e-06,
+      "loss": 0.1336,
+      "step": 21003
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 16.358102798461914,
+      "learning_rate": 2.426473664393591e-06,
+      "loss": 0.5794,
+      "step": 21004
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 59.56919860839844,
+      "learning_rate": 2.4256369493369038e-06,
+      "loss": 1.5782,
+      "step": 21005
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 14.38819408416748,
+      "learning_rate": 2.424800234280216e-06,
+      "loss": 0.6115,
+      "step": 21006
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 77.99490356445312,
+      "learning_rate": 2.4239635192235285e-06,
+      "loss": 1.1944,
+      "step": 21007
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 15.622925758361816,
+      "learning_rate": 2.4231268041668413e-06,
+      "loss": 1.0147,
+      "step": 21008
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 6.72713565826416,
+      "learning_rate": 2.4222900891101537e-06,
+      "loss": 0.7304,
+      "step": 21009
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.2990753650665283,
+      "learning_rate": 2.421453374053466e-06,
+      "loss": 0.0797,
+      "step": 21010
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 8.245819091796875,
+      "learning_rate": 2.420616658996779e-06,
+      "loss": 0.4123,
+      "step": 21011
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 15.5343599319458,
+      "learning_rate": 2.4197799439400917e-06,
+      "loss": 0.7004,
+      "step": 21012
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 6.714312553405762,
+      "learning_rate": 2.418943228883404e-06,
+      "loss": 0.4894,
+      "step": 21013
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 193.41500854492188,
+      "learning_rate": 2.4181065138267164e-06,
+      "loss": 0.8122,
+      "step": 21014
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 13.690166473388672,
+      "learning_rate": 2.4172697987700292e-06,
+      "loss": 0.607,
+      "step": 21015
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 12.655191421508789,
+      "learning_rate": 2.4164330837133416e-06,
+      "loss": 0.3633,
+      "step": 21016
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 20.513376235961914,
+      "learning_rate": 2.415596368656654e-06,
+      "loss": 0.5502,
+      "step": 21017
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 13.626429557800293,
+      "learning_rate": 2.414759653599967e-06,
+      "loss": 0.9843,
+      "step": 21018
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 60.88151550292969,
+      "learning_rate": 2.4139229385432796e-06,
+      "loss": 1.2858,
+      "step": 21019
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.7170236110687256,
+      "learning_rate": 2.413086223486592e-06,
+      "loss": 0.0817,
+      "step": 21020
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 15.460129737854004,
+      "learning_rate": 2.4122495084299044e-06,
+      "loss": 0.9185,
+      "step": 21021
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 6.197540283203125,
+      "learning_rate": 2.411412793373217e-06,
+      "loss": 0.2762,
+      "step": 21022
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 21.68492889404297,
+      "learning_rate": 2.4105760783165295e-06,
+      "loss": 0.9793,
+      "step": 21023
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 23.647262573242188,
+      "learning_rate": 2.409739363259842e-06,
+      "loss": 0.9037,
+      "step": 21024
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.1407809257507324,
+      "learning_rate": 2.4089026482031547e-06,
+      "loss": 0.1567,
+      "step": 21025
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 13.150588035583496,
+      "learning_rate": 2.4080659331464675e-06,
+      "loss": 1.1496,
+      "step": 21026
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 7.087491035461426,
+      "learning_rate": 2.4072292180897795e-06,
+      "loss": 0.4358,
+      "step": 21027
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 76.25025939941406,
+      "learning_rate": 2.4063925030330923e-06,
+      "loss": 3.1025,
+      "step": 21028
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.9695963859558105,
+      "learning_rate": 2.405555787976405e-06,
+      "loss": 0.1382,
+      "step": 21029
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 24.534563064575195,
+      "learning_rate": 2.4047190729197175e-06,
+      "loss": 0.8972,
+      "step": 21030
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 9.088956832885742,
+      "learning_rate": 2.40388235786303e-06,
+      "loss": 0.4364,
+      "step": 21031
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 11.973158836364746,
+      "learning_rate": 2.4030456428063426e-06,
+      "loss": 0.0905,
+      "step": 21032
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.274954080581665,
+      "learning_rate": 2.402208927749655e-06,
+      "loss": 0.1071,
+      "step": 21033
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 15.855847358703613,
+      "learning_rate": 2.4013722126929674e-06,
+      "loss": 1.3971,
+      "step": 21034
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 22.961088180541992,
+      "learning_rate": 2.40053549763628e-06,
+      "loss": 1.0131,
+      "step": 21035
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 9.95785903930664,
+      "learning_rate": 2.399698782579593e-06,
+      "loss": 1.2831,
+      "step": 21036
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 18.962446212768555,
+      "learning_rate": 2.3988620675229054e-06,
+      "loss": 0.5042,
+      "step": 21037
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 18.83690643310547,
+      "learning_rate": 2.3980253524662177e-06,
+      "loss": 1.9688,
+      "step": 21038
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 25.99830436706543,
+      "learning_rate": 2.3971886374095305e-06,
+      "loss": 0.9632,
+      "step": 21039
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 56.00041198730469,
+      "learning_rate": 2.396351922352843e-06,
+      "loss": 0.6948,
+      "step": 21040
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.097028732299805,
+      "learning_rate": 2.3955152072961553e-06,
+      "loss": 0.2113,
+      "step": 21041
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 13.911191940307617,
+      "learning_rate": 2.394678492239468e-06,
+      "loss": 0.9646,
+      "step": 21042
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 57.96830368041992,
+      "learning_rate": 2.393841777182781e-06,
+      "loss": 0.9468,
+      "step": 21043
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 73.54779815673828,
+      "learning_rate": 2.393005062126093e-06,
+      "loss": 1.08,
+      "step": 21044
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 17.078927993774414,
+      "learning_rate": 2.3921683470694057e-06,
+      "loss": 0.5569,
+      "step": 21045
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 9.047167778015137,
+      "learning_rate": 2.3913316320127185e-06,
+      "loss": 1.7128,
+      "step": 21046
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 17.860641479492188,
+      "learning_rate": 2.390494916956031e-06,
+      "loss": 0.6911,
+      "step": 21047
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 21.053466796875,
+      "learning_rate": 2.3896582018993432e-06,
+      "loss": 1.0133,
+      "step": 21048
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 35.85791778564453,
+      "learning_rate": 2.388821486842656e-06,
+      "loss": 0.7643,
+      "step": 21049
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 17.7092227935791,
+      "learning_rate": 2.3879847717859684e-06,
+      "loss": 1.0582,
+      "step": 21050
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 10.426857948303223,
+      "learning_rate": 2.3871480567292808e-06,
+      "loss": 0.5564,
+      "step": 21051
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 779.8558349609375,
+      "learning_rate": 2.3863113416725936e-06,
+      "loss": 0.3771,
+      "step": 21052
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 19.85243034362793,
+      "learning_rate": 2.3854746266159064e-06,
+      "loss": 1.423,
+      "step": 21053
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 25.018346786499023,
+      "learning_rate": 2.3846379115592188e-06,
+      "loss": 0.5543,
+      "step": 21054
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 36.439334869384766,
+      "learning_rate": 2.383801196502531e-06,
+      "loss": 3.2907,
+      "step": 21055
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 21.46392822265625,
+      "learning_rate": 2.382964481445844e-06,
+      "loss": 1.368,
+      "step": 21056
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 31.040111541748047,
+      "learning_rate": 2.3821277663891563e-06,
+      "loss": 1.1571,
+      "step": 21057
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 112.70236206054688,
+      "learning_rate": 2.3812910513324687e-06,
+      "loss": 0.8701,
+      "step": 21058
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 61.37261199951172,
+      "learning_rate": 2.3804543362757815e-06,
+      "loss": 1.4428,
+      "step": 21059
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 21.433765411376953,
+      "learning_rate": 2.3796176212190943e-06,
+      "loss": 1.0511,
+      "step": 21060
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 27.52485466003418,
+      "learning_rate": 2.3787809061624067e-06,
+      "loss": 1.7669,
+      "step": 21061
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.9846388101577759,
+      "learning_rate": 2.377944191105719e-06,
+      "loss": 0.0667,
+      "step": 21062
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 8.852572441101074,
+      "learning_rate": 2.377107476049032e-06,
+      "loss": 0.3681,
+      "step": 21063
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 39.31391525268555,
+      "learning_rate": 2.3762707609923442e-06,
+      "loss": 0.408,
+      "step": 21064
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 9.55923843383789,
+      "learning_rate": 2.3754340459356566e-06,
+      "loss": 0.8344,
+      "step": 21065
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 16.293432235717773,
+      "learning_rate": 2.3745973308789694e-06,
+      "loss": 1.4,
+      "step": 21066
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 29.0277042388916,
+      "learning_rate": 2.373760615822282e-06,
+      "loss": 0.486,
+      "step": 21067
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 14.477266311645508,
+      "learning_rate": 2.372923900765594e-06,
+      "loss": 1.2147,
+      "step": 21068
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 26.399272918701172,
+      "learning_rate": 2.372087185708907e-06,
+      "loss": 0.6965,
+      "step": 21069
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 9.183669090270996,
+      "learning_rate": 2.3712504706522198e-06,
+      "loss": 0.9176,
+      "step": 21070
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 165.87615966796875,
+      "learning_rate": 2.370413755595532e-06,
+      "loss": 1.3961,
+      "step": 21071
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 13.78731918334961,
+      "learning_rate": 2.3695770405388445e-06,
+      "loss": 1.2472,
+      "step": 21072
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 15.470413208007812,
+      "learning_rate": 2.3687403254821573e-06,
+      "loss": 0.6179,
+      "step": 21073
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 11.213066101074219,
+      "learning_rate": 2.3679036104254697e-06,
+      "loss": 0.7374,
+      "step": 21074
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 8.551079750061035,
+      "learning_rate": 2.367066895368782e-06,
+      "loss": 0.6043,
+      "step": 21075
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 11.665029525756836,
+      "learning_rate": 2.366230180312095e-06,
+      "loss": 0.3494,
+      "step": 21076
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 116.39488983154297,
+      "learning_rate": 2.3653934652554077e-06,
+      "loss": 2.3552,
+      "step": 21077
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 19.45047378540039,
+      "learning_rate": 2.36455675019872e-06,
+      "loss": 0.6457,
+      "step": 21078
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 11.663397789001465,
+      "learning_rate": 2.3637200351420325e-06,
+      "loss": 0.4456,
+      "step": 21079
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 12.752456665039062,
+      "learning_rate": 2.3628833200853453e-06,
+      "loss": 1.113,
+      "step": 21080
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 17.248798370361328,
+      "learning_rate": 2.3620466050286576e-06,
+      "loss": 2.3385,
+      "step": 21081
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 12.785979270935059,
+      "learning_rate": 2.36120988997197e-06,
+      "loss": 0.4756,
+      "step": 21082
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 28.546894073486328,
+      "learning_rate": 2.360373174915283e-06,
+      "loss": 0.211,
+      "step": 21083
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 21.222902297973633,
+      "learning_rate": 2.3595364598585956e-06,
+      "loss": 0.75,
+      "step": 21084
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 8.43933391571045,
+      "learning_rate": 2.3586997448019076e-06,
+      "loss": 0.2291,
+      "step": 21085
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 7.80899715423584,
+      "learning_rate": 2.3578630297452204e-06,
+      "loss": 0.5475,
+      "step": 21086
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 20.870101928710938,
+      "learning_rate": 2.357026314688533e-06,
+      "loss": 0.9009,
+      "step": 21087
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 14.971521377563477,
+      "learning_rate": 2.3561895996318456e-06,
+      "loss": 1.3156,
+      "step": 21088
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 12.223374366760254,
+      "learning_rate": 2.355352884575158e-06,
+      "loss": 1.7749,
+      "step": 21089
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 19.237293243408203,
+      "learning_rate": 2.3545161695184707e-06,
+      "loss": 2.4655,
+      "step": 21090
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 11.605509757995605,
+      "learning_rate": 2.353679454461783e-06,
+      "loss": 0.442,
+      "step": 21091
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.847765922546387,
+      "learning_rate": 2.3528427394050955e-06,
+      "loss": 0.6636,
+      "step": 21092
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 16.315454483032227,
+      "learning_rate": 2.3520060243484083e-06,
+      "loss": 0.5457,
+      "step": 21093
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 14.820706367492676,
+      "learning_rate": 2.351169309291721e-06,
+      "loss": 0.6809,
+      "step": 21094
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 16.140243530273438,
+      "learning_rate": 2.3503325942350335e-06,
+      "loss": 0.712,
+      "step": 21095
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 17.508155822753906,
+      "learning_rate": 2.349495879178346e-06,
+      "loss": 0.7529,
+      "step": 21096
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 15.05785846710205,
+      "learning_rate": 2.3486591641216587e-06,
+      "loss": 0.3504,
+      "step": 21097
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 11.685197830200195,
+      "learning_rate": 2.347822449064971e-06,
+      "loss": 0.8119,
+      "step": 21098
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.483983039855957,
+      "learning_rate": 2.3469857340082834e-06,
+      "loss": 0.1311,
+      "step": 21099
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 9.965291976928711,
+      "learning_rate": 2.3461490189515962e-06,
+      "loss": 0.3623,
+      "step": 21100
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 12.08174991607666,
+      "learning_rate": 2.345312303894909e-06,
+      "loss": 0.2287,
+      "step": 21101
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 18.51045799255371,
+      "learning_rate": 2.3444755888382214e-06,
+      "loss": 0.6785,
+      "step": 21102
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 28.313142776489258,
+      "learning_rate": 2.3436388737815338e-06,
+      "loss": 1.2543,
+      "step": 21103
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 7.365633010864258,
+      "learning_rate": 2.3428021587248466e-06,
+      "loss": 0.4875,
+      "step": 21104
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 100.35225677490234,
+      "learning_rate": 2.341965443668159e-06,
+      "loss": 1.3184,
+      "step": 21105
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.2970550060272217,
+      "learning_rate": 2.3411287286114713e-06,
+      "loss": 0.0375,
+      "step": 21106
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 20.789562225341797,
+      "learning_rate": 2.340292013554784e-06,
+      "loss": 0.6848,
+      "step": 21107
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 17.4787654876709,
+      "learning_rate": 2.3394552984980965e-06,
+      "loss": 1.5577,
+      "step": 21108
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 27.095874786376953,
+      "learning_rate": 2.3386185834414093e-06,
+      "loss": 0.9173,
+      "step": 21109
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 42.789302825927734,
+      "learning_rate": 2.3377818683847217e-06,
+      "loss": 0.8784,
+      "step": 21110
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 11.568604469299316,
+      "learning_rate": 2.3369451533280345e-06,
+      "loss": 0.5209,
+      "step": 21111
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 8.733729362487793,
+      "learning_rate": 2.336108438271347e-06,
+      "loss": 1.4767,
+      "step": 21112
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 23.612058639526367,
+      "learning_rate": 2.3352717232146592e-06,
+      "loss": 0.7964,
+      "step": 21113
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 8.344063758850098,
+      "learning_rate": 2.334435008157972e-06,
+      "loss": 0.5217,
+      "step": 21114
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 10.633986473083496,
+      "learning_rate": 2.3335982931012844e-06,
+      "loss": 0.6476,
+      "step": 21115
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 21.941987991333008,
+      "learning_rate": 2.3327615780445972e-06,
+      "loss": 1.5961,
+      "step": 21116
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 31.771562576293945,
+      "learning_rate": 2.3319248629879096e-06,
+      "loss": 1.9412,
+      "step": 21117
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 15.069620132446289,
+      "learning_rate": 2.3310881479312224e-06,
+      "loss": 1.5819,
+      "step": 21118
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 16.771032333374023,
+      "learning_rate": 2.3302514328745348e-06,
+      "loss": 0.2783,
+      "step": 21119
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 6.584004878997803,
+      "learning_rate": 2.329414717817847e-06,
+      "loss": 0.9431,
+      "step": 21120
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.5742485523223877,
+      "learning_rate": 2.32857800276116e-06,
+      "loss": 0.0997,
+      "step": 21121
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 85.23136138916016,
+      "learning_rate": 2.3277412877044723e-06,
+      "loss": 0.7781,
+      "step": 21122
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 19.807178497314453,
+      "learning_rate": 2.3269045726477847e-06,
+      "loss": 0.9071,
+      "step": 21123
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.632420063018799,
+      "learning_rate": 2.3260678575910975e-06,
+      "loss": 0.1671,
+      "step": 21124
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 30.62589454650879,
+      "learning_rate": 2.3252311425344103e-06,
+      "loss": 1.0105,
+      "step": 21125
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 13.17794132232666,
+      "learning_rate": 2.3243944274777227e-06,
+      "loss": 0.4428,
+      "step": 21126
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 287.6689147949219,
+      "learning_rate": 2.323557712421035e-06,
+      "loss": 0.7331,
+      "step": 21127
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 32.80656814575195,
+      "learning_rate": 2.322720997364348e-06,
+      "loss": 3.8538,
+      "step": 21128
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 17.656213760375977,
+      "learning_rate": 2.3218842823076603e-06,
+      "loss": 0.3552,
+      "step": 21129
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 18.911216735839844,
+      "learning_rate": 2.3210475672509726e-06,
+      "loss": 0.8092,
+      "step": 21130
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 8.57754898071289,
+      "learning_rate": 2.3202108521942854e-06,
+      "loss": 0.2883,
+      "step": 21131
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 12.58278751373291,
+      "learning_rate": 2.319374137137598e-06,
+      "loss": 1.1585,
+      "step": 21132
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 22.420452117919922,
+      "learning_rate": 2.3185374220809106e-06,
+      "loss": 1.0776,
+      "step": 21133
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 28.474044799804688,
+      "learning_rate": 2.317700707024223e-06,
+      "loss": 0.7962,
+      "step": 21134
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.378330707550049,
+      "learning_rate": 2.316863991967536e-06,
+      "loss": 0.1699,
+      "step": 21135
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 13.12812328338623,
+      "learning_rate": 2.316027276910848e-06,
+      "loss": 0.5099,
+      "step": 21136
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 7.315151214599609,
+      "learning_rate": 2.3151905618541606e-06,
+      "loss": 0.7315,
+      "step": 21137
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 9.438087463378906,
+      "learning_rate": 2.3143538467974734e-06,
+      "loss": 0.7399,
+      "step": 21138
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 32.87921142578125,
+      "learning_rate": 2.3135171317407857e-06,
+      "loss": 2.3377,
+      "step": 21139
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 17.833864212036133,
+      "learning_rate": 2.3126804166840985e-06,
+      "loss": 2.3104,
+      "step": 21140
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 21.223148345947266,
+      "learning_rate": 2.311843701627411e-06,
+      "loss": 0.6581,
+      "step": 21141
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 12.318990707397461,
+      "learning_rate": 2.3110069865707237e-06,
+      "loss": 1.3616,
+      "step": 21142
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 7.584165096282959,
+      "learning_rate": 2.310170271514036e-06,
+      "loss": 0.2593,
+      "step": 21143
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 11.725146293640137,
+      "learning_rate": 2.3093335564573485e-06,
+      "loss": 0.3027,
+      "step": 21144
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 7.892516136169434,
+      "learning_rate": 2.3084968414006613e-06,
+      "loss": 0.2908,
+      "step": 21145
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.382558822631836,
+      "learning_rate": 2.3076601263439737e-06,
+      "loss": 0.0627,
+      "step": 21146
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 9.933945655822754,
+      "learning_rate": 2.3068234112872865e-06,
+      "loss": 0.2756,
+      "step": 21147
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 8.98509693145752,
+      "learning_rate": 2.305986696230599e-06,
+      "loss": 1.4634,
+      "step": 21148
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 8.52401065826416,
+      "learning_rate": 2.3051499811739112e-06,
+      "loss": 0.4213,
+      "step": 21149
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 6.180952548980713,
+      "learning_rate": 2.304313266117224e-06,
+      "loss": 0.4455,
+      "step": 21150
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 7.034838676452637,
+      "learning_rate": 2.3034765510605364e-06,
+      "loss": 0.6509,
+      "step": 21151
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 29.16041374206543,
+      "learning_rate": 2.302639836003849e-06,
+      "loss": 0.9128,
+      "step": 21152
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 32.40922927856445,
+      "learning_rate": 2.3018031209471616e-06,
+      "loss": 1.5813,
+      "step": 21153
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 15.488125801086426,
+      "learning_rate": 2.3009664058904744e-06,
+      "loss": 1.5634,
+      "step": 21154
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 19.64240837097168,
+      "learning_rate": 2.3001296908337868e-06,
+      "loss": 1.8619,
+      "step": 21155
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 12.897624015808105,
+      "learning_rate": 2.299292975777099e-06,
+      "loss": 0.5688,
+      "step": 21156
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 10.610186576843262,
+      "learning_rate": 2.298456260720412e-06,
+      "loss": 1.3531,
+      "step": 21157
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 77.85247039794922,
+      "learning_rate": 2.2976195456637243e-06,
+      "loss": 1.2388,
+      "step": 21158
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 43.8265495300293,
+      "learning_rate": 2.296782830607037e-06,
+      "loss": 1.5372,
+      "step": 21159
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 8.644686698913574,
+      "learning_rate": 2.2959461155503495e-06,
+      "loss": 0.7571,
+      "step": 21160
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 19.352209091186523,
+      "learning_rate": 2.2951094004936623e-06,
+      "loss": 2.5901,
+      "step": 21161
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.2310041189193726,
+      "learning_rate": 2.2942726854369747e-06,
+      "loss": 0.0243,
+      "step": 21162
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 9.61866283416748,
+      "learning_rate": 2.293435970380287e-06,
+      "loss": 0.3114,
+      "step": 21163
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 35.355979919433594,
+      "learning_rate": 2.2925992553236e-06,
+      "loss": 0.8666,
+      "step": 21164
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 11.301298141479492,
+      "learning_rate": 2.2917625402669122e-06,
+      "loss": 1.0186,
+      "step": 21165
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 47.20576858520508,
+      "learning_rate": 2.290925825210225e-06,
+      "loss": 0.8651,
+      "step": 21166
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.567676067352295,
+      "learning_rate": 2.2900891101535374e-06,
+      "loss": 0.5583,
+      "step": 21167
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 22.5992488861084,
+      "learning_rate": 2.28925239509685e-06,
+      "loss": 1.4118,
+      "step": 21168
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 7.403890132904053,
+      "learning_rate": 2.2884156800401626e-06,
+      "loss": 0.3636,
+      "step": 21169
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 14.168628692626953,
+      "learning_rate": 2.287578964983475e-06,
+      "loss": 0.4622,
+      "step": 21170
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 15.81656265258789,
+      "learning_rate": 2.2867422499267878e-06,
+      "loss": 1.4615,
+      "step": 21171
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.977477073669434,
+      "learning_rate": 2.2859055348701e-06,
+      "loss": 0.2417,
+      "step": 21172
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.2455973625183105,
+      "learning_rate": 2.2850688198134125e-06,
+      "loss": 1.464,
+      "step": 21173
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 10.898738861083984,
+      "learning_rate": 2.2842321047567253e-06,
+      "loss": 0.5189,
+      "step": 21174
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 7.893359184265137,
+      "learning_rate": 2.2833953897000377e-06,
+      "loss": 1.1675,
+      "step": 21175
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 19.0368709564209,
+      "learning_rate": 2.2825586746433505e-06,
+      "loss": 0.7819,
+      "step": 21176
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 13.245115280151367,
+      "learning_rate": 2.281721959586663e-06,
+      "loss": 0.7157,
+      "step": 21177
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 21.507944107055664,
+      "learning_rate": 2.2808852445299757e-06,
+      "loss": 1.327,
+      "step": 21178
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 16.07476234436035,
+      "learning_rate": 2.280048529473288e-06,
+      "loss": 0.7543,
+      "step": 21179
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 16.660404205322266,
+      "learning_rate": 2.2792118144166005e-06,
+      "loss": 0.9788,
+      "step": 21180
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 13.21631908416748,
+      "learning_rate": 2.2783750993599133e-06,
+      "loss": 2.407,
+      "step": 21181
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 46.83183288574219,
+      "learning_rate": 2.2775383843032256e-06,
+      "loss": 0.7512,
+      "step": 21182
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 17.08712387084961,
+      "learning_rate": 2.2767016692465384e-06,
+      "loss": 0.6045,
+      "step": 21183
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 2.616086006164551,
+      "learning_rate": 2.275864954189851e-06,
+      "loss": 0.0414,
+      "step": 21184
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 96.06143951416016,
+      "learning_rate": 2.2750282391331636e-06,
+      "loss": 1.3606,
+      "step": 21185
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 16.728219985961914,
+      "learning_rate": 2.274191524076476e-06,
+      "loss": 1.6106,
+      "step": 21186
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 22.86241912841797,
+      "learning_rate": 2.2733548090197884e-06,
+      "loss": 1.0409,
+      "step": 21187
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.369406223297119,
+      "learning_rate": 2.272518093963101e-06,
+      "loss": 0.2209,
+      "step": 21188
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 9.571803092956543,
+      "learning_rate": 2.2716813789064135e-06,
+      "loss": 0.7741,
+      "step": 21189
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 7.1296491622924805,
+      "learning_rate": 2.270844663849726e-06,
+      "loss": 0.6491,
+      "step": 21190
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 9.656634330749512,
+      "learning_rate": 2.2700079487930387e-06,
+      "loss": 0.3507,
+      "step": 21191
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 19.73517417907715,
+      "learning_rate": 2.2691712337363515e-06,
+      "loss": 0.6804,
+      "step": 21192
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 200.69479370117188,
+      "learning_rate": 2.268334518679664e-06,
+      "loss": 1.956,
+      "step": 21193
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 24.948909759521484,
+      "learning_rate": 2.2674978036229763e-06,
+      "loss": 0.6231,
+      "step": 21194
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 12.767256736755371,
+      "learning_rate": 2.266661088566289e-06,
+      "loss": 1.2729,
+      "step": 21195
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 76.8818130493164,
+      "learning_rate": 2.2658243735096015e-06,
+      "loss": 1.7502,
+      "step": 21196
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 13.32066535949707,
+      "learning_rate": 2.264987658452914e-06,
+      "loss": 1.0797,
+      "step": 21197
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 30.614362716674805,
+      "learning_rate": 2.2641509433962266e-06,
+      "loss": 1.2337,
+      "step": 21198
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 22.417505264282227,
+      "learning_rate": 2.2633142283395395e-06,
+      "loss": 1.1799,
+      "step": 21199
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 41.29404830932617,
+      "learning_rate": 2.262477513282852e-06,
+      "loss": 1.8972,
+      "step": 21200
+    },
+    {
+      "epoch": 2.66,
+      "eval_loss": 0.07736419886350632,
+      "eval_runtime": 95.8867,
+      "eval_samples_per_second": 36.939,
+      "eval_steps_per_second": 36.939,
+      "step": 21200
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 10.624398231506348,
+      "learning_rate": 2.261640798226164e-06,
+      "loss": 1.2197,
+      "step": 21201
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.1262216567993164,
+      "learning_rate": 2.260804083169477e-06,
+      "loss": 0.3585,
+      "step": 21202
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 13.62775707244873,
+      "learning_rate": 2.2599673681127894e-06,
+      "loss": 1.4098,
+      "step": 21203
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 26.97039222717285,
+      "learning_rate": 2.2591306530561018e-06,
+      "loss": 1.6124,
+      "step": 21204
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 87.18284606933594,
+      "learning_rate": 2.2582939379994146e-06,
+      "loss": 1.0743,
+      "step": 21205
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 11.793691635131836,
+      "learning_rate": 2.257457222942727e-06,
+      "loss": 0.4472,
+      "step": 21206
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 14.459900856018066,
+      "learning_rate": 2.2566205078860397e-06,
+      "loss": 0.491,
+      "step": 21207
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.120659351348877,
+      "learning_rate": 2.255783792829352e-06,
+      "loss": 1.3269,
+      "step": 21208
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.0259759426116943,
+      "learning_rate": 2.254947077772665e-06,
+      "loss": 0.1151,
+      "step": 21209
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 43.027565002441406,
+      "learning_rate": 2.2541103627159773e-06,
+      "loss": 0.9673,
+      "step": 21210
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 19.61045265197754,
+      "learning_rate": 2.2532736476592897e-06,
+      "loss": 2.2392,
+      "step": 21211
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 21.75444221496582,
+      "learning_rate": 2.2524369326026025e-06,
+      "loss": 0.3609,
+      "step": 21212
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 19.857431411743164,
+      "learning_rate": 2.251600217545915e-06,
+      "loss": 1.7332,
+      "step": 21213
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 7.568485736846924,
+      "learning_rate": 2.2507635024892272e-06,
+      "loss": 0.3944,
+      "step": 21214
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.792768478393555,
+      "learning_rate": 2.24992678743254e-06,
+      "loss": 1.1603,
+      "step": 21215
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 8.036165237426758,
+      "learning_rate": 2.249090072375853e-06,
+      "loss": 0.4277,
+      "step": 21216
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 18.618942260742188,
+      "learning_rate": 2.2482533573191652e-06,
+      "loss": 1.1808,
+      "step": 21217
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 16.693944931030273,
+      "learning_rate": 2.2474166422624776e-06,
+      "loss": 0.7119,
+      "step": 21218
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 76.36405944824219,
+      "learning_rate": 2.2465799272057904e-06,
+      "loss": 2.6666,
+      "step": 21219
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 163.22959899902344,
+      "learning_rate": 2.2457432121491028e-06,
+      "loss": 1.1867,
+      "step": 21220
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 15.051604270935059,
+      "learning_rate": 2.244906497092415e-06,
+      "loss": 1.2664,
+      "step": 21221
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 8.50916862487793,
+      "learning_rate": 2.244069782035728e-06,
+      "loss": 0.6828,
+      "step": 21222
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 11.24258041381836,
+      "learning_rate": 2.2432330669790408e-06,
+      "loss": 0.2977,
+      "step": 21223
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 35.36621856689453,
+      "learning_rate": 2.242396351922353e-06,
+      "loss": 2.1577,
+      "step": 21224
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 6.245741367340088,
+      "learning_rate": 2.2415596368656655e-06,
+      "loss": 0.4049,
+      "step": 21225
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 12.00324821472168,
+      "learning_rate": 2.2407229218089783e-06,
+      "loss": 0.4334,
+      "step": 21226
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 23.8895320892334,
+      "learning_rate": 2.2398862067522907e-06,
+      "loss": 0.9146,
+      "step": 21227
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 8.3075532913208,
+      "learning_rate": 2.239049491695603e-06,
+      "loss": 0.3673,
+      "step": 21228
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 31.701955795288086,
+      "learning_rate": 2.238212776638916e-06,
+      "loss": 2.4295,
+      "step": 21229
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 52.2243537902832,
+      "learning_rate": 2.2373760615822287e-06,
+      "loss": 0.9841,
+      "step": 21230
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 18.255245208740234,
+      "learning_rate": 2.2365393465255406e-06,
+      "loss": 1.609,
+      "step": 21231
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 331.67742919921875,
+      "learning_rate": 2.2357026314688534e-06,
+      "loss": 2.2401,
+      "step": 21232
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 21.084026336669922,
+      "learning_rate": 2.2348659164121662e-06,
+      "loss": 1.3826,
+      "step": 21233
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 41.60992431640625,
+      "learning_rate": 2.2340292013554786e-06,
+      "loss": 0.5638,
+      "step": 21234
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 25.75510597229004,
+      "learning_rate": 2.233192486298791e-06,
+      "loss": 2.0345,
+      "step": 21235
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 9.333986282348633,
+      "learning_rate": 2.232355771242104e-06,
+      "loss": 0.7398,
+      "step": 21236
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 13.220513343811035,
+      "learning_rate": 2.231519056185416e-06,
+      "loss": 0.7039,
+      "step": 21237
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 13.639309883117676,
+      "learning_rate": 2.2306823411287286e-06,
+      "loss": 0.7037,
+      "step": 21238
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 39.554649353027344,
+      "learning_rate": 2.2298456260720414e-06,
+      "loss": 1.9648,
+      "step": 21239
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 8.830999374389648,
+      "learning_rate": 2.229008911015354e-06,
+      "loss": 1.0308,
+      "step": 21240
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.17583703994751,
+      "learning_rate": 2.2281721959586665e-06,
+      "loss": 0.1988,
+      "step": 21241
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 15.095453262329102,
+      "learning_rate": 2.227335480901979e-06,
+      "loss": 0.77,
+      "step": 21242
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 35.32551956176758,
+      "learning_rate": 2.2264987658452917e-06,
+      "loss": 3.1314,
+      "step": 21243
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 18.713911056518555,
+      "learning_rate": 2.225662050788604e-06,
+      "loss": 0.9851,
+      "step": 21244
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 203.83648681640625,
+      "learning_rate": 2.2248253357319165e-06,
+      "loss": 1.8396,
+      "step": 21245
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 14.608433723449707,
+      "learning_rate": 2.2239886206752293e-06,
+      "loss": 1.2817,
+      "step": 21246
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 14.23134994506836,
+      "learning_rate": 2.223151905618542e-06,
+      "loss": 0.4104,
+      "step": 21247
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 10.220736503601074,
+      "learning_rate": 2.222315190561854e-06,
+      "loss": 0.369,
+      "step": 21248
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 26.177026748657227,
+      "learning_rate": 2.221478475505167e-06,
+      "loss": 0.8393,
+      "step": 21249
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 52.15340805053711,
+      "learning_rate": 2.2206417604484796e-06,
+      "loss": 1.0064,
+      "step": 21250
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 20.59659194946289,
+      "learning_rate": 2.219805045391792e-06,
+      "loss": 0.7332,
+      "step": 21251
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 13.785306930541992,
+      "learning_rate": 2.2189683303351044e-06,
+      "loss": 0.6688,
+      "step": 21252
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 16.041227340698242,
+      "learning_rate": 2.218131615278417e-06,
+      "loss": 0.7687,
+      "step": 21253
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 26.35457420349121,
+      "learning_rate": 2.2172949002217296e-06,
+      "loss": 0.5819,
+      "step": 21254
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 24.650827407836914,
+      "learning_rate": 2.216458185165042e-06,
+      "loss": 0.9493,
+      "step": 21255
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 34.303810119628906,
+      "learning_rate": 2.2156214701083548e-06,
+      "loss": 1.2207,
+      "step": 21256
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 208.72035217285156,
+      "learning_rate": 2.2147847550516676e-06,
+      "loss": 1.478,
+      "step": 21257
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 49.18112564086914,
+      "learning_rate": 2.21394803999498e-06,
+      "loss": 0.9865,
+      "step": 21258
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 23.931133270263672,
+      "learning_rate": 2.2131113249382923e-06,
+      "loss": 1.3532,
+      "step": 21259
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.023369312286377,
+      "learning_rate": 2.212274609881605e-06,
+      "loss": 0.4093,
+      "step": 21260
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 29.752965927124023,
+      "learning_rate": 2.2114378948249175e-06,
+      "loss": 0.7628,
+      "step": 21261
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 7.689481258392334,
+      "learning_rate": 2.21060117976823e-06,
+      "loss": 0.2462,
+      "step": 21262
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 27.474773406982422,
+      "learning_rate": 2.2097644647115427e-06,
+      "loss": 0.7475,
+      "step": 21263
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 21.958890914916992,
+      "learning_rate": 2.2089277496548555e-06,
+      "loss": 0.4677,
+      "step": 21264
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 14.977560997009277,
+      "learning_rate": 2.208091034598168e-06,
+      "loss": 0.9122,
+      "step": 21265
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 8.61684513092041,
+      "learning_rate": 2.2072543195414802e-06,
+      "loss": 0.3372,
+      "step": 21266
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 16.10511589050293,
+      "learning_rate": 2.206417604484793e-06,
+      "loss": 0.5668,
+      "step": 21267
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 15.77026653289795,
+      "learning_rate": 2.2055808894281054e-06,
+      "loss": 0.8589,
+      "step": 21268
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 11.518630027770996,
+      "learning_rate": 2.2047441743714178e-06,
+      "loss": 0.2575,
+      "step": 21269
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.5974454879760742,
+      "learning_rate": 2.2039074593147306e-06,
+      "loss": 0.0352,
+      "step": 21270
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 8.891850471496582,
+      "learning_rate": 2.2030707442580434e-06,
+      "loss": 0.8898,
+      "step": 21271
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 23.711355209350586,
+      "learning_rate": 2.2022340292013553e-06,
+      "loss": 0.5634,
+      "step": 21272
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 23.55339813232422,
+      "learning_rate": 2.201397314144668e-06,
+      "loss": 0.7932,
+      "step": 21273
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 24.76299285888672,
+      "learning_rate": 2.200560599087981e-06,
+      "loss": 1.3911,
+      "step": 21274
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 14.951611518859863,
+      "learning_rate": 2.1997238840312933e-06,
+      "loss": 1.411,
+      "step": 21275
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 13.73678207397461,
+      "learning_rate": 2.1988871689746057e-06,
+      "loss": 0.9345,
+      "step": 21276
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 16.367549896240234,
+      "learning_rate": 2.1980504539179185e-06,
+      "loss": 0.5163,
+      "step": 21277
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 11.205177307128906,
