diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,39731 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.0,
+  "eval_steps": 500,
+  "global_step": 5604,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 1443.5744044969551,
+      "learning_rate": 5.91715976331361e-08,
+      "loss": 12.2344,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1534.8176448250863,
+      "learning_rate": 1.183431952662722e-07,
+      "loss": 11.7344,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1326.3321019398716,
+      "learning_rate": 1.775147928994083e-07,
+      "loss": 11.6875,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1288.3465989587996,
+      "learning_rate": 2.366863905325444e-07,
+      "loss": 11.0938,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1341.804033668729,
+      "learning_rate": 2.958579881656805e-07,
+      "loss": 12.3125,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1329.3145585197879,
+      "learning_rate": 3.550295857988166e-07,
+      "loss": 11.0938,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1130.227729481188,
+      "learning_rate": 4.1420118343195276e-07,
+      "loss": 11.375,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1160.6261885441586,
+      "learning_rate": 4.733727810650888e-07,
+      "loss": 9.8906,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1039.7040454148366,
+      "learning_rate": 5.32544378698225e-07,
+      "loss": 9.8984,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 835.4908181410407,
+      "learning_rate": 5.91715976331361e-07,
+      "loss": 8.6406,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 687.6534892135122,
+      "learning_rate": 6.50887573964497e-07,
+      "loss": 7.5391,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 354.312170952273,
+      "learning_rate": 7.100591715976332e-07,
+      "loss": 7.4141,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 365.8199705384397,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 7.0469,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1216.9260067540185,
+      "learning_rate": 8.284023668639055e-07,
+      "loss": 8.3516,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1474.4218345737838,
+      "learning_rate": 8.875739644970415e-07,
+      "loss": 8.1016,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1393.9506087028988,
+      "learning_rate": 9.467455621301776e-07,
+      "loss": 8.0,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1475.954883625004,
+      "learning_rate": 1.0059171597633138e-06,
+      "loss": 8.7891,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1219.677012756364,
+      "learning_rate": 1.06508875739645e-06,
+      "loss": 6.4844,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 938.9090098183609,
+      "learning_rate": 1.1242603550295859e-06,
+      "loss": 4.9531,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 732.6809380140464,
+      "learning_rate": 1.183431952662722e-06,
+      "loss": 3.7578,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 312.90545925368025,
+      "learning_rate": 1.242603550295858e-06,
+      "loss": 2.3359,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 131.42857798633014,
+      "learning_rate": 1.301775147928994e-06,
+      "loss": 2.5195,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 204.75611392809807,
+      "learning_rate": 1.3609467455621303e-06,
+      "loss": 2.1074,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 281.53999585238597,
+      "learning_rate": 1.4201183431952664e-06,
+      "loss": 2.3008,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 407.1340542086892,
+      "learning_rate": 1.4792899408284026e-06,
+      "loss": 2.3633,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 513.0873413511963,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 2.457,
+      "step": 26
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 461.55622845849354,
+      "learning_rate": 1.5976331360946749e-06,
+      "loss": 2.1602,
+      "step": 27
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 478.542794042002,
+      "learning_rate": 1.656804733727811e-06,
+      "loss": 2.1387,
+      "step": 28
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 344.6282465781462,
+      "learning_rate": 1.7159763313609468e-06,
+      "loss": 1.7637,
+      "step": 29
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 362.57393430326465,
+      "learning_rate": 1.775147928994083e-06,
+      "loss": 1.6602,
+      "step": 30
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 168.67532989078214,
+      "learning_rate": 1.834319526627219e-06,
+      "loss": 1.3105,
+      "step": 31
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 78.2335154608893,
+      "learning_rate": 1.8934911242603552e-06,
+      "loss": 1.0537,
+      "step": 32
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 74.86312041211706,
+      "learning_rate": 1.952662721893491e-06,
+      "loss": 0.9717,
+      "step": 33
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 173.72567437330426,
+      "learning_rate": 2.0118343195266275e-06,
+      "loss": 1.0166,
+      "step": 34
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 244.678248816213,
+      "learning_rate": 2.0710059171597635e-06,
+      "loss": 1.1279,
+      "step": 35
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 289.70307970146496,
+      "learning_rate": 2.1301775147929e-06,
+      "loss": 1.1523,
+      "step": 36