+      "learning_rate": 2.197213738861231e-06,
+      "loss": 0.7954,
+      "step": 21278
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 12.95816707611084,
+      "learning_rate": 2.1963770238045433e-06,
+      "loss": 0.5146,
+      "step": 21279
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 11.08674430847168,
+      "learning_rate": 2.195540308747856e-06,
+      "loss": 1.5952,
+      "step": 21280
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 23.338382720947266,
+      "learning_rate": 2.194703593691169e-06,
+      "loss": 1.3663,
+      "step": 21281
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 6.167308330535889,
+      "learning_rate": 2.1938668786344812e-06,
+      "loss": 0.209,
+      "step": 21282
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 33.593170166015625,
+      "learning_rate": 2.1930301635777936e-06,
+      "loss": 0.759,
+      "step": 21283
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 15.196028709411621,
+      "learning_rate": 2.1921934485211064e-06,
+      "loss": 1.0326,
+      "step": 21284
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 13.69805908203125,
+      "learning_rate": 2.191356733464419e-06,
+      "loss": 3.6335,
+      "step": 21285
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 18.835142135620117,
+      "learning_rate": 2.190520018407731e-06,
+      "loss": 0.5714,
+      "step": 21286
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 13.020919799804688,
+      "learning_rate": 2.189683303351044e-06,
+      "loss": 1.5028,
+      "step": 21287
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 38.04112243652344,
+      "learning_rate": 2.1888465882943568e-06,
+      "loss": 3.1631,
+      "step": 21288
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 100.92923736572266,
+      "learning_rate": 2.188009873237669e-06,
+      "loss": 1.2095,
+      "step": 21289
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 155.2234344482422,
+      "learning_rate": 2.1871731581809815e-06,
+      "loss": 1.6139,
+      "step": 21290
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 18.216903686523438,
+      "learning_rate": 2.1863364431242943e-06,
+      "loss": 1.0611,
+      "step": 21291
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 16.728137969970703,
+      "learning_rate": 2.1854997280676067e-06,
+      "loss": 0.6507,
+      "step": 21292
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 11.784873962402344,
+      "learning_rate": 2.184663013010919e-06,
+      "loss": 2.4959,
+      "step": 21293
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 17.049114227294922,
+      "learning_rate": 2.183826297954232e-06,
+      "loss": 0.7656,
+      "step": 21294
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 26.675931930541992,
+      "learning_rate": 2.1829895828975443e-06,
+      "loss": 1.5826,
+      "step": 21295
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 10.614181518554688,
+      "learning_rate": 2.1821528678408567e-06,
+      "loss": 0.8793,
+      "step": 21296
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 11.736658096313477,
+      "learning_rate": 2.1813161527841695e-06,
+      "loss": 0.8583,
+      "step": 21297
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 10.700706481933594,
+      "learning_rate": 2.1804794377274823e-06,
+      "loss": 1.0716,
+      "step": 21298
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 12.090224266052246,
+      "learning_rate": 2.1796427226707946e-06,
+      "loss": 0.6494,
+      "step": 21299
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 87.24824523925781,
+      "learning_rate": 2.178806007614107e-06,
+      "loss": 2.3423,
+      "step": 21300
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 32.511173248291016,
+      "learning_rate": 2.17796929255742e-06,
+      "loss": 1.1127,
+      "step": 21301
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 19.083251953125,
+      "learning_rate": 2.177132577500732e-06,
+      "loss": 0.5119,
+      "step": 21302
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.403653383255005,
+      "learning_rate": 2.1762958624440446e-06,
+      "loss": 0.183,
+      "step": 21303
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 9.263798713684082,
+      "learning_rate": 2.1754591473873574e-06,
+      "loss": 0.5555,
+      "step": 21304
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 30.40312957763672,
+      "learning_rate": 2.17462243233067e-06,
+      "loss": 0.7933,
+      "step": 21305
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 718.2808227539062,
+      "learning_rate": 2.1737857172739826e-06,
+      "loss": 1.9912,
+      "step": 21306
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 10.145508766174316,
+      "learning_rate": 2.172949002217295e-06,
+      "loss": 0.523,
+      "step": 21307
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 7.460542678833008,
+      "learning_rate": 2.1721122871606077e-06,
+      "loss": 0.6282,
+      "step": 21308
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 10.18742847442627,
+      "learning_rate": 2.17127557210392e-06,
+      "loss": 0.4166,
+      "step": 21309
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 14.787338256835938,
+      "learning_rate": 2.1704388570472325e-06,
+      "loss": 0.526,
+      "step": 21310
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 8.858278274536133,
+      "learning_rate": 2.1696021419905453e-06,
+      "loss": 1.3333,
+      "step": 21311
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 6.033418655395508,
+      "learning_rate": 2.1687654269338577e-06,
+      "loss": 0.3257,
+      "step": 21312
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 19.701229095458984,
+      "learning_rate": 2.1679287118771705e-06,
+      "loss": 0.9941,
+      "step": 21313
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 25.754854202270508,
+      "learning_rate": 2.167091996820483e-06,
+      "loss": 1.102,
+      "step": 21314
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 17.089696884155273,
+      "learning_rate": 2.1662552817637957e-06,
+      "loss": 0.6877,
+      "step": 21315
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 47.599430084228516,
+      "learning_rate": 2.165418566707108e-06,
+      "loss": 3.1583,
+      "step": 21316
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 29.585224151611328,
+      "learning_rate": 2.1645818516504204e-06,
+      "loss": 1.3397,
+      "step": 21317
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 27.140111923217773,
+      "learning_rate": 2.1637451365937332e-06,
+      "loss": 2.3643,
+      "step": 21318
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 11.597197532653809,
+      "learning_rate": 2.1629084215370456e-06,
+      "loss": 1.4418,
+      "step": 21319
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.406264781951904,
+      "learning_rate": 2.1620717064803584e-06,
+      "loss": 0.115,
+      "step": 21320
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 70.19218444824219,
+      "learning_rate": 2.1612349914236708e-06,
+      "loss": 1.7635,
+      "step": 21321
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 7.70452356338501,
+      "learning_rate": 2.1603982763669836e-06,
+      "loss": 0.6965,
+      "step": 21322
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 35.344058990478516,
+      "learning_rate": 2.159561561310296e-06,
+      "loss": 0.8861,
+      "step": 21323
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.616994619369507,
+      "learning_rate": 2.1587248462536083e-06,
+      "loss": 0.1458,
+      "step": 21324
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 8.057775497436523,
+      "learning_rate": 2.157888131196921e-06,
+      "loss": 0.5616,
+      "step": 21325
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 192.39141845703125,
+      "learning_rate": 2.1570514161402335e-06,
+      "loss": 1.0416,
+      "step": 21326
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.729534149169922,
+      "learning_rate": 2.1562147010835463e-06,
+      "loss": 0.2792,
+      "step": 21327
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 25.381532669067383,
+      "learning_rate": 2.1553779860268587e-06,
+      "loss": 1.9781,
+      "step": 21328
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 7.036624431610107,
+      "learning_rate": 2.1545412709701715e-06,
+      "loss": 0.547,
+      "step": 21329
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.12929368019104,
+      "learning_rate": 2.153704555913484e-06,
+      "loss": 0.1703,
+      "step": 21330
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 10.610880851745605,
+      "learning_rate": 2.1528678408567963e-06,
+      "loss": 1.7418,
+      "step": 21331
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.1598801612854,
+      "learning_rate": 2.152031125800109e-06,
+      "loss": 0.2733,
+      "step": 21332
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 11.032546997070312,
+      "learning_rate": 2.1511944107434214e-06,
+      "loss": 0.5282,
+      "step": 21333
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 43.82704162597656,
+      "learning_rate": 2.1503576956867342e-06,
+      "loss": 0.9411,
+      "step": 21334
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 6.436817169189453,
+      "learning_rate": 2.1495209806300466e-06,
+      "loss": 0.2611,
+      "step": 21335
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 97.83805847167969,
+      "learning_rate": 2.148684265573359e-06,
+      "loss": 2.1338,
+      "step": 21336
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 11.422772407531738,
+      "learning_rate": 2.147847550516672e-06,
+      "loss": 0.6259,
+      "step": 21337
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 16.130887985229492,
+      "learning_rate": 2.147010835459984e-06,
+      "loss": 0.8244,
+      "step": 21338
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 81.12645721435547,
+      "learning_rate": 2.146174120403297e-06,
+      "loss": 2.4464,
+      "step": 21339
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 35.82655715942383,
+      "learning_rate": 2.1453374053466094e-06,
+      "loss": 0.8004,
+      "step": 21340
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 70.28784942626953,
+      "learning_rate": 2.1445006902899217e-06,
+      "loss": 1.1992,
+      "step": 21341
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 20.804058074951172,
+      "learning_rate": 2.1436639752332345e-06,
+      "loss": 1.8358,
+      "step": 21342
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 87.13873291015625,
+      "learning_rate": 2.142827260176547e-06,
+      "loss": 0.8842,
+      "step": 21343
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 14.349838256835938,
+      "learning_rate": 2.1419905451198597e-06,
+      "loss": 0.8715,
+      "step": 21344
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 8.820053100585938,
+      "learning_rate": 2.141153830063172e-06,
+      "loss": 1.5338,
+      "step": 21345
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 8.869967460632324,
+      "learning_rate": 2.140317115006485e-06,
+      "loss": 0.4037,
+      "step": 21346
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 23.704086303710938,
+      "learning_rate": 2.1394803999497973e-06,
+      "loss": 0.6011,
+      "step": 21347
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 10.375609397888184,
+      "learning_rate": 2.1386436848931096e-06,
+      "loss": 0.5108,
+      "step": 21348
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 22.28101921081543,
+      "learning_rate": 2.1378069698364225e-06,
+      "loss": 1.2317,
+      "step": 21349
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 14.879716873168945,
+      "learning_rate": 2.136970254779735e-06,
+      "loss": 1.5005,
+      "step": 21350
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 8.093547821044922,
+      "learning_rate": 2.1361335397230476e-06,
+      "loss": 0.5505,
+      "step": 21351
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 26.884418487548828,
+      "learning_rate": 2.13529682466636e-06,
+      "loss": 1.0462,
+      "step": 21352
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 50.99755859375,
+      "learning_rate": 2.1344601096096724e-06,
+      "loss": 2.7212,
+      "step": 21353
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.737620830535889,
+      "learning_rate": 2.133623394552985e-06,
+      "loss": 0.0827,
+      "step": 21354
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 182.28158569335938,
+      "learning_rate": 2.1327866794962976e-06,
+      "loss": 2.0251,
+      "step": 21355
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 12.041132926940918,
+      "learning_rate": 2.1319499644396104e-06,
+      "loss": 1.6925,
+      "step": 21356
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 46.95587158203125,
+      "learning_rate": 2.1311132493829227e-06,
+      "loss": 2.1608,
+      "step": 21357
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 12.917547225952148,
+      "learning_rate": 2.1302765343262355e-06,
+      "loss": 1.1931,
+      "step": 21358
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 10.681229591369629,
+      "learning_rate": 2.129439819269548e-06,
+      "loss": 1.2388,
+      "step": 21359
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 20.28715705871582,
+      "learning_rate": 2.1286031042128603e-06,
+      "loss": 0.7494,
+      "step": 21360
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.435354232788086,
+      "learning_rate": 2.127766389156173e-06,
+      "loss": 1.407,
+      "step": 21361
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 21.793575286865234,
+      "learning_rate": 2.1269296740994855e-06,
+      "loss": 0.791,
+      "step": 21362
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 14.35647201538086,
+      "learning_rate": 2.1260929590427983e-06,
+      "loss": 0.7086,
+      "step": 21363
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 7.733112335205078,
+      "learning_rate": 2.1252562439861107e-06,
+      "loss": 0.2449,
+      "step": 21364
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.2984607219696045,
+      "learning_rate": 2.1244195289294235e-06,
+      "loss": 0.2176,
+      "step": 21365
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.907418727874756,
+      "learning_rate": 2.123582813872736e-06,
+      "loss": 0.564,
+      "step": 21366
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 22.646575927734375,
+      "learning_rate": 2.1227460988160482e-06,
+      "loss": 2.2823,
+      "step": 21367
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 32.62713623046875,
+      "learning_rate": 2.121909383759361e-06,
+      "loss": 2.7619,
+      "step": 21368
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.670522689819336,
+      "learning_rate": 2.1210726687026734e-06,
+      "loss": 0.2738,
+      "step": 21369
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.159451007843018,
+      "learning_rate": 2.120235953645986e-06,
+      "loss": 0.1918,
+      "step": 21370
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 18.758514404296875,
+      "learning_rate": 2.1193992385892986e-06,
+      "loss": 1.2502,
+      "step": 21371
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 20.235401153564453,
+      "learning_rate": 2.1185625235326114e-06,
+      "loss": 1.0483,
+      "step": 21372
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 68.90306091308594,
+      "learning_rate": 2.1177258084759238e-06,
+      "loss": 2.1111,
+      "step": 21373
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 8.712653160095215,
+      "learning_rate": 2.116889093419236e-06,
+      "loss": 0.4773,
+      "step": 21374
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 27.22140884399414,
+      "learning_rate": 2.116052378362549e-06,
+      "loss": 2.1803,
+      "step": 21375
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.715273857116699,
+      "learning_rate": 2.1152156633058613e-06,
+      "loss": 0.2097,
+      "step": 21376
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 34.50177764892578,
+      "learning_rate": 2.1143789482491737e-06,
+      "loss": 1.2874,
+      "step": 21377
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 100.45948028564453,
+      "learning_rate": 2.1135422331924865e-06,
+      "loss": 1.0679,
+      "step": 21378
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 13.128931999206543,
+      "learning_rate": 2.1127055181357993e-06,
+      "loss": 0.8483,
+      "step": 21379
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 23.208322525024414,
+      "learning_rate": 2.1118688030791117e-06,
+      "loss": 1.6605,
+      "step": 21380
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 17.86162757873535,
+      "learning_rate": 2.111032088022424e-06,
+      "loss": 0.9252,
+      "step": 21381
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 12.019597053527832,
+      "learning_rate": 2.110195372965737e-06,
+      "loss": 0.685,
+      "step": 21382
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 9.456475257873535,
+      "learning_rate": 2.1093586579090492e-06,
+      "loss": 0.6459,
+      "step": 21383
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 16.749839782714844,
+      "learning_rate": 2.1085219428523616e-06,
+      "loss": 1.6394,
+      "step": 21384
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 728.4263305664062,
+      "learning_rate": 2.1076852277956744e-06,
+      "loss": 1.6383,
+      "step": 21385
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 16.955371856689453,
+      "learning_rate": 2.106848512738987e-06,
+      "loss": 0.7841,
+      "step": 21386
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 35.67090606689453,
+      "learning_rate": 2.1060117976822996e-06,
+      "loss": 2.771,
+      "step": 21387
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 13.471112251281738,
+      "learning_rate": 2.105175082625612e-06,
+      "loss": 1.3309,
+      "step": 21388
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 12.307778358459473,
+      "learning_rate": 2.1043383675689248e-06,
+      "loss": 0.7766,
+      "step": 21389
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 30.031293869018555,
+      "learning_rate": 2.103501652512237e-06,
+      "loss": 1.2038,
+      "step": 21390
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 8.911970138549805,
+      "learning_rate": 2.1026649374555495e-06,
+      "loss": 0.6402,
+      "step": 21391
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 21.951520919799805,
+      "learning_rate": 2.1018282223988623e-06,
+      "loss": 0.8277,
+      "step": 21392
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.777732849121094,
+      "learning_rate": 2.1009915073421747e-06,
+      "loss": 0.3424,
+      "step": 21393
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 17.507043838500977,
+      "learning_rate": 2.100154792285487e-06,
+      "loss": 1.7706,
+      "step": 21394
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 47.57414627075195,
+      "learning_rate": 2.0993180772288e-06,
+      "loss": 3.0551,
+      "step": 21395
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 30.226314544677734,
+      "learning_rate": 2.0984813621721127e-06,
+      "loss": 0.5564,
+      "step": 21396
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 82.02247619628906,
+      "learning_rate": 2.097644647115425e-06,
+      "loss": 1.1771,
+      "step": 21397
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 17.59989356994629,
+      "learning_rate": 2.0968079320587375e-06,
+      "loss": 0.6229,
+      "step": 21398
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 13.957846641540527,
+      "learning_rate": 2.0959712170020503e-06,
+      "loss": 0.3632,
+      "step": 21399
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 7.702903747558594,
+      "learning_rate": 2.0951345019453626e-06,
+      "loss": 0.5295,
+      "step": 21400
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 34.69055938720703,
+      "learning_rate": 2.094297786888675e-06,
+      "loss": 1.3101,
+      "step": 21401
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 22.140445709228516,
+      "learning_rate": 2.093461071831988e-06,
+      "loss": 0.8607,
+      "step": 21402
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 12.767571449279785,
+      "learning_rate": 2.0926243567753006e-06,
+      "loss": 0.5277,
+      "step": 21403
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 6.7225775718688965,
+      "learning_rate": 2.091787641718613e-06,
+      "loss": 0.1126,
+      "step": 21404
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 21.867769241333008,
+      "learning_rate": 2.0909509266619254e-06,
+      "loss": 1.0007,
+      "step": 21405
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.484789848327637,
+      "learning_rate": 2.090114211605238e-06,
+      "loss": 0.2842,
+      "step": 21406
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 10.440987586975098,
+      "learning_rate": 2.0892774965485506e-06,
+      "loss": 0.4244,
+      "step": 21407
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 18.795623779296875,
+      "learning_rate": 2.088440781491863e-06,
+      "loss": 0.6584,
+      "step": 21408
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 14.028873443603516,
+      "learning_rate": 2.0876040664351757e-06,
+      "loss": 0.7828,
+      "step": 21409
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 9.235030174255371,
+      "learning_rate": 2.0867673513784885e-06,
+      "loss": 0.3285,
+      "step": 21410
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 10.508769035339355,
+      "learning_rate": 2.085930636321801e-06,
+      "loss": 0.2151,
+      "step": 21411
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 10.280550003051758,
+      "learning_rate": 2.0850939212651133e-06,
+      "loss": 0.6408,
+      "step": 21412
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 11.279296875,
+      "learning_rate": 2.084257206208426e-06,
+      "loss": 0.49,
+      "step": 21413
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.540591716766357,
+      "learning_rate": 2.0834204911517385e-06,
+      "loss": 0.2417,
+      "step": 21414
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 43.89884948730469,
+      "learning_rate": 2.082583776095051e-06,
+      "loss": 1.1984,
+      "step": 21415
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 11.392531394958496,
+      "learning_rate": 2.0817470610383637e-06,
+      "loss": 0.4732,
+      "step": 21416
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 21.991235733032227,
+      "learning_rate": 2.080910345981676e-06,
+      "loss": 2.2995,
+      "step": 21417
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 12.45938491821289,
+      "learning_rate": 2.0800736309249884e-06,
+      "loss": 0.8814,
+      "step": 21418
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 12.782137870788574,
+      "learning_rate": 2.079236915868301e-06,
+      "loss": 0.8995,
+      "step": 21419
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 23.23676300048828,
+      "learning_rate": 2.078400200811614e-06,
+      "loss": 1.4313,
+      "step": 21420
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 21.540143966674805,
+      "learning_rate": 2.0775634857549264e-06,
+      "loss": 0.9224,
+      "step": 21421
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 14.929125785827637,
+      "learning_rate": 2.0767267706982388e-06,
+      "loss": 1.0599,
+      "step": 21422
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 6.553576469421387,
+      "learning_rate": 2.0758900556415516e-06,
+      "loss": 0.4106,
+      "step": 21423
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 83.32090759277344,
+      "learning_rate": 2.075053340584864e-06,
+      "loss": 2.2362,
+      "step": 21424
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 37.81772994995117,
+      "learning_rate": 2.0742166255281763e-06,
+      "loss": 1.1683,
+      "step": 21425
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 16.711854934692383,
+      "learning_rate": 2.073379910471489e-06,
+      "loss": 0.4657,
+      "step": 21426
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 17.520536422729492,
+      "learning_rate": 2.072543195414802e-06,
+      "loss": 0.4987,
+      "step": 21427
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 21.420486450195312,
+      "learning_rate": 2.0717064803581143e-06,
+      "loss": 1.0809,
+      "step": 21428
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 13.875296592712402,
+      "learning_rate": 2.0708697653014267e-06,
+      "loss": 0.8191,
+      "step": 21429
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 13.885878562927246,
+      "learning_rate": 2.0700330502447395e-06,
+      "loss": 1.0701,
+      "step": 21430
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 14.665055274963379,
+      "learning_rate": 2.069196335188052e-06,
+      "loss": 0.6985,
+      "step": 21431
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 9.924116134643555,
+      "learning_rate": 2.0683596201313642e-06,
+      "loss": 0.5416,
+      "step": 21432
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 38.15719985961914,
+      "learning_rate": 2.067522905074677e-06,
+      "loss": 0.415,
+      "step": 21433
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 6.417823791503906,
+      "learning_rate": 2.06668619001799e-06,
+      "loss": 1.634,
+      "step": 21434
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 14.809370040893555,
+      "learning_rate": 2.065849474961302e-06,
+      "loss": 0.4852,
+      "step": 21435
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 23.902238845825195,
+      "learning_rate": 2.0650127599046146e-06,
+      "loss": 1.0686,
+      "step": 21436
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 12.463006019592285,
+      "learning_rate": 2.0641760448479274e-06,
+      "loss": 0.564,
+      "step": 21437
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.596797943115234,
+      "learning_rate": 2.0633393297912398e-06,
+      "loss": 0.521,
+      "step": 21438
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 19.941085815429688,
+      "learning_rate": 2.062502614734552e-06,
+      "loss": 0.5877,
+      "step": 21439
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 12.362852096557617,
+      "learning_rate": 2.061665899677865e-06,
+      "loss": 0.7027,
+      "step": 21440
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 32.76074981689453,
+      "learning_rate": 2.0608291846211773e-06,
+      "loss": 0.7459,
+      "step": 21441
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 7.287831783294678,
+      "learning_rate": 2.0599924695644897e-06,
+      "loss": 0.3511,
+      "step": 21442
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.938597679138184,
+      "learning_rate": 2.0591557545078025e-06,
+      "loss": 0.5302,
+      "step": 21443
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 7.694718837738037,
+      "learning_rate": 2.0583190394511153e-06,
+      "loss": 0.1274,
+      "step": 21444
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 21.389583587646484,
+      "learning_rate": 2.0574823243944277e-06,
+      "loss": 1.3318,
+      "step": 21445
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 16.10489845275879,
+      "learning_rate": 2.05664560933774e-06,
+      "loss": 1.4647,
+      "step": 21446
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 19.475576400756836,
+      "learning_rate": 2.055808894281053e-06,
+      "loss": 0.9934,
+      "step": 21447
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.483088970184326,
+      "learning_rate": 2.0549721792243653e-06,
+      "loss": 0.0673,
+      "step": 21448
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 9.39845085144043,
+      "learning_rate": 2.0541354641676776e-06,
+      "loss": 0.52,
+      "step": 21449
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 24.6356201171875,
+      "learning_rate": 2.0532987491109904e-06,
+      "loss": 1.1128,
+      "step": 21450
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 7.2979512214660645,
+      "learning_rate": 2.0524620340543032e-06,
+      "loss": 0.5031,
+      "step": 21451
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 7.22651481628418,
+      "learning_rate": 2.051625318997615e-06,
+      "loss": 0.4692,
+      "step": 21452
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 13.737506866455078,
+      "learning_rate": 2.050788603940928e-06,
+      "loss": 1.4457,
+      "step": 21453
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 11.899345397949219,
+      "learning_rate": 2.049951888884241e-06,
+      "loss": 0.4482,
+      "step": 21454
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 14.08261489868164,
+      "learning_rate": 2.049115173827553e-06,
+      "loss": 0.4664,
+      "step": 21455
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 15.5072021484375,
+      "learning_rate": 2.0482784587708656e-06,
+      "loss": 0.3376,
+      "step": 21456
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 14.26341724395752,
+      "learning_rate": 2.0474417437141784e-06,
+      "loss": 0.8661,
+      "step": 21457
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 11.781574249267578,
+      "learning_rate": 2.0466050286574907e-06,
+      "loss": 2.138,
+      "step": 21458
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 21.719669342041016,
+      "learning_rate": 2.045768313600803e-06,
+      "loss": 0.4253,
+      "step": 21459
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 9.645719528198242,
+      "learning_rate": 2.044931598544116e-06,
+      "loss": 0.5508,
+      "step": 21460
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.674021244049072,
+      "learning_rate": 2.0440948834874287e-06,
+      "loss": 0.0924,
+      "step": 21461
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 15.2584228515625,
+      "learning_rate": 2.043258168430741e-06,
+      "loss": 1.799,
+      "step": 21462
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.222739219665527,
+      "learning_rate": 2.0424214533740535e-06,
+      "loss": 0.2849,
+      "step": 21463
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 25.117767333984375,
+      "learning_rate": 2.0415847383173663e-06,
+      "loss": 1.5771,
+      "step": 21464
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 24.659822463989258,
+      "learning_rate": 2.0407480232606787e-06,
+      "loss": 0.7105,
+      "step": 21465
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 64.56285095214844,
+      "learning_rate": 2.039911308203991e-06,
+      "loss": 2.2335,
+      "step": 21466
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 7.914119243621826,
+      "learning_rate": 2.039074593147304e-06,
+      "loss": 0.6704,
+      "step": 21467
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 19.407703399658203,
+      "learning_rate": 2.0382378780906166e-06,
+      "loss": 1.5496,
+      "step": 21468
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 106.05070495605469,
+      "learning_rate": 2.037401163033929e-06,
+      "loss": 1.4463,
+      "step": 21469
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 7.93324089050293,
+      "learning_rate": 2.0365644479772414e-06,
+      "loss": 0.5607,
+      "step": 21470
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 18.987442016601562,
+      "learning_rate": 2.035727732920554e-06,
+      "loss": 0.6445,
+      "step": 21471
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 36.77301788330078,
+      "learning_rate": 2.0348910178638666e-06,
+      "loss": 2.6085,
+      "step": 21472
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 21.309961318969727,
+      "learning_rate": 2.034054302807179e-06,
+      "loss": 1.2199,
+      "step": 21473
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 22.824777603149414,
+      "learning_rate": 2.0332175877504918e-06,
+      "loss": 2.7914,
+      "step": 21474
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 28.599380493164062,
+      "learning_rate": 2.0323808726938046e-06,
+      "loss": 0.6091,
+      "step": 21475
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 13.5480375289917,
+      "learning_rate": 2.0315441576371165e-06,
+      "loss": 1.1295,
+      "step": 21476
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 7.5399956703186035,
+      "learning_rate": 2.0307074425804293e-06,
+      "loss": 0.3935,
+      "step": 21477
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 12.060484886169434,
+      "learning_rate": 2.029870727523742e-06,
+      "loss": 0.8903,
+      "step": 21478
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 14.38305950164795,
+      "learning_rate": 2.0290340124670545e-06,
+      "loss": 1.5363,
+      "step": 21479
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 22.1859130859375,
+      "learning_rate": 2.028197297410367e-06,
+      "loss": 1.9556,
+      "step": 21480
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 15.168512344360352,
+      "learning_rate": 2.0273605823536797e-06,
+      "loss": 0.8704,
+      "step": 21481
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 12.938074111938477,
+      "learning_rate": 2.026523867296992e-06,
+      "loss": 0.9819,
+      "step": 21482
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 15.477502822875977,
+      "learning_rate": 2.0256871522403044e-06,
+      "loss": 0.4481,
+      "step": 21483
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 21.59099578857422,
+      "learning_rate": 2.0248504371836172e-06,
+      "loss": 2.4869,
+      "step": 21484
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 28.2064151763916,
+      "learning_rate": 2.02401372212693e-06,
+      "loss": 1.1286,
+      "step": 21485
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 10.416912078857422,
+      "learning_rate": 2.0231770070702424e-06,
+      "loss": 0.4991,
+      "step": 21486
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 11.99328899383545,
+      "learning_rate": 2.022340292013555e-06,
+      "loss": 1.4455,
+      "step": 21487
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 82.79418182373047,
+      "learning_rate": 2.0215035769568676e-06,
+      "loss": 2.1569,
+      "step": 21488
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 17.955158233642578,
+      "learning_rate": 2.02066686190018e-06,
+      "loss": 2.2086,
+      "step": 21489
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 6.074029922485352,
+      "learning_rate": 2.0198301468434924e-06,
+      "loss": 0.1896,
+      "step": 21490
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 18.251638412475586,
+      "learning_rate": 2.018993431786805e-06,
+      "loss": 1.5237,
+      "step": 21491
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 10.850665092468262,
+      "learning_rate": 2.018156716730118e-06,
+      "loss": 0.7379,
+      "step": 21492
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 16.91484260559082,
+      "learning_rate": 2.0173200016734303e-06,
+      "loss": 0.9855,
+      "step": 21493
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 22.841798782348633,
+      "learning_rate": 2.0164832866167427e-06,
+      "loss": 0.374,
+      "step": 21494
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 43.07855224609375,
+      "learning_rate": 2.0156465715600555e-06,
+      "loss": 1.9408,
+      "step": 21495
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 14.059098243713379,
+      "learning_rate": 2.014809856503368e-06,
+      "loss": 0.7753,
+      "step": 21496
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.776729583740234,
+      "learning_rate": 2.0139731414466803e-06,
+      "loss": 0.1441,
+      "step": 21497
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 9.423643112182617,
+      "learning_rate": 2.013136426389993e-06,
+      "loss": 0.2007,
+      "step": 21498
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 14.687823295593262,
+      "learning_rate": 2.0122997113333055e-06,
+      "loss": 1.5296,
+      "step": 21499
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 15.722042083740234,
+      "learning_rate": 2.0114629962766183e-06,
+      "loss": 1.5207,
+      "step": 21500
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 17.29703140258789,
+      "learning_rate": 2.0106262812199306e-06,
+      "loss": 1.5851,
+      "step": 21501
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 16.280162811279297,
+      "learning_rate": 2.0097895661632434e-06,
+      "loss": 0.9692,
+      "step": 21502
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 14.40085506439209,
+      "learning_rate": 2.008952851106556e-06,
+      "loss": 1.2911,
+      "step": 21503
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.068446159362793,
+      "learning_rate": 2.008116136049868e-06,
+      "loss": 1.4361,
+      "step": 21504
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 11.074398040771484,
+      "learning_rate": 2.007279420993181e-06,
+      "loss": 0.4624,
+      "step": 21505
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 136.94422912597656,
+      "learning_rate": 2.0064427059364934e-06,
+      "loss": 2.9412,
+      "step": 21506
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.199703216552734,
+      "learning_rate": 2.005605990879806e-06,
+      "loss": 1.524,
+      "step": 21507
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 26.445039749145508,
+      "learning_rate": 2.0047692758231185e-06,
+      "loss": 1.0059,
+      "step": 21508
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 19.995515823364258,
+      "learning_rate": 2.0039325607664314e-06,
+      "loss": 1.796,
+      "step": 21509
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 136.7841796875,
+      "learning_rate": 2.0030958457097437e-06,
+      "loss": 1.4174,
+      "step": 21510
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 26.809480667114258,
+      "learning_rate": 2.002259130653056e-06,
+      "loss": 0.6017,
+      "step": 21511
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 21.66718292236328,
+      "learning_rate": 2.001422415596369e-06,
+      "loss": 1.6301,
+      "step": 21512
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 47.82748794555664,
+      "learning_rate": 2.0005857005396813e-06,
+      "loss": 1.1217,
+      "step": 21513
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 32.09178161621094,
+      "learning_rate": 1.999748985482994e-06,
+      "loss": 1.5899,
+      "step": 21514
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 17.64548110961914,
+      "learning_rate": 1.9989122704263065e-06,
+      "loss": 1.1916,
+      "step": 21515
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 13.294170379638672,
+      "learning_rate": 1.9980755553696193e-06,
+      "loss": 0.4883,
+      "step": 21516
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 32.86570739746094,
+      "learning_rate": 1.9972388403129316e-06,
+      "loss": 1.4325,
+      "step": 21517
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 15.223159790039062,
+      "learning_rate": 1.996402125256244e-06,
+      "loss": 0.4239,
+      "step": 21518
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 6.6393938064575195,
+      "learning_rate": 1.995565410199557e-06,
+      "loss": 0.4592,
+      "step": 21519
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 9.456892967224121,
+      "learning_rate": 1.994728695142869e-06,
+      "loss": 0.8297,
+      "step": 21520
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 35.015174865722656,
+      "learning_rate": 1.9938919800861816e-06,
+      "loss": 1.4686,
+      "step": 21521
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 15.150407791137695,
+      "learning_rate": 1.9930552650294944e-06,
+      "loss": 0.7951,
+      "step": 21522
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 175.0820770263672,
+      "learning_rate": 1.9922185499728068e-06,
+      "loss": 1.808,
+      "step": 21523
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 7.424928188323975,
+      "learning_rate": 1.9913818349161196e-06,
+      "loss": 0.266,
+      "step": 21524
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 298.6964111328125,
+      "learning_rate": 1.990545119859432e-06,
+      "loss": 0.8698,
+      "step": 21525
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 35.207061767578125,
+      "learning_rate": 1.9897084048027447e-06,
+      "loss": 1.2134,
+      "step": 21526
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.058426856994629,
+      "learning_rate": 1.988871689746057e-06,
+      "loss": 0.4285,
+      "step": 21527
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 13.8907470703125,
+      "learning_rate": 1.9880349746893695e-06,
+      "loss": 0.6789,
+      "step": 21528
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 27.26032829284668,
+      "learning_rate": 1.9871982596326823e-06,
+      "loss": 0.8729,
+      "step": 21529
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 11.517545700073242,
+      "learning_rate": 1.9863615445759947e-06,
+      "loss": 1.5727,
+      "step": 21530
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 38.588294982910156,
+      "learning_rate": 1.9855248295193075e-06,
+      "loss": 2.6482,
+      "step": 21531
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 10.834522247314453,
+      "learning_rate": 1.98468811446262e-06,
+      "loss": 1.2162,
+      "step": 21532
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 15.166847229003906,
+      "learning_rate": 1.9838513994059327e-06,
+      "loss": 0.7046,
+      "step": 21533
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.857996463775635,
+      "learning_rate": 1.983014684349245e-06,
+      "loss": 0.5158,
+      "step": 21534
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 6.261816024780273,
+      "learning_rate": 1.9821779692925574e-06,
+      "loss": 0.4164,
+      "step": 21535
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 11.289036750793457,
+      "learning_rate": 1.9813412542358702e-06,
+      "loss": 0.5289,
+      "step": 21536
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 12.965081214904785,
+      "learning_rate": 1.9805045391791826e-06,
+      "loss": 1.0809,
+      "step": 21537
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 12.697193145751953,
+      "learning_rate": 1.9796678241224954e-06,
+      "loss": 1.7587,
+      "step": 21538
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 13.06324577331543,
+      "learning_rate": 1.9788311090658078e-06,
+      "loss": 1.1945,
+      "step": 21539
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 89.4460220336914,
+      "learning_rate": 1.97799439400912e-06,
+      "loss": 3.2284,
+      "step": 21540
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.918099403381348,
+      "learning_rate": 1.977157678952433e-06,
+      "loss": 0.346,
+      "step": 21541
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 24.16585922241211,
+      "learning_rate": 1.9763209638957453e-06,
+      "loss": 0.6387,
+      "step": 21542
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 34.83650588989258,
+      "learning_rate": 1.975484248839058e-06,
+      "loss": 1.1662,
+      "step": 21543
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 30.044612884521484,
+      "learning_rate": 1.9746475337823705e-06,
+      "loss": 1.5414,
+      "step": 21544
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 13.216431617736816,
+      "learning_rate": 1.9738108187256833e-06,
+      "loss": 1.3495,
+      "step": 21545
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.502356052398682,
+      "learning_rate": 1.9729741036689957e-06,
+      "loss": 0.4508,
+      "step": 21546
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 20.345603942871094,
+      "learning_rate": 1.972137388612308e-06,
+      "loss": 1.4439,
+      "step": 21547
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 61.608367919921875,
+      "learning_rate": 1.971300673555621e-06,
+      "loss": 0.8592,
+      "step": 21548
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 13.994182586669922,
+      "learning_rate": 1.9704639584989333e-06,
+      "loss": 0.5676,
+      "step": 21549
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 8.34164047241211,
+      "learning_rate": 1.969627243442246e-06,
+      "loss": 0.3883,
+      "step": 21550
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 19.863996505737305,
+      "learning_rate": 1.9687905283855584e-06,
+      "loss": 0.8862,
+      "step": 21551
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.623724937438965,
+      "learning_rate": 1.9679538133288712e-06,
+      "loss": 0.2411,
+      "step": 21552
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 11.14766788482666,
+      "learning_rate": 1.9671170982721836e-06,
+      "loss": 0.7617,
+      "step": 21553