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 261.404070412537,
+      "learning_rate": 2.1893491124260358e-06,
+      "loss": 1.0615,
+      "step": 37
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 248.72543936315603,
+      "learning_rate": 2.2485207100591717e-06,
+      "loss": 0.96,
+      "step": 38
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 175.19387912539813,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 0.8223,
+      "step": 39
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 129.57230819430404,
+      "learning_rate": 2.366863905325444e-06,
+      "loss": 0.7617,
+      "step": 40
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 64.75648312193796,
+      "learning_rate": 2.42603550295858e-06,
+      "loss": 0.6279,
+      "step": 41
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 40.345496542665984,
+      "learning_rate": 2.485207100591716e-06,
+      "loss": 0.6426,
+      "step": 42
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 112.96546034948359,
+      "learning_rate": 2.5443786982248527e-06,
+      "loss": 0.6133,
+      "step": 43
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 113.11724928693071,
+      "learning_rate": 2.603550295857988e-06,
+      "loss": 0.6172,
+      "step": 44
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 170.2583256015434,
+      "learning_rate": 2.6627218934911246e-06,
+      "loss": 0.6826,
+      "step": 45
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 181.13824595705816,
+      "learning_rate": 2.7218934911242605e-06,
+      "loss": 0.6807,
+      "step": 46
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 127.83994169123808,
+      "learning_rate": 2.7810650887573965e-06,
+      "loss": 0.5532,
+      "step": 47
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 160.92696853722236,
+      "learning_rate": 2.840236686390533e-06,
+      "loss": 0.6377,
+      "step": 48
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 83.85759914507452,
+      "learning_rate": 2.8994082840236688e-06,
+      "loss": 0.5171,
+      "step": 49
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 34.82212793694129,
+      "learning_rate": 2.958579881656805e-06,
+      "loss": 0.4097,
+      "step": 50
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 25.956713751406728,
+      "learning_rate": 3.017751479289941e-06,
+      "loss": 0.4434,
+      "step": 51
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 78.75487723337257,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 0.394,
+      "step": 52
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 154.51220671292447,
+      "learning_rate": 3.1360946745562134e-06,
+      "loss": 0.5396,
+      "step": 53
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 167.45386551318256,
+      "learning_rate": 3.1952662721893497e-06,
+      "loss": 0.5225,
+      "step": 54
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 153.7596657181405,
+      "learning_rate": 3.2544378698224853e-06,
+      "loss": 0.4897,
+      "step": 55
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 104.90584941846187,
+      "learning_rate": 3.313609467455622e-06,
+      "loss": 0.4272,
+      "step": 56
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 80.49574128984705,
+      "learning_rate": 3.3727810650887576e-06,
+      "loss": 0.395,
+      "step": 57
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 44.08211560352312,
+      "learning_rate": 3.4319526627218935e-06,
+      "loss": 0.3936,
+      "step": 58
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 74.39313269747478,
+      "learning_rate": 3.49112426035503e-06,
+      "loss": 0.4253,
+      "step": 59
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 69.8169714038155,
+      "learning_rate": 3.550295857988166e-06,
+      "loss": 0.3569,
+      "step": 60
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 89.54284154509011,
+      "learning_rate": 3.609467455621302e-06,
+      "loss": 0.3887,
+      "step": 61
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 103.65697090523831,
+      "learning_rate": 3.668639053254438e-06,
+      "loss": 0.3853,
+      "step": 62
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 77.98627320346968,
+      "learning_rate": 3.7278106508875745e-06,
+      "loss": 0.3628,
+      "step": 63
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 42.66416290101905,
+      "learning_rate": 3.7869822485207104e-06,
+      "loss": 0.3237,
+      "step": 64
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 33.27388124719927,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 0.3032,
+      "step": 65
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 36.86492361076572,
+      "learning_rate": 3.905325443786982e-06,
+      "loss": 0.3418,
+      "step": 66
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 66.58133496209655,
+      "learning_rate": 3.964497041420119e-06,
+      "loss": 0.3433,
+      "step": 67
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 34.40834590692991,
+      "learning_rate": 4.023668639053255e-06,
+      "loss": 0.3169,
+      "step": 68
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 89.33579913199095,
+      "learning_rate": 4.0828402366863906e-06,
+      "loss": 0.3481,
+      "step": 69
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 75.01768722175284,
+      "learning_rate": 4.142011834319527e-06,
+      "loss": 0.3252,
+      "step": 70
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.173648614704415,
+      "learning_rate": 4.201183431952663e-06,
+      "loss": 0.2759,
+      "step": 71
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 18.990275484108082,
+      "learning_rate": 4.2603550295858e-06,
+      "loss": 0.2705,
+      "step": 72
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 38.76375650821992,
+      "learning_rate": 4.319526627218935e-06,
+      "loss": 0.2842,
+      "step": 73
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 56.67502763908607,
+      "learning_rate": 4.3786982248520715e-06,
+      "loss": 0.3154,
+      "step": 74
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 66.38202402483932,
+      "learning_rate": 4.437869822485207e-06,
+      "loss": 0.334,
+      "step": 75
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 26.659837428204344,
+      "learning_rate": 4.497041420118343e-06,
+      "loss": 0.2437,
+      "step": 76
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 49.89366358764825,
+      "learning_rate": 4.55621301775148e-06,
+      "loss": 0.2893,
+      "step": 77
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 22.724969702099184,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 0.2725,
+      "step": 78
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 28.80561993848226,
+      "learning_rate": 4.674556213017752e-06,
+      "loss": 0.3044,
+      "step": 79
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 29.286530669940298,
+      "learning_rate": 4.733727810650888e-06,
+      "loss": 0.2842,
+      "step": 80
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 14.16660585102508,
+      "learning_rate": 4.792899408284024e-06,
+      "loss": 0.2437,
+      "step": 81
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 73.9926353523121,
+      "learning_rate": 4.85207100591716e-06,
+      "loss": 0.2827,
+      "step": 82
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 29.878951527569956,
+      "learning_rate": 4.911242603550296e-06,
+      "loss": 0.2458,
+      "step": 83
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.484475177918583,
+      "learning_rate": 4.970414201183432e-06,
+      "loss": 0.2476,
+      "step": 84
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 55.59592780267639,
+      "learning_rate": 5.029585798816569e-06,
+      "loss": 0.2793,
+      "step": 85
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 50.21555063218247,
+      "learning_rate": 5.088757396449705e-06,
+      "loss": 0.2581,
+      "step": 86
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.528834407068818,
+      "learning_rate": 5.14792899408284e-06,
+      "loss": 0.239,
+      "step": 87
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.041769271462492,
+      "learning_rate": 5.207100591715976e-06,
+      "loss": 0.2402,
+      "step": 88
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 21.52430533549206,
+      "learning_rate": 5.266272189349113e-06,
+      "loss": 0.2417,
+      "step": 89
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.677037672727517,
+      "learning_rate": 5.325443786982249e-06,
+      "loss": 0.2317,
+      "step": 90
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 38.59157227506834,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 0.2336,
+      "step": 91
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 50.86665965832706,
+      "learning_rate": 5.443786982248521e-06,
+      "loss": 0.2283,
+      "step": 92
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.565157782160531,
+      "learning_rate": 5.502958579881657e-06,
+      "loss": 0.229,
+      "step": 93
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.612894516114126,
+      "learning_rate": 5.562130177514793e-06,
+      "loss": 0.2034,
+      "step": 94
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 41.99675072884753,
+      "learning_rate": 5.621301775147929e-06,
+      "loss": 0.218,
+      "step": 95
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 21.606516634808887,
+      "learning_rate": 5.680473372781066e-06,
+      "loss": 0.238,
+      "step": 96
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 38.612734423443584,
+      "learning_rate": 5.739644970414202e-06,
+      "loss": 0.2727,
+      "step": 97
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.14173463955849,
+      "learning_rate": 5.7988165680473375e-06,
+      "loss": 0.2144,
+      "step": 98
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.726835683501431,
+      "learning_rate": 5.857988165680474e-06,
+      "loss": 0.2217,
+      "step": 99
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 33.271941153238885,
+      "learning_rate": 5.91715976331361e-06,
+      "loss": 0.2334,
+      "step": 100
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 28.312522667414854,
+      "learning_rate": 5.976331360946747e-06,
+      "loss": 0.229,
+      "step": 101
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 39.47150040026952,
+      "learning_rate": 6.035502958579882e-06,
+      "loss": 0.2241,
+      "step": 102
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.387184326679442,
+      "learning_rate": 6.0946745562130185e-06,
+      "loss": 0.1914,
+      "step": 103
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.393251230339942,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.1992,
+      "step": 104
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 43.36911762706155,
+      "learning_rate": 6.21301775147929e-06,
+      "loss": 0.2244,
+      "step": 105