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 104.04212188720703,
+      "learning_rate": 1.966280383215496e-06,
+      "loss": 0.7672,
+      "step": 21554
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 17.19343376159668,
+      "learning_rate": 1.965443668158809e-06,
+      "loss": 1.7867,
+      "step": 21555
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 18.00482177734375,
+      "learning_rate": 1.964606953102121e-06,
+      "loss": 1.5705,
+      "step": 21556
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 28.12812614440918,
+      "learning_rate": 1.9637702380454336e-06,
+      "loss": 1.6228,
+      "step": 21557
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 15.558965682983398,
+      "learning_rate": 1.9629335229887464e-06,
+      "loss": 0.7861,
+      "step": 21558
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 12.45241928100586,
+      "learning_rate": 1.962096807932059e-06,
+      "loss": 1.613,
+      "step": 21559
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 7.432572841644287,
+      "learning_rate": 1.9612600928753715e-06,
+      "loss": 0.3244,
+      "step": 21560
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 9.60072135925293,
+      "learning_rate": 1.960423377818684e-06,
+      "loss": 0.4073,
+      "step": 21561
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 44.53190231323242,
+      "learning_rate": 1.9595866627619967e-06,
+      "loss": 2.8229,
+      "step": 21562
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 13.338286399841309,
+      "learning_rate": 1.958749947705309e-06,
+      "loss": 1.023,
+      "step": 21563
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 13.374138832092285,
+      "learning_rate": 1.9579132326486215e-06,
+      "loss": 1.9172,
+      "step": 21564
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 98.44874572753906,
+      "learning_rate": 1.9570765175919343e-06,
+      "loss": 0.9126,
+      "step": 21565
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 17.078516006469727,
+      "learning_rate": 1.9562398025352467e-06,
+      "loss": 1.1122,
+      "step": 21566
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 21.637590408325195,
+      "learning_rate": 1.9554030874785595e-06,
+      "loss": 1.1725,
+      "step": 21567
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 12.546353340148926,
+      "learning_rate": 1.954566372421872e-06,
+      "loss": 0.5145,
+      "step": 21568
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 52.77216720581055,
+      "learning_rate": 1.9537296573651846e-06,
+      "loss": 0.7235,
+      "step": 21569
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 18.392202377319336,
+      "learning_rate": 1.952892942308497e-06,
+      "loss": 0.8054,
+      "step": 21570
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 21.224721908569336,
+      "learning_rate": 1.9520562272518094e-06,
+      "loss": 1.5617,
+      "step": 21571
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 20.848115921020508,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 0.674,
+      "step": 21572
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 13.857409477233887,
+      "learning_rate": 1.9503827971384346e-06,
+      "loss": 1.3511,
+      "step": 21573
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 10.296735763549805,
+      "learning_rate": 1.9495460820817474e-06,
+      "loss": 1.0816,
+      "step": 21574
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 13.349283218383789,
+      "learning_rate": 1.9487093670250598e-06,
+      "loss": 1.3952,
+      "step": 21575
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.422610759735107,
+      "learning_rate": 1.9478726519683726e-06,
+      "loss": 0.6176,
+      "step": 21576
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 9.50144100189209,
+      "learning_rate": 1.947035936911685e-06,
+      "loss": 0.3522,
+      "step": 21577
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 15.972548484802246,
+      "learning_rate": 1.9461992218549973e-06,
+      "loss": 0.8747,
+      "step": 21578
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 16.36379051208496,
+      "learning_rate": 1.94536250679831e-06,
+      "loss": 1.5866,
+      "step": 21579
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 18.411258697509766,
+      "learning_rate": 1.9445257917416225e-06,
+      "loss": 1.6996,
+      "step": 21580
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 6.302036285400391,
+      "learning_rate": 1.943689076684935e-06,
+      "loss": 0.6719,
+      "step": 21581
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 6.9534406661987305,
+      "learning_rate": 1.9428523616282477e-06,
+      "loss": 0.2405,
+      "step": 21582
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 13.45012378692627,
+      "learning_rate": 1.9420156465715605e-06,
+      "loss": 1.1696,
+      "step": 21583
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.985976219177246,
+      "learning_rate": 1.941178931514873e-06,
+      "loss": 0.1027,
+      "step": 21584
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 19.764118194580078,
+      "learning_rate": 1.9403422164581852e-06,
+      "loss": 2.5148,
+      "step": 21585
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 15.650075912475586,
+      "learning_rate": 1.939505501401498e-06,
+      "loss": 1.9206,
+      "step": 21586
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 7.171905994415283,
+      "learning_rate": 1.9386687863448104e-06,
+      "loss": 0.5164,
+      "step": 21587
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 10.496377944946289,
+      "learning_rate": 1.9378320712881228e-06,
+      "loss": 1.1155,
+      "step": 21588
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 30.06625747680664,
+      "learning_rate": 1.9369953562314356e-06,
+      "loss": 0.9575,
+      "step": 21589
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 13.560931205749512,
+      "learning_rate": 1.9361586411747484e-06,
+      "loss": 0.7953,
+      "step": 21590
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 21.100141525268555,
+      "learning_rate": 1.9353219261180608e-06,
+      "loss": 1.099,
+      "step": 21591
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 11.04590129852295,
+      "learning_rate": 1.934485211061373e-06,
+      "loss": 0.6509,
+      "step": 21592
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 33.39369583129883,
+      "learning_rate": 1.933648496004686e-06,
+      "loss": 1.4008,
+      "step": 21593
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.745856285095215,
+      "learning_rate": 1.9328117809479983e-06,
+      "loss": 0.3466,
+      "step": 21594
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 10.121007919311523,
+      "learning_rate": 1.9319750658913107e-06,
+      "loss": 0.5584,
+      "step": 21595
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 35.45912551879883,
+      "learning_rate": 1.9311383508346235e-06,
+      "loss": 0.7605,
+      "step": 21596
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 10.28868579864502,
+      "learning_rate": 1.9303016357779363e-06,
+      "loss": 0.5701,
+      "step": 21597
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 34.35209274291992,
+      "learning_rate": 1.9294649207212483e-06,
+      "loss": 0.795,
+      "step": 21598
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 33.060420989990234,
+      "learning_rate": 1.928628205664561e-06,
+      "loss": 2.3566,
+      "step": 21599
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 91.64486694335938,
+      "learning_rate": 1.927791490607874e-06,
+      "loss": 1.8895,
+      "step": 21600
+    },
+    {
+      "epoch": 2.71,
+      "eval_loss": 0.07624088227748871,
+      "eval_runtime": 95.1929,
+      "eval_samples_per_second": 37.209,
+      "eval_steps_per_second": 37.209,
+      "step": 21600
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 8.517383575439453,
+      "learning_rate": 1.9269547755511862e-06,
+      "loss": 1.8898,
+      "step": 21601
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 6.557233810424805,
+      "learning_rate": 1.9261180604944986e-06,
+      "loss": 0.0612,
+      "step": 21602
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 18.834272384643555,
+      "learning_rate": 1.9252813454378114e-06,
+      "loss": 1.7287,
+      "step": 21603
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 14.700809478759766,
+      "learning_rate": 1.924444630381124e-06,
+      "loss": 0.6263,
+      "step": 21604
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 8.74665641784668,
+      "learning_rate": 1.923607915324436e-06,
+      "loss": 0.3854,
+      "step": 21605
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 20.223066329956055,
+      "learning_rate": 1.922771200267749e-06,
+      "loss": 1.1869,
+      "step": 21606
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 26.129592895507812,
+      "learning_rate": 1.9219344852110618e-06,
+      "loss": 0.783,
+      "step": 21607
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 6.939517498016357,
+      "learning_rate": 1.921097770154374e-06,
+      "loss": 0.3401,
+      "step": 21608
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 15.920483589172363,
+      "learning_rate": 1.9202610550976865e-06,
+      "loss": 2.7957,
+      "step": 21609
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 11.822491645812988,
+      "learning_rate": 1.9194243400409993e-06,
+      "loss": 1.9297,
+      "step": 21610
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 8.13525104522705,
+      "learning_rate": 1.9185876249843117e-06,
+      "loss": 0.1707,
+      "step": 21611
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 15.861075401306152,
+      "learning_rate": 1.917750909927624e-06,
+      "loss": 0.7218,
+      "step": 21612
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 14.167207717895508,
+      "learning_rate": 1.916914194870937e-06,
+      "loss": 0.354,
+      "step": 21613
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 7.59569787979126,
+      "learning_rate": 1.9160774798142497e-06,
+      "loss": 0.1944,
+      "step": 21614
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 29.29405975341797,
+      "learning_rate": 1.915240764757562e-06,
+      "loss": 0.7861,
+      "step": 21615
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 19.714954376220703,
+      "learning_rate": 1.9144040497008745e-06,
+      "loss": 0.9029,
+      "step": 21616
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 9.297428131103516,
+      "learning_rate": 1.9135673346441873e-06,
+      "loss": 0.291,
+      "step": 21617
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 79.9549560546875,
+      "learning_rate": 1.9127306195874996e-06,
+      "loss": 2.0887,
+      "step": 21618
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 21.961437225341797,
+      "learning_rate": 1.911893904530812e-06,
+      "loss": 0.6198,
+      "step": 21619
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.938629627227783,
+      "learning_rate": 1.911057189474125e-06,
+      "loss": 0.0518,
+      "step": 21620
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 99.69220733642578,
+      "learning_rate": 1.910220474417437e-06,
+      "loss": 1.0809,
+      "step": 21621
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 26.269498825073242,
+      "learning_rate": 1.9093837593607496e-06,
+      "loss": 1.1937,
+      "step": 21622
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 14.736011505126953,
+      "learning_rate": 1.9085470443040624e-06,
+      "loss": 0.4535,
+      "step": 21623
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 21.275299072265625,
+      "learning_rate": 1.907710329247375e-06,
+      "loss": 0.8934,
+      "step": 21624
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 8.874500274658203,
+      "learning_rate": 1.9068736141906873e-06,
+      "loss": 0.8592,
+      "step": 21625
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 2.27887225151062,
+      "learning_rate": 1.906036899134e-06,
+      "loss": 0.0699,
+      "step": 21626
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 132.70150756835938,
+      "learning_rate": 1.9052001840773127e-06,
+      "loss": 1.296,
+      "step": 21627
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.818041801452637,
+      "learning_rate": 1.9043634690206253e-06,
+      "loss": 0.1964,
+      "step": 21628
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 79.94800567626953,
+      "learning_rate": 1.9035267539639377e-06,
+      "loss": 0.5249,
+      "step": 21629
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.631694793701172,
+      "learning_rate": 1.9026900389072503e-06,
+      "loss": 0.2513,
+      "step": 21630
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 11.695740699768066,
+      "learning_rate": 1.9018533238505629e-06,
+      "loss": 1.6826,
+      "step": 21631
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 9.75593376159668,
+      "learning_rate": 1.9010166087938753e-06,
+      "loss": 0.4342,
+      "step": 21632
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 30.43581771850586,
+      "learning_rate": 1.9001798937371879e-06,
+      "loss": 0.7867,
+      "step": 21633
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 34.77962112426758,
+      "learning_rate": 1.8993431786805007e-06,
+      "loss": 1.3018,
+      "step": 21634
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.737854480743408,
+      "learning_rate": 1.8985064636238132e-06,
+      "loss": 1.0135,
+      "step": 21635
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 17.90479850769043,
+      "learning_rate": 1.8976697485671256e-06,
+      "loss": 1.1262,
+      "step": 21636
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.772305011749268,
+      "learning_rate": 1.8968330335104382e-06,
+      "loss": 0.7722,
+      "step": 21637
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 20.269916534423828,
+      "learning_rate": 1.8959963184537508e-06,
+      "loss": 1.4899,
+      "step": 21638
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 7.383659362792969,
+      "learning_rate": 1.8951596033970632e-06,
+      "loss": 0.3814,
+      "step": 21639
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 13.869362831115723,
+      "learning_rate": 1.8943228883403758e-06,
+      "loss": 1.1662,
+      "step": 21640
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 9.787107467651367,
+      "learning_rate": 1.8934861732836884e-06,
+      "loss": 1.9305,
+      "step": 21641
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 16.482925415039062,
+      "learning_rate": 1.8926494582270012e-06,
+      "loss": 0.3837,
+      "step": 21642
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 25.612733840942383,
+      "learning_rate": 1.8918127431703133e-06,
+      "loss": 0.7088,
+      "step": 21643
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 22.032100677490234,
+      "learning_rate": 1.8909760281136261e-06,
+      "loss": 2.0197,
+      "step": 21644
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 24.02498435974121,
+      "learning_rate": 1.8901393130569387e-06,
+      "loss": 1.9074,
+      "step": 21645
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 30.454744338989258,
+      "learning_rate": 1.889302598000251e-06,
+      "loss": 0.9754,
+      "step": 21646
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 82.28011322021484,
+      "learning_rate": 1.8884658829435637e-06,
+      "loss": 1.1165,
+      "step": 21647
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.8160178661346436,
+      "learning_rate": 1.8876291678868763e-06,
+      "loss": 0.8781,
+      "step": 21648
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 29.909202575683594,
+      "learning_rate": 1.8867924528301889e-06,
+      "loss": 0.7232,
+      "step": 21649
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 6.371339321136475,
+      "learning_rate": 1.8859557377735013e-06,
+      "loss": 0.2522,
+      "step": 21650
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 20.851184844970703,
+      "learning_rate": 1.885119022716814e-06,
+      "loss": 0.6746,
+      "step": 21651
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 35.381099700927734,
+      "learning_rate": 1.8842823076601266e-06,
+      "loss": 0.7025,
+      "step": 21652
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 9.720460891723633,
+      "learning_rate": 1.883445592603439e-06,
+      "loss": 0.7627,
+      "step": 21653
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 25.254837036132812,
+      "learning_rate": 1.8826088775467516e-06,
+      "loss": 0.7075,
+      "step": 21654
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 14.657265663146973,
+      "learning_rate": 1.8817721624900642e-06,
+      "loss": 0.5678,
+      "step": 21655
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 12.222607612609863,
+      "learning_rate": 1.8809354474333766e-06,
+      "loss": 1.016,
+      "step": 21656
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 11.868119239807129,
+      "learning_rate": 1.8800987323766892e-06,
+      "loss": 1.553,
+      "step": 21657
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 35.980133056640625,
+      "learning_rate": 1.8792620173200018e-06,
+      "loss": 1.4554,
+      "step": 21658
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 9.9580659866333,
+      "learning_rate": 1.8784253022633146e-06,
+      "loss": 0.472,
+      "step": 21659
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 8.805882453918457,
+      "learning_rate": 1.877588587206627e-06,
+      "loss": 0.2537,
+      "step": 21660
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 10.853120803833008,
+      "learning_rate": 1.8767518721499395e-06,
+      "loss": 0.534,
+      "step": 21661
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 14.03399658203125,
+      "learning_rate": 1.8759151570932521e-06,
+      "loss": 1.2223,
+      "step": 21662
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 7.437746047973633,
+      "learning_rate": 1.8750784420365645e-06,
+      "loss": 0.3798,
+      "step": 21663
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 11.152474403381348,
+      "learning_rate": 1.874241726979877e-06,
+      "loss": 0.5058,
+      "step": 21664
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 14.955697059631348,
+      "learning_rate": 1.8734050119231897e-06,
+      "loss": 1.0318,
+      "step": 21665
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 19.80023765563965,
+      "learning_rate": 1.8725682968665025e-06,
+      "loss": 0.8431,
+      "step": 21666
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 24.065574645996094,
+      "learning_rate": 1.8717315818098146e-06,
+      "loss": 0.8752,
+      "step": 21667
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 21.037424087524414,
+      "learning_rate": 1.8708948667531274e-06,
+      "loss": 1.5721,
+      "step": 21668
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 7.22127628326416,
+      "learning_rate": 1.87005815169644e-06,
+      "loss": 0.6915,
+      "step": 21669
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 30.11231803894043,
+      "learning_rate": 1.8692214366397524e-06,
+      "loss": 1.088,
+      "step": 21670
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 11.352420806884766,
+      "learning_rate": 1.868384721583065e-06,
+      "loss": 0.6094,
+      "step": 21671
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 12.00910758972168,
+      "learning_rate": 1.8675480065263776e-06,
+      "loss": 0.8504,
+      "step": 21672
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 13.739583015441895,
+      "learning_rate": 1.8667112914696902e-06,
+      "loss": 1.2051,
+      "step": 21673
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 7.9926676750183105,
+      "learning_rate": 1.8658745764130026e-06,
+      "loss": 0.7858,
+      "step": 21674
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 20.544986724853516,
+      "learning_rate": 1.8650378613563152e-06,
+      "loss": 1.1847,
+      "step": 21675
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 19.822420120239258,
+      "learning_rate": 1.864201146299628e-06,
+      "loss": 0.983,
+      "step": 21676
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 8.64639663696289,
+      "learning_rate": 1.8633644312429403e-06,
+      "loss": 0.7687,
+      "step": 21677
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 10.299410820007324,
+      "learning_rate": 1.862527716186253e-06,
+      "loss": 0.4984,
+      "step": 21678
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 9.877063751220703,
+      "learning_rate": 1.8616910011295655e-06,
+      "loss": 0.4443,
+      "step": 21679
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 6.195727348327637,
+      "learning_rate": 1.8608542860728781e-06,
+      "loss": 0.7731,
+      "step": 21680
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 25.56712532043457,
+      "learning_rate": 1.8600175710161905e-06,
+      "loss": 1.5167,
+      "step": 21681
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 13.832436561584473,
+      "learning_rate": 1.859180855959503e-06,
+      "loss": 1.0589,
+      "step": 21682
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 21.81914710998535,
+      "learning_rate": 1.8583441409028159e-06,
+      "loss": 0.9787,
+      "step": 21683
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 11.977666854858398,
+      "learning_rate": 1.857507425846128e-06,
+      "loss": 0.8455,
+      "step": 21684
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 28.767717361450195,
+      "learning_rate": 1.8566707107894408e-06,
+      "loss": 1.0714,
+      "step": 21685
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 40.01023483276367,
+      "learning_rate": 1.8558339957327534e-06,
+      "loss": 0.736,
+      "step": 21686
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 30.824186325073242,
+      "learning_rate": 1.854997280676066e-06,
+      "loss": 2.5069,
+      "step": 21687
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 18.87152862548828,
+      "learning_rate": 1.8541605656193784e-06,
+      "loss": 0.9405,
+      "step": 21688
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 35.233551025390625,
+      "learning_rate": 1.853323850562691e-06,
+      "loss": 1.3706,
+      "step": 21689
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 48.823116302490234,
+      "learning_rate": 1.8524871355060036e-06,
+      "loss": 1.7343,
+      "step": 21690
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 12.798388481140137,
+      "learning_rate": 1.851650420449316e-06,
+      "loss": 0.9201,
+      "step": 21691
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 11.396600723266602,
+      "learning_rate": 1.8508137053926288e-06,
+      "loss": 0.4124,
+      "step": 21692
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 7.024505138397217,
+      "learning_rate": 1.8499769903359414e-06,
+      "loss": 0.354,
+      "step": 21693
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 7.042710304260254,
+      "learning_rate": 1.849140275279254e-06,
+      "loss": 0.1709,
+      "step": 21694
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 37.1627082824707,
+      "learning_rate": 1.8483035602225663e-06,
+      "loss": 1.0533,
+      "step": 21695
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 19.921981811523438,
+      "learning_rate": 1.847466845165879e-06,
+      "loss": 0.8625,
+      "step": 21696
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 27.523574829101562,
+      "learning_rate": 1.8466301301091915e-06,
+      "loss": 2.2229,
+      "step": 21697
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 24.74726676940918,
+      "learning_rate": 1.8457934150525039e-06,
+      "loss": 0.5465,
+      "step": 21698
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 36.155052185058594,
+      "learning_rate": 1.8449566999958165e-06,
+      "loss": 0.1337,
+      "step": 21699
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.531887054443359,
+      "learning_rate": 1.8441199849391293e-06,
+      "loss": 0.5163,
+      "step": 21700
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 7.0997138023376465,
+      "learning_rate": 1.8432832698824416e-06,
+      "loss": 0.2779,
+      "step": 21701
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 13.166228294372559,
+      "learning_rate": 1.8424465548257542e-06,
+      "loss": 0.447,
+      "step": 21702
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 14.30151653289795,
+      "learning_rate": 1.8416098397690668e-06,
+      "loss": 0.8173,
+      "step": 21703
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 18.93376350402832,
+      "learning_rate": 1.8407731247123794e-06,
+      "loss": 1.1971,
+      "step": 21704
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.382608890533447,
+      "learning_rate": 1.8399364096556918e-06,
+      "loss": 0.3352,
+      "step": 21705
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 19.910825729370117,
+      "learning_rate": 1.8390996945990044e-06,
+      "loss": 0.8676,
+      "step": 21706
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 63.348533630371094,
+      "learning_rate": 1.838262979542317e-06,
+      "loss": 2.9701,
+      "step": 21707
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 15.751937866210938,
+      "learning_rate": 1.8374262644856294e-06,
+      "loss": 0.9916,
+      "step": 21708
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 8.19721508026123,
+      "learning_rate": 1.8365895494289422e-06,
+      "loss": 0.5512,
+      "step": 21709
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 66.76535034179688,
+      "learning_rate": 1.8357528343722547e-06,
+      "loss": 1.0022,
+      "step": 21710
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.930474281311035,
+      "learning_rate": 1.8349161193155673e-06,
+      "loss": 0.1584,
+      "step": 21711
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 36.89176559448242,
+      "learning_rate": 1.8340794042588797e-06,
+      "loss": 3.0869,
+      "step": 21712
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 33.027488708496094,
+      "learning_rate": 1.8332426892021923e-06,
+      "loss": 1.591,
+      "step": 21713
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 12.010252952575684,
+      "learning_rate": 1.832405974145505e-06,
+      "loss": 2.052,
+      "step": 21714
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 8.362545013427734,
+      "learning_rate": 1.8315692590888173e-06,
+      "loss": 1.0886,
+      "step": 21715
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 11.814370155334473,
+      "learning_rate": 1.8307325440321299e-06,
+      "loss": 0.4781,
+      "step": 21716
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 5.486781120300293,
+      "learning_rate": 1.8298958289754427e-06,
+      "loss": 0.3103,
+      "step": 21717
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 50.36022186279297,
+      "learning_rate": 1.8290591139187553e-06,
+      "loss": 2.5289,
+      "step": 21718
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 11.407174110412598,
+      "learning_rate": 1.8282223988620676e-06,
+      "loss": 0.5349,
+      "step": 21719
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 25.968992233276367,
+      "learning_rate": 1.8273856838053802e-06,
+      "loss": 1.0201,
+      "step": 21720
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 11.296031951904297,
+      "learning_rate": 1.8265489687486928e-06,
+      "loss": 0.8723,
+      "step": 21721
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 65.60367584228516,
+      "learning_rate": 1.8257122536920052e-06,
+      "loss": 1.6013,
+      "step": 21722
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 43.265079498291016,
+      "learning_rate": 1.8248755386353178e-06,
+      "loss": 1.2649,
+      "step": 21723
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 37.025516510009766,
+      "learning_rate": 1.8240388235786306e-06,
+      "loss": 0.2077,
+      "step": 21724
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 78.99784851074219,
+      "learning_rate": 1.8232021085219432e-06,
+      "loss": 1.0799,
+      "step": 21725
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 18.098739624023438,
+      "learning_rate": 1.8223653934652556e-06,
+      "loss": 1.2167,
+      "step": 21726
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 15.547638893127441,
+      "learning_rate": 1.8215286784085681e-06,
+      "loss": 1.6061,
+      "step": 21727
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 9.266261100769043,
+      "learning_rate": 1.8206919633518807e-06,
+      "loss": 1.6266,
+      "step": 21728
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 8.35934066772461,
+      "learning_rate": 1.8198552482951931e-06,
+      "loss": 2.1372,
+      "step": 21729
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 16.34247398376465,
+      "learning_rate": 1.8190185332385057e-06,
+      "loss": 0.9927,
+      "step": 21730
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 15.102217674255371,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 1.725,
+      "step": 21731
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 61.07167053222656,
+      "learning_rate": 1.817345103125131e-06,
+      "loss": 1.3133,
+      "step": 21732
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 2.809201240539551,
+      "learning_rate": 1.8165083880684435e-06,
+      "loss": 0.1029,
+      "step": 21733
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 20.1793212890625,
+      "learning_rate": 1.815671673011756e-06,
+      "loss": 1.6741,
+      "step": 21734
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 9.742021560668945,
+      "learning_rate": 1.8148349579550687e-06,
+      "loss": 0.3113,
+      "step": 21735
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 20.455286026000977,
+      "learning_rate": 1.813998242898381e-06,
+      "loss": 1.5275,
+      "step": 21736
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 13.391501426696777,
+      "learning_rate": 1.8131615278416936e-06,
+      "loss": 0.6146,
+      "step": 21737
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 7.942105293273926,
+      "learning_rate": 1.8123248127850062e-06,
+      "loss": 0.6434,
+      "step": 21738
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 49.41905975341797,
+      "learning_rate": 1.8114880977283188e-06,
+      "loss": 1.6082,
+      "step": 21739
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 11.774738311767578,
+      "learning_rate": 1.8106513826716312e-06,
+      "loss": 0.5761,
+      "step": 21740
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 8.68564510345459,
+      "learning_rate": 1.809814667614944e-06,
+      "loss": 1.0904,
+      "step": 21741
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 8.028971672058105,
+      "learning_rate": 1.8089779525582566e-06,
+      "loss": 0.2022,
+      "step": 21742
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 10.16524600982666,
+      "learning_rate": 1.808141237501569e-06,
+      "loss": 0.6065,
+      "step": 21743
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 7.454187393188477,
+      "learning_rate": 1.8073045224448815e-06,
+      "loss": 0.3915,
+      "step": 21744
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 11.614712715148926,
+      "learning_rate": 1.8064678073881941e-06,
+      "loss": 1.6263,
+      "step": 21745
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.0575791597366333,
+      "learning_rate": 1.8056310923315065e-06,
+      "loss": 0.0253,
+      "step": 21746
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 5.01806640625,
+      "learning_rate": 1.804794377274819e-06,
+      "loss": 0.6434,
+      "step": 21747
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.5121804475784302,
+      "learning_rate": 1.8039576622181317e-06,
+      "loss": 0.0145,
+      "step": 21748
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 36.74553298950195,
+      "learning_rate": 1.8031209471614445e-06,
+      "loss": 1.5289,
+      "step": 21749
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 13.009232521057129,
+      "learning_rate": 1.8022842321047569e-06,
+      "loss": 0.5034,
+      "step": 21750
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.547818660736084,
+      "learning_rate": 1.8014475170480695e-06,
+      "loss": 0.5641,
+      "step": 21751
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 31.20549201965332,
+      "learning_rate": 1.800610801991382e-06,
+      "loss": 1.0199,
+      "step": 21752
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 8.372740745544434,
+      "learning_rate": 1.7997740869346944e-06,
+      "loss": 0.7164,
+      "step": 21753
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 7.5386834144592285,
+      "learning_rate": 1.798937371878007e-06,
+      "loss": 1.4045,
+      "step": 21754
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 9.312443733215332,
+      "learning_rate": 1.7981006568213196e-06,
+      "loss": 0.3174,
+      "step": 21755
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 13.363283157348633,
+      "learning_rate": 1.7972639417646324e-06,
+      "loss": 0.9533,
+      "step": 21756
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 9.282561302185059,
+      "learning_rate": 1.7964272267079446e-06,
+      "loss": 1.0512,
+      "step": 21757
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 12.2020845413208,
+      "learning_rate": 1.7955905116512574e-06,
+      "loss": 0.7064,
+      "step": 21758
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 9.255086898803711,
+      "learning_rate": 1.79475379659457e-06,
+      "loss": 0.9093,
+      "step": 21759
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 22.05156135559082,
+      "learning_rate": 1.7939170815378823e-06,
+      "loss": 0.8288,
+      "step": 21760
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 15.910368919372559,
+      "learning_rate": 1.793080366481195e-06,
+      "loss": 0.5122,
+      "step": 21761
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 17.45794105529785,
+      "learning_rate": 1.7922436514245075e-06,
+      "loss": 1.8548,
+      "step": 21762
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 9.76082706451416,
+      "learning_rate": 1.7914069363678201e-06,
+      "loss": 1.6464,
+      "step": 21763
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 36.96340560913086,
+      "learning_rate": 1.7905702213111325e-06,
+      "loss": 1.2681,
+      "step": 21764
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 12.535592079162598,
+      "learning_rate": 1.7897335062544453e-06,
+      "loss": 1.5499,
+      "step": 21765
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 157.78836059570312,
+      "learning_rate": 1.7888967911977579e-06,
+      "loss": 1.6722,
+      "step": 21766
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 17.902124404907227,
+      "learning_rate": 1.7880600761410703e-06,
+      "loss": 0.9337,
+      "step": 21767
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 13.388311386108398,
+      "learning_rate": 1.7872233610843829e-06,
+      "loss": 0.7366,
+      "step": 21768
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 36.524497985839844,
+      "learning_rate": 1.7863866460276954e-06,
+      "loss": 1.4949,
+      "step": 21769
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 26.989593505859375,
+      "learning_rate": 1.785549930971008e-06,
+      "loss": 0.8982,
+      "step": 21770
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 13.714203834533691,
+      "learning_rate": 1.7847132159143204e-06,
+      "loss": 0.5994,
+      "step": 21771
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.962925910949707,
+      "learning_rate": 1.783876500857633e-06,
+      "loss": 0.1842,
+      "step": 21772
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 28.30582046508789,
+      "learning_rate": 1.7830397858009458e-06,
+      "loss": 2.5369,
+      "step": 21773
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 7.63728666305542,
+      "learning_rate": 1.7822030707442582e-06,
+      "loss": 0.6249,
+      "step": 21774
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 51.46284103393555,
+      "learning_rate": 1.7813663556875708e-06,
+      "loss": 1.8673,
+      "step": 21775
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 32.90729904174805,
+      "learning_rate": 1.7805296406308834e-06,
+      "loss": 2.6009,
+      "step": 21776
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 10.800202369689941,
+      "learning_rate": 1.779692925574196e-06,
+      "loss": 0.64,
+      "step": 21777
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 6.670138359069824,
+      "learning_rate": 1.7788562105175083e-06,
+      "loss": 0.3723,
+      "step": 21778
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 150.59474182128906,
+      "learning_rate": 1.778019495460821e-06,
+      "loss": 3.0521,
+      "step": 21779
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.466110944747925,
+      "learning_rate": 1.7771827804041335e-06,
+      "loss": 0.205,
+      "step": 21780
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 33.088748931884766,
+      "learning_rate": 1.7763460653474459e-06,
+      "loss": 0.7049,
+      "step": 21781
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 27.63764190673828,
+      "learning_rate": 1.7755093502907587e-06,
+      "loss": 0.6086,
+      "step": 21782
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 17.60592269897461,
+      "learning_rate": 1.7746726352340713e-06,
+      "loss": 1.0391,
+      "step": 21783
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 33.07358169555664,
+      "learning_rate": 1.7738359201773839e-06,
+      "loss": 2.4924,
+      "step": 21784
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.072381019592285,
+      "learning_rate": 1.7729992051206962e-06,
+      "loss": 0.0847,
+      "step": 21785
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 16.366477966308594,
+      "learning_rate": 1.7721624900640088e-06,
+      "loss": 1.4918,
+      "step": 21786
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 13.376606941223145,
+      "learning_rate": 1.7713257750073214e-06,
+      "loss": 1.6522,
+      "step": 21787
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 9.831418991088867,
+      "learning_rate": 1.7704890599506338e-06,
+      "loss": 0.5395,
+      "step": 21788
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 15.025633811950684,
+      "learning_rate": 1.7696523448939464e-06,
+      "loss": 1.3841,
+      "step": 21789
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 10.509660720825195,
+      "learning_rate": 1.7688156298372592e-06,
+      "loss": 0.1849,
+      "step": 21790
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 9.489770889282227,
+      "learning_rate": 1.7679789147805716e-06,
+      "loss": 0.4074,
+      "step": 21791
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 5.313810348510742,
+      "learning_rate": 1.7671421997238842e-06,
+      "loss": 0.5014,
+      "step": 21792
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 19.57631492614746,
+      "learning_rate": 1.7663054846671968e-06,
+      "loss": 0.7097,
+      "step": 21793
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 28.672203063964844,
+      "learning_rate": 1.7654687696105093e-06,
+      "loss": 2.6373,
+      "step": 21794
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 17.289979934692383,
+      "learning_rate": 1.7646320545538217e-06,
+      "loss": 1.1245,
+      "step": 21795
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 39.91363525390625,
+      "learning_rate": 1.7637953394971343e-06,
+      "loss": 1.1635,
+      "step": 21796
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 26.618194580078125,
+      "learning_rate": 1.7629586244404471e-06,
+      "loss": 1.1167,
+      "step": 21797
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 302.7001953125,
+      "learning_rate": 1.7621219093837593e-06,
+      "loss": 1.2704,
+      "step": 21798
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.354804992675781,
+      "learning_rate": 1.761285194327072e-06,
+      "loss": 0.2161,
+      "step": 21799
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 33.09197998046875,
+      "learning_rate": 1.7604484792703847e-06,
+      "loss": 1.3867,
+      "step": 21800
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 12.664949417114258,
+      "learning_rate": 1.7596117642136973e-06,
+      "loss": 0.5833,
+      "step": 21801
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 29.650514602661133,
+      "learning_rate": 1.7587750491570096e-06,
+      "loss": 1.6012,
+      "step": 21802
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 20.913442611694336,
+      "learning_rate": 1.7579383341003222e-06,
+      "loss": 0.822,
+      "step": 21803
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 12.550580024719238,
+      "learning_rate": 1.7571016190436348e-06,
+      "loss": 0.2997,
+      "step": 21804
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.505816459655762,
+      "learning_rate": 1.7562649039869472e-06,
+      "loss": 0.9827,
+      "step": 21805
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 15.06021785736084,
+      "learning_rate": 1.75542818893026e-06,
+      "loss": 0.6605,
+      "step": 21806
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 10.036102294921875,
+      "learning_rate": 1.7545914738735726e-06,
+      "loss": 0.3061,
+      "step": 21807
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 25.392353057861328,
+      "learning_rate": 1.7537547588168852e-06,
+      "loss": 0.874,
+      "step": 21808
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 42.024314880371094,
+      "learning_rate": 1.7529180437601976e-06,
+      "loss": 0.5915,
+      "step": 21809
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 14.7062349319458,
+      "learning_rate": 1.7520813287035102e-06,
+      "loss": 1.5098,
+      "step": 21810
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 20.205429077148438,
+      "learning_rate": 1.7512446136468227e-06,
+      "loss": 0.5976,
+      "step": 21811
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 14.482434272766113,
+      "learning_rate": 1.7504078985901351e-06,
+      "loss": 0.8241,
+      "step": 21812
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.756473541259766,
+      "learning_rate": 1.7495711835334477e-06,
+      "loss": 0.2505,
+      "step": 21813
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 26.58550262451172,
+      "learning_rate": 1.7487344684767605e-06,
+      "loss": 2.0213,
+      "step": 21814
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 15.44678783416748,
+      "learning_rate": 1.747897753420073e-06,
+      "loss": 0.9022,
+      "step": 21815
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 66.37962341308594,
+      "learning_rate": 1.7470610383633855e-06,
+      "loss": 1.9281,
+      "step": 21816
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 22.037757873535156,
+      "learning_rate": 1.746224323306698e-06,
+      "loss": 1.2474,
+      "step": 21817
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 85.15798950195312,
+      "learning_rate": 1.7453876082500107e-06,
+      "loss": 2.9829,
+      "step": 21818
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 17.27855682373047,
+      "learning_rate": 1.744550893193323e-06,
+      "loss": 0.7159,
+      "step": 21819
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 21.975770950317383,
+      "learning_rate": 1.7437141781366356e-06,
+      "loss": 0.9852,
+      "step": 21820
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.015631198883057,
+      "learning_rate": 1.7428774630799482e-06,
+      "loss": 0.7902,
+      "step": 21821
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 19.04303741455078,
+      "learning_rate": 1.742040748023261e-06,
+      "loss": 0.6132,
+      "step": 21822
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 14.610198974609375,
+      "learning_rate": 1.7412040329665734e-06,
+      "loss": 0.2462,
+      "step": 21823
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 9.195034980773926,
+      "learning_rate": 1.740367317909886e-06,
+      "loss": 0.6449,
+      "step": 21824
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 10.353257179260254,
+      "learning_rate": 1.7395306028531986e-06,
+      "loss": 0.4132,
+      "step": 21825
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 16.202743530273438,
+      "learning_rate": 1.738693887796511e-06,
+      "loss": 0.6932,
+      "step": 21826
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 12.876373291015625,
+      "learning_rate": 1.7378571727398235e-06,
+      "loss": 1.0852,
+      "step": 21827
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 68.7238540649414,
+      "learning_rate": 1.7370204576831361e-06,
+      "loss": 1.1054,
+      "step": 21828
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 13.575844764709473,
+      "learning_rate": 1.736183742626449e-06,
+      "loss": 1.2549,
+      "step": 21829
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 29.486371994018555,
+      "learning_rate": 1.7353470275697611e-06,
+      "loss": 0.5472,
+      "step": 21830
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 14.410101890563965,
+      "learning_rate": 1.734510312513074e-06,
+      "loss": 0.4578,
+      "step": 21831
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 2.11220645904541,
+      "learning_rate": 1.7336735974563865e-06,
+      "loss": 0.0596,
+      "step": 21832
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 10.411478996276855,
+      "learning_rate": 1.7328368823996989e-06,
+      "loss": 0.8638,
+      "step": 21833
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 17.41630744934082,
+      "learning_rate": 1.7320001673430115e-06,
+      "loss": 0.7133,
+      "step": 21834
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 7.677325248718262,
+      "learning_rate": 1.731163452286324e-06,
+      "loss": 0.7754,
+      "step": 21835
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 9.375128746032715,
+      "learning_rate": 1.7303267372296364e-06,
+      "loss": 0.6212,
+      "step": 21836
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 13.899431228637695,
+      "learning_rate": 1.729490022172949e-06,
+      "loss": 1.0943,
+      "step": 21837
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 9.955965042114258,
+      "learning_rate": 1.7286533071162618e-06,
+      "loss": 0.1808,
+      "step": 21838
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 21.23792266845703,
+      "learning_rate": 1.7278165920595744e-06,
+      "loss": 0.8996,
+      "step": 21839
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 18.382326126098633,
+      "learning_rate": 1.7269798770028868e-06,
+      "loss": 0.8314,
+      "step": 21840
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 18.497270584106445,
+      "learning_rate": 1.7261431619461994e-06,
+      "loss": 0.7368,
+      "step": 21841
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 13.710346221923828,
+      "learning_rate": 1.725306446889512e-06,
+      "loss": 0.2386,
+      "step": 21842
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 10.088309288024902,
+      "learning_rate": 1.7244697318328244e-06,
+      "loss": 0.6079,
+      "step": 21843
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 13.157638549804688,
+      "learning_rate": 1.723633016776137e-06,
+      "loss": 0.5675,
+      "step": 21844
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 15.954193115234375,
+      "learning_rate": 1.7227963017194495e-06,
+      "loss": 0.7772,
+      "step": 21845
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 13.904277801513672,
+      "learning_rate": 1.7219595866627623e-06,
+      "loss": 0.7493,
+      "step": 21846
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 33.891056060791016,
+      "learning_rate": 1.7211228716060747e-06,
+      "loss": 1.3535,
+      "step": 21847
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 9.177570343017578,
+      "learning_rate": 1.7202861565493873e-06,
+      "loss": 0.7781,
+      "step": 21848
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 7.171292781829834,
+      "learning_rate": 1.7194494414926999e-06,
+      "loss": 0.184,
+      "step": 21849
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 17.63407325744629,
+      "learning_rate": 1.7186127264360123e-06,
+      "loss": 1.3083,
+      "step": 21850
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 34.11215591430664,
+      "learning_rate": 1.7177760113793249e-06,
+      "loss": 2.8637,
+      "step": 21851
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 19.464448928833008,
+      "learning_rate": 1.7169392963226375e-06,
+      "loss": 1.291,
+      "step": 21852
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 17.207942962646484,
+      "learning_rate": 1.71610258126595e-06,
+      "loss": 0.7999,
+      "step": 21853
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 9.438970565795898,
+      "learning_rate": 1.7152658662092624e-06,
+      "loss": 0.5002,
+      "step": 21854
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 14.350874900817871,
+      "learning_rate": 1.7144291511525752e-06,
+      "loss": 0.6589,
+      "step": 21855
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.654712200164795,
+      "learning_rate": 1.7135924360958878e-06,
+      "loss": 0.9977,
+      "step": 21856
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 106.78719329833984,
+      "learning_rate": 1.7127557210392002e-06,
+      "loss": 0.518,
+      "step": 21857
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 19.357154846191406,
+      "learning_rate": 1.7119190059825128e-06,
+      "loss": 0.889,
+      "step": 21858
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.225431442260742,
+      "learning_rate": 1.7110822909258254e-06,
+      "loss": 0.2714,
+      "step": 21859
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 15.32987117767334,
+      "learning_rate": 1.710245575869138e-06,
+      "loss": 1.762,
+      "step": 21860
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 96.23703002929688,
+      "learning_rate": 1.7094088608124503e-06,
+      "loss": 1.4382,
+      "step": 21861
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 180.13279724121094,
+      "learning_rate": 1.708572145755763e-06,
+      "loss": 1.5237,
+      "step": 21862
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 70.8868408203125,
+      "learning_rate": 1.7077354306990757e-06,
+      "loss": 1.6646,
+      "step": 21863
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 25.734073638916016,
+      "learning_rate": 1.7068987156423881e-06,
+      "loss": 0.5257,
+      "step": 21864
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 17.142911911010742,
+      "learning_rate": 1.7060620005857007e-06,
+      "loss": 2.0709,
+      "step": 21865
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 28.677814483642578,
+      "learning_rate": 1.7052252855290133e-06,
+      "loss": 1.3275,
+      "step": 21866
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 10.25939655303955,
+      "learning_rate": 1.7043885704723259e-06,
+      "loss": 1.1485,
+      "step": 21867
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 22.28053092956543,
+      "learning_rate": 1.7035518554156383e-06,
+      "loss": 1.6201,
+      "step": 21868
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 9.923303604125977,
+      "learning_rate": 1.7027151403589508e-06,
+      "loss": 0.705,
+      "step": 21869
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 10.079099655151367,
+      "learning_rate": 1.7018784253022636e-06,
+      "loss": 2.1897,
+      "step": 21870
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 11.395952224731445,
+      "learning_rate": 1.7010417102455758e-06,
+      "loss": 1.1989,
+      "step": 21871
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 19.015522003173828,
+      "learning_rate": 1.7002049951888886e-06,
+      "loss": 0.8393,
+      "step": 21872
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 7.965001583099365,
+      "learning_rate": 1.6993682801322012e-06,
+      "loss": 1.1799,
+      "step": 21873
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 2.6550655364990234,
+      "learning_rate": 1.6985315650755138e-06,
+      "loss": 0.0839,
+      "step": 21874
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 7.228896141052246,
+      "learning_rate": 1.6976948500188262e-06,
+      "loss": 0.2871,
+      "step": 21875
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 21.15328598022461,
+      "learning_rate": 1.6968581349621388e-06,
+      "loss": 0.4294,
+      "step": 21876
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 42.655616760253906,
+      "learning_rate": 1.6960214199054514e-06,
+      "loss": 1.9843,
+      "step": 21877
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 56.5605583190918,
+      "learning_rate": 1.6951847048487637e-06,
+      "loss": 1.3758,
+      "step": 21878
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 15.730677604675293,
+      "learning_rate": 1.6943479897920765e-06,
+      "loss": 0.4272,
+      "step": 21879
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.9887042045593262,
+      "learning_rate": 1.6935112747353891e-06,
+      "loss": 0.0324,
+      "step": 21880
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 15.896023750305176,
+      "learning_rate": 1.6926745596787015e-06,
+      "loss": 1.0028,
+      "step": 21881
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 11.582592010498047,
+      "learning_rate": 1.691837844622014e-06,
+      "loss": 1.0274,
+      "step": 21882
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 13.835699081420898,
+      "learning_rate": 1.6910011295653267e-06,
+      "loss": 1.5205,
+      "step": 21883
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 82.54322052001953,
+      "learning_rate": 1.6901644145086393e-06,
+      "loss": 1.3526,
+      "step": 21884
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 19.077165603637695,
+      "learning_rate": 1.6893276994519517e-06,
+      "loss": 0.6383,
+      "step": 21885
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 19.55196762084961,
+      "learning_rate": 1.6884909843952642e-06,
+      "loss": 1.022,
+      "step": 21886
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 8.582889556884766,
+      "learning_rate": 1.687654269338577e-06,
+      "loss": 0.4232,
+      "step": 21887
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 15.185364723205566,
+      "learning_rate": 1.6868175542818892e-06,
+      "loss": 0.5666,
+      "step": 21888
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 37.80705642700195,
+      "learning_rate": 1.685980839225202e-06,
+      "loss": 1.2841,
+      "step": 21889
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 18.197996139526367,
+      "learning_rate": 1.6851441241685146e-06,
+      "loss": 0.4401,
+      "step": 21890
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7867083549499512,
+      "learning_rate": 1.6843074091118272e-06,
+      "loss": 0.0157,
+      "step": 21891
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 19.772058486938477,
+      "learning_rate": 1.6834706940551396e-06,
+      "loss": 0.5584,
+      "step": 21892
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 11.649335861206055,
+      "learning_rate": 1.6826339789984522e-06,
+      "loss": 1.1179,
+      "step": 21893
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 33.918922424316406,
+      "learning_rate": 1.6817972639417648e-06,
+      "loss": 2.8612,
+      "step": 21894
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 401.5687561035156,
+      "learning_rate": 1.6809605488850771e-06,
+      "loss": 1.3614,
+      "step": 21895
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 13.494359016418457,
+      "learning_rate": 1.68012383382839e-06,
+      "loss": 1.6599,
+      "step": 21896
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 326.7059631347656,
+      "learning_rate": 1.6792871187717025e-06,
+      "loss": 0.3422,
+      "step": 21897
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 65.26554107666016,
+      "learning_rate": 1.6784504037150151e-06,
+      "loss": 1.9373,
+      "step": 21898
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 8.846159934997559,
+      "learning_rate": 1.6776136886583275e-06,
+      "loss": 0.1525,
+      "step": 21899
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 54.872589111328125,
+      "learning_rate": 1.67677697360164e-06,
+      "loss": 0.8604,
+      "step": 21900
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 15.160786628723145,
+      "learning_rate": 1.6759402585449527e-06,
+      "loss": 1.3821,
+      "step": 21901
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 92.08221435546875,
+      "learning_rate": 1.675103543488265e-06,
+      "loss": 2.8299,
+      "step": 21902
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 2.968740940093994,
+      "learning_rate": 1.6742668284315776e-06,
+      "loss": 0.1037,
+      "step": 21903
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 2.996655225753784,
+      "learning_rate": 1.6734301133748904e-06,
+      "loss": 0.0593,
+      "step": 21904
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 10.670302391052246,
+      "learning_rate": 1.672593398318203e-06,
+      "loss": 0.2745,
+      "step": 21905
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 103.45101165771484,
+      "learning_rate": 1.6717566832615154e-06,
+      "loss": 1.8951,
+      "step": 21906
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 11.243412017822266,
+      "learning_rate": 1.670919968204828e-06,
+      "loss": 0.5089,
+      "step": 21907
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 35.79745101928711,
+      "learning_rate": 1.6700832531481406e-06,
+      "loss": 0.6931,
+      "step": 21908
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 38.157169342041016,
+      "learning_rate": 1.669246538091453e-06,
+      "loss": 2.0333,
+      "step": 21909
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 7.276881217956543,
+      "learning_rate": 1.6684098230347656e-06,
+      "loss": 0.3027,
+      "step": 21910
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 10.117690086364746,
+      "learning_rate": 1.6675731079780784e-06,
+      "loss": 0.4207,
+      "step": 21911
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.078430414199829,
+      "learning_rate": 1.666736392921391e-06,
+      "loss": 0.1118,
+      "step": 21912
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 15.570343971252441,
+      "learning_rate": 1.6658996778647033e-06,
+      "loss": 0.4665,
+      "step": 21913
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.487558841705322,
+      "learning_rate": 1.665062962808016e-06,
+      "loss": 0.2549,
+      "step": 21914
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 22.4284610748291,
+      "learning_rate": 1.6642262477513285e-06,
+      "loss": 1.4474,
+      "step": 21915
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 11.733662605285645,
+      "learning_rate": 1.6633895326946409e-06,
+      "loss": 0.7138,
+      "step": 21916
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 26.12148666381836,
+      "learning_rate": 1.6625528176379535e-06,
+      "loss": 2.1125,
+      "step": 21917
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 11.106646537780762,
+      "learning_rate": 1.661716102581266e-06,
+      "loss": 0.7808,
+      "step": 21918
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 15.67684555053711,
+      "learning_rate": 1.6608793875245789e-06,
+      "loss": 0.4385,
+      "step": 21919
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 288.1107482910156,
+      "learning_rate": 1.660042672467891e-06,
+      "loss": 1.8863,
+      "step": 21920
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 91.8941650390625,
+      "learning_rate": 1.6592059574112038e-06,
+      "loss": 1.952,
+      "step": 21921
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 14.114540100097656,
+      "learning_rate": 1.6583692423545164e-06,
+      "loss": 0.3145,
+      "step": 21922
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 59.77403259277344,
+      "learning_rate": 1.6575325272978288e-06,
+      "loss": 1.2466,
+      "step": 21923
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 32.73985290527344,
+      "learning_rate": 1.6566958122411414e-06,
+      "loss": 1.3254,
+      "step": 21924
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 22.290546417236328,
+      "learning_rate": 1.655859097184454e-06,
+      "loss": 2.5204,
+      "step": 21925
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 21.034669876098633,
+      "learning_rate": 1.6550223821277664e-06,
+      "loss": 2.2211,
+      "step": 21926
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 56.31417465209961,
+      "learning_rate": 1.654185667071079e-06,
+      "loss": 1.0157,
+      "step": 21927
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 11.791213035583496,
+      "learning_rate": 1.6533489520143918e-06,
+      "loss": 0.4559,
+      "step": 21928
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 16.289724349975586,
+      "learning_rate": 1.6525122369577043e-06,
+      "loss": 2.1772,
+      "step": 21929
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 6.940343856811523,
+      "learning_rate": 1.6516755219010167e-06,
+      "loss": 1.2079,
+      "step": 21930
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 18.676956176757812,
+      "learning_rate": 1.6508388068443293e-06,
+      "loss": 1.9555,
+      "step": 21931
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 9.926246643066406,
+      "learning_rate": 1.650002091787642e-06,
+      "loss": 2.177,
+      "step": 21932
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 8.71130657196045,
+      "learning_rate": 1.6491653767309543e-06,
+      "loss": 0.4256,
+      "step": 21933
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 65.7220458984375,
+      "learning_rate": 1.6483286616742669e-06,
+      "loss": 1.3896,
+      "step": 21934
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 23.7147216796875,
+      "learning_rate": 1.6474919466175795e-06,
+      "loss": 0.8591,
+      "step": 21935
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 13.173413276672363,
+      "learning_rate": 1.6466552315608923e-06,
+      "loss": 0.5658,
+      "step": 21936
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 7.923994064331055,
+      "learning_rate": 1.6458185165042046e-06,
+      "loss": 1.5625,
+      "step": 21937
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 15.859750747680664,
+      "learning_rate": 1.6449818014475172e-06,
+      "loss": 2.2383,
+      "step": 21938
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 2.3159897327423096,
+      "learning_rate": 1.6441450863908298e-06,
+      "loss": 0.1943,
+      "step": 21939
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 20.445648193359375,
+      "learning_rate": 1.6433083713341422e-06,
+      "loss": 1.4534,
+      "step": 21940
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 14.416092872619629,
+      "learning_rate": 1.6424716562774548e-06,
+      "loss": 0.372,
+      "step": 21941
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 26.19920539855957,
+      "learning_rate": 1.6416349412207674e-06,
+      "loss": 1.3652,
+      "step": 21942
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 9.537945747375488,
+      "learning_rate": 1.6407982261640802e-06,
+      "loss": 1.5486,
+      "step": 21943
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 13.7022705078125,
+      "learning_rate": 1.6399615111073923e-06,
+      "loss": 0.3865,
+      "step": 21944
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 17.28717803955078,
+      "learning_rate": 1.6391247960507051e-06,
+      "loss": 1.0219,
+      "step": 21945
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 13.585247993469238,
+      "learning_rate": 1.6382880809940177e-06,
+      "loss": 0.5534,
+      "step": 21946
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 14.266397476196289,
+      "learning_rate": 1.6374513659373301e-06,
+      "loss": 1.9522,
+      "step": 21947
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 2.218893051147461,
+      "learning_rate": 1.6366146508806427e-06,
+      "loss": 0.0877,
+      "step": 21948
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 13.300233840942383,
+      "learning_rate": 1.6357779358239553e-06,
+      "loss": 1.1357,
+      "step": 21949
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 8.30359935760498,
+      "learning_rate": 1.6349412207672679e-06,
+      "loss": 0.268,
+      "step": 21950
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 6.496923446655273,
+      "learning_rate": 1.6341045057105803e-06,
+      "loss": 1.0002,
+      "step": 21951
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 11.52630615234375,
+      "learning_rate": 1.6332677906538929e-06,
+      "loss": 0.5624,
+      "step": 21952
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 8.756043434143066,
+      "learning_rate": 1.6324310755972057e-06,
+      "loss": 0.7606,
+      "step": 21953
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 11.567495346069336,
+      "learning_rate": 1.631594360540518e-06,
+      "loss": 0.3646,
+      "step": 21954
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 5.404824256896973,
+      "learning_rate": 1.6307576454838306e-06,
+      "loss": 0.2808,
+      "step": 21955
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.953155755996704,
+      "learning_rate": 1.6299209304271432e-06,
+      "loss": 0.0587,
+      "step": 21956
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 6.900728225708008,
+      "learning_rate": 1.6290842153704558e-06,
+      "loss": 0.3259,
+      "step": 21957
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 21.54115104675293,
+      "learning_rate": 1.6282475003137682e-06,
+      "loss": 0.9631,
+      "step": 21958
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 11.71082592010498,
+      "learning_rate": 1.6274107852570808e-06,
+      "loss": 0.6494,
+      "step": 21959
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 23.017696380615234,
+      "learning_rate": 1.6265740702003936e-06,
+      "loss": 1.0764,
+      "step": 21960
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 16.811050415039062,
+      "learning_rate": 1.6257373551437057e-06,
+      "loss": 0.562,
+      "step": 21961
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 10.228431701660156,
+      "learning_rate": 1.6249006400870185e-06,
+      "loss": 0.4482,
+      "step": 21962
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 12.603243827819824,
+      "learning_rate": 1.6240639250303311e-06,
+      "loss": 0.775,
+      "step": 21963
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 11.481263160705566,
+      "learning_rate": 1.6232272099736437e-06,
+      "loss": 1.3212,
+      "step": 21964
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 10.63491439819336,
+      "learning_rate": 1.622390494916956e-06,
+      "loss": 0.3953,
+      "step": 21965
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 21.720531463623047,
+      "learning_rate": 1.6215537798602687e-06,
+      "loss": 2.258,
+      "step": 21966
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 9.512266159057617,
+      "learning_rate": 1.6207170648035813e-06,
+      "loss": 0.7471,
+      "step": 21967
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 7.566118240356445,
+      "learning_rate": 1.6198803497468937e-06,
+      "loss": 0.2139,
+      "step": 21968
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 9.541152000427246,
+      "learning_rate": 1.6190436346902065e-06,
+      "loss": 0.6269,
+      "step": 21969
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 19.328577041625977,
+      "learning_rate": 1.618206919633519e-06,
+      "loss": 0.5275,
+      "step": 21970
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 32.173343658447266,
+      "learning_rate": 1.6173702045768314e-06,
+      "loss": 1.8638,
+      "step": 21971
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 36.626922607421875,
+      "learning_rate": 1.616533489520144e-06,
+      "loss": 0.6485,
+      "step": 21972
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 6.538182258605957,
+      "learning_rate": 1.6156967744634566e-06,
+      "loss": 0.5285,
+      "step": 21973
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 15.379257202148438,
+      "learning_rate": 1.6148600594067692e-06,
+      "loss": 0.7523,
+      "step": 21974
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 5.955650806427002,
+      "learning_rate": 1.6140233443500816e-06,
+      "loss": 0.3615,
+      "step": 21975
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 17.628507614135742,
+      "learning_rate": 1.6131866292933942e-06,
+      "loss": 0.7373,
+      "step": 21976
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 14.149693489074707,
+      "learning_rate": 1.612349914236707e-06,
+      "loss": 0.2518,
+      "step": 21977
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 13.153942108154297,
+      "learning_rate": 1.6115131991800193e-06,
+      "loss": 1.4344,
+      "step": 21978
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 10.648983001708984,
+      "learning_rate": 1.610676484123332e-06,
+      "loss": 1.6093,
+      "step": 21979
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 20.930753707885742,
+      "learning_rate": 1.6098397690666445e-06,
+      "loss": 0.6113,
+      "step": 21980
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 10.933890342712402,
+      "learning_rate": 1.6090030540099571e-06,
+      "loss": 0.8931,
+      "step": 21981
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 9.867168426513672,
+      "learning_rate": 1.6081663389532695e-06,
+      "loss": 0.8762,
+      "step": 21982
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 13.132024765014648,
+      "learning_rate": 1.607329623896582e-06,
+      "loss": 2.9462,
+      "step": 21983
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 9.387228012084961,
+      "learning_rate": 1.6064929088398947e-06,
+      "loss": 0.2168,
+      "step": 21984
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 10.731705665588379,
+      "learning_rate": 1.605656193783207e-06,
+      "loss": 0.617,
+      "step": 21985
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 10.764081001281738,
+      "learning_rate": 1.6048194787265199e-06,
+      "loss": 0.7568,
+      "step": 21986
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 7.572868824005127,
+      "learning_rate": 1.6039827636698324e-06,
+      "loss": 1.1191,
+      "step": 21987
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 9.564583778381348,
+      "learning_rate": 1.603146048613145e-06,
+      "loss": 0.5623,
+      "step": 21988
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 9.795342445373535,
+      "learning_rate": 1.6023093335564574e-06,
+      "loss": 1.1223,
+      "step": 21989
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 16.631587982177734,
+      "learning_rate": 1.60147261849977e-06,
+      "loss": 0.9183,
+      "step": 21990
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 3.764885425567627,
+      "learning_rate": 1.6006359034430826e-06,
+      "loss": 0.1322,
+      "step": 21991
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 11.128994941711426,
+      "learning_rate": 1.599799188386395e-06,
+      "loss": 0.7086,
+      "step": 21992
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 28.502336502075195,
+      "learning_rate": 1.5989624733297076e-06,
+      "loss": 2.1299,
+      "step": 21993
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 6.179806232452393,
+      "learning_rate": 1.5981257582730204e-06,
+      "loss": 0.5095,
+      "step": 21994
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 94.47281646728516,
+      "learning_rate": 1.597289043216333e-06,
+      "loss": 1.3763,
+      "step": 21995
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 5.738989353179932,
+      "learning_rate": 1.5964523281596453e-06,
+      "loss": 0.2774,
+      "step": 21996
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 9.436177253723145,
+      "learning_rate": 1.595615613102958e-06,
+      "loss": 1.0281,
+      "step": 21997
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 28.344717025756836,
+      "learning_rate": 1.5947788980462705e-06,
+      "loss": 1.0338,
+      "step": 21998
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 16.288572311401367,
+      "learning_rate": 1.5939421829895829e-06,
+      "loss": 0.8427,
+      "step": 21999
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 14.561866760253906,
+      "learning_rate": 1.5931054679328955e-06,
+      "loss": 0.7721,
+      "step": 22000
+    },
+    {
+      "epoch": 2.76,
+      "eval_loss": 0.07404118776321411,
+      "eval_runtime": 95.399,
+      "eval_samples_per_second": 37.128,
+      "eval_steps_per_second": 37.128,
+      "step": 22000
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 12.305974960327148,
+      "learning_rate": 1.5922687528762083e-06,
+      "loss": 1.7453,
+      "step": 22001
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 7.190908432006836,
+      "learning_rate": 1.5914320378195209e-06,
+      "loss": 0.229,
+      "step": 22002
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 21.130447387695312,
+      "learning_rate": 1.5905953227628333e-06,
+      "loss": 0.8447,
+      "step": 22003
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 5.994531631469727,
+      "learning_rate": 1.5897586077061458e-06,
+      "loss": 0.2864,
+      "step": 22004
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 12.654057502746582,
+      "learning_rate": 1.5889218926494584e-06,
+      "loss": 1.6754,
+      "step": 22005
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 15.838152885437012,
+      "learning_rate": 1.5880851775927708e-06,
+      "loss": 0.6666,
+      "step": 22006
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 6.544735431671143,
+      "learning_rate": 1.5872484625360834e-06,
+      "loss": 0.7359,
+      "step": 22007
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 18.814769744873047,
+      "learning_rate": 1.586411747479396e-06,
+      "loss": 1.3601,
+      "step": 22008
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 34.54404067993164,
+      "learning_rate": 1.5855750324227084e-06,
+      "loss": 0.6953,
+      "step": 22009
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 60.191776275634766,
+      "learning_rate": 1.5847383173660212e-06,
+      "loss": 2.1437,
+      "step": 22010
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 12.865761756896973,
+      "learning_rate": 1.5839016023093338e-06,
+      "loss": 0.5915,
+      "step": 22011
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 55.058258056640625,
+      "learning_rate": 1.5830648872526464e-06,
+      "loss": 1.368,
+      "step": 22012
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 6.351691722869873,
+      "learning_rate": 1.5822281721959587e-06,
+      "loss": 0.5392,
+      "step": 22013
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.2716728448867798,
+      "learning_rate": 1.5813914571392713e-06,
+      "loss": 0.0164,
+      "step": 22014
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 13.238324165344238,
+      "learning_rate": 1.580554742082584e-06,
+      "loss": 0.666,
+      "step": 22015
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 10.876042366027832,
+      "learning_rate": 1.5797180270258963e-06,
+      "loss": 0.6721,
+      "step": 22016
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 8.264463424682617,
+      "learning_rate": 1.5788813119692089e-06,
+      "loss": 0.9602,
+      "step": 22017
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 5.361449241638184,
+      "learning_rate": 1.5780445969125217e-06,
+      "loss": 0.2794,
+      "step": 22018
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 35.89775085449219,
+      "learning_rate": 1.5772078818558343e-06,
+      "loss": 0.8674,
+      "step": 22019
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 10.809699058532715,
+      "learning_rate": 1.5763711667991466e-06,
+      "loss": 0.3178,
+      "step": 22020
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 26.81633186340332,
+      "learning_rate": 1.5755344517424592e-06,
+      "loss": 1.2973,
+      "step": 22021
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 97.02884674072266,
+      "learning_rate": 1.5746977366857718e-06,
+      "loss": 1.5536,
+      "step": 22022
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 37.21788787841797,
+      "learning_rate": 1.5738610216290842e-06,
+      "loss": 1.3391,
+      "step": 22023
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 29.007442474365234,
+      "learning_rate": 1.5730243065723968e-06,
+      "loss": 1.586,
+      "step": 22024
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 12.787467956542969,
+      "learning_rate": 1.5721875915157094e-06,
+      "loss": 0.9574,
+      "step": 22025
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 14.426722526550293,
+      "learning_rate": 1.5713508764590222e-06,
+      "loss": 0.4947,
+      "step": 22026
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 18.206111907958984,
+      "learning_rate": 1.5705141614023346e-06,
+      "loss": 1.6256,
+      "step": 22027
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 11.78557014465332,
+      "learning_rate": 1.5696774463456472e-06,
+      "loss": 0.2601,
+      "step": 22028
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 9.279020309448242,
+      "learning_rate": 1.5688407312889597e-06,
+      "loss": 0.8065,
+      "step": 22029
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 7.007302761077881,
+      "learning_rate": 1.5680040162322721e-06,
+      "loss": 0.7756,
+      "step": 22030
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 19.91046905517578,
+      "learning_rate": 1.5671673011755847e-06,
+      "loss": 0.7541,
+      "step": 22031
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 57.409095764160156,
+      "learning_rate": 1.5663305861188973e-06,
+      "loss": 4.0752,
+      "step": 22032
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 12.127786636352539,
+      "learning_rate": 1.5654938710622101e-06,
+      "loss": 2.0843,
+      "step": 22033
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 5.797756671905518,
+      "learning_rate": 1.5646571560055223e-06,
+      "loss": 0.2609,
+      "step": 22034
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 11.170928955078125,
+      "learning_rate": 1.563820440948835e-06,
+      "loss": 0.2006,
+      "step": 22035
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 10.267579078674316,
+      "learning_rate": 1.5629837258921477e-06,
+      "loss": 1.2304,
+      "step": 22036
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 16.07809829711914,
+      "learning_rate": 1.56214701083546e-06,
+      "loss": 1.8852,
+      "step": 22037
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 9.74322509765625,
+      "learning_rate": 1.5613102957787726e-06,
+      "loss": 1.6937,
+      "step": 22038
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 9.701142311096191,
+      "learning_rate": 1.5604735807220852e-06,
+      "loss": 2.0227,
+      "step": 22039
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 31.289939880371094,
+      "learning_rate": 1.5596368656653978e-06,
+      "loss": 2.159,
+      "step": 22040
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 7.308387756347656,
+      "learning_rate": 1.5588001506087102e-06,
+      "loss": 1.11,
+      "step": 22041
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.7404890060424805,
+      "learning_rate": 1.557963435552023e-06,
+      "loss": 0.0592,
+      "step": 22042
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 18.605871200561523,
+      "learning_rate": 1.5571267204953356e-06,
+      "loss": 2.3566,
+      "step": 22043
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.8233323097229,
+      "learning_rate": 1.556290005438648e-06,
+      "loss": 0.3014,
+      "step": 22044
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 10.767655372619629,
+      "learning_rate": 1.5554532903819606e-06,
+      "loss": 0.9742,
+      "step": 22045
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 8.517369270324707,
+      "learning_rate": 1.5546165753252731e-06,
+      "loss": 0.3589,
+      "step": 22046
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 15.145514488220215,
+      "learning_rate": 1.5537798602685857e-06,
+      "loss": 1.6709,
+      "step": 22047
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 17.403635025024414,
+      "learning_rate": 1.5529431452118981e-06,
+      "loss": 1.2472,
+      "step": 22048
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 8.701679229736328,
+      "learning_rate": 1.5521064301552107e-06,
+      "loss": 0.2516,
+      "step": 22049
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 17.895038604736328,
+      "learning_rate": 1.5512697150985235e-06,
+      "loss": 1.2595,
+      "step": 22050
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 13.750988006591797,
+      "learning_rate": 1.5504330000418359e-06,
+      "loss": 0.3344,
+      "step": 22051
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 6.85670804977417,
+      "learning_rate": 1.5495962849851485e-06,
+      "loss": 0.1961,
+      "step": 22052
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 41.78623962402344,
+      "learning_rate": 1.548759569928461e-06,
+      "loss": 0.9157,
+      "step": 22053
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 14.114798545837402,
+      "learning_rate": 1.5479228548717734e-06,
+      "loss": 1.2567,
+      "step": 22054
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 7.244355201721191,
+      "learning_rate": 1.547086139815086e-06,
+      "loss": 0.1725,
+      "step": 22055
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 8.193955421447754,
+      "learning_rate": 1.5462494247583986e-06,
+      "loss": 0.5188,
+      "step": 22056
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 6.3143134117126465,
+      "learning_rate": 1.5454127097017112e-06,
+      "loss": 1.343,
+      "step": 22057
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.948602199554443,
+      "learning_rate": 1.5445759946450236e-06,
+      "loss": 0.2202,
+      "step": 22058
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 3.0916144847869873,
+      "learning_rate": 1.5437392795883364e-06,
+      "loss": 0.1855,
+      "step": 22059
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 15.833561897277832,
+      "learning_rate": 1.542902564531649e-06,
+      "loss": 1.0426,
+      "step": 22060
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 33.184932708740234,
+      "learning_rate": 1.5420658494749614e-06,
+      "loss": 4.4033,
+      "step": 22061
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 15.196234703063965,
+      "learning_rate": 1.541229134418274e-06,
+      "loss": 0.7092,
+      "step": 22062
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 9.446250915527344,
+      "learning_rate": 1.5403924193615865e-06,
+      "loss": 0.7741,
+      "step": 22063
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 7.459219932556152,
+      "learning_rate": 1.5395557043048991e-06,
+      "loss": 0.8118,
+      "step": 22064
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 9.842741012573242,
+      "learning_rate": 1.5387189892482115e-06,
+      "loss": 0.9376,
+      "step": 22065
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 9.335184097290039,
+      "learning_rate": 1.537882274191524e-06,
+      "loss": 2.2261,
+      "step": 22066
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 16.755369186401367,
+      "learning_rate": 1.537045559134837e-06,
+      "loss": 0.866,
+      "step": 22067
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 20.118492126464844,
+      "learning_rate": 1.5362088440781493e-06,
+      "loss": 1.2994,
+      "step": 22068
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 15.074871063232422,
+      "learning_rate": 1.5353721290214619e-06,
+      "loss": 0.7052,
+      "step": 22069
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 11.162530899047852,
+      "learning_rate": 1.5345354139647745e-06,
+      "loss": 0.2569,
+      "step": 22070
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 16.171817779541016,
+      "learning_rate": 1.533698698908087e-06,
+      "loss": 0.7483,
+      "step": 22071
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 12.387211799621582,
+      "learning_rate": 1.5328619838513994e-06,
+      "loss": 0.3203,
+      "step": 22072
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 49.55227279663086,
+      "learning_rate": 1.532025268794712e-06,
+      "loss": 1.2671,
+      "step": 22073
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 14.436878204345703,
+      "learning_rate": 1.5311885537380248e-06,
+      "loss": 0.6779,
+      "step": 22074
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 9.842228889465332,
+      "learning_rate": 1.530351838681337e-06,
+      "loss": 1.8578,
+      "step": 22075
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 7.118929386138916,
+      "learning_rate": 1.5295151236246498e-06,
+      "loss": 0.3359,
+      "step": 22076
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.044647455215454,
+      "learning_rate": 1.5286784085679624e-06,
+      "loss": 0.0278,
+      "step": 22077
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 9.421809196472168,
+      "learning_rate": 1.527841693511275e-06,
+      "loss": 1.5462,
+      "step": 22078
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 14.961891174316406,
+      "learning_rate": 1.5270049784545873e-06,
+      "loss": 0.8989,
+      "step": 22079
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 30.62376594543457,
+      "learning_rate": 1.5261682633979e-06,
+      "loss": 0.742,
+      "step": 22080
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 36.153297424316406,
+      "learning_rate": 1.5253315483412125e-06,
+      "loss": 0.8372,
+      "step": 22081
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 64.26068878173828,
+      "learning_rate": 1.524494833284525e-06,
+      "loss": 2.5712,
+      "step": 22082
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 14.459300994873047,
+      "learning_rate": 1.5236581182278377e-06,
+      "loss": 0.4504,
+      "step": 22083
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 3.5654618740081787,
+      "learning_rate": 1.5228214031711503e-06,
+      "loss": 0.2064,
+      "step": 22084
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 10.165255546569824,
+      "learning_rate": 1.5219846881144629e-06,
+      "loss": 2.0243,
+      "step": 22085
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 26.57514762878418,
+      "learning_rate": 1.5211479730577753e-06,
+      "loss": 1.5504,
+      "step": 22086
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 56.33488845825195,
+      "learning_rate": 1.5203112580010879e-06,
+      "loss": 2.0848,
+      "step": 22087
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 34.85393142700195,
+      "learning_rate": 1.5194745429444004e-06,
+      "loss": 1.8452,
+      "step": 22088
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 13.139994621276855,
+      "learning_rate": 1.5186378278877128e-06,
+      "loss": 0.3163,
+      "step": 22089
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 10.634119987487793,
+      "learning_rate": 1.5178011128310254e-06,
+      "loss": 0.827,
+      "step": 22090
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 18.775665283203125,
+      "learning_rate": 1.5169643977743382e-06,
+      "loss": 1.0176,
+      "step": 22091
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 6.013891696929932,
+      "learning_rate": 1.5161276827176508e-06,
+      "loss": 0.1807,
+      "step": 22092
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 22.8477840423584,
+      "learning_rate": 1.5152909676609632e-06,
+      "loss": 2.253,
+      "step": 22093
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 2.667479991912842,
+      "learning_rate": 1.5144542526042758e-06,
+      "loss": 0.2037,
+      "step": 22094
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 8.795310974121094,
+      "learning_rate": 1.5136175375475884e-06,
+      "loss": 0.4072,
+      "step": 22095
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 8.953936576843262,
+      "learning_rate": 1.5127808224909007e-06,
+      "loss": 0.5833,
+      "step": 22096
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 31.658977508544922,
+      "learning_rate": 1.5119441074342133e-06,
+      "loss": 2.745,
+      "step": 22097
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 14.953171730041504,
+      "learning_rate": 1.511107392377526e-06,
+      "loss": 1.4642,
+      "step": 22098
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 8.55324935913086,
+      "learning_rate": 1.5102706773208383e-06,
+      "loss": 0.4538,
+      "step": 22099
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 34.371070861816406,
+      "learning_rate": 1.509433962264151e-06,
+      "loss": 1.0861,
+      "step": 22100
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 16.9543399810791,
+      "learning_rate": 1.5085972472074637e-06,
+      "loss": 0.5991,
+      "step": 22101
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 9.909595489501953,