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 54.08032944347241,
+      "learning_rate": 6.272189349112427e-06,
+      "loss": 0.2598,
+      "step": 106
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 34.12457211661539,
+      "learning_rate": 6.331360946745563e-06,
+      "loss": 0.2126,
+      "step": 107
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.87841011317557,
+      "learning_rate": 6.3905325443786995e-06,
+      "loss": 0.1987,
+      "step": 108
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 45.90938123438191,
+      "learning_rate": 6.449704142011834e-06,
+      "loss": 0.2549,
+      "step": 109
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 43.25371651313902,
+      "learning_rate": 6.5088757396449705e-06,
+      "loss": 0.239,
+      "step": 110
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 37.74000118926859,
+      "learning_rate": 6.568047337278107e-06,
+      "loss": 0.2085,
+      "step": 111
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.833634104200893,
+      "learning_rate": 6.627218934911244e-06,
+      "loss": 0.1685,
+      "step": 112
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 36.89708866713719,
+      "learning_rate": 6.686390532544379e-06,
+      "loss": 0.2686,
+      "step": 113
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 50.89388673898367,
+      "learning_rate": 6.745562130177515e-06,
+      "loss": 0.2107,
+      "step": 114
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 44.94360423791429,
+      "learning_rate": 6.8047337278106515e-06,
+      "loss": 0.2383,
+      "step": 115
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.889886449107443,
+      "learning_rate": 6.863905325443787e-06,
+      "loss": 0.2217,
+      "step": 116
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 51.161996812885185,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 0.1895,
+      "step": 117
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 48.48659614955822,
+      "learning_rate": 6.98224852071006e-06,
+      "loss": 0.2227,
+      "step": 118
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.885655124875624,
+      "learning_rate": 7.041420118343196e-06,
+      "loss": 0.1958,
+      "step": 119
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 27.132063430841765,
+      "learning_rate": 7.100591715976332e-06,
+      "loss": 0.2239,
+      "step": 120
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 18.86057434855715,
+      "learning_rate": 7.159763313609468e-06,
+      "loss": 0.1902,
+      "step": 121
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.637608877658806,
+      "learning_rate": 7.218934911242604e-06,
+      "loss": 0.2109,
+      "step": 122
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.047033973462321,
+      "learning_rate": 7.278106508875741e-06,
+      "loss": 0.1904,
+      "step": 123
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.709619680883495,
+      "learning_rate": 7.337278106508876e-06,
+      "loss": 0.1729,
+      "step": 124
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 44.025278094830405,
+      "learning_rate": 7.396449704142013e-06,
+      "loss": 0.2012,
+      "step": 125
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.579611221108369,
+      "learning_rate": 7.455621301775149e-06,
+      "loss": 0.1807,
+      "step": 126
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.178080204391332,
+      "learning_rate": 7.5147928994082845e-06,
+      "loss": 0.1792,
+      "step": 127
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.70187830067804,
+      "learning_rate": 7.573964497041421e-06,
+      "loss": 0.1648,
+      "step": 128
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 41.67469230519406,
+      "learning_rate": 7.633136094674556e-06,
+      "loss": 0.189,
+      "step": 129
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 56.145533909909716,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 0.1772,
+      "step": 130
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.886508074747201,
+      "learning_rate": 7.751479289940829e-06,
+      "loss": 0.2095,
+      "step": 131
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 92.16611785910685,
+      "learning_rate": 7.810650887573965e-06,
+      "loss": 0.2383,
+      "step": 132
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 40.64519507288005,
+      "learning_rate": 7.869822485207102e-06,
+      "loss": 0.2139,
+      "step": 133
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 14.527350282989591,
+      "learning_rate": 7.928994082840237e-06,
+      "loss": 0.2041,
+      "step": 134
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 82.46586621777837,
+      "learning_rate": 7.988165680473373e-06,
+      "loss": 0.2083,
+      "step": 135
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 60.35861295571972,
+      "learning_rate": 8.04733727810651e-06,
+      "loss": 0.2388,
+      "step": 136
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.976620441569606,
+      "learning_rate": 8.106508875739646e-06,
+      "loss": 0.1814,
+      "step": 137
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 76.28662037424316,
+      "learning_rate": 8.165680473372781e-06,
+      "loss": 0.218,
+      "step": 138
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 49.990057565286385,
+      "learning_rate": 8.224852071005918e-06,
+      "loss": 0.2263,
+      "step": 139