+      "learning_rate": 1.5077605321507763e-06,
+      "loss": 0.4898,
+      "step": 22102
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 82.10369110107422,
+      "learning_rate": 1.5069238170940887e-06,
+      "loss": 2.7447,
+      "step": 22103
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 7.9112324714660645,
+      "learning_rate": 1.5060871020374012e-06,
+      "loss": 1.4893,
+      "step": 22104
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.8757380247116089,
+      "learning_rate": 1.5052503869807138e-06,
+      "loss": 0.0745,
+      "step": 22105
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 18.111446380615234,
+      "learning_rate": 1.5044136719240262e-06,
+      "loss": 1.2961,
+      "step": 22106
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 21.040069580078125,
+      "learning_rate": 1.5035769568673388e-06,
+      "loss": 0.4997,
+      "step": 22107
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 12.237754821777344,
+      "learning_rate": 1.5027402418106516e-06,
+      "loss": 0.9439,
+      "step": 22108
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 38.84874725341797,
+      "learning_rate": 1.5019035267539642e-06,
+      "loss": 1.1505,
+      "step": 22109
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 15.144665718078613,
+      "learning_rate": 1.5010668116972766e-06,
+      "loss": 0.5755,
+      "step": 22110
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 194.444091796875,
+      "learning_rate": 1.5002300966405892e-06,
+      "loss": 1.1618,
+      "step": 22111
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 90.46455383300781,
+      "learning_rate": 1.4993933815839018e-06,
+      "loss": 1.0779,
+      "step": 22112
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 25.218923568725586,
+      "learning_rate": 1.4985566665272141e-06,
+      "loss": 0.7475,
+      "step": 22113
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 14.284053802490234,
+      "learning_rate": 1.4977199514705267e-06,
+      "loss": 0.9119,
+      "step": 22114
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.315624713897705,
+      "learning_rate": 1.4968832364138395e-06,
+      "loss": 0.2522,
+      "step": 22115
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 10.262313842773438,
+      "learning_rate": 1.4960465213571521e-06,
+      "loss": 0.7692,
+      "step": 22116
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 10.80276870727539,
+      "learning_rate": 1.4952098063004645e-06,
+      "loss": 0.2312,
+      "step": 22117
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 11.167405128479004,
+      "learning_rate": 1.494373091243777e-06,
+      "loss": 1.6331,
+      "step": 22118
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.160336971282959,
+      "learning_rate": 1.4935363761870897e-06,
+      "loss": 0.1514,
+      "step": 22119
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 9.625428199768066,
+      "learning_rate": 1.492699661130402e-06,
+      "loss": 0.6352,
+      "step": 22120
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.5974812507629395,
+      "learning_rate": 1.4918629460737146e-06,
+      "loss": 0.2554,
+      "step": 22121
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 55.964046478271484,
+      "learning_rate": 1.4910262310170272e-06,
+      "loss": 1.9656,
+      "step": 22122
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 33.85047149658203,
+      "learning_rate": 1.49018951596034e-06,
+      "loss": 2.8984,
+      "step": 22123
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 20.248414993286133,
+      "learning_rate": 1.4893528009036524e-06,
+      "loss": 0.7146,
+      "step": 22124
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 3.2281851768493652,
+      "learning_rate": 1.488516085846965e-06,
+      "loss": 0.0871,
+      "step": 22125
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 17.509889602661133,
+      "learning_rate": 1.4876793707902776e-06,
+      "loss": 1.3073,
+      "step": 22126
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 3.7109663486480713,
+      "learning_rate": 1.48684265573359e-06,
+      "loss": 0.177,
+      "step": 22127
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 33.91141891479492,
+      "learning_rate": 1.4860059406769026e-06,
+      "loss": 1.4068,
+      "step": 22128
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.998182773590088,
+      "learning_rate": 1.4851692256202152e-06,
+      "loss": 0.3091,
+      "step": 22129
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 99.66908264160156,
+      "learning_rate": 1.4843325105635277e-06,
+      "loss": 4.5171,
+      "step": 22130
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 12.617664337158203,
+      "learning_rate": 1.4834957955068401e-06,
+      "loss": 0.5151,
+      "step": 22131
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 2.63997220993042,
+      "learning_rate": 1.482659080450153e-06,
+      "loss": 0.0602,
+      "step": 22132
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 28.04462242126465,
+      "learning_rate": 1.4818223653934655e-06,
+      "loss": 1.3028,
+      "step": 22133
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 15.393304824829102,
+      "learning_rate": 1.4809856503367779e-06,
+      "loss": 0.7099,
+      "step": 22134
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 8.952462196350098,
+      "learning_rate": 1.4801489352800905e-06,
+      "loss": 0.232,
+      "step": 22135
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.731597423553467,
+      "learning_rate": 1.479312220223403e-06,
+      "loss": 0.2321,
+      "step": 22136
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 13.162818908691406,
+      "learning_rate": 1.4784755051667157e-06,
+      "loss": 0.7053,
+      "step": 22137
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 12.984853744506836,
+      "learning_rate": 1.477638790110028e-06,
+      "loss": 0.4655,
+      "step": 22138
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 25.313459396362305,
+      "learning_rate": 1.4768020750533406e-06,
+      "loss": 1.4556,
+      "step": 22139
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 69.79540252685547,
+      "learning_rate": 1.4759653599966534e-06,
+      "loss": 1.2859,
+      "step": 22140
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 22.690126419067383,
+      "learning_rate": 1.4751286449399658e-06,
+      "loss": 1.272,
+      "step": 22141
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 3.0274555683135986,
+      "learning_rate": 1.4742919298832784e-06,
+      "loss": 0.11,
+      "step": 22142
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 34.69709014892578,
+      "learning_rate": 1.473455214826591e-06,
+      "loss": 3.0681,
+      "step": 22143
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 18.169361114501953,
+      "learning_rate": 1.4726184997699034e-06,
+      "loss": 0.3721,
+      "step": 22144
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 14.725988388061523,
+      "learning_rate": 1.471781784713216e-06,
+      "loss": 0.6978,
+      "step": 22145
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 33.30964660644531,
+      "learning_rate": 1.4709450696565285e-06,
+      "loss": 0.3039,
+      "step": 22146
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 22.753398895263672,
+      "learning_rate": 1.4701083545998413e-06,
+      "loss": 0.7515,
+      "step": 22147
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 11.723492622375488,
+      "learning_rate": 1.4692716395431535e-06,
+      "loss": 0.3227,
+      "step": 22148
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 12.993480682373047,
+      "learning_rate": 1.4684349244864663e-06,
+      "loss": 0.3641,
+      "step": 22149
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 37.284122467041016,
+      "learning_rate": 1.467598209429779e-06,
+      "loss": 1.941,
+      "step": 22150
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 6.983309268951416,
+      "learning_rate": 1.4667614943730913e-06,
+      "loss": 0.9879,
+      "step": 22151
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 12.566207885742188,
+      "learning_rate": 1.4659247793164039e-06,
+      "loss": 0.5651,
+      "step": 22152
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 27.132915496826172,
+      "learning_rate": 1.4650880642597165e-06,
+      "loss": 2.1663,
+      "step": 22153
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 105.89178466796875,
+      "learning_rate": 1.464251349203029e-06,
+      "loss": 0.6645,
+      "step": 22154
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 15.870180130004883,
+      "learning_rate": 1.4634146341463414e-06,
+      "loss": 1.048,
+      "step": 22155
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 101.65707397460938,
+      "learning_rate": 1.4625779190896542e-06,
+      "loss": 1.8495,
+      "step": 22156
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 487.60479736328125,
+      "learning_rate": 1.4617412040329668e-06,
+      "loss": 0.9429,
+      "step": 22157
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 16.204265594482422,
+      "learning_rate": 1.4609044889762792e-06,
+      "loss": 1.0264,
+      "step": 22158
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 21.302675247192383,
+      "learning_rate": 1.4600677739195918e-06,
+      "loss": 1.2089,
+      "step": 22159
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 18.612960815429688,
+      "learning_rate": 1.4592310588629044e-06,
+      "loss": 0.7613,
+      "step": 22160
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 6.147254943847656,
+      "learning_rate": 1.458394343806217e-06,
+      "loss": 0.2872,
+      "step": 22161
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 102.18993377685547,
+      "learning_rate": 1.4575576287495294e-06,
+      "loss": 1.4707,
+      "step": 22162
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 10.75220012664795,
+      "learning_rate": 1.456720913692842e-06,
+      "loss": 0.6927,
+      "step": 22163
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 13.363828659057617,
+      "learning_rate": 1.4558841986361547e-06,
+      "loss": 0.708,
+      "step": 22164
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 15.768172264099121,
+      "learning_rate": 1.455047483579467e-06,
+      "loss": 1.7297,
+      "step": 22165
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 25.005382537841797,
+      "learning_rate": 1.4542107685227797e-06,
+      "loss": 1.0066,
+      "step": 22166
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 6.12571382522583,
+      "learning_rate": 1.4533740534660923e-06,
+      "loss": 0.1812,
+      "step": 22167
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 18.252429962158203,
+      "learning_rate": 1.4525373384094049e-06,
+      "loss": 1.1004,
+      "step": 22168
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 14.02508544921875,
+      "learning_rate": 1.4517006233527173e-06,
+      "loss": 0.4741,
+      "step": 22169
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 25.932781219482422,
+      "learning_rate": 1.4508639082960299e-06,
+      "loss": 1.9042,
+      "step": 22170
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 111.95731353759766,
+      "learning_rate": 1.4500271932393425e-06,
+      "loss": 0.8828,
+      "step": 22171
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 71.82391357421875,
+      "learning_rate": 1.4491904781826548e-06,
+      "loss": 1.8012,
+      "step": 22172
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 8.690864562988281,
+      "learning_rate": 1.4483537631259676e-06,
+      "loss": 0.8402,
+      "step": 22173
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 9.595281600952148,
+      "learning_rate": 1.4475170480692802e-06,
+      "loss": 1.3443,
+      "step": 22174
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 9.36928653717041,
+      "learning_rate": 1.4466803330125928e-06,
+      "loss": 1.1032,
+      "step": 22175
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 15.006649017333984,
+      "learning_rate": 1.4458436179559052e-06,
+      "loss": 0.9495,
+      "step": 22176
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 73.973388671875,
+      "learning_rate": 1.4450069028992178e-06,
+      "loss": 0.3064,
+      "step": 22177
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 3.6623029708862305,
+      "learning_rate": 1.4441701878425304e-06,
+      "loss": 0.2425,
+      "step": 22178
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 7.811987400054932,
+      "learning_rate": 1.4433334727858427e-06,
+      "loss": 0.4153,
+      "step": 22179
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 92.86666107177734,
+      "learning_rate": 1.4424967577291553e-06,
+      "loss": 1.6914,
+      "step": 22180
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 23.44353485107422,
+      "learning_rate": 1.4416600426724681e-06,
+      "loss": 0.8927,
+      "step": 22181
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.478631496429443,
+      "learning_rate": 1.4408233276157807e-06,
+      "loss": 0.1989,
+      "step": 22182
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 109.07917785644531,
+      "learning_rate": 1.4399866125590931e-06,
+      "loss": 2.1958,
+      "step": 22183
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 10.12816047668457,
+      "learning_rate": 1.4391498975024057e-06,
+      "loss": 0.4367,
+      "step": 22184
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 46.511898040771484,
+      "learning_rate": 1.4383131824457183e-06,
+      "loss": 1.5326,
+      "step": 22185
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.351166725158691,
+      "learning_rate": 1.4374764673890307e-06,
+      "loss": 0.4717,
+      "step": 22186
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 7.09501314163208,
+      "learning_rate": 1.4366397523323433e-06,
+      "loss": 0.5723,
+      "step": 22187
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.911225318908691,
+      "learning_rate": 1.435803037275656e-06,
+      "loss": 0.9178,
+      "step": 22188
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 36.53730392456055,
+      "learning_rate": 1.4349663222189682e-06,
+      "loss": 1.3609,
+      "step": 22189
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 10.447989463806152,
+      "learning_rate": 1.434129607162281e-06,
+      "loss": 0.8304,
+      "step": 22190
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 15.028526306152344,
+      "learning_rate": 1.4332928921055936e-06,
+      "loss": 1.0632,
+      "step": 22191
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 26.80242156982422,
+      "learning_rate": 1.4324561770489062e-06,
+      "loss": 1.0022,
+      "step": 22192
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 55.87772750854492,
+      "learning_rate": 1.4316194619922186e-06,
+      "loss": 0.5473,
+      "step": 22193
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 25.055089950561523,
+      "learning_rate": 1.4307827469355312e-06,
+      "loss": 1.14,
+      "step": 22194
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 13.183613777160645,
+      "learning_rate": 1.4299460318788438e-06,
+      "loss": 0.6271,
+      "step": 22195
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.73961067199707,
+      "learning_rate": 1.4291093168221561e-06,
+      "loss": 0.0372,
+      "step": 22196
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 13.73011302947998,
+      "learning_rate": 1.4282726017654687e-06,
+      "loss": 0.1936,
+      "step": 22197
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 22.286022186279297,
+      "learning_rate": 1.4274358867087815e-06,
+      "loss": 1.1696,
+      "step": 22198
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 31.08314323425293,
+      "learning_rate": 1.4265991716520941e-06,
+      "loss": 1.0551,
+      "step": 22199
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 18.57979965209961,
+      "learning_rate": 1.4257624565954065e-06,
+      "loss": 1.5874,
+      "step": 22200
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 45.38309097290039,
+      "learning_rate": 1.424925741538719e-06,
+      "loss": 1.421,
+      "step": 22201
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 11.388998985290527,
+      "learning_rate": 1.4240890264820317e-06,
+      "loss": 0.9901,
+      "step": 22202
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 11.183924674987793,
+      "learning_rate": 1.423252311425344e-06,
+      "loss": 0.45,
+      "step": 22203
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 10.512621879577637,
+      "learning_rate": 1.4224155963686567e-06,
+      "loss": 0.6383,
+      "step": 22204
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 20.923755645751953,
+      "learning_rate": 1.4215788813119695e-06,
+      "loss": 1.6652,
+      "step": 22205
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 22.1677188873291,
+      "learning_rate": 1.420742166255282e-06,
+      "loss": 2.2149,
+      "step": 22206
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 16.348461151123047,
+      "learning_rate": 1.4199054511985944e-06,
+      "loss": 0.9331,
+      "step": 22207
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 13.549220085144043,
+      "learning_rate": 1.419068736141907e-06,
+      "loss": 0.3617,
+      "step": 22208
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 32.703369140625,
+      "learning_rate": 1.4182320210852196e-06,
+      "loss": 2.7096,
+      "step": 22209
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 11.458334922790527,
+      "learning_rate": 1.417395306028532e-06,
+      "loss": 0.5415,
+      "step": 22210
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 31.576841354370117,
+      "learning_rate": 1.4165585909718446e-06,
+      "loss": 0.7802,
+      "step": 22211
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 16.505142211914062,
+      "learning_rate": 1.4157218759151572e-06,
+      "loss": 1.3228,
+      "step": 22212
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 10.127742767333984,
+      "learning_rate": 1.41488516085847e-06,
+      "loss": 0.22,
+      "step": 22213
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 88.93213653564453,
+      "learning_rate": 1.4140484458017823e-06,
+      "loss": 1.3596,
+      "step": 22214
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 16.036808013916016,
+      "learning_rate": 1.413211730745095e-06,
+      "loss": 1.5984,
+      "step": 22215
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 81.969970703125,
+      "learning_rate": 1.4123750156884075e-06,
+      "loss": 1.7192,
+      "step": 22216
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 9.316056251525879,
+      "learning_rate": 1.41153830063172e-06,
+      "loss": 0.2607,
+      "step": 22217
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 9.6272554397583,
+      "learning_rate": 1.4107015855750325e-06,
+      "loss": 1.7587,
+      "step": 22218
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 7.52526330947876,
+      "learning_rate": 1.409864870518345e-06,
+      "loss": 0.7524,
+      "step": 22219
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 48.97482681274414,
+      "learning_rate": 1.4090281554616579e-06,
+      "loss": 2.1355,
+      "step": 22220
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 15.028366088867188,
+      "learning_rate": 1.40819144040497e-06,
+      "loss": 0.4171,
+      "step": 22221
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 14.43149185180664,
+      "learning_rate": 1.4073547253482828e-06,
+      "loss": 0.8771,
+      "step": 22222
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 10.106249809265137,
+      "learning_rate": 1.4065180102915954e-06,
+      "loss": 0.3454,
+      "step": 22223
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 14.374917984008789,
+      "learning_rate": 1.4056812952349078e-06,
+      "loss": 0.7194,
+      "step": 22224
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 19.454383850097656,
+      "learning_rate": 1.4048445801782204e-06,
+      "loss": 0.8213,
+      "step": 22225
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 10.458917617797852,
+      "learning_rate": 1.404007865121533e-06,
+      "loss": 0.3617,
+      "step": 22226
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 11.162657737731934,
+      "learning_rate": 1.4031711500648456e-06,
+      "loss": 0.8042,
+      "step": 22227
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 19.725736618041992,
+      "learning_rate": 1.402334435008158e-06,
+      "loss": 0.8581,
+      "step": 22228
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 9.259628295898438,
+      "learning_rate": 1.4014977199514706e-06,
+      "loss": 0.5202,
+      "step": 22229
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 13.855989456176758,
+      "learning_rate": 1.4006610048947834e-06,
+      "loss": 0.539,
+      "step": 22230
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.412811756134033,
+      "learning_rate": 1.3998242898380957e-06,
+      "loss": 0.2478,
+      "step": 22231
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 12.529213905334473,
+      "learning_rate": 1.3989875747814083e-06,
+      "loss": 1.2361,
+      "step": 22232
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 16.189722061157227,
+      "learning_rate": 1.398150859724721e-06,
+      "loss": 0.3163,
+      "step": 22233
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 25.4317626953125,
+      "learning_rate": 1.3973141446680333e-06,
+      "loss": 2.3149,
+      "step": 22234
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.911479473114014,
+      "learning_rate": 1.3964774296113459e-06,
+      "loss": 0.718,
+      "step": 22235
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 14.407179832458496,
+      "learning_rate": 1.3956407145546585e-06,
+      "loss": 0.4233,
+      "step": 22236
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 152.7520294189453,
+      "learning_rate": 1.3948039994979713e-06,
+      "loss": 3.3073,
+      "step": 22237
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 26.118106842041016,
+      "learning_rate": 1.3939672844412834e-06,
+      "loss": 1.4666,
+      "step": 22238
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 7.672816276550293,
+      "learning_rate": 1.3931305693845962e-06,
+      "loss": 0.3535,
+      "step": 22239
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 10.650636672973633,
+      "learning_rate": 1.3922938543279088e-06,
+      "loss": 0.7928,
+      "step": 22240
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 9.059195518493652,
+      "learning_rate": 1.3914571392712212e-06,
+      "loss": 0.8067,
+      "step": 22241
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 7.795120716094971,
+      "learning_rate": 1.3906204242145338e-06,
+      "loss": 0.4376,
+      "step": 22242
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 16.20629119873047,
+      "learning_rate": 1.3897837091578464e-06,
+      "loss": 0.5432,
+      "step": 22243
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 8.30618667602539,
+      "learning_rate": 1.388946994101159e-06,
+      "loss": 0.5956,
+      "step": 22244
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.620741844177246,
+      "learning_rate": 1.3881102790444714e-06,
+      "loss": 0.2403,
+      "step": 22245
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 90.4376220703125,
+      "learning_rate": 1.3872735639877842e-06,
+      "loss": 0.9716,
+      "step": 22246
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 2.850315809249878,
+      "learning_rate": 1.3864368489310968e-06,
+      "loss": 0.3054,
+      "step": 22247
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 8.010313034057617,
+      "learning_rate": 1.3856001338744091e-06,
+      "loss": 0.2131,
+      "step": 22248
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 9.284424781799316,
+      "learning_rate": 1.3847634188177217e-06,
+      "loss": 0.2982,
+      "step": 22249
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 20.03277015686035,
+      "learning_rate": 1.3839267037610343e-06,
+      "loss": 1.3185,
+      "step": 22250
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 24.733230590820312,
+      "learning_rate": 1.383089988704347e-06,
+      "loss": 0.5493,
+      "step": 22251
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 13.268937110900879,
+      "learning_rate": 1.3822532736476593e-06,
+      "loss": 1.1436,
+      "step": 22252
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 14.504895210266113,
+      "learning_rate": 1.3814165585909719e-06,
+      "loss": 0.6736,
+      "step": 22253
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 7.637774467468262,
+      "learning_rate": 1.3805798435342847e-06,
+      "loss": 0.4366,
+      "step": 22254
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 50.78046798706055,
+      "learning_rate": 1.379743128477597e-06,
+      "loss": 0.9827,
+      "step": 22255
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 25.723501205444336,
+      "learning_rate": 1.3789064134209096e-06,
+      "loss": 0.8503,
+      "step": 22256
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 15.402575492858887,
+      "learning_rate": 1.3780696983642222e-06,
+      "loss": 0.5884,
+      "step": 22257
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 23.20662498474121,
+      "learning_rate": 1.3772329833075348e-06,
+      "loss": 2.1696,
+      "step": 22258
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 12.273051261901855,
+      "learning_rate": 1.3763962682508472e-06,
+      "loss": 2.4901,
+      "step": 22259
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 24.371145248413086,
+      "learning_rate": 1.3755595531941598e-06,
+      "loss": 1.1444,
+      "step": 22260
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 6.011651039123535,
+      "learning_rate": 1.3747228381374724e-06,
+      "loss": 0.4818,
+      "step": 22261
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 16.847021102905273,
+      "learning_rate": 1.3738861230807848e-06,
+      "loss": 0.4117,
+      "step": 22262
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 6.79587984085083,
+      "learning_rate": 1.3730494080240976e-06,
+      "loss": 0.7666,
+      "step": 22263
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 26.817100524902344,
+      "learning_rate": 1.3722126929674101e-06,
+      "loss": 1.3122,
+      "step": 22264
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 23.336999893188477,
+      "learning_rate": 1.3713759779107227e-06,
+      "loss": 2.0669,
+      "step": 22265
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 7.358304500579834,
+      "learning_rate": 1.3705392628540351e-06,
+      "loss": 0.3368,
+      "step": 22266
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 26.694063186645508,
+      "learning_rate": 1.3697025477973477e-06,
+      "loss": 0.399,
+      "step": 22267
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 23.938920974731445,
+      "learning_rate": 1.3688658327406603e-06,
+      "loss": 1.0198,
+      "step": 22268
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 21.425411224365234,
+      "learning_rate": 1.3680291176839727e-06,
+      "loss": 0.7446,
+      "step": 22269
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 11.50066089630127,
+      "learning_rate": 1.3671924026272853e-06,
+      "loss": 0.9048,
+      "step": 22270
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 9.022086143493652,
+      "learning_rate": 1.366355687570598e-06,
+      "loss": 0.205,
+      "step": 22271
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 113.79344940185547,
+      "learning_rate": 1.3655189725139107e-06,
+      "loss": 2.931,
+      "step": 22272
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 12.153200149536133,
+      "learning_rate": 1.364682257457223e-06,
+      "loss": 0.7875,
+      "step": 22273
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 30.668458938598633,
+      "learning_rate": 1.3638455424005356e-06,
+      "loss": 0.8056,
+      "step": 22274
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 39.18879318237305,
+      "learning_rate": 1.3630088273438482e-06,
+      "loss": 1.0483,
+      "step": 22275
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 11.61953353881836,
+      "learning_rate": 1.3621721122871606e-06,
+      "loss": 0.9745,
+      "step": 22276
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 10.363846778869629,
+      "learning_rate": 1.3613353972304732e-06,
+      "loss": 0.2994,
+      "step": 22277
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 10.151551246643066,
+      "learning_rate": 1.360498682173786e-06,
+      "loss": 0.571,
+      "step": 22278
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 17.394407272338867,
+      "learning_rate": 1.3596619671170982e-06,
+      "loss": 1.6727,
+      "step": 22279
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 19.183019638061523,
+      "learning_rate": 1.358825252060411e-06,
+      "loss": 1.3546,
+      "step": 22280
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 12.273883819580078,
+      "learning_rate": 1.3579885370037235e-06,
+      "loss": 1.5346,
+      "step": 22281
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 11.285651206970215,
+      "learning_rate": 1.3571518219470361e-06,
+      "loss": 2.372,
+      "step": 22282
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 27.698572158813477,
+      "learning_rate": 1.3563151068903485e-06,
+      "loss": 2.4062,
+      "step": 22283
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 18.60379409790039,
+      "learning_rate": 1.355478391833661e-06,
+      "loss": 0.8674,
+      "step": 22284
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 23.93038558959961,
+      "learning_rate": 1.3546416767769737e-06,
+      "loss": 1.1512,
+      "step": 22285
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 13.560357093811035,
+      "learning_rate": 1.353804961720286e-06,
+      "loss": 1.3127,
+      "step": 22286
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 41.231143951416016,
+      "learning_rate": 1.3529682466635989e-06,
+      "loss": 0.8257,
+      "step": 22287
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 6.1269731521606445,
+      "learning_rate": 1.3521315316069115e-06,
+      "loss": 0.295,
+      "step": 22288
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 38.05921936035156,
+      "learning_rate": 1.351294816550224e-06,
+      "loss": 0.5179,
+      "step": 22289
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 7.305657863616943,
+      "learning_rate": 1.3504581014935364e-06,
+      "loss": 0.1885,
+      "step": 22290
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 58.65455627441406,
+      "learning_rate": 1.349621386436849e-06,
+      "loss": 0.8713,
+      "step": 22291
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 7.845381259918213,
+      "learning_rate": 1.3487846713801616e-06,
+      "loss": 0.4956,
+      "step": 22292
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 6.775365352630615,
+      "learning_rate": 1.347947956323474e-06,
+      "loss": 0.5024,
+      "step": 22293
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 28.21062469482422,
+      "learning_rate": 1.3471112412667866e-06,
+      "loss": 1.7029,
+      "step": 22294
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 22.696487426757812,
+      "learning_rate": 1.3462745262100994e-06,
+      "loss": 1.1901,
+      "step": 22295
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 8.468282699584961,
+      "learning_rate": 1.345437811153412e-06,
+      "loss": 1.3211,
+      "step": 22296
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 8.390238761901855,
+      "learning_rate": 1.3446010960967243e-06,
+      "loss": 0.5668,
+      "step": 22297
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 10.155052185058594,
+      "learning_rate": 1.343764381040037e-06,
+      "loss": 0.3998,
+      "step": 22298
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 22.792644500732422,
+      "learning_rate": 1.3429276659833495e-06,
+      "loss": 0.4127,
+      "step": 22299
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 6.965851783752441,
+      "learning_rate": 1.342090950926662e-06,
+      "loss": 0.6262,
+      "step": 22300
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 28.035646438598633,
+      "learning_rate": 1.3412542358699745e-06,
+      "loss": 0.9244,
+      "step": 22301
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 17.093761444091797,
+      "learning_rate": 1.340417520813287e-06,
+      "loss": 0.9921,
+      "step": 22302
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 42.652687072753906,
+      "learning_rate": 1.3395808057565999e-06,
+      "loss": 0.8933,
+      "step": 22303
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 8.79464054107666,
+      "learning_rate": 1.3387440906999123e-06,
+      "loss": 0.3885,
+      "step": 22304
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 7.84977912902832,
+      "learning_rate": 1.3379073756432249e-06,
+      "loss": 0.4833,
+      "step": 22305
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 8.520516395568848,
+      "learning_rate": 1.3370706605865374e-06,
+      "loss": 1.4362,
+      "step": 22306
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 9.85045337677002,
+      "learning_rate": 1.3362339455298498e-06,
+      "loss": 1.3511,
+      "step": 22307
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 9.924347877502441,
+      "learning_rate": 1.3353972304731624e-06,
+      "loss": 0.9138,
+      "step": 22308
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 64.50171661376953,
+      "learning_rate": 1.334560515416475e-06,
+      "loss": 1.7466,
+      "step": 22309
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 69.96508026123047,
+      "learning_rate": 1.3337238003597878e-06,
+      "loss": 1.3254,
+      "step": 22310
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 17.21268081665039,
+      "learning_rate": 1.3328870853031e-06,
+      "loss": 1.1874,
+      "step": 22311
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 20.13076400756836,
+      "learning_rate": 1.3320503702464128e-06,
+      "loss": 1.0865,
+      "step": 22312
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 3.721005916595459,
+      "learning_rate": 1.3312136551897254e-06,
+      "loss": 0.1926,
+      "step": 22313
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 14.235478401184082,
+      "learning_rate": 1.3303769401330377e-06,
+      "loss": 0.7383,
+      "step": 22314
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 14.641931533813477,
+      "learning_rate": 1.3295402250763503e-06,
+      "loss": 1.0961,
+      "step": 22315
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 12.710472106933594,
+      "learning_rate": 1.328703510019663e-06,
+      "loss": 0.7244,
+      "step": 22316
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 19.97145652770996,
+      "learning_rate": 1.3278667949629755e-06,
+      "loss": 1.1615,
+      "step": 22317
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 8.688436508178711,
+      "learning_rate": 1.3270300799062879e-06,
+      "loss": 1.3011,
+      "step": 22318
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 6.827741622924805,
+      "learning_rate": 1.3261933648496007e-06,
+      "loss": 0.4903,
+      "step": 22319
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 10.897211074829102,
+      "learning_rate": 1.3253566497929133e-06,
+      "loss": 0.5886,
+      "step": 22320
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 101.98963928222656,
+      "learning_rate": 1.3245199347362257e-06,
+      "loss": 2.3322,
+      "step": 22321
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 51.10252380371094,
+      "learning_rate": 1.3236832196795383e-06,
+      "loss": 0.8964,
+      "step": 22322
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 11.098581314086914,
+      "learning_rate": 1.3228465046228508e-06,
+      "loss": 0.7318,
+      "step": 22323
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 63.52899169921875,
+      "learning_rate": 1.3220097895661632e-06,
+      "loss": 1.7013,
+      "step": 22324
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 7.581141471862793,
+      "learning_rate": 1.3211730745094758e-06,
+      "loss": 1.4585,
+      "step": 22325
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 19.211050033569336,
+      "learning_rate": 1.3203363594527884e-06,
+      "loss": 1.2471,
+      "step": 22326
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 6.085424900054932,
+      "learning_rate": 1.3194996443961012e-06,
+      "loss": 0.6361,
+      "step": 22327
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 9.494768142700195,
+      "learning_rate": 1.3186629293394136e-06,
+      "loss": 1.0115,
+      "step": 22328
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 12.25251293182373,
+      "learning_rate": 1.3178262142827262e-06,
+      "loss": 0.277,
+      "step": 22329
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 10.640769004821777,
+      "learning_rate": 1.3169894992260388e-06,
+      "loss": 1.3467,
+      "step": 22330
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 349.9528503417969,
+      "learning_rate": 1.3161527841693511e-06,
+      "loss": 1.6585,
+      "step": 22331
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 32.06199645996094,
+      "learning_rate": 1.3153160691126637e-06,
+      "loss": 2.6387,
+      "step": 22332
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 16.242982864379883,
+      "learning_rate": 1.3144793540559763e-06,
+      "loss": 0.5804,
+      "step": 22333
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 7.629415035247803,
+      "learning_rate": 1.313642638999289e-06,
+      "loss": 0.6288,
+      "step": 22334
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 26.868345260620117,
+      "learning_rate": 1.3128059239426013e-06,
+      "loss": 1.5236,
+      "step": 22335
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 7.834210395812988,
+      "learning_rate": 1.311969208885914e-06,
+      "loss": 0.3088,
+      "step": 22336
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 8.49197769165039,
+      "learning_rate": 1.3111324938292267e-06,
+      "loss": 0.5462,
+      "step": 22337
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 15.133360862731934,
+      "learning_rate": 1.310295778772539e-06,
+      "loss": 2.2611,
+      "step": 22338
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 3.2791247367858887,
+      "learning_rate": 1.3094590637158516e-06,
+      "loss": 0.2174,
+      "step": 22339
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 2.5591752529144287,
+      "learning_rate": 1.3086223486591642e-06,
+      "loss": 0.1529,
+      "step": 22340
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 24.870407104492188,
+      "learning_rate": 1.3077856336024768e-06,
+      "loss": 0.9089,
+      "step": 22341
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 20.577219009399414,
+      "learning_rate": 1.3069489185457892e-06,
+      "loss": 1.445,
+      "step": 22342
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 17.739303588867188,
+      "learning_rate": 1.3061122034891018e-06,
+      "loss": 0.8651,
+      "step": 22343
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 30.139007568359375,
+      "learning_rate": 1.3052754884324146e-06,
+      "loss": 0.9499,
+      "step": 22344
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 19.0602970123291,
+      "learning_rate": 1.304438773375727e-06,
+      "loss": 1.2218,
+      "step": 22345
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 19.527162551879883,
+      "learning_rate": 1.3036020583190396e-06,
+      "loss": 0.6184,
+      "step": 22346
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 7.0958099365234375,
+      "learning_rate": 1.3027653432623522e-06,
+      "loss": 0.2741,
+      "step": 22347
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.984929084777832,
+      "learning_rate": 1.3019286282056647e-06,
+      "loss": 0.2121,
+      "step": 22348
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 15.76543140411377,
+      "learning_rate": 1.3010919131489771e-06,
+      "loss": 1.2545,
+      "step": 22349
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.139402389526367,
+      "learning_rate": 1.3002551980922897e-06,
+      "loss": 0.2187,
+      "step": 22350
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 8.716383934020996,
+      "learning_rate": 1.2994184830356025e-06,
+      "loss": 1.4159,
+      "step": 22351
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 91.85513305664062,
+      "learning_rate": 1.2985817679789147e-06,
+      "loss": 2.825,
+      "step": 22352
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 16.891550064086914,
+      "learning_rate": 1.2977450529222275e-06,
+      "loss": 0.548,
+      "step": 22353
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 27.4136962890625,
+      "learning_rate": 1.29690833786554e-06,
+      "loss": 0.7632,
+      "step": 22354
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 15.752585411071777,
+      "learning_rate": 1.2960716228088527e-06,
+      "loss": 1.5803,
+      "step": 22355
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 5.520205974578857,
+      "learning_rate": 1.295234907752165e-06,
+      "loss": 0.3901,
+      "step": 22356
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 35.271385192871094,
+      "learning_rate": 1.2943981926954776e-06,
+      "loss": 0.5498,
+      "step": 22357
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 11.174712181091309,
+      "learning_rate": 1.2935614776387902e-06,
+      "loss": 0.938,
+      "step": 22358
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 6.589173793792725,
+      "learning_rate": 1.2927247625821026e-06,
+      "loss": 0.5047,
+      "step": 22359
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 68.62174224853516,
+      "learning_rate": 1.2918880475254154e-06,
+      "loss": 0.5637,
+      "step": 22360
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 9.69046401977539,
+      "learning_rate": 1.291051332468728e-06,
+      "loss": 0.5623,
+      "step": 22361
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 34.01222610473633,
+      "learning_rate": 1.2902146174120406e-06,
+      "loss": 0.5374,
+      "step": 22362
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 3.726040840148926,
+      "learning_rate": 1.289377902355353e-06,
+      "loss": 0.339,
+      "step": 22363
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 29.50176239013672,
+      "learning_rate": 1.2885411872986656e-06,
+      "loss": 2.0108,
+      "step": 22364
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 16.313310623168945,
+      "learning_rate": 1.2877044722419781e-06,
+      "loss": 0.6736,
+      "step": 22365
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 7.732245445251465,
+      "learning_rate": 1.2868677571852905e-06,
+      "loss": 1.1685,
+      "step": 22366
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 14.34288501739502,
+      "learning_rate": 1.2860310421286031e-06,
+      "loss": 0.5316,
+      "step": 22367
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 20.638566970825195,
+      "learning_rate": 1.285194327071916e-06,
+      "loss": 1.3243,
+      "step": 22368
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 13.029252052307129,
+      "learning_rate": 1.284357612015228e-06,
+      "loss": 1.3098,
+      "step": 22369
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 19.74635887145996,
+      "learning_rate": 1.2835208969585409e-06,
+      "loss": 1.8711,
+      "step": 22370
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 22.1148738861084,
+      "learning_rate": 1.2826841819018535e-06,
+      "loss": 0.9669,
+      "step": 22371
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 17.366498947143555,
+      "learning_rate": 1.281847466845166e-06,
+      "loss": 0.7144,
+      "step": 22372
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 9.708720207214355,
+      "learning_rate": 1.2810107517884784e-06,
+      "loss": 0.272,
+      "step": 22373
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 14.64260482788086,
+      "learning_rate": 1.280174036731791e-06,
+      "loss": 0.9609,
+      "step": 22374
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 38.372615814208984,
+      "learning_rate": 1.2793373216751036e-06,
+      "loss": 1.3878,
+      "step": 22375
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 6.752889633178711,
+      "learning_rate": 1.278500606618416e-06,
+      "loss": 0.0768,
+      "step": 22376
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 64.34168243408203,
+      "learning_rate": 1.2776638915617288e-06,
+      "loss": 0.9934,
+      "step": 22377
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 8.978327751159668,
+      "learning_rate": 1.2768271765050414e-06,
+      "loss": 2.2121,
+      "step": 22378
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 80.38655853271484,
+      "learning_rate": 1.275990461448354e-06,
+      "loss": 1.3939,
+      "step": 22379
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 7.601134300231934,
+      "learning_rate": 1.2751537463916664e-06,
+      "loss": 0.8712,
+      "step": 22380
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 12.885122299194336,
+      "learning_rate": 1.274317031334979e-06,
+      "loss": 0.5965,
+      "step": 22381
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 60.07326126098633,
+      "learning_rate": 1.2734803162782915e-06,
+      "loss": 1.6777,
+      "step": 22382
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 33.034019470214844,
+      "learning_rate": 1.272643601221604e-06,
+      "loss": 0.5645,
+      "step": 22383
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 8.753458023071289,
+      "learning_rate": 1.2718068861649165e-06,
+      "loss": 0.2947,
+      "step": 22384
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 7.3992156982421875,
+      "learning_rate": 1.2709701711082293e-06,
+      "loss": 0.2016,
+      "step": 22385
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 31.3106689453125,
+      "learning_rate": 1.270133456051542e-06,
+      "loss": 1.7178,
+      "step": 22386
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 17.78363037109375,
+      "learning_rate": 1.2692967409948543e-06,
+      "loss": 0.6034,
+      "step": 22387
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 11.05273151397705,
+      "learning_rate": 1.2684600259381669e-06,
+      "loss": 1.2865,
+      "step": 22388
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 6.84112024307251,
+      "learning_rate": 1.2676233108814795e-06,
+      "loss": 0.2924,
+      "step": 22389
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 27.54936408996582,
+      "learning_rate": 1.2667865958247918e-06,
+      "loss": 1.6276,
+      "step": 22390
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 40.96695327758789,
+      "learning_rate": 1.2659498807681044e-06,
+      "loss": 1.1847,
+      "step": 22391
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 22.234130859375,
+      "learning_rate": 1.2651131657114172e-06,
+      "loss": 0.8053,
+      "step": 22392
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 21.5286865234375,
+      "learning_rate": 1.2642764506547298e-06,
+      "loss": 2.3074,
+      "step": 22393
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 23.00366973876953,
+      "learning_rate": 1.2634397355980422e-06,
+      "loss": 1.866,
+      "step": 22394
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 11.963624954223633,
+      "learning_rate": 1.2626030205413548e-06,
+      "loss": 0.525,
+      "step": 22395
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 7.500345230102539,
+      "learning_rate": 1.2617663054846674e-06,
+      "loss": 0.2791,
+      "step": 22396
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 40.571903228759766,
+      "learning_rate": 1.2609295904279798e-06,
+      "loss": 1.9009,
+      "step": 22397
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 10.458895683288574,
+      "learning_rate": 1.2600928753712923e-06,
+      "loss": 0.7065,
+      "step": 22398
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 21.273595809936523,
+      "learning_rate": 1.259256160314605e-06,
+      "loss": 0.6852,
+      "step": 22399
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 9.218423843383789,
+      "learning_rate": 1.2584194452579177e-06,
+      "loss": 0.3528,
+      "step": 22400
+    },
+    {
+      "epoch": 2.81,
+      "eval_loss": 0.07806830108165741,
+      "eval_runtime": 96.4114,
+      "eval_samples_per_second": 36.738,
+      "eval_steps_per_second": 36.738,
+      "step": 22400
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 23.908506393432617,
+      "learning_rate": 1.25758273020123e-06,
+      "loss": 1.2257,
+      "step": 22401
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.9419846534729,
+      "learning_rate": 1.2567460151445427e-06,
+      "loss": 0.072,
+      "step": 22402
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 19.66534423828125,
+      "learning_rate": 1.2559093000878553e-06,
+      "loss": 2.4977,
+      "step": 22403
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 11.625102043151855,
+      "learning_rate": 1.2550725850311677e-06,
+      "loss": 0.7169,
+      "step": 22404
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 11.85889720916748,
+      "learning_rate": 1.2542358699744803e-06,
+      "loss": 0.22,
+      "step": 22405
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 22.91364288330078,
+      "learning_rate": 1.2533991549177929e-06,
+      "loss": 0.5825,
+      "step": 22406
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 14.040718078613281,
+      "learning_rate": 1.2525624398611054e-06,
+      "loss": 0.7661,
+      "step": 22407
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 14.2266263961792,
+      "learning_rate": 1.2517257248044178e-06,
+      "loss": 1.4066,
+      "step": 22408
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 105.7550048828125,
+      "learning_rate": 1.2508890097477306e-06,
+      "loss": 2.0166,
+      "step": 22409
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 18.059402465820312,
+      "learning_rate": 1.2500522946910432e-06,
+      "loss": 0.7641,
+      "step": 22410
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 8.43260669708252,
+      "learning_rate": 1.2492155796343556e-06,
+      "loss": 0.362,
+      "step": 22411
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 20.3315372467041,
+      "learning_rate": 1.2483788645776682e-06,
+      "loss": 1.883,
+      "step": 22412
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 134.1221923828125,
+      "learning_rate": 1.2475421495209808e-06,
+      "loss": 0.3431,
+      "step": 22413
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 24.616275787353516,
+      "learning_rate": 1.2467054344642934e-06,
+      "loss": 1.5184,
+      "step": 22414
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 10.436813354492188,
+      "learning_rate": 1.2458687194076057e-06,
+      "loss": 0.7298,
+      "step": 22415
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 14.31775188446045,
+      "learning_rate": 1.2450320043509183e-06,
+      "loss": 0.9045,
+      "step": 22416
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 29.62679672241211,
+      "learning_rate": 1.244195289294231e-06,
+      "loss": 0.5825,
+      "step": 22417
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 3.6407570838928223,
+      "learning_rate": 1.2433585742375435e-06,
+      "loss": 0.1671,
+      "step": 22418
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 16.576555252075195,
+      "learning_rate": 1.242521859180856e-06,
+      "loss": 0.4805,
+      "step": 22419
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 12.0615234375,
+      "learning_rate": 1.2416851441241687e-06,
+      "loss": 0.6741,
+      "step": 22420
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 18.188894271850586,
+      "learning_rate": 1.2408484290674813e-06,
+      "loss": 0.4955,
+      "step": 22421
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 10.029412269592285,
+      "learning_rate": 1.2400117140107937e-06,
+      "loss": 0.853,
+      "step": 22422
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 3.855475902557373,
+      "learning_rate": 1.2391749989541062e-06,
+      "loss": 0.3004,
+      "step": 22423
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 111.1025619506836,
+      "learning_rate": 1.2383382838974188e-06,
+      "loss": 2.0545,
+      "step": 22424
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 58.22159194946289,
+      "learning_rate": 1.2375015688407314e-06,
+      "loss": 1.6608,
+      "step": 22425
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 30.532649993896484,
+      "learning_rate": 1.236664853784044e-06,
+      "loss": 1.4307,
+      "step": 22426
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 12.674906730651855,
+      "learning_rate": 1.2358281387273564e-06,
+      "loss": 0.8895,
+      "step": 22427
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 30.858558654785156,
+      "learning_rate": 1.234991423670669e-06,
+      "loss": 0.9602,
+      "step": 22428
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 95.74943542480469,
+      "learning_rate": 1.2341547086139816e-06,
+      "loss": 1.8953,
+      "step": 22429
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 7.380533218383789,
+      "learning_rate": 1.2333179935572942e-06,
+      "loss": 2.7785,
+      "step": 22430
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 3.821453809738159,
+      "learning_rate": 1.2324812785006068e-06,
+      "loss": 0.2935,
+      "step": 22431
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 12.30856990814209,
+      "learning_rate": 1.2316445634439193e-06,
+      "loss": 1.5786,
+      "step": 22432
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 8.218766212463379,
+      "learning_rate": 1.230807848387232e-06,
+      "loss": 0.0842,
+      "step": 22433
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 9.76360034942627,
+      "learning_rate": 1.2299711333305443e-06,
+      "loss": 0.7282,
+      "step": 22434
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 11.711906433105469,
+      "learning_rate": 1.229134418273857e-06,
+      "loss": 0.5736,
+      "step": 22435
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.1550116539001465,
+      "learning_rate": 1.2282977032171695e-06,
+      "loss": 0.1218,
+      "step": 22436
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 41.38620376586914,
+      "learning_rate": 1.227460988160482e-06,
+      "loss": 1.7576,
+      "step": 22437
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 19.813682556152344,
+      "learning_rate": 1.2266242731037947e-06,
+      "loss": 0.6346,
+      "step": 22438
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 5.613452434539795,
+      "learning_rate": 1.2257875580471073e-06,
+      "loss": 0.4866,
+      "step": 22439
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 18.13245391845703,
+      "learning_rate": 1.2249508429904196e-06,
+      "loss": 1.4789,
+      "step": 22440
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 11.131356239318848,
+      "learning_rate": 1.2241141279337322e-06,
+      "loss": 0.5981,
+      "step": 22441
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 19.741561889648438,
+      "learning_rate": 1.2232774128770448e-06,
+      "loss": 0.8035,
+      "step": 22442
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.2421314716339111,
+      "learning_rate": 1.2224406978203574e-06,
+      "loss": 0.0455,
+      "step": 22443
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 3.521859884262085,
+      "learning_rate": 1.22160398276367e-06,
+      "loss": 0.0857,
+      "step": 22444
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 91.45506286621094,
+      "learning_rate": 1.2207672677069824e-06,
+      "loss": 2.0198,
+      "step": 22445
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 25.695064544677734,
+      "learning_rate": 1.219930552650295e-06,
+      "loss": 0.8741,
+      "step": 22446
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 20.033838272094727,
+      "learning_rate": 1.2190938375936076e-06,
+      "loss": 0.6432,
+      "step": 22447
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 7.642991065979004,
+      "learning_rate": 1.2182571225369202e-06,
+      "loss": 0.6744,
+      "step": 22448
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 14.59307861328125,
+      "learning_rate": 1.2174204074802327e-06,
+      "loss": 2.1747,
+      "step": 22449
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 2.3779585361480713,
+      "learning_rate": 1.2165836924235453e-06,
+      "loss": 0.069,
+      "step": 22450
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 22.025114059448242,
+      "learning_rate": 1.215746977366858e-06,
+      "loss": 0.9735,
+      "step": 22451
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 12.39184856414795,
+      "learning_rate": 1.2149102623101703e-06,
+      "loss": 1.1784,
+      "step": 22452
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 21.25431251525879,
+      "learning_rate": 1.2140735472534829e-06,
+      "loss": 1.6685,
+      "step": 22453
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 17.156089782714844,
+      "learning_rate": 1.2132368321967955e-06,
+      "loss": 0.6898,
+      "step": 22454
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 14.869438171386719,
+      "learning_rate": 1.212400117140108e-06,
+      "loss": 0.7735,
+      "step": 22455
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 16.35450553894043,
+      "learning_rate": 1.2115634020834207e-06,
+      "loss": 1.148,
+      "step": 22456
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 5.551679611206055,
+      "learning_rate": 1.210726687026733e-06,
+      "loss": 0.331,
+      "step": 22457
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 333.46478271484375,
+      "learning_rate": 1.2098899719700458e-06,
+      "loss": 0.785,
+      "step": 22458
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 10.77422046661377,
+      "learning_rate": 1.2090532569133582e-06,
+      "loss": 0.238,
+      "step": 22459
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 9.586767196655273,
+      "learning_rate": 1.2082165418566708e-06,
+      "loss": 1.4508,
+      "step": 22460
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 11.997692108154297,
+      "learning_rate": 1.2073798267999834e-06,
+      "loss": 2.2217,
+      "step": 22461
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 11.284348487854004,
+      "learning_rate": 1.206543111743296e-06,
+      "loss": 0.2131,
+      "step": 22462
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 9.041314125061035,
+      "learning_rate": 1.2057063966866086e-06,
+      "loss": 1.3704,
+      "step": 22463
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 95.8100814819336,
+      "learning_rate": 1.204869681629921e-06,
+      "loss": 0.6106,
+      "step": 22464
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 9.691165924072266,
+      "learning_rate": 1.2040329665732338e-06,
+      "loss": 0.3612,
+      "step": 22465
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 9.955235481262207,
+      "learning_rate": 1.2031962515165461e-06,
+      "loss": 0.5314,
+      "step": 22466
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 26.02245330810547,
+      "learning_rate": 1.2023595364598587e-06,
+      "loss": 3.2644,
+      "step": 22467
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 6.988763332366943,
+      "learning_rate": 1.2015228214031713e-06,
+      "loss": 0.9097,
+      "step": 22468
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 6.152822494506836,
+      "learning_rate": 1.2006861063464837e-06,
+      "loss": 0.4453,
+      "step": 22469
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 21.691246032714844,
+      "learning_rate": 1.1998493912897965e-06,
+      "loss": 0.8282,
+      "step": 22470
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 7.050600051879883,
+      "learning_rate": 1.1990126762331089e-06,
+      "loss": 0.1907,
+      "step": 22471
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 74.11097717285156,
+      "learning_rate": 1.1981759611764215e-06,
+      "loss": 0.5265,
+      "step": 22472
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 9.140049934387207,
+      "learning_rate": 1.197339246119734e-06,
+      "loss": 0.6186,
+      "step": 22473
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 16.83196449279785,
+      "learning_rate": 1.1965025310630464e-06,
+      "loss": 0.4343,
+      "step": 22474
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 19.166473388671875,
+      "learning_rate": 1.1956658160063592e-06,
+      "loss": 1.0129,
+      "step": 22475
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 27.66779327392578,
+      "learning_rate": 1.1948291009496716e-06,
+      "loss": 0.8765,
+      "step": 22476
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 6.122854232788086,
+      "learning_rate": 1.1939923858929842e-06,
+      "loss": 0.2845,
+      "step": 22477
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 13.45590877532959,
+      "learning_rate": 1.1931556708362968e-06,
+      "loss": 1.699,
+      "step": 22478
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 19.627155303955078,
+      "learning_rate": 1.1923189557796094e-06,
+      "loss": 0.7282,
+      "step": 22479
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 7.944036483764648,
+      "learning_rate": 1.191482240722922e-06,
+      "loss": 1.7327,
+      "step": 22480
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 19.45163345336914,
+      "learning_rate": 1.1906455256662344e-06,
+      "loss": 0.7671,
+      "step": 22481
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 9.033167839050293,
+      "learning_rate": 1.1898088106095472e-06,
+      "loss": 0.2838,
+      "step": 22482
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 40.726802825927734,
+      "learning_rate": 1.1889720955528595e-06,
+      "loss": 1.4657,
+      "step": 22483
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 7.455557823181152,
+      "learning_rate": 1.1881353804961721e-06,
+      "loss": 0.4454,
+      "step": 22484
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 9.799588203430176,
+      "learning_rate": 1.1872986654394847e-06,
+      "loss": 1.0009,
+      "step": 22485
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 11.131878852844238,
+      "learning_rate": 1.186461950382797e-06,
+      "loss": 1.3634,
+      "step": 22486
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 72.7829818725586,
+      "learning_rate": 1.1856252353261099e-06,
+      "loss": 1.2031,
+      "step": 22487
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 17.33318519592285,
+      "learning_rate": 1.1847885202694223e-06,
+      "loss": 0.3565,
+      "step": 22488
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 7.278853416442871,
+      "learning_rate": 1.1839518052127349e-06,
+      "loss": 0.326,
+      "step": 22489
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 16.62299346923828,
+      "learning_rate": 1.1831150901560474e-06,
+      "loss": 0.5297,
+      "step": 22490
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 24.486923217773438,
+      "learning_rate": 1.18227837509936e-06,
+      "loss": 1.6779,
+      "step": 22491
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 14.43045711517334,
+      "learning_rate": 1.1814416600426726e-06,
+      "loss": 1.7082,
+      "step": 22492
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 18.77744483947754,
+      "learning_rate": 1.180604944985985e-06,
+      "loss": 0.8335,
+      "step": 22493
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 8.909994125366211,
+      "learning_rate": 1.1797682299292978e-06,
+      "loss": 1.7434,
+      "step": 22494
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 2.27986741065979,
+      "learning_rate": 1.1789315148726102e-06,
+      "loss": 0.1914,
+      "step": 22495
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 11.619105339050293,
+      "learning_rate": 1.1780947998159228e-06,
+      "loss": 1.1619,
+      "step": 22496
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 34.9727783203125,
+      "learning_rate": 1.1772580847592354e-06,
+      "loss": 1.296,
+      "step": 22497
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 7.1914167404174805,
+      "learning_rate": 1.1764213697025477e-06,
+      "loss": 0.1431,
+      "step": 22498
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 13.69640064239502,
+      "learning_rate": 1.1755846546458605e-06,
+      "loss": 0.7162,
+      "step": 22499
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 28.79456329345703,
+      "learning_rate": 1.174747939589173e-06,
+      "loss": 1.167,
+      "step": 22500
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 42.08561706542969,
+      "learning_rate": 1.1739112245324855e-06,
+      "loss": 1.3198,
+      "step": 22501
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 20.690153121948242,
+      "learning_rate": 1.1730745094757981e-06,
+      "loss": 2.3716,
+      "step": 22502
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 48.5506591796875,
+      "learning_rate": 1.1722377944191107e-06,
+      "loss": 1.4877,
+      "step": 22503
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 17.76114845275879,
+      "learning_rate": 1.1714010793624233e-06,
+      "loss": 0.366,
+      "step": 22504
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 20.296478271484375,
+      "learning_rate": 1.1705643643057357e-06,
+      "loss": 1.1978,
+      "step": 22505
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 5.036375999450684,
+      "learning_rate": 1.1697276492490483e-06,
+      "loss": 0.1301,
+      "step": 22506
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 31.5634822845459,
+      "learning_rate": 1.1688909341923608e-06,
+      "loss": 1.8502,
+      "step": 22507
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 14.203628540039062,
+      "learning_rate": 1.1680542191356734e-06,
+      "loss": 0.7201,
+      "step": 22508
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 25.401737213134766,
+      "learning_rate": 1.167217504078986e-06,
+      "loss": 1.0927,
+      "step": 22509
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 26.5362606048584,
+      "learning_rate": 1.1663807890222986e-06,
+      "loss": 1.4557,
+      "step": 22510
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 39.88875961303711,
+      "learning_rate": 1.1655440739656112e-06,
+      "loss": 4.3725,
+      "step": 22511
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 11.070425033569336,
+      "learning_rate": 1.1647073589089236e-06,
+      "loss": 0.9716,
+      "step": 22512
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 31.722532272338867,
+      "learning_rate": 1.1638706438522362e-06,
+      "loss": 1.8316,
+      "step": 22513
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 28.786523818969727,
+      "learning_rate": 1.1630339287955488e-06,
+      "loss": 2.101,
+      "step": 22514
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 17.756635665893555,
+      "learning_rate": 1.1621972137388614e-06,
+      "loss": 1.3284,
+      "step": 22515
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 18.357152938842773,
+      "learning_rate": 1.161360498682174e-06,
+      "loss": 1.6907,
+      "step": 22516
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 21.013147354125977,
+      "learning_rate": 1.1605237836254863e-06,
+      "loss": 0.8034,
+      "step": 22517
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 14.976155281066895,
+      "learning_rate": 1.159687068568799e-06,
+      "loss": 1.068,
+      "step": 22518
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 102.81354522705078,
+      "learning_rate": 1.1588503535121115e-06,
+      "loss": 1.9243,
+      "step": 22519
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 14.17033863067627,
+      "learning_rate": 1.158013638455424e-06,
+      "loss": 1.1028,
+      "step": 22520
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 133.36195373535156,
+      "learning_rate": 1.1571769233987367e-06,
+      "loss": 1.0503,
+      "step": 22521
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 18.195566177368164,
+      "learning_rate": 1.1563402083420493e-06,
+      "loss": 1.143,
+      "step": 22522
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 15.236781120300293,
+      "learning_rate": 1.1555034932853619e-06,
+      "loss": 0.828,
+      "step": 22523
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 23.67279624938965,
+      "learning_rate": 1.1546667782286742e-06,
+      "loss": 0.6793,
+      "step": 22524
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 27.306930541992188,
+      "learning_rate": 1.1538300631719868e-06,
+      "loss": 1.3672,
+      "step": 22525
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 9.656767845153809,
+      "learning_rate": 1.1529933481152994e-06,
+      "loss": 0.627,
+      "step": 22526
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 14.961053848266602,
+      "learning_rate": 1.152156633058612e-06,
+      "loss": 1.361,
+      "step": 22527
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 20.277849197387695,
+      "learning_rate": 1.1513199180019246e-06,
+      "loss": 1.0639,
+      "step": 22528
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 22.297962188720703,
+      "learning_rate": 1.1504832029452372e-06,
+      "loss": 2.0858,
+      "step": 22529
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 2.163911819458008,
+      "learning_rate": 1.1496464878885496e-06,
+      "loss": 0.1031,
+      "step": 22530
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 15.617265701293945,
+      "learning_rate": 1.1488097728318622e-06,
+      "loss": 0.9074,
+      "step": 22531
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 2.05405855178833,
+      "learning_rate": 1.1479730577751747e-06,
+      "loss": 0.1178,
+      "step": 22532
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 10.521383285522461,
+      "learning_rate": 1.1471363427184873e-06,
+      "loss": 1.549,
+      "step": 22533
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 13.366665840148926,
+      "learning_rate": 1.1462996276618e-06,
+      "loss": 0.9132,
+      "step": 22534
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 14.899330139160156,
+      "learning_rate": 1.1454629126051125e-06,
+      "loss": 0.7934,
+      "step": 22535
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 51.658470153808594,
+      "learning_rate": 1.144626197548425e-06,
+      "loss": 1.1244,
+      "step": 22536
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 235.84877014160156,
+      "learning_rate": 1.1437894824917375e-06,
+      "loss": 1.9939,
+      "step": 22537
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 11.595799446105957,
+      "learning_rate": 1.14295276743505e-06,
+      "loss": 0.2692,
+      "step": 22538
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 8.764069557189941,
+      "learning_rate": 1.1421160523783627e-06,
+      "loss": 0.6718,
+      "step": 22539
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 12.100363731384277,
+      "learning_rate": 1.1412793373216753e-06,
+      "loss": 0.4135,
+      "step": 22540
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 12.80016803741455,
+      "learning_rate": 1.1404426222649878e-06,
+      "loss": 0.4654,
+      "step": 22541
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 74.28779602050781,
+      "learning_rate": 1.1396059072083002e-06,
+      "loss": 1.9398,
+      "step": 22542
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 13.961068153381348,
+      "learning_rate": 1.1387691921516128e-06,
+      "loss": 0.1284,
+      "step": 22543
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 6.4551615715026855,
+      "learning_rate": 1.1379324770949254e-06,
+      "loss": 0.2141,
+      "step": 22544
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 10.471248626708984,
+      "learning_rate": 1.137095762038238e-06,
+      "loss": 0.7794,
+      "step": 22545
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 10.314981460571289,
+      "learning_rate": 1.1362590469815506e-06,
+      "loss": 1.3724,
+      "step": 22546
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 31.218461990356445,
+      "learning_rate": 1.135422331924863e-06,
+      "loss": 2.968,
+      "step": 22547
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 19.79244041442871,
+      "learning_rate": 1.1345856168681758e-06,
+      "loss": 0.7614,
+      "step": 22548
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 8.343755722045898,
+      "learning_rate": 1.1337489018114881e-06,
+      "loss": 0.5436,
+      "step": 22549
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 13.522882461547852,
+      "learning_rate": 1.1329121867548007e-06,
+      "loss": 0.447,
+      "step": 22550
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 8.117951393127441,
+      "learning_rate": 1.1320754716981133e-06,
+      "loss": 0.7153,
+      "step": 22551
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 19.595169067382812,
+      "learning_rate": 1.131238756641426e-06,
+      "loss": 0.8592,
+      "step": 22552
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 13.410231590270996,
+      "learning_rate": 1.1304020415847385e-06,
+      "loss": 0.4868,
+      "step": 22553
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 19.711048126220703,
+      "learning_rate": 1.1295653265280509e-06,
+      "loss": 0.6378,
+      "step": 22554
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 20.844606399536133,
+      "learning_rate": 1.1287286114713635e-06,
+      "loss": 0.7601,
+      "step": 22555
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 8.204733848571777,
+      "learning_rate": 1.127891896414676e-06,
+      "loss": 0.2567,
+      "step": 22556
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 25.223560333251953,
+      "learning_rate": 1.1270551813579887e-06,
+      "loss": 2.1085,
+      "step": 22557
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 13.40188217163086,
+      "learning_rate": 1.1262184663013012e-06,
+      "loss": 1.0065,
+      "step": 22558
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 7.276838779449463,
+      "learning_rate": 1.1253817512446136e-06,
+      "loss": 0.7332,
+      "step": 22559
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 105.237548828125,
+      "learning_rate": 1.1245450361879264e-06,
+      "loss": 1.9138,
+      "step": 22560
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 19.08675765991211,
+      "learning_rate": 1.1237083211312388e-06,
+      "loss": 0.4783,
+      "step": 22561
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 15.11357593536377,
+      "learning_rate": 1.1228716060745514e-06,
+      "loss": 1.7871,
+      "step": 22562
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 31.744043350219727,
+      "learning_rate": 1.122034891017864e-06,
+      "loss": 2.8459,
+      "step": 22563
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.599182605743408,
+      "learning_rate": 1.1211981759611766e-06,
+      "loss": 0.2267,
+      "step": 22564
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 17.8436336517334,
+      "learning_rate": 1.1203614609044892e-06,
+      "loss": 0.8453,
+      "step": 22565
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 13.657493591308594,
+      "learning_rate": 1.1195247458478015e-06,
+      "loss": 0.7263,
+      "step": 22566
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 12.992782592773438,
+      "learning_rate": 1.1186880307911143e-06,
+      "loss": 0.8964,
+      "step": 22567
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 15.952777862548828,
+      "learning_rate": 1.1178513157344267e-06,
+      "loss": 1.2787,
+      "step": 22568
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 21.19894790649414,
+      "learning_rate": 1.1170146006777393e-06,
+      "loss": 2.0199,
+      "step": 22569
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 17.431442260742188,
+      "learning_rate": 1.116177885621052e-06,
+      "loss": 1.0392,
+      "step": 22570
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 2.0603222846984863,
+      "learning_rate": 1.1153411705643643e-06,
+      "loss": 0.1267,
+      "step": 22571
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 11.604832649230957,
+      "learning_rate": 1.114504455507677e-06,
+      "loss": 1.8294,
+      "step": 22572
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.897436141967773,
+      "learning_rate": 1.1136677404509895e-06,
+      "loss": 0.4455,
+      "step": 22573
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 9.812992095947266,
+      "learning_rate": 1.112831025394302e-06,
+      "loss": 1.161,
+      "step": 22574
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 8.65078353881836,
+      "learning_rate": 1.1119943103376146e-06,
+      "loss": 1.5561,
+      "step": 22575
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 7.049508094787598,
+      "learning_rate": 1.111157595280927e-06,
+      "loss": 0.238,
+      "step": 22576
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 17.952756881713867,
+      "learning_rate": 1.1103208802242398e-06,
+      "loss": 0.7611,
+      "step": 22577
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 20.81865692138672,
+      "learning_rate": 1.1094841651675522e-06,
+      "loss": 0.9094,
+      "step": 22578
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 30.030189514160156,
+      "learning_rate": 1.1086474501108648e-06,
+      "loss": 0.8968,
+      "step": 22579
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 18.892066955566406,
+      "learning_rate": 1.1078107350541774e-06,
+      "loss": 0.725,
+      "step": 22580
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 11.122385025024414,
+      "learning_rate": 1.10697401999749e-06,
+      "loss": 0.8928,
+      "step": 22581
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 15.550915718078613,
+      "learning_rate": 1.1061373049408026e-06,
+      "loss": 0.9142,
+      "step": 22582
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 107.0487060546875,
+      "learning_rate": 1.105300589884115e-06,
+      "loss": 0.951,
+      "step": 22583
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 6.754660606384277,
+      "learning_rate": 1.1044638748274277e-06,
+      "loss": 0.7648,
+      "step": 22584
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 10.755863189697266,
+      "learning_rate": 1.1036271597707401e-06,
+      "loss": 1.1334,
+      "step": 22585
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 11.957283020019531,
+      "learning_rate": 1.1027904447140527e-06,
+      "loss": 1.8996,
+      "step": 22586
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 13.54029655456543,
+      "learning_rate": 1.1019537296573653e-06,
+      "loss": 0.5176,
+      "step": 22587
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 142.4005584716797,
+      "learning_rate": 1.1011170146006777e-06,
+      "loss": 1.6767,
+      "step": 22588
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 11.815372467041016,
+      "learning_rate": 1.1002802995439905e-06,
+      "loss": 1.6228,
+      "step": 22589
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.578014373779297,
+      "learning_rate": 1.0994435844873029e-06,
+      "loss": 0.1797,
+      "step": 22590
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 10.217596054077148,
+      "learning_rate": 1.0986068694306154e-06,
+      "loss": 0.2287,
+      "step": 22591
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 7.644987106323242,
+      "learning_rate": 1.097770154373928e-06,
+      "loss": 0.8052,
+      "step": 22592
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 45.09164810180664,
+      "learning_rate": 1.0969334393172406e-06,
+      "loss": 2.0139,
+      "step": 22593
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.5308732986450195,
+      "learning_rate": 1.0960967242605532e-06,
+      "loss": 0.1777,
+      "step": 22594
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 13.71830940246582,
+      "learning_rate": 1.0952600092038656e-06,
+      "loss": 0.4998,
+      "step": 22595
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 13.78543472290039,
+      "learning_rate": 1.0944232941471784e-06,
+      "loss": 0.3428,
+      "step": 22596
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 57.147552490234375,
+      "learning_rate": 1.0935865790904908e-06,
+      "loss": 2.1839,
+      "step": 22597
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 25.24407386779785,
+      "learning_rate": 1.0927498640338034e-06,
+      "loss": 1.0217,
+      "step": 22598
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 10.55384349822998,
+      "learning_rate": 1.091913148977116e-06,
+      "loss": 0.4198,
+      "step": 22599
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 8.443134307861328,
+      "learning_rate": 1.0910764339204283e-06,
+      "loss": 0.9573,
+      "step": 22600
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 10.019526481628418,
+      "learning_rate": 1.0902397188637411e-06,
+      "loss": 1.2772,
+      "step": 22601
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.7345339059829712,
+      "learning_rate": 1.0894030038070535e-06,
+      "loss": 0.1394,
+      "step": 22602
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 24.403596878051758,
+      "learning_rate": 1.088566288750366e-06,
+      "loss": 0.8839,
+      "step": 22603
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 9.105947494506836,
+      "learning_rate": 1.0877295736936787e-06,
+      "loss": 0.871,
+      "step": 22604
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 6.4465413093566895,
+      "learning_rate": 1.0868928586369913e-06,
+      "loss": 0.3138,
+      "step": 22605
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 9.141624450683594,
+      "learning_rate": 1.0860561435803039e-06,
+      "loss": 0.0722,
+      "step": 22606
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 8.823845863342285,
+      "learning_rate": 1.0852194285236162e-06,
+      "loss": 0.8992,
+      "step": 22607
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 10.580303192138672,
+      "learning_rate": 1.0843827134669288e-06,
+      "loss": 0.8326,
+      "step": 22608
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.165613651275635,
+      "learning_rate": 1.0835459984102414e-06,
+      "loss": 1.5931,
+      "step": 22609
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 5.843441963195801,
+      "learning_rate": 1.082709283353554e-06,
+      "loss": 0.2336,
+      "step": 22610
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 11.892224311828613,
+      "learning_rate": 1.0818725682968666e-06,
+      "loss": 1.7391,
+      "step": 22611
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.86954927444458,
+      "learning_rate": 1.0810358532401792e-06,
+      "loss": 1.0329,
+      "step": 22612
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 12.3053560256958,
+      "learning_rate": 1.0801991381834918e-06,
+      "loss": 0.4214,
+      "step": 22613
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 7.288525581359863,
+      "learning_rate": 1.0793624231268042e-06,
+      "loss": 0.4891,
+      "step": 22614
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 11.251693725585938,
+      "learning_rate": 1.0785257080701168e-06,
+      "loss": 0.3047,
+      "step": 22615
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 21.362695693969727,
+      "learning_rate": 1.0776889930134293e-06,
+      "loss": 1.0763,
+      "step": 22616
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 5.845518589019775,
+      "learning_rate": 1.076852277956742e-06,
+      "loss": 0.798,
+      "step": 22617
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 8.721627235412598,
+      "learning_rate": 1.0760155629000545e-06,
+      "loss": 0.2793,
+      "step": 22618
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 16.777753829956055,
+      "learning_rate": 1.0751788478433671e-06,
+      "loss": 1.5033,
+      "step": 22619
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 23.648386001586914,
+      "learning_rate": 1.0743421327866795e-06,
+      "loss": 2.3692,
+      "step": 22620
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 14.79354476928711,
+      "learning_rate": 1.073505417729992e-06,
+      "loss": 0.6241,
+      "step": 22621
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 77.8792953491211,
+      "learning_rate": 1.0726687026733047e-06,
+      "loss": 2.4211,
+      "step": 22622
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.98190450668335,
+      "learning_rate": 1.0718319876166173e-06,
+      "loss": 0.2484,
+      "step": 22623
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 10.655683517456055,
+      "learning_rate": 1.0709952725599299e-06,
+      "loss": 0.9693,
+      "step": 22624
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 34.36240005493164,
+      "learning_rate": 1.0701585575032424e-06,
+      "loss": 1.4315,
+      "step": 22625
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 10.894150733947754,
+      "learning_rate": 1.0693218424465548e-06,
+      "loss": 1.6335,
+      "step": 22626
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 6.933768272399902,
+      "learning_rate": 1.0684851273898674e-06,
+      "loss": 0.5767,
+      "step": 22627
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 18.84404945373535,
+      "learning_rate": 1.06764841233318e-06,
+      "loss": 1.5802,
+      "step": 22628
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 34.67753219604492,
+      "learning_rate": 1.0668116972764926e-06,
+      "loss": 4.0518,
+      "step": 22629
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 96.99250793457031,
+      "learning_rate": 1.0659749822198052e-06,
+      "loss": 3.8837,
+      "step": 22630
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 18.811664581298828,
+      "learning_rate": 1.0651382671631178e-06,
+      "loss": 1.5722,
+      "step": 22631
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 8.915626525878906,
+      "learning_rate": 1.0643015521064302e-06,
+      "loss": 1.6112,
+      "step": 22632
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 34.93801498413086,
+      "learning_rate": 1.0634648370497427e-06,
+      "loss": 1.2345,
+      "step": 22633
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 24.479141235351562,
+      "learning_rate": 1.0626281219930553e-06,
+      "loss": 1.2939,
+      "step": 22634
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 8.075628280639648,
+      "learning_rate": 1.061791406936368e-06,
+      "loss": 1.6627,
+      "step": 22635
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 13.343048095703125,
+      "learning_rate": 1.0609546918796805e-06,
+      "loss": 0.6281,
+      "step": 22636
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 17.33895492553711,
+      "learning_rate": 1.060117976822993e-06,
+      "loss": 0.6769,
+      "step": 22637
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 11.125850677490234,
+      "learning_rate": 1.0592812617663057e-06,
+      "loss": 0.9356,
+      "step": 22638
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 18.014148712158203,
+      "learning_rate": 1.058444546709618e-06,
+      "loss": 0.3273,
+      "step": 22639
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 8.335336685180664,
+      "learning_rate": 1.0576078316529307e-06,
+      "loss": 0.5336,
+      "step": 22640
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 14.801827430725098,
+      "learning_rate": 1.0567711165962433e-06,
+      "loss": 0.5302,
+      "step": 22641
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 14.088603019714355,
+      "learning_rate": 1.0559344015395558e-06,
+      "loss": 0.3378,
+      "step": 22642
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 384.4151306152344,
+      "learning_rate": 1.0550976864828684e-06,
+      "loss": 1.3484,
+      "step": 22643
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 11.850312232971191,
+      "learning_rate": 1.0542609714261808e-06,
+      "loss": 0.4568,
+      "step": 22644
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 21.090885162353516,
+      "learning_rate": 1.0534242563694934e-06,
+      "loss": 1.4683,
+      "step": 22645
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 25.674741744995117,
+      "learning_rate": 1.052587541312806e-06,
+      "loss": 1.5322,
+      "step": 22646
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 14.398693084716797,
+      "learning_rate": 1.0517508262561186e-06,
+      "loss": 0.7668,
+      "step": 22647
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 13.833110809326172,
+      "learning_rate": 1.0509141111994312e-06,
+      "loss": 0.4308,
+      "step": 22648
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 28.17066764831543,
+      "learning_rate": 1.0500773961427435e-06,
+      "loss": 4.24,
+      "step": 22649
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 51.966556549072266,
+      "learning_rate": 1.0492406810860563e-06,
+      "loss": 2.0793,
+      "step": 22650
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 73.22601318359375,
+      "learning_rate": 1.0484039660293687e-06,
+      "loss": 2.8554,
+      "step": 22651