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.12483970210935,
+      "learning_rate": 8.284023668639054e-06,
+      "loss": 0.1628,
+      "step": 140
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 34.40581343675325,
+      "learning_rate": 8.343195266272191e-06,
+      "loss": 0.2046,
+      "step": 141
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 86.73910967064404,
+      "learning_rate": 8.402366863905327e-06,
+      "loss": 0.2832,
+      "step": 142
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 44.998603587663474,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 0.2,
+      "step": 143
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 60.632795608186285,
+      "learning_rate": 8.5207100591716e-06,
+      "loss": 0.163,
+      "step": 144
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 139.66966654303818,
+      "learning_rate": 8.579881656804735e-06,
+      "loss": 0.356,
+      "step": 145
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 21.720375790615257,
+      "learning_rate": 8.63905325443787e-06,
+      "loss": 0.1716,
+      "step": 146
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 93.50974348316393,
+      "learning_rate": 8.698224852071006e-06,
+      "loss": 0.2695,
+      "step": 147
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 77.34688239840453,
+      "learning_rate": 8.757396449704143e-06,
+      "loss": 0.2104,
+      "step": 148
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 38.11609599313387,
+      "learning_rate": 8.816568047337279e-06,
+      "loss": 0.1985,
+      "step": 149
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 34.022306259789644,
+      "learning_rate": 8.875739644970414e-06,
+      "loss": 0.1743,
+      "step": 150
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 86.92363217035724,
+      "learning_rate": 8.934911242603551e-06,
+      "loss": 0.2617,
+      "step": 151
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 43.71673888744464,
+      "learning_rate": 8.994082840236687e-06,
+      "loss": 0.2148,
+      "step": 152
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 31.570603625815373,
+      "learning_rate": 9.053254437869822e-06,
+      "loss": 0.1858,
+      "step": 153
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 61.077459467399706,
+      "learning_rate": 9.11242603550296e-06,
+      "loss": 0.2427,
+      "step": 154
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 50.10622818300115,
+      "learning_rate": 9.171597633136095e-06,
+      "loss": 0.2407,
+      "step": 155
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 29.673041634897242,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 0.1736,
+      "step": 156
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 34.729480106635634,
+      "learning_rate": 9.289940828402368e-06,
+      "loss": 0.1782,
+      "step": 157
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 44.993817096293725,
+      "learning_rate": 9.349112426035503e-06,
+      "loss": 0.1809,
+      "step": 158
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.899448788867884,
+      "learning_rate": 9.40828402366864e-06,
+      "loss": 0.1951,
+      "step": 159
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 18.286850457168683,
+      "learning_rate": 9.467455621301776e-06,
+      "loss": 0.1606,
+      "step": 160
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 56.53072371599443,
+      "learning_rate": 9.526627218934912e-06,
+      "loss": 0.2,
+      "step": 161
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.264038479497928,
+      "learning_rate": 9.585798816568049e-06,
+      "loss": 0.1448,
+      "step": 162
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 43.211370059501874,
+      "learning_rate": 9.644970414201184e-06,
+      "loss": 0.1975,
+      "step": 163
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.273967210767667,
+      "learning_rate": 9.70414201183432e-06,
+      "loss": 0.1809,
+      "step": 164
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 32.98648840761662,
+      "learning_rate": 9.763313609467457e-06,
+      "loss": 0.1685,
+      "step": 165
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 33.93062311124414,
+      "learning_rate": 9.822485207100593e-06,
+      "loss": 0.1956,
+      "step": 166
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 51.5363090967256,
+      "learning_rate": 9.88165680473373e-06,
+      "loss": 0.187,
+      "step": 167
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.60396802401412,
+      "learning_rate": 9.940828402366864e-06,
+      "loss": 0.1592,
+      "step": 168
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 25.22452245463198,
+      "learning_rate": 1e-05,
+      "loss": 0.1738,
+      "step": 169
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 57.51781109179427,
+      "learning_rate": 9.999999164703534e-06,
+      "loss": 0.21,
+      "step": 170
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 19.002582848688764,
+      "learning_rate": 9.999996658814406e-06,
+      "loss": 0.1428,
+      "step": 171
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.276224778176067,
+      "learning_rate": 9.999992482333461e-06,
+      "loss": 0.1465,
+      "step": 172
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 44.678917780683705,
+      "learning_rate": 9.99998663526209e-06,
+      "loss": 0.1617,
+      "step": 173
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.960442853339805,
+      "learning