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 8.28471851348877,
+      "learning_rate": 1.0475672509726813e-06,
+      "loss": 0.2304,
+      "step": 22652
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 92.43878936767578,
+      "learning_rate": 1.046730535915994e-06,
+      "loss": 2.0176,
+      "step": 22653
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 15.221357345581055,
+      "learning_rate": 1.0458938208593065e-06,
+      "loss": 0.5011,
+      "step": 22654
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 5.658198356628418,
+      "learning_rate": 1.045057105802619e-06,
+      "loss": 0.3352,
+      "step": 22655
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 19.164915084838867,
+      "learning_rate": 1.0442203907459315e-06,
+      "loss": 1.2764,
+      "step": 22656
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 16.541826248168945,
+      "learning_rate": 1.0433836756892443e-06,
+      "loss": 0.6415,
+      "step": 22657
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.7284843921661377,
+      "learning_rate": 1.0425469606325566e-06,
+      "loss": 0.0434,
+      "step": 22658
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 10.829341888427734,
+      "learning_rate": 1.0417102455758692e-06,
+      "loss": 0.4431,
+      "step": 22659
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 14.03040599822998,
+      "learning_rate": 1.0408735305191818e-06,
+      "loss": 1.3381,
+      "step": 22660
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 10.154394149780273,
+      "learning_rate": 1.0400368154624942e-06,
+      "loss": 1.0825,
+      "step": 22661
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.6120656728744507,
+      "learning_rate": 1.039200100405807e-06,
+      "loss": 0.0268,
+      "step": 22662
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 22.520339965820312,
+      "learning_rate": 1.0383633853491194e-06,
+      "loss": 0.4964,
+      "step": 22663
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 16.420804977416992,
+      "learning_rate": 1.037526670292432e-06,
+      "loss": 2.3717,
+      "step": 22664
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 10.640663146972656,
+      "learning_rate": 1.0366899552357446e-06,
+      "loss": 1.0823,
+      "step": 22665
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 8.779507637023926,
+      "learning_rate": 1.0358532401790572e-06,
+      "loss": 0.4319,
+      "step": 22666
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 7.7672648429870605,
+      "learning_rate": 1.0350165251223697e-06,
+      "loss": 0.8595,
+      "step": 22667
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 30.301015853881836,
+      "learning_rate": 1.0341798100656821e-06,
+      "loss": 0.4864,
+      "step": 22668
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 15.309416770935059,
+      "learning_rate": 1.033343095008995e-06,
+      "loss": 0.7949,
+      "step": 22669
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 18.696205139160156,
+      "learning_rate": 1.0325063799523073e-06,
+      "loss": 0.5772,
+      "step": 22670
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 11.885797500610352,
+      "learning_rate": 1.0316696648956199e-06,
+      "loss": 1.4711,
+      "step": 22671
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 14.371444702148438,
+      "learning_rate": 1.0308329498389325e-06,
+      "loss": 0.701,
+      "step": 22672
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 10.62415885925293,
+      "learning_rate": 1.0299962347822449e-06,
+      "loss": 0.6706,
+      "step": 22673
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 12.693033218383789,
+      "learning_rate": 1.0291595197255577e-06,
+      "loss": 1.2575,
+      "step": 22674
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 22.45041275024414,
+      "learning_rate": 1.02832280466887e-06,
+      "loss": 0.8786,
+      "step": 22675
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 12.779298782348633,
+      "learning_rate": 1.0274860896121826e-06,
+      "loss": 0.874,
+      "step": 22676
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 71.88469696044922,
+      "learning_rate": 1.0266493745554952e-06,
+      "loss": 1.2904,
+      "step": 22677
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 18.182397842407227,
+      "learning_rate": 1.0258126594988076e-06,
+      "loss": 1.5542,
+      "step": 22678
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 23.903799057006836,
+      "learning_rate": 1.0249759444421204e-06,
+      "loss": 0.6968,
+      "step": 22679
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 37.96636962890625,
+      "learning_rate": 1.0241392293854328e-06,
+      "loss": 0.8339,
+      "step": 22680
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 17.603233337402344,
+      "learning_rate": 1.0233025143287454e-06,
+      "loss": 0.8701,
+      "step": 22681
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.552093029022217,
+      "learning_rate": 1.022465799272058e-06,
+      "loss": 0.121,
+      "step": 22682
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 16.59980583190918,
+      "learning_rate": 1.0216290842153706e-06,
+      "loss": 1.0351,
+      "step": 22683
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 98.66775512695312,
+      "learning_rate": 1.0207923691586831e-06,
+      "loss": 0.7874,
+      "step": 22684
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 18.55583953857422,
+      "learning_rate": 1.0199556541019955e-06,
+      "loss": 1.0839,
+      "step": 22685
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 27.879377365112305,
+      "learning_rate": 1.0191189390453083e-06,
+      "loss": 0.5498,
+      "step": 22686
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 11.789697647094727,
+      "learning_rate": 1.0182822239886207e-06,
+      "loss": 0.5332,
+      "step": 22687
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 21.460121154785156,
+      "learning_rate": 1.0174455089319333e-06,
+      "loss": 0.9124,
+      "step": 22688
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 13.511993408203125,
+      "learning_rate": 1.0166087938752459e-06,
+      "loss": 0.3465,
+      "step": 22689
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 10.475018501281738,
+      "learning_rate": 1.0157720788185583e-06,
+      "loss": 0.3422,
+      "step": 22690
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 5.866156101226807,
+      "learning_rate": 1.014935363761871e-06,
+      "loss": 0.2938,
+      "step": 22691
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 12.232178688049316,
+      "learning_rate": 1.0140986487051834e-06,
+      "loss": 0.3215,
+      "step": 22692
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 12.380698204040527,
+      "learning_rate": 1.013261933648496e-06,
+      "loss": 0.4729,
+      "step": 22693
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 3.1376395225524902,
+      "learning_rate": 1.0124252185918086e-06,
+      "loss": 0.2062,
+      "step": 22694
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 15.682345390319824,
+      "learning_rate": 1.0115885035351212e-06,
+      "loss": 0.3961,
+      "step": 22695
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 27.296039581298828,
+      "learning_rate": 1.0107517884784338e-06,
+      "loss": 0.5643,
+      "step": 22696
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 28.356821060180664,
+      "learning_rate": 1.0099150734217462e-06,
+      "loss": 1.7059,
+      "step": 22697
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 24.347946166992188,
+      "learning_rate": 1.009078358365059e-06,
+      "loss": 0.7439,
+      "step": 22698
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 25.663373947143555,
+      "learning_rate": 1.0082416433083714e-06,
+      "loss": 1.4367,
+      "step": 22699
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 14.530847549438477,
+      "learning_rate": 1.007404928251684e-06,
+      "loss": 0.6064,
+      "step": 22700
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 6.665618896484375,
+      "learning_rate": 1.0065682131949965e-06,
+      "loss": 0.4526,
+      "step": 22701
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 21.13688850402832,
+      "learning_rate": 1.0057314981383091e-06,
+      "loss": 1.0899,
+      "step": 22702
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 9.309040069580078,
+      "learning_rate": 1.0048947830816217e-06,
+      "loss": 0.8862,
+      "step": 22703
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 8.414344787597656,
+      "learning_rate": 1.004058068024934e-06,
+      "loss": 0.2494,
+      "step": 22704
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 8.917380332946777,
+      "learning_rate": 1.0032213529682467e-06,
+      "loss": 1.2896,
+      "step": 22705
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 3.6235129833221436,
+      "learning_rate": 1.0023846379115593e-06,
+      "loss": 0.1717,
+      "step": 22706
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 3.6737544536590576,
+      "learning_rate": 1.0015479228548719e-06,
+      "loss": 0.0908,
+      "step": 22707
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 38.907203674316406,
+      "learning_rate": 1.0007112077981845e-06,
+      "loss": 4.4776,
+      "step": 22708
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 13.30609130859375,
+      "learning_rate": 9.99874492741497e-07,
+      "loss": 0.3695,
+      "step": 22709
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 10.25177001953125,
+      "learning_rate": 9.990377776848096e-07,
+      "loss": 0.3207,
+      "step": 22710
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 13.686180114746094,
+      "learning_rate": 9.98201062628122e-07,
+      "loss": 0.344,
+      "step": 22711
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.232295036315918,
+      "learning_rate": 9.973643475714346e-07,
+      "loss": 0.2562,
+      "step": 22712
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 17.139768600463867,
+      "learning_rate": 9.965276325147472e-07,
+      "loss": 0.6248,
+      "step": 22713
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 63.723995208740234,
+      "learning_rate": 9.956909174580598e-07,
+      "loss": 1.3205,
+      "step": 22714
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.056235313415527,
+      "learning_rate": 9.948542024013724e-07,
+      "loss": 0.1356,
+      "step": 22715
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 6.38409423828125,
+      "learning_rate": 9.940174873446848e-07,
+      "loss": 0.1036,
+      "step": 22716
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 10.570390701293945,
+      "learning_rate": 9.931807722879973e-07,
+      "loss": 0.9784,
+      "step": 22717
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 43.72049331665039,
+      "learning_rate": 9.9234405723131e-07,
+      "loss": 1.594,
+      "step": 22718
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 7.340564727783203,
+      "learning_rate": 9.915073421746225e-07,
+      "loss": 1.9042,
+      "step": 22719
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 11.679526329040527,
+      "learning_rate": 9.906706271179351e-07,
+      "loss": 0.519,
+      "step": 22720
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 6.5310893058776855,
+      "learning_rate": 9.898339120612477e-07,
+      "loss": 0.2962,
+      "step": 22721
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 18.232358932495117,
+      "learning_rate": 9.8899719700456e-07,
+      "loss": 0.885,
+      "step": 22722
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 10.807291030883789,
+      "learning_rate": 9.881604819478727e-07,
+      "loss": 0.3318,
+      "step": 22723
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 9.099920272827148,
+      "learning_rate": 9.873237668911853e-07,
+      "loss": 0.8537,
+      "step": 22724
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 8.183937072753906,
+      "learning_rate": 9.864870518344978e-07,
+      "loss": 0.4723,
+      "step": 22725
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 6.987006187438965,
+      "learning_rate": 9.856503367778104e-07,
+      "loss": 0.5541,
+      "step": 22726
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 22.417749404907227,
+      "learning_rate": 9.84813621721123e-07,
+      "loss": 1.1988,
+      "step": 22727
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 2.788111925125122,
+      "learning_rate": 9.839769066644356e-07,
+      "loss": 0.1956,
+      "step": 22728
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 14.666352272033691,
+      "learning_rate": 9.83140191607748e-07,
+      "loss": 0.6124,
+      "step": 22729
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 16.536130905151367,
+      "learning_rate": 9.823034765510606e-07,
+      "loss": 0.6982,
+      "step": 22730
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 20.315258026123047,
+      "learning_rate": 9.814667614943732e-07,
+      "loss": 0.4652,
+      "step": 22731
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 7.736295223236084,
+      "learning_rate": 9.806300464376858e-07,
+      "loss": 0.4856,
+      "step": 22732
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 6.603104591369629,
+      "learning_rate": 9.797933313809984e-07,
+      "loss": 0.2176,
+      "step": 22733
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 29.130878448486328,
+      "learning_rate": 9.789566163243107e-07,
+      "loss": 1.2782,
+      "step": 22734
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 26.576641082763672,
+      "learning_rate": 9.781199012676233e-07,
+      "loss": 0.2887,
+      "step": 22735
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 9.380056381225586,
+      "learning_rate": 9.77283186210936e-07,
+      "loss": 1.7271,
+      "step": 22736
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 110.59015655517578,
+      "learning_rate": 9.764464711542485e-07,
+      "loss": 0.7674,
+      "step": 22737
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 32.303466796875,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 1.5897,
+      "step": 22738
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 59.81349563598633,
+      "learning_rate": 9.747730410408737e-07,
+      "loss": 2.2013,
+      "step": 22739
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 321.8476257324219,
+      "learning_rate": 9.739363259841863e-07,
+      "loss": 1.7899,
+      "step": 22740
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 11.349199295043945,
+      "learning_rate": 9.730996109274987e-07,
+      "loss": 0.5554,
+      "step": 22741
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 13.518997192382812,
+      "learning_rate": 9.722628958708112e-07,
+      "loss": 0.8175,
+      "step": 22742
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 12.089187622070312,
+      "learning_rate": 9.714261808141238e-07,
+      "loss": 0.2888,
+      "step": 22743
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 20.820369720458984,
+      "learning_rate": 9.705894657574364e-07,
+      "loss": 0.88,
+      "step": 22744
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 35.164302825927734,
+      "learning_rate": 9.69752750700749e-07,
+      "loss": 2.1835,
+      "step": 22745
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 9.61741828918457,
+      "learning_rate": 9.689160356440614e-07,
+      "loss": 0.7723,
+      "step": 22746
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 9.48019027709961,
+      "learning_rate": 9.680793205873742e-07,
+      "loss": 0.3809,
+      "step": 22747
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 12.2240571975708,
+      "learning_rate": 9.672426055306866e-07,
+      "loss": 0.3868,
+      "step": 22748
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 3.512730836868286,
+      "learning_rate": 9.664058904739992e-07,
+      "loss": 0.1358,
+      "step": 22749
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 38.790340423583984,
+      "learning_rate": 9.655691754173118e-07,
+      "loss": 1.7435,
+      "step": 22750
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 10.633916854858398,
+      "learning_rate": 9.647324603606241e-07,
+      "loss": 0.7846,
+      "step": 22751
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.6371921300888062,
+      "learning_rate": 9.63895745303937e-07,
+      "loss": 0.0126,
+      "step": 22752
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 17.8377742767334,
+      "learning_rate": 9.630590302472493e-07,
+      "loss": 0.9767,
+      "step": 22753
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 36.62273025512695,
+      "learning_rate": 9.62222315190562e-07,
+      "loss": 2.6855,
+      "step": 22754
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 15.418661117553711,
+      "learning_rate": 9.613856001338745e-07,
+      "loss": 0.4603,
+      "step": 22755
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 3.746643304824829,
+      "learning_rate": 9.60548885077187e-07,
+      "loss": 0.1215,
+      "step": 22756
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 9.70486831665039,
+      "learning_rate": 9.597121700204997e-07,
+      "loss": 0.3539,
+      "step": 22757
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 182.56591796875,
+      "learning_rate": 9.58875454963812e-07,
+      "loss": 1.801,
+      "step": 22758
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 22.018207550048828,
+      "learning_rate": 9.580387399071249e-07,
+      "loss": 1.7881,
+      "step": 22759
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 21.515419006347656,
+      "learning_rate": 9.572020248504372e-07,
+      "loss": 0.8439,
+      "step": 22760
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 12.654033660888672,
+      "learning_rate": 9.563653097937498e-07,
+      "loss": 1.161,
+      "step": 22761
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 83.0149917602539,
+      "learning_rate": 9.555285947370624e-07,
+      "loss": 1.7713,
+      "step": 22762
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 10.134322166442871,
+      "learning_rate": 9.546918796803748e-07,
+      "loss": 1.1104,
+      "step": 22763
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 6.967130184173584,
+      "learning_rate": 9.538551646236876e-07,
+      "loss": 0.4005,
+      "step": 22764
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 10.932229042053223,
+      "learning_rate": 9.53018449567e-07,
+      "loss": 0.4291,
+      "step": 22765
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 13.56113338470459,
+      "learning_rate": 9.521817345103127e-07,
+      "loss": 0.5479,
+      "step": 22766
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 8.211050033569336,
+      "learning_rate": 9.513450194536251e-07,
+      "loss": 0.2898,
+      "step": 22767
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 42.85634994506836,
+      "learning_rate": 9.505083043969376e-07,
+      "loss": 2.0145,
+      "step": 22768
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 23.32672691345215,
+      "learning_rate": 9.496715893402503e-07,
+      "loss": 1.4944,
+      "step": 22769
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 23.611173629760742,
+      "learning_rate": 9.488348742835628e-07,
+      "loss": 1.2648,
+      "step": 22770
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 18.620038986206055,
+      "learning_rate": 9.479981592268754e-07,
+      "loss": 1.098,
+      "step": 22771
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 22.640104293823242,
+      "learning_rate": 9.471614441701879e-07,
+      "loss": 2.1549,
+      "step": 22772
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 23.428749084472656,
+      "learning_rate": 9.463247291135006e-07,
+      "loss": 1.4812,
+      "step": 22773
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 7.496427536010742,
+      "learning_rate": 9.454880140568131e-07,
+      "loss": 0.3486,
+      "step": 22774
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 14.45265007019043,
+      "learning_rate": 9.446512990001256e-07,
+      "loss": 1.7724,
+      "step": 22775
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 13.942972183227539,
+      "learning_rate": 9.438145839434381e-07,
+      "loss": 0.9157,
+      "step": 22776
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 30.090429306030273,
+      "learning_rate": 9.429778688867506e-07,
+      "loss": 0.9834,
+      "step": 22777
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 9.631400108337402,
+      "learning_rate": 9.421411538300633e-07,
+      "loss": 1.4044,
+      "step": 22778
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 11.13482666015625,
+      "learning_rate": 9.413044387733758e-07,
+      "loss": 0.7159,
+      "step": 22779
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 29.129552841186523,
+      "learning_rate": 9.404677237166883e-07,
+      "loss": 1.5577,
+      "step": 22780
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 34.3688850402832,
+      "learning_rate": 9.396310086600009e-07,
+      "loss": 1.8399,
+      "step": 22781
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.2850165367126465,
+      "learning_rate": 9.387942936033135e-07,
+      "loss": 0.0655,
+      "step": 22782
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 21.041540145874023,
+      "learning_rate": 9.379575785466261e-07,
+      "loss": 0.3858,
+      "step": 22783
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 8.379650115966797,
+      "learning_rate": 9.371208634899385e-07,
+      "loss": 0.6699,
+      "step": 22784
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 16.045167922973633,
+      "learning_rate": 9.362841484332512e-07,
+      "loss": 0.9175,
+      "step": 22785
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 36.10418701171875,
+      "learning_rate": 9.354474333765637e-07,
+      "loss": 0.7872,
+      "step": 22786
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 2.9977262020111084,
+      "learning_rate": 9.346107183198762e-07,
+      "loss": 0.1647,
+      "step": 22787
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 12.359587669372559,
+      "learning_rate": 9.337740032631888e-07,
+      "loss": 0.7004,
+      "step": 22788
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 23.762327194213867,
+      "learning_rate": 9.329372882065013e-07,
+      "loss": 1.3242,
+      "step": 22789
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 24.016380310058594,
+      "learning_rate": 9.32100573149814e-07,
+      "loss": 0.9295,
+      "step": 22790
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 340.8974304199219,
+      "learning_rate": 9.312638580931265e-07,
+      "loss": 1.8105,
+      "step": 22791
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 12.684603691101074,
+      "learning_rate": 9.304271430364391e-07,
+      "loss": 0.2586,
+      "step": 22792
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 49.162139892578125,
+      "learning_rate": 9.295904279797515e-07,
+      "loss": 1.6199,
+      "step": 22793
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 11.601898193359375,
+      "learning_rate": 9.28753712923064e-07,
+      "loss": 0.3166,
+      "step": 22794
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 16.88113021850586,
+      "learning_rate": 9.279169978663767e-07,
+      "loss": 0.3754,
+      "step": 22795
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 21.27098846435547,
+      "learning_rate": 9.270802828096892e-07,
+      "loss": 1.9876,
+      "step": 22796
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 17.28947639465332,
+      "learning_rate": 9.262435677530018e-07,
+      "loss": 1.0031,
+      "step": 22797
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 11.878498077392578,
+      "learning_rate": 9.254068526963144e-07,
+      "loss": 0.6122,
+      "step": 22798
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 24.625106811523438,
+      "learning_rate": 9.24570137639627e-07,
+      "loss": 0.6079,
+      "step": 22799
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 11.562703132629395,
+      "learning_rate": 9.237334225829395e-07,
+      "loss": 1.325,
+      "step": 22800
+    },
+    {
+      "epoch": 2.86,
+      "eval_loss": 0.07697702199220657,
+      "eval_runtime": 95.4787,
+      "eval_samples_per_second": 37.097,
+      "eval_steps_per_second": 37.097,
+      "step": 22800
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 45.269893646240234,
+      "learning_rate": 9.228967075262519e-07,
+      "loss": 2.3619,
+      "step": 22801
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 28.223426818847656,
+      "learning_rate": 9.220599924695646e-07,
+      "loss": 1.3014,
+      "step": 22802
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 15.729945182800293,
+      "learning_rate": 9.212232774128771e-07,
+      "loss": 1.9507,
+      "step": 22803
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 53.05438232421875,
+      "learning_rate": 9.203865623561897e-07,
+      "loss": 1.2993,
+      "step": 22804
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 10.385889053344727,
+      "learning_rate": 9.195498472995022e-07,
+      "loss": 1.204,
+      "step": 22805
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 21.43562889099121,
+      "learning_rate": 9.187131322428147e-07,
+      "loss": 0.6698,
+      "step": 22806
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 6.3664093017578125,
+      "learning_rate": 9.178764171861274e-07,
+      "loss": 1.0888,
+      "step": 22807
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 14.827553749084473,
+      "learning_rate": 9.170397021294399e-07,
+      "loss": 0.4481,
+      "step": 22808
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 5.577608108520508,
+      "learning_rate": 9.162029870727524e-07,
+      "loss": 0.5304,
+      "step": 22809
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 13.780040740966797,
+      "learning_rate": 9.153662720160649e-07,
+      "loss": 0.5358,
+      "step": 22810
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 9.930581092834473,
+      "learning_rate": 9.145295569593776e-07,
+      "loss": 0.6939,
+      "step": 22811
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 27.25922393798828,
+      "learning_rate": 9.136928419026901e-07,
+      "loss": 1.7797,
+      "step": 22812
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 24.871353149414062,
+      "learning_rate": 9.128561268460026e-07,
+      "loss": 1.393,
+      "step": 22813
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 6.4324212074279785,
+      "learning_rate": 9.120194117893153e-07,
+      "loss": 0.2362,
+      "step": 22814
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 44.0561408996582,
+      "learning_rate": 9.111826967326278e-07,
+      "loss": 2.3997,
+      "step": 22815
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 26.62943458557129,
+      "learning_rate": 9.103459816759404e-07,
+      "loss": 0.4729,
+      "step": 22816
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 9.564854621887207,
+      "learning_rate": 9.095092666192529e-07,
+      "loss": 0.3948,
+      "step": 22817
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 6.366318225860596,
+      "learning_rate": 9.086725515625655e-07,
+      "loss": 0.403,
+      "step": 22818
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 6.157698154449463,
+      "learning_rate": 9.07835836505878e-07,
+      "loss": 0.1733,
+      "step": 22819
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 5.42472505569458,
+      "learning_rate": 9.069991214491905e-07,
+      "loss": 0.3922,
+      "step": 22820
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 9.176149368286133,
+      "learning_rate": 9.061624063925031e-07,
+      "loss": 0.6081,
+      "step": 22821
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 24.528722763061523,
+      "learning_rate": 9.053256913358156e-07,
+      "loss": 2.3687,
+      "step": 22822
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 13.149932861328125,
+      "learning_rate": 9.044889762791283e-07,
+      "loss": 1.6374,
+      "step": 22823
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 12.362624168395996,
+      "learning_rate": 9.036522612224408e-07,
+      "loss": 0.5837,
+      "step": 22824
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 9.579484939575195,
+      "learning_rate": 9.028155461657533e-07,
+      "loss": 1.7579,
+      "step": 22825
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 15.601700782775879,
+      "learning_rate": 9.019788311090658e-07,
+      "loss": 0.5001,
+      "step": 22826
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.9154582023620605,
+      "learning_rate": 9.011421160523784e-07,
+      "loss": 0.0562,
+      "step": 22827
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 22.330734252929688,
+      "learning_rate": 9.00305400995691e-07,
+      "loss": 1.3774,
+      "step": 22828
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 26.706493377685547,
+      "learning_rate": 8.994686859390035e-07,
+      "loss": 1.2342,
+      "step": 22829
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 31.059598922729492,
+      "learning_rate": 8.986319708823162e-07,
+      "loss": 0.5336,
+      "step": 22830
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.273304462432861,
+      "learning_rate": 8.977952558256287e-07,
+      "loss": 0.2168,
+      "step": 22831
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 6.656005859375,
+      "learning_rate": 8.969585407689412e-07,
+      "loss": 0.3318,
+      "step": 22832
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 33.49717330932617,
+      "learning_rate": 8.961218257122538e-07,
+      "loss": 0.4419,
+      "step": 22833
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 12.423048973083496,
+      "learning_rate": 8.952851106555662e-07,
+      "loss": 1.1112,
+      "step": 22834
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 32.052955627441406,
+      "learning_rate": 8.944483955988789e-07,
+      "loss": 1.0875,
+      "step": 22835
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 29.69832420349121,
+      "learning_rate": 8.936116805421914e-07,
+      "loss": 1.3687,
+      "step": 22836
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.517812252044678,
+      "learning_rate": 8.92774965485504e-07,
+      "loss": 1.0006,
+      "step": 22837
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 66.99334716796875,
+      "learning_rate": 8.919382504288165e-07,
+      "loss": 1.8536,
+      "step": 22838
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 13.752928733825684,
+      "learning_rate": 8.911015353721291e-07,
+      "loss": 0.5339,
+      "step": 22839
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 16.232959747314453,
+      "learning_rate": 8.902648203154417e-07,
+      "loss": 1.1483,
+      "step": 22840
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 64.5910873413086,
+      "learning_rate": 8.894281052587542e-07,
+      "loss": 1.1875,
+      "step": 22841
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 13.074568748474121,
+      "learning_rate": 8.885913902020668e-07,
+      "loss": 1.5426,
+      "step": 22842
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 12.445037841796875,
+      "learning_rate": 8.877546751453793e-07,
+      "loss": 0.3852,
+      "step": 22843
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 13.778498649597168,
+      "learning_rate": 8.869179600886919e-07,
+      "loss": 1.0955,
+      "step": 22844
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.992719650268555,
+      "learning_rate": 8.860812450320044e-07,
+      "loss": 0.3819,
+      "step": 22845
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 15.028879165649414,
+      "learning_rate": 8.852445299753169e-07,
+      "loss": 1.5503,
+      "step": 22846
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 14.219990730285645,
+      "learning_rate": 8.844078149186296e-07,
+      "loss": 0.4838,
+      "step": 22847
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 9.950181007385254,
+      "learning_rate": 8.835710998619421e-07,
+      "loss": 0.681,
+      "step": 22848
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 13.02139663696289,
+      "learning_rate": 8.827343848052547e-07,
+      "loss": 0.2376,
+      "step": 22849
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 15.152532577514648,
+      "learning_rate": 8.818976697485672e-07,
+      "loss": 0.9848,
+      "step": 22850
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 108.89427185058594,
+      "learning_rate": 8.810609546918796e-07,
+      "loss": 1.4961,
+      "step": 22851
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 33.63383102416992,
+      "learning_rate": 8.802242396351923e-07,
+      "loss": 1.314,
+      "step": 22852
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 11.617888450622559,
+      "learning_rate": 8.793875245785048e-07,
+      "loss": 0.256,
+      "step": 22853
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 41.808021545410156,
+      "learning_rate": 8.785508095218174e-07,
+      "loss": 2.0645,
+      "step": 22854
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 3.3971142768859863,
+      "learning_rate": 8.7771409446513e-07,
+      "loss": 0.0616,
+      "step": 22855
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 16.903583526611328,
+      "learning_rate": 8.768773794084426e-07,
+      "loss": 0.4121,
+      "step": 22856
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 14.798884391784668,
+      "learning_rate": 8.760406643517551e-07,
+      "loss": 0.5728,
+      "step": 22857
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 16.237289428710938,
+      "learning_rate": 8.752039492950676e-07,
+      "loss": 0.5617,
+      "step": 22858
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 14.962675094604492,
+      "learning_rate": 8.743672342383803e-07,
+      "loss": 0.3719,
+      "step": 22859
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 12.828699111938477,
+      "learning_rate": 8.735305191816927e-07,
+      "loss": 0.892,
+      "step": 22860
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 2.6358883380889893,
+      "learning_rate": 8.726938041250053e-07,
+      "loss": 0.0822,
+      "step": 22861
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 16.86612319946289,
+      "learning_rate": 8.718570890683178e-07,
+      "loss": 0.7652,
+      "step": 22862
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 11.26570987701416,
+      "learning_rate": 8.710203740116305e-07,
+      "loss": 1.0248,
+      "step": 22863
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 3.653964042663574,
+      "learning_rate": 8.70183658954943e-07,
+      "loss": 0.8155,
+      "step": 22864
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 13.583930969238281,
+      "learning_rate": 8.693469438982555e-07,
+      "loss": 0.9328,
+      "step": 22865
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 12.397669792175293,
+      "learning_rate": 8.685102288415681e-07,
+      "loss": 1.7724,
+      "step": 22866
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 16.839582443237305,
+      "learning_rate": 8.676735137848806e-07,
+      "loss": 0.7151,
+      "step": 22867
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 17.016077041625977,
+      "learning_rate": 8.668367987281932e-07,
+      "loss": 1.2868,
+      "step": 22868
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 39.06998062133789,
+      "learning_rate": 8.660000836715057e-07,
+      "loss": 1.0109,
+      "step": 22869
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 2.039919853210449,
+      "learning_rate": 8.651633686148182e-07,
+      "loss": 0.0619,
+      "step": 22870
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 27.89397430419922,
+      "learning_rate": 8.643266535581309e-07,
+      "loss": 0.8545,
+      "step": 22871
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 8.406256675720215,
+      "learning_rate": 8.634899385014434e-07,
+      "loss": 0.553,
+      "step": 22872
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 12.104752540588379,
+      "learning_rate": 8.62653223444756e-07,
+      "loss": 0.4784,
+      "step": 22873
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 13.156915664672852,
+      "learning_rate": 8.618165083880685e-07,
+      "loss": 0.8673,
+      "step": 22874
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 150.2317657470703,
+      "learning_rate": 8.609797933313812e-07,
+      "loss": 2.7136,
+      "step": 22875
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 11.70964241027832,
+      "learning_rate": 8.601430782746937e-07,
+      "loss": 1.5755,
+      "step": 22876
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 25.56842803955078,
+      "learning_rate": 8.593063632180061e-07,
+      "loss": 1.2766,
+      "step": 22877
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.928918838500977,
+      "learning_rate": 8.584696481613187e-07,
+      "loss": 0.391,
+      "step": 22878
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.155158042907715,
+      "learning_rate": 8.576329331046312e-07,
+      "loss": 0.2252,
+      "step": 22879
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 37.66831588745117,
+      "learning_rate": 8.567962180479439e-07,
+      "loss": 1.4692,
+      "step": 22880
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 50.934024810791016,
+      "learning_rate": 8.559595029912564e-07,
+      "loss": 2.9425,
+      "step": 22881
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 37.8398323059082,
+      "learning_rate": 8.55122787934569e-07,
+      "loss": 1.5301,
+      "step": 22882
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 16.43904685974121,
+      "learning_rate": 8.542860728778815e-07,
+      "loss": 0.9197,
+      "step": 22883
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 29.151037216186523,
+      "learning_rate": 8.534493578211941e-07,
+      "loss": 3.2912,
+      "step": 22884
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 3.3625638484954834,
+      "learning_rate": 8.526126427645066e-07,
+      "loss": 0.0577,
+      "step": 22885
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 17.012426376342773,
+      "learning_rate": 8.517759277078191e-07,
+      "loss": 0.5314,
+      "step": 22886
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 6.2187628746032715,
+      "learning_rate": 8.509392126511318e-07,
+      "loss": 0.7742,
+      "step": 22887
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 2.5364348888397217,
+      "learning_rate": 8.501024975944443e-07,
+      "loss": 0.0683,
+      "step": 22888
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 10.555633544921875,
+      "learning_rate": 8.492657825377569e-07,
+      "loss": 1.6642,
+      "step": 22889
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 24.966533660888672,
+      "learning_rate": 8.484290674810694e-07,
+      "loss": 0.5993,
+      "step": 22890
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 131.3794403076172,
+      "learning_rate": 8.475923524243819e-07,
+      "loss": 1.2467,
+      "step": 22891
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.629092693328857,
+      "learning_rate": 8.467556373676946e-07,
+      "loss": 0.6006,
+      "step": 22892
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 11.908555030822754,
+      "learning_rate": 8.45918922311007e-07,
+      "loss": 0.2282,
+      "step": 22893
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 50.8621940612793,
+      "learning_rate": 8.450822072543196e-07,
+      "loss": 4.1995,
+      "step": 22894
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.015515327453613,
+      "learning_rate": 8.442454921976321e-07,
+      "loss": 0.1471,
+      "step": 22895
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 279.75604248046875,
+      "learning_rate": 8.434087771409446e-07,
+      "loss": 1.1202,
+      "step": 22896
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 15.043505668640137,
+      "learning_rate": 8.425720620842573e-07,
+      "loss": 0.406,
+      "step": 22897
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 9.919713973999023,
+      "learning_rate": 8.417353470275698e-07,
+      "loss": 0.3008,
+      "step": 22898
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 11.66871166229248,
+      "learning_rate": 8.408986319708824e-07,
+      "loss": 0.3378,
+      "step": 22899
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 9.860762596130371,
+      "learning_rate": 8.40061916914195e-07,
+      "loss": 0.2331,
+      "step": 22900
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 21.140554428100586,
+      "learning_rate": 8.392252018575076e-07,
+      "loss": 2.2822,
+      "step": 22901
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 6.3714823722839355,
+      "learning_rate": 8.3838848680082e-07,
+      "loss": 1.3731,
+      "step": 22902
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 8.593062400817871,
+      "learning_rate": 8.375517717441325e-07,
+      "loss": 0.923,
+      "step": 22903
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 6.22232723236084,
+      "learning_rate": 8.367150566874452e-07,
+      "loss": 0.1529,
+      "step": 22904
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 18.07470703125,
+      "learning_rate": 8.358783416307577e-07,
+      "loss": 0.7162,
+      "step": 22905
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 12.35638427734375,
+      "learning_rate": 8.350416265740703e-07,
+      "loss": 0.2988,
+      "step": 22906
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 11.684820175170898,
+      "learning_rate": 8.342049115173828e-07,
+      "loss": 0.5593,
+      "step": 22907
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 39.3046760559082,
+      "learning_rate": 8.333681964606955e-07,
+      "loss": 1.6821,
+      "step": 22908
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 12.850509643554688,
+      "learning_rate": 8.32531481404008e-07,
+      "loss": 0.6752,
+      "step": 22909
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 15.1165189743042,
+      "learning_rate": 8.316947663473204e-07,
+      "loss": 0.1947,
+      "step": 22910
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 13.694976806640625,
+      "learning_rate": 8.30858051290633e-07,
+      "loss": 1.8173,
+      "step": 22911
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 76.21894836425781,
+      "learning_rate": 8.300213362339455e-07,
+      "loss": 1.1363,
+      "step": 22912
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 20.433887481689453,
+      "learning_rate": 8.291846211772582e-07,
+      "loss": 1.649,
+      "step": 22913
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 16.72854232788086,
+      "learning_rate": 8.283479061205707e-07,
+      "loss": 0.7471,
+      "step": 22914
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 41.515411376953125,
+      "learning_rate": 8.275111910638832e-07,
+      "loss": 0.8162,
+      "step": 22915
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 10.91340160369873,
+      "learning_rate": 8.266744760071959e-07,
+      "loss": 0.2465,
+      "step": 22916
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.5930448770523071,
+      "learning_rate": 8.258377609505084e-07,
+      "loss": 0.0162,
+      "step": 22917
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 24.409202575683594,
+      "learning_rate": 8.25001045893821e-07,
+      "loss": 1.3316,
+      "step": 22918
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 15.13890552520752,
+      "learning_rate": 8.241643308371334e-07,
+      "loss": 1.8263,
+      "step": 22919
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 29.22232437133789,
+      "learning_rate": 8.233276157804461e-07,
+      "loss": 1.3149,
+      "step": 22920
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 22.16623878479004,
+      "learning_rate": 8.224909007237586e-07,
+      "loss": 1.1153,
+      "step": 22921
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 5.999701023101807,
+      "learning_rate": 8.216541856670711e-07,
+      "loss": 0.2013,
+      "step": 22922
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 34.92793273925781,
+      "learning_rate": 8.208174706103837e-07,
+      "loss": 1.1674,
+      "step": 22923
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 10.559051513671875,
+      "learning_rate": 8.199807555536962e-07,
+      "loss": 0.1636,
+      "step": 22924
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 12.62888240814209,
+      "learning_rate": 8.191440404970089e-07,
+      "loss": 0.9471,
+      "step": 22925
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 34.148990631103516,
+      "learning_rate": 8.183073254403214e-07,
+      "loss": 1.0717,
+      "step": 22926
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 13.00924301147461,
+      "learning_rate": 8.174706103836339e-07,
+      "loss": 0.7507,
+      "step": 22927
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 15.056406021118164,
+      "learning_rate": 8.166338953269464e-07,
+      "loss": 0.2511,
+      "step": 22928
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 15.062324523925781,
+      "learning_rate": 8.15797180270259e-07,
+      "loss": 0.6409,
+      "step": 22929
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 18.487070083618164,
+      "learning_rate": 8.149604652135716e-07,
+      "loss": 1.1899,
+      "step": 22930
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 15.771196365356445,
+      "learning_rate": 8.141237501568841e-07,
+      "loss": 1.087,
+      "step": 22931
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 6.037031173706055,
+      "learning_rate": 8.132870351001968e-07,
+      "loss": 0.303,
+      "step": 22932
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 26.775182723999023,
+      "learning_rate": 8.124503200435093e-07,
+      "loss": 1.1372,
+      "step": 22933
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 15.645440101623535,
+      "learning_rate": 8.116136049868219e-07,
+      "loss": 0.6334,
+      "step": 22934
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 15.815534591674805,
+      "learning_rate": 8.107768899301343e-07,
+      "loss": 0.7138,
+      "step": 22935
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 7.859343528747559,
+      "learning_rate": 8.099401748734468e-07,
+      "loss": 0.801,
+      "step": 22936
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 2.0807950496673584,
+      "learning_rate": 8.091034598167595e-07,
+      "loss": 0.0397,
+      "step": 22937
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 25.227819442749023,
+      "learning_rate": 8.08266744760072e-07,
+      "loss": 1.1612,
+      "step": 22938
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 8.388067245483398,
+      "learning_rate": 8.074300297033846e-07,
+      "loss": 1.2875,
+      "step": 22939
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 20.248489379882812,
+      "learning_rate": 8.065933146466971e-07,
+      "loss": 0.9424,
+      "step": 22940
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 16.774770736694336,
+      "learning_rate": 8.057565995900097e-07,
+      "loss": 0.7271,
+      "step": 22941
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 86.81754302978516,
+      "learning_rate": 8.049198845333223e-07,
+      "loss": 0.5,
+      "step": 22942
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 11.418489456176758,
+      "learning_rate": 8.040831694766347e-07,
+      "loss": 0.5448,
+      "step": 22943
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.479661464691162,
+      "learning_rate": 8.032464544199473e-07,
+      "loss": 0.0863,
+      "step": 22944
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 10.140596389770508,
+      "learning_rate": 8.024097393632599e-07,
+      "loss": 0.3461,
+      "step": 22945
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 10.982282638549805,
+      "learning_rate": 8.015730243065725e-07,
+      "loss": 0.435,
+      "step": 22946
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 33.08202362060547,
+      "learning_rate": 8.00736309249885e-07,
+      "loss": 1.4527,
+      "step": 22947
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 23.6079044342041,
+      "learning_rate": 7.998995941931975e-07,
+      "loss": 0.8683,
+      "step": 22948
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 6.480509281158447,
+      "learning_rate": 7.990628791365102e-07,
+      "loss": 0.8845,
+      "step": 22949
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 16.13175392150879,
+      "learning_rate": 7.982261640798227e-07,
+      "loss": 0.5201,
+      "step": 22950
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.348287582397461,
+      "learning_rate": 7.973894490231353e-07,
+      "loss": 0.1425,
+      "step": 22951
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 8.483288764953613,
+      "learning_rate": 7.965527339664477e-07,
+      "loss": 0.8687,
+      "step": 22952
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 23.86382484436035,
+      "learning_rate": 7.957160189097604e-07,
+      "loss": 0.444,
+      "step": 22953
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 12.939167976379395,
+      "learning_rate": 7.948793038530729e-07,
+      "loss": 0.7317,
+      "step": 22954
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 14.58227825164795,
+      "learning_rate": 7.940425887963854e-07,
+      "loss": 0.3941,
+      "step": 22955
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 56.66604232788086,
+      "learning_rate": 7.93205873739698e-07,
+      "loss": 2.5979,
+      "step": 22956
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 13.80915355682373,
+      "learning_rate": 7.923691586830106e-07,
+      "loss": 1.4544,
+      "step": 22957
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 3.105476140975952,
+      "learning_rate": 7.915324436263232e-07,
+      "loss": 0.1329,
+      "step": 22958
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 29.131650924682617,
+      "learning_rate": 7.906957285696357e-07,
+      "loss": 0.6263,
+      "step": 22959
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 9.802521705627441,
+      "learning_rate": 7.898590135129481e-07,
+      "loss": 0.592,
+      "step": 22960
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 9.273344039916992,
+      "learning_rate": 7.890222984562608e-07,
+      "loss": 0.2079,
+      "step": 22961
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 5.975650787353516,
+      "learning_rate": 7.881855833995733e-07,
+      "loss": 0.6229,
+      "step": 22962
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 126.74982452392578,
+      "learning_rate": 7.873488683428859e-07,
+      "loss": 2.1825,
+      "step": 22963
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 11.358890533447266,
+      "learning_rate": 7.865121532861984e-07,
+      "loss": 0.4745,
+      "step": 22964
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 24.60085105895996,
+      "learning_rate": 7.856754382295111e-07,
+      "loss": 1.0467,
+      "step": 22965
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 7.778239727020264,
+      "learning_rate": 7.848387231728236e-07,
+      "loss": 0.6853,
+      "step": 22966
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 26.256319046020508,
+      "learning_rate": 7.840020081161361e-07,
+      "loss": 1.8415,
+      "step": 22967
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 11.514891624450684,
+      "learning_rate": 7.831652930594487e-07,
+      "loss": 0.3687,
+      "step": 22968
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 554.5360717773438,
+      "learning_rate": 7.823285780027611e-07,
+      "loss": 1.308,
+      "step": 22969
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 14.025053024291992,
+      "learning_rate": 7.814918629460738e-07,
+      "loss": 0.9177,
+      "step": 22970
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 16.900421142578125,
+      "learning_rate": 7.806551478893863e-07,
+      "loss": 0.9049,
+      "step": 22971
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 19.033496856689453,
+      "learning_rate": 7.798184328326989e-07,
+      "loss": 0.6428,
+      "step": 22972
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 8.236845970153809,
+      "learning_rate": 7.789817177760115e-07,
+      "loss": 0.4224,
+      "step": 22973
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 18.759428024291992,
+      "learning_rate": 7.78145002719324e-07,
+      "loss": 0.9096,
+      "step": 22974
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 7.645739555358887,
+      "learning_rate": 7.773082876626366e-07,
+      "loss": 0.509,
+      "step": 22975
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 21.72773551940918,
+      "learning_rate": 7.764715726059491e-07,
+      "loss": 1.2064,
+      "step": 22976
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.316758155822754,
+      "learning_rate": 7.756348575492618e-07,
+      "loss": 0.2389,
+      "step": 22977
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 13.951539993286133,
+      "learning_rate": 7.747981424925742e-07,
+      "loss": 0.4588,
+      "step": 22978
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 5.033867359161377,
+      "learning_rate": 7.739614274358867e-07,
+      "loss": 0.769,
+      "step": 22979
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 19.416301727294922,
+      "learning_rate": 7.731247123791993e-07,
+      "loss": 3.3767,
+      "step": 22980
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 18.113895416259766,
+      "learning_rate": 7.722879973225118e-07,
+      "loss": 0.5824,
+      "step": 22981
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 23.813232421875,
+      "learning_rate": 7.714512822658245e-07,
+      "loss": 0.7504,
+      "step": 22982
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 19.223020553588867,
+      "learning_rate": 7.70614567209137e-07,
+      "loss": 1.0433,
+      "step": 22983
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 6.926052093505859,
+      "learning_rate": 7.697778521524496e-07,
+      "loss": 0.5064,
+      "step": 22984
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 9.047198295593262,
+      "learning_rate": 7.68941137095762e-07,
+      "loss": 1.1277,
+      "step": 22985
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 5.469225883483887,
+      "learning_rate": 7.681044220390746e-07,
+      "loss": 0.258,
+      "step": 22986
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 10.57030200958252,
+      "learning_rate": 7.672677069823872e-07,
+      "loss": 0.8368,
+      "step": 22987
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 21.639989852905273,
+      "learning_rate": 7.664309919256997e-07,
+      "loss": 2.6035,
+      "step": 22988
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 13.677913665771484,
+      "learning_rate": 7.655942768690124e-07,
+      "loss": 0.5595,
+      "step": 22989
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.7012652158737183,
+      "learning_rate": 7.647575618123249e-07,
+      "loss": 0.0864,
+      "step": 22990
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 22.131879806518555,
+      "learning_rate": 7.639208467556375e-07,
+      "loss": 2.1319,
+      "step": 22991
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 20.789060592651367,
+      "learning_rate": 7.6308413169895e-07,
+      "loss": 0.8272,
+      "step": 22992
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 13.723470687866211,
+      "learning_rate": 7.622474166422625e-07,
+      "loss": 0.8894,
+      "step": 22993
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 16.123985290527344,
+      "learning_rate": 7.614107015855751e-07,
+      "loss": 0.7385,
+      "step": 22994
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 6.176941871643066,
+      "learning_rate": 7.605739865288876e-07,
+      "loss": 1.6784,
+      "step": 22995
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 20.015087127685547,
+      "learning_rate": 7.597372714722002e-07,
+      "loss": 1.2517,
+      "step": 22996
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 24.278839111328125,
+      "learning_rate": 7.589005564155127e-07,
+      "loss": 1.9441,
+      "step": 22997
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 8.983095169067383,
+      "learning_rate": 7.580638413588254e-07,
+      "loss": 0.687,
+      "step": 22998
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.301553726196289,
+      "learning_rate": 7.572271263021379e-07,
+      "loss": 0.4347,
+      "step": 22999
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 15.785185813903809,
+      "learning_rate": 7.563904112454504e-07,
+      "loss": 1.4665,
+      "step": 23000
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 30.19902801513672,
+      "learning_rate": 7.55553696188763e-07,
+      "loss": 1.1447,
+      "step": 23001
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 12.137067794799805,
+      "learning_rate": 7.547169811320755e-07,
+      "loss": 0.5485,
+      "step": 23002
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 25.211074829101562,
+      "learning_rate": 7.538802660753881e-07,
+      "loss": 2.0282,
+      "step": 23003
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 9.797856330871582,
+      "learning_rate": 7.530435510187006e-07,
+      "loss": 0.5698,
+      "step": 23004
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 20.488197326660156,
+      "learning_rate": 7.522068359620131e-07,
+      "loss": 0.782,
+      "step": 23005
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 35.42162322998047,
+      "learning_rate": 7.513701209053258e-07,
+      "loss": 1.8208,
+      "step": 23006
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 12.245278358459473,
+      "learning_rate": 7.505334058486383e-07,
+      "loss": 1.6764,
+      "step": 23007
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 29.070436477661133,
+      "learning_rate": 7.496966907919509e-07,
+      "loss": 1.7391,
+      "step": 23008
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 485.6443176269531,
+      "learning_rate": 7.488599757352634e-07,
+      "loss": 1.8613,
+      "step": 23009
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 6.379110813140869,
+      "learning_rate": 7.480232606785761e-07,
+      "loss": 0.3097,
+      "step": 23010
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 21.31460189819336,
+      "learning_rate": 7.471865456218885e-07,
+      "loss": 2.8207,
+      "step": 23011
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 20.664499282836914,
+      "learning_rate": 7.46349830565201e-07,
+      "loss": 2.4824,
+      "step": 23012
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 40.1668815612793,
+      "learning_rate": 7.455131155085136e-07,
+      "loss": 3.3693,
+      "step": 23013
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 21.943397521972656,
+      "learning_rate": 7.446764004518262e-07,
+      "loss": 0.3427,
+      "step": 23014
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 9.698728561401367,
+      "learning_rate": 7.438396853951388e-07,
+      "loss": 0.5873,
+      "step": 23015
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 17.50680160522461,
+      "learning_rate": 7.430029703384513e-07,
+      "loss": 1.2329,
+      "step": 23016
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 20.397890090942383,
+      "learning_rate": 7.421662552817639e-07,
+      "loss": 2.4361,
+      "step": 23017
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 6.906713962554932,
+      "learning_rate": 7.413295402250765e-07,
+      "loss": 0.2127,
+      "step": 23018
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 7.076891899108887,
+      "learning_rate": 7.404928251683889e-07,
+      "loss": 1.7859,
+      "step": 23019
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 14.63764476776123,
+      "learning_rate": 7.396561101117015e-07,
+      "loss": 0.3831,
+      "step": 23020
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 6.63559627532959,
+      "learning_rate": 7.38819395055014e-07,
+      "loss": 1.1545,
+      "step": 23021
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 20.352502822875977,
+      "learning_rate": 7.379826799983267e-07,
+      "loss": 0.8208,
+      "step": 23022
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 15.528678894042969,
+      "learning_rate": 7.371459649416392e-07,
+      "loss": 0.539,
+      "step": 23023
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 2.7121853828430176,
+      "learning_rate": 7.363092498849517e-07,
+      "loss": 0.1615,
+      "step": 23024
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 14.02507209777832,
+      "learning_rate": 7.354725348282643e-07,
+      "loss": 0.5469,
+      "step": 23025
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 31.693456649780273,
+      "learning_rate": 7.346358197715768e-07,
+      "loss": 1.0745,
+      "step": 23026
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 35.96394348144531,
+      "learning_rate": 7.337991047148895e-07,
+      "loss": 1.2756,
+      "step": 23027
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 13.627449989318848,
+      "learning_rate": 7.329623896582019e-07,
+      "loss": 0.9638,
+      "step": 23028
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 14.99082088470459,
+      "learning_rate": 7.321256746015145e-07,
+      "loss": 1.5079,
+      "step": 23029
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 11.57680606842041,
+      "learning_rate": 7.312889595448271e-07,
+      "loss": 0.282,
+      "step": 23030
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 8.11372184753418,
+      "learning_rate": 7.304522444881396e-07,
+      "loss": 0.6722,
+      "step": 23031
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 17.07073402404785,
+      "learning_rate": 7.296155294314522e-07,
+      "loss": 1.3778,
+      "step": 23032
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 19.17386817932129,
+      "learning_rate": 7.287788143747647e-07,
+      "loss": 1.5862,
+      "step": 23033
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 30.804306030273438,
+      "learning_rate": 7.279420993180774e-07,
+      "loss": 1.9452,
+      "step": 23034
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 6.854452133178711,
+      "learning_rate": 7.271053842613899e-07,
+      "loss": 0.231,
+      "step": 23035
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 9.98362922668457,
+      "learning_rate": 7.262686692047024e-07,
+      "loss": 0.3267,
+      "step": 23036
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 10.96481704711914,
+      "learning_rate": 7.254319541480149e-07,
+      "loss": 0.6027,
+      "step": 23037
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 21.294233322143555,
+      "learning_rate": 7.245952390913274e-07,
+      "loss": 0.8939,
+      "step": 23038
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 15.081424713134766,
+      "learning_rate": 7.237585240346401e-07,
+      "loss": 0.5742,
+      "step": 23039
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 11.576356887817383,
+      "learning_rate": 7.229218089779526e-07,
+      "loss": 1.1172,
+      "step": 23040
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 306.48291015625,
+      "learning_rate": 7.220850939212652e-07,
+      "loss": 3.5268,
+      "step": 23041
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 13.041169166564941,
+      "learning_rate": 7.212483788645777e-07,
+      "loss": 0.3596,
+      "step": 23042
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 3.398914337158203,
+      "learning_rate": 7.204116638078904e-07,
+      "loss": 0.2012,
+      "step": 23043
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 9.716375350952148,
+      "learning_rate": 7.195749487512028e-07,
+      "loss": 0.3987,
+      "step": 23044
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 19.549253463745117,
+      "learning_rate": 7.187382336945153e-07,
+      "loss": 1.277,
+      "step": 23045
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 27.09036636352539,
+      "learning_rate": 7.17901518637828e-07,
+      "loss": 1.2798,
+      "step": 23046
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.830048561096191,
+      "learning_rate": 7.170648035811405e-07,
+      "loss": 0.2446,
+      "step": 23047
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 126.25982666015625,
+      "learning_rate": 7.162280885244531e-07,
+      "loss": 1.1631,
+      "step": 23048
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 38.02354049682617,
+      "learning_rate": 7.153913734677656e-07,
+      "loss": 0.8995,
+      "step": 23049
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 9.047268867492676,
+      "learning_rate": 7.145546584110781e-07,
+      "loss": 0.3694,
+      "step": 23050
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 21.358596801757812,
+      "learning_rate": 7.137179433543908e-07,
+      "loss": 1.0105,
+      "step": 23051
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 9.097389221191406,
+      "learning_rate": 7.128812282977033e-07,
+      "loss": 0.5755,
+      "step": 23052
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 25.98590850830078,
+      "learning_rate": 7.120445132410158e-07,
+      "loss": 0.4946,
+      "step": 23053
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 10.612198829650879,
+      "learning_rate": 7.112077981843283e-07,
+      "loss": 0.684,
+      "step": 23054
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 10.165337562561035,
+      "learning_rate": 7.10371083127641e-07,
+      "loss": 0.9295,
+      "step": 23055
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 3.873164415359497,
+      "learning_rate": 7.095343680709535e-07,
+      "loss": 0.1756,
+      "step": 23056
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 16.203018188476562,
+      "learning_rate": 7.08697653014266e-07,
+      "loss": 1.1577,
+      "step": 23057
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 6.412531852722168,
+      "learning_rate": 7.078609379575786e-07,
+      "loss": 0.2554,
+      "step": 23058
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 45.4237174987793,
+      "learning_rate": 7.070242229008912e-07,
+      "loss": 1.6425,
+      "step": 23059
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 15.77699089050293,
+      "learning_rate": 7.061875078442038e-07,
+      "loss": 0.9989,
+      "step": 23060
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 7.485088348388672,
+      "learning_rate": 7.053507927875162e-07,
+      "loss": 0.3462,
+      "step": 23061
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 11.70754337310791,
+      "learning_rate": 7.045140777308289e-07,
+      "loss": 0.6414,
+      "step": 23062
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 22.87289047241211,
+      "learning_rate": 7.036773626741414e-07,
+      "loss": 0.3964,
+      "step": 23063
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 9.068693161010742,
+      "learning_rate": 7.028406476174539e-07,
+      "loss": 0.6969,
+      "step": 23064
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 302.4141845703125,
+      "learning_rate": 7.020039325607665e-07,
+      "loss": 1.5057,
+      "step": 23065
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 8.26516056060791,
+      "learning_rate": 7.01167217504079e-07,
+      "loss": 0.3393,
+      "step": 23066
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 13.177416801452637,
+      "learning_rate": 7.003305024473917e-07,
+      "loss": 1.6422,
+      "step": 23067
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 18.07768440246582,
+      "learning_rate": 6.994937873907042e-07,
+      "loss": 0.8876,
+      "step": 23068
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 25.200332641601562,
+      "learning_rate": 6.986570723340166e-07,
+      "loss": 0.5052,
+      "step": 23069
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 8.668069839477539,
+      "learning_rate": 6.978203572773292e-07,
+      "loss": 0.3231,
+      "step": 23070
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 12.362847328186035,
+      "learning_rate": 6.969836422206417e-07,
+      "loss": 0.9623,
+      "step": 23071
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 7.626944065093994,
+      "learning_rate": 6.961469271639544e-07,
+      "loss": 0.6111,
+      "step": 23072
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 8.494346618652344,
+      "learning_rate": 6.953102121072669e-07,
+      "loss": 0.2686,
+      "step": 23073
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 39.15300369262695,
+      "learning_rate": 6.944734970505795e-07,
+      "loss": 0.6873,
+      "step": 23074
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 11.612045288085938,
+      "learning_rate": 6.936367819938921e-07,
+      "loss": 1.364,
+      "step": 23075
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 11.220239639282227,
+      "learning_rate": 6.928000669372046e-07,
+      "loss": 0.6776,
+      "step": 23076
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 17.403223037719727,
+      "learning_rate": 6.919633518805172e-07,
+      "loss": 0.8078,
+      "step": 23077
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 9.634208679199219,
+      "learning_rate": 6.911266368238296e-07,
+      "loss": 0.2619,
+      "step": 23078
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 16.1423282623291,
+      "learning_rate": 6.902899217671423e-07,
+      "loss": 0.6613,
+      "step": 23079
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 15.611533164978027,
+      "learning_rate": 6.894532067104548e-07,
+      "loss": 0.9895,
+      "step": 23080
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 29.376800537109375,
+      "learning_rate": 6.886164916537674e-07,
+      "loss": 1.5025,
+      "step": 23081
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 22.50613021850586,
+      "learning_rate": 6.877797765970799e-07,
+      "loss": 1.1317,
+      "step": 23082
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 296.834716796875,
+      "learning_rate": 6.869430615403924e-07,
+      "loss": 1.7144,
+      "step": 23083
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 7.580447673797607,
+      "learning_rate": 6.861063464837051e-07,
+      "loss": 1.5265,
+      "step": 23084
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 26.02106475830078,
+      "learning_rate": 6.852696314270176e-07,
+      "loss": 1.6396,
+      "step": 23085
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 5.650911331176758,
+      "learning_rate": 6.844329163703301e-07,
+      "loss": 0.1539,
+      "step": 23086
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 15.165826797485352,
+      "learning_rate": 6.835962013136426e-07,
+      "loss": 0.5427,
+      "step": 23087
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 8.462101936340332,
+      "learning_rate": 6.827594862569553e-07,
+      "loss": 0.9209,
+      "step": 23088
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 18.40264892578125,
+      "learning_rate": 6.819227712002678e-07,
+      "loss": 1.3245,
+      "step": 23089
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 10.905312538146973,
+      "learning_rate": 6.810860561435803e-07,
+      "loss": 0.5452,
+      "step": 23090
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 29.667407989501953,
+      "learning_rate": 6.80249341086893e-07,
+      "loss": 1.4181,
+      "step": 23091
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 6.203871726989746,
+      "learning_rate": 6.794126260302055e-07,
+      "loss": 0.3344,
+      "step": 23092
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 21.73906707763672,
+      "learning_rate": 6.785759109735181e-07,
+      "loss": 0.9113,
+      "step": 23093
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 23.47955322265625,
+      "learning_rate": 6.777391959168306e-07,
+      "loss": 0.0693,
+      "step": 23094
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 9.022835731506348,
+      "learning_rate": 6.76902480860143e-07,
+      "loss": 1.6314,
+      "step": 23095
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 14.276798248291016,
+      "learning_rate": 6.760657658034557e-07,
+      "loss": 0.7597,
+      "step": 23096
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 11.97426986694336,
+      "learning_rate": 6.752290507467682e-07,
+      "loss": 0.2431,
+      "step": 23097
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 7.717976093292236,
+      "learning_rate": 6.743923356900808e-07,
+      "loss": 0.8563,
+      "step": 23098
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 15.67991828918457,
+      "learning_rate": 6.735556206333933e-07,
+      "loss": 1.0141,
+      "step": 23099
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 5.017505645751953,
+      "learning_rate": 6.72718905576706e-07,
+      "loss": 0.6121,
+      "step": 23100
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 6.8837151527404785,
+      "learning_rate": 6.718821905200185e-07,
+      "loss": 0.2891,
+      "step": 23101
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 14.142083168029785,
+      "learning_rate": 6.71045475463331e-07,
+      "loss": 1.5944,
+      "step": 23102
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 18.120759963989258,
+      "learning_rate": 6.702087604066435e-07,
+      "loss": 1.2532,
+      "step": 23103
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 16.809154510498047,
+      "learning_rate": 6.693720453499561e-07,
+      "loss": 1.4368,
+      "step": 23104
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 15.845349311828613,
+      "learning_rate": 6.685353302932687e-07,
+      "loss": 0.829,
+      "step": 23105
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 15.640596389770508,
+      "learning_rate": 6.676986152365812e-07,
+      "loss": 1.0846,
+      "step": 23106
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 11.92232894897461,
+      "learning_rate": 6.668619001798939e-07,
+      "loss": 0.4998,
+      "step": 23107
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 14.522953987121582,
+      "learning_rate": 6.660251851232064e-07,
+      "loss": 0.672,
+      "step": 23108
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 18.471738815307617,
+      "learning_rate": 6.651884700665189e-07,
+      "loss": 0.7952,
+      "step": 23109
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 11.424614906311035,
+      "learning_rate": 6.643517550098315e-07,
+      "loss": 0.6084,
+      "step": 23110
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 16.51778221130371,
+      "learning_rate": 6.635150399531439e-07,
+      "loss": 2.2651,
+      "step": 23111
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 23.68822479248047,
+      "learning_rate": 6.626783248964566e-07,
+      "loss": 0.3113,
+      "step": 23112
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 9.63133430480957,
+      "learning_rate": 6.618416098397691e-07,
+      "loss": 0.1978,
+      "step": 23113
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 45.005104064941406,
+      "learning_rate": 6.610048947830816e-07,
+      "loss": 1.5576,
+      "step": 23114
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 15.488126754760742,
+      "learning_rate": 6.601681797263942e-07,
+      "loss": 0.7878,
+      "step": 23115
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.7126505374908447,
+      "learning_rate": 6.593314646697068e-07,
+      "loss": 0.1323,
+      "step": 23116
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 6.407670497894287,
+      "learning_rate": 6.584947496130194e-07,
+      "loss": 0.1611,
+      "step": 23117
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 34.24800491333008,
+      "learning_rate": 6.576580345563319e-07,
+      "loss": 0.7007,
+      "step": 23118
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 231.00238037109375,
+      "learning_rate": 6.568213194996445e-07,
+      "loss": 2.245,
+      "step": 23119
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 34.10979080200195,
+      "learning_rate": 6.55984604442957e-07,
+      "loss": 2.3559,
+      "step": 23120
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 11.2767972946167,
+      "learning_rate": 6.551478893862695e-07,
+      "loss": 1.0564,
+      "step": 23121
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 26.728120803833008,
+      "learning_rate": 6.543111743295821e-07,
+      "loss": 1.647,
+      "step": 23122
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 13.866838455200195,
+      "learning_rate": 6.534744592728946e-07,
+      "loss": 0.576,
+      "step": 23123
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 10.672852516174316,
+      "learning_rate": 6.526377442162073e-07,
+      "loss": 1.7061,
+      "step": 23124
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 6.957767009735107,
+      "learning_rate": 6.518010291595198e-07,
+      "loss": 0.4779,
+      "step": 23125
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 21.99214744567871,
+      "learning_rate": 6.509643141028324e-07,
+      "loss": 1.5281,
+      "step": 23126
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 25.584535598754883,
+      "learning_rate": 6.501275990461449e-07,
+      "loss": 0.9848,
+      "step": 23127
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 8.245368003845215,
+      "learning_rate": 6.492908839894573e-07,
+      "loss": 0.0568,
+      "step": 23128
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 16.42379379272461,
+      "learning_rate": 6.4845416893277e-07,
+      "loss": 1.4548,
+      "step": 23129
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 12.99228286743164,
+      "learning_rate": 6.476174538760825e-07,
+      "loss": 0.3139,
+      "step": 23130
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 3.258230447769165,
+      "learning_rate": 6.467807388193951e-07,
+      "loss": 0.1501,
+      "step": 23131
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 23.172542572021484,
+      "learning_rate": 6.459440237627077e-07,
+      "loss": 0.8088,
+      "step": 23132
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 3.2802579402923584,
+      "learning_rate": 6.451073087060203e-07,
+      "loss": 0.0514,
+      "step": 23133
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 11.514467239379883,
+      "learning_rate": 6.442705936493328e-07,
+      "loss": 0.3247,
+      "step": 23134
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 8.98300552368164,
+      "learning_rate": 6.434338785926453e-07,
+      "loss": 1.1936,
+      "step": 23135
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 26.330307006835938,
+      "learning_rate": 6.42597163535958e-07,
+      "loss": 1.6227,
+      "step": 23136
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 17.73244857788086,
+      "learning_rate": 6.417604484792704e-07,
+      "loss": 0.9954,
+      "step": 23137
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 15.236470222473145,
+      "learning_rate": 6.40923733422583e-07,
+      "loss": 0.7773,
+      "step": 23138
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 68.36768341064453,
+      "learning_rate": 6.400870183658955e-07,
+      "loss": 2.8554,
+      "step": 23139
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 8.454971313476562,
+      "learning_rate": 6.39250303309208e-07,
+      "loss": 0.7537,
+      "step": 23140
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 10.701157569885254,
+      "learning_rate": 6.384135882525207e-07,
+      "loss": 0.7045,
+      "step": 23141
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 13.48745059967041,
+      "learning_rate": 6.375768731958332e-07,
+      "loss": 0.2927,
+      "step": 23142
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 32.00803756713867,
+      "learning_rate": 6.367401581391458e-07,
+      "loss": 3.1151,
+      "step": 23143
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 10.901165962219238,
+      "learning_rate": 6.359034430824583e-07,
+      "loss": 0.5296,
+      "step": 23144
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 14.928459167480469,
+      "learning_rate": 6.35066728025771e-07,
+      "loss": 1.511,
+      "step": 23145
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 7.383054733276367,
+      "learning_rate": 6.342300129690834e-07,
+      "loss": 0.3177,
+      "step": 23146
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 7.198001861572266,
+      "learning_rate": 6.333932979123959e-07,
+      "loss": 0.2922,
+      "step": 23147
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 6.210714340209961,
+      "learning_rate": 6.325565828557086e-07,
+      "loss": 0.8739,
+      "step": 23148
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 43.29275894165039,
+      "learning_rate": 6.317198677990211e-07,
+      "loss": 1.1105,
+      "step": 23149
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 15.920175552368164,
+      "learning_rate": 6.308831527423337e-07,
+      "loss": 1.0372,
+      "step": 23150
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 8.938474655151367,
+      "learning_rate": 6.300464376856462e-07,
+      "loss": 0.8599,
+      "step": 23151
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 13.253548622131348,
+      "learning_rate": 6.292097226289589e-07,
+      "loss": 1.8187,
+      "step": 23152
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 14.02438735961914,
+      "learning_rate": 6.283730075722714e-07,
+      "loss": 0.7358,
+      "step": 23153
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 12.823583602905273,
+      "learning_rate": 6.275362925155838e-07,
+      "loss": 0.6888,
+      "step": 23154
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 18.59247398376465,
+      "learning_rate": 6.266995774588964e-07,
+      "loss": 0.6411,
+      "step": 23155
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 12.513319969177246,
+      "learning_rate": 6.258628624022089e-07,
+      "loss": 1.2862,
+      "step": 23156
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 185.5263671875,
+      "learning_rate": 6.250261473455216e-07,
+      "loss": 2.3999,
+      "step": 23157
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 11.595245361328125,
+      "learning_rate": 6.241894322888341e-07,
+      "loss": 1.5035,
+      "step": 23158
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 17.20775604248047,
+      "learning_rate": 6.233527172321467e-07,
+      "loss": 0.8866,
+      "step": 23159
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 68.4419174194336,
+      "learning_rate": 6.225160021754592e-07,
+      "loss": 0.2079,
+      "step": 23160
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 6.003311634063721,
+      "learning_rate": 6.216792871187718e-07,
+      "loss": 0.1097,
+      "step": 23161
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 55.0208854675293,
+      "learning_rate": 6.208425720620843e-07,
+      "loss": 2.2685,
+      "step": 23162
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 11.7134428024292,
+      "learning_rate": 6.200058570053968e-07,
+      "loss": 0.8954,
+      "step": 23163
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 19.41765022277832,
+      "learning_rate": 6.191691419487094e-07,
+      "loss": 2.6211,
+      "step": 23164
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 14.811298370361328,
+      "learning_rate": 6.18332426892022e-07,
+      "loss": 0.5761,
+      "step": 23165
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 7.661105632781982,
+      "learning_rate": 6.174957118353345e-07,
+      "loss": 0.5413,
+      "step": 23166
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 11.37485408782959,
+      "learning_rate": 6.166589967786471e-07,
+      "loss": 0.409,
+      "step": 23167
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 7.339240550994873,
+      "learning_rate": 6.158222817219597e-07,
+      "loss": 0.8778,
+      "step": 23168
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 13.293191909790039,
+      "learning_rate": 6.149855666652722e-07,
+      "loss": 0.8464,
+      "step": 23169
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 5.022184371948242,
+      "learning_rate": 6.141488516085847e-07,
+      "loss": 0.1938,
+      "step": 23170
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 11.445906639099121,
+      "learning_rate": 6.133121365518973e-07,
+      "loss": 1.5553,
+      "step": 23171
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 17.587867736816406,
+      "learning_rate": 6.124754214952098e-07,
+      "loss": 0.9976,
+      "step": 23172
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 26.690670013427734,
+      "learning_rate": 6.116387064385224e-07,
+      "loss": 0.836,
+      "step": 23173
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 20.364336013793945,
+      "learning_rate": 6.10801991381835e-07,
+      "loss": 1.8699,
+      "step": 23174
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 15.160350799560547,
+      "learning_rate": 6.099652763251475e-07,
+      "loss": 0.4064,
+      "step": 23175
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 288.4300842285156,
+      "learning_rate": 6.091285612684601e-07,
+      "loss": 0.857,
+      "step": 23176
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.267444610595703,
+      "learning_rate": 6.082918462117727e-07,
+      "loss": 0.3047,
+      "step": 23177
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 23.117055892944336,
+      "learning_rate": 6.074551311550851e-07,
+      "loss": 0.9334,
+      "step": 23178
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.963648796081543,
+      "learning_rate": 6.066184160983977e-07,
+      "loss": 0.639,
+      "step": 23179
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 81.25936126708984,
+      "learning_rate": 6.057817010417103e-07,
+      "loss": 1.1255,
+      "step": 23180
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 205.37783813476562,
+      "learning_rate": 6.049449859850229e-07,
+      "loss": 1.3728,
+      "step": 23181
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 13.395986557006836,
+      "learning_rate": 6.041082709283354e-07,
+      "loss": 1.2588,
+      "step": 23182
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 14.80759334564209,
+      "learning_rate": 6.03271555871648e-07,
+      "loss": 1.804,
+      "step": 23183
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 8.79738712310791,
+      "learning_rate": 6.024348408149605e-07,
+      "loss": 0.223,
+      "step": 23184
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 18.706571578979492,
+      "learning_rate": 6.015981257582731e-07,
+      "loss": 0.8289,
+      "step": 23185
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 13.420476913452148,
+      "learning_rate": 6.007614107015857e-07,
+      "loss": 1.9154,
+      "step": 23186
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 10.114333152770996,
+      "learning_rate": 5.999246956448982e-07,
+      "loss": 0.3806,
+      "step": 23187
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 5.580753803253174,
+      "learning_rate": 5.990879805882107e-07,
+      "loss": 0.219,
+      "step": 23188
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 15.963912963867188,
+      "learning_rate": 5.982512655315232e-07,
+      "loss": 0.7323,
+      "step": 23189
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 14.737431526184082,
+      "learning_rate": 5.974145504748358e-07,
+      "loss": 0.672,
+      "step": 23190
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 10.397017478942871,
+      "learning_rate": 5.965778354181484e-07,
+      "loss": 0.2714,
+      "step": 23191
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 78.07597351074219,
+      "learning_rate": 5.95741120361461e-07,
+      "loss": 1.9645,
+      "step": 23192
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 10.356441497802734,
+      "learning_rate": 5.949044053047736e-07,
+      "loss": 0.6972,
+      "step": 23193
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 174.6509552001953,
+      "learning_rate": 5.940676902480861e-07,
+      "loss": 1.2032,
+      "step": 23194
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 19.284364700317383,
+      "learning_rate": 5.932309751913985e-07,
+      "loss": 1.3186,
+      "step": 23195
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 3.266484260559082,
+      "learning_rate": 5.923942601347111e-07,
+      "loss": 0.0663,
+      "step": 23196
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 11.210550308227539,
+      "learning_rate": 5.915575450780237e-07,
+      "loss": 0.6201,
+      "step": 23197
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 99.94712829589844,
+      "learning_rate": 5.907208300213363e-07,
+      "loss": 2.5873,
+      "step": 23198
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 72.69721984863281,
+      "learning_rate": 5.898841149646489e-07,
+      "loss": 0.947,
+      "step": 23199
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.3036818504333496,
+      "learning_rate": 5.890473999079614e-07,
+      "loss": 0.0282,
+      "step": 23200
+    },
+    {
+      "epoch": 2.91,
+      "eval_loss": 0.07849589735269547,
+      "eval_runtime": 96.2443,
+      "eval_samples_per_second": 36.802,
+      "eval_steps_per_second": 36.802,
+      "step": 23200
     }
   ],
   "logging_steps": 1,
-  "max_steps": 9978,
+  "max_steps": 23904,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 3,
-  "save_steps": 120,
-  "total_flos": 435510745758720.0,
+  "save_steps": 400,
+  "total_flos": 3277316217323520.0,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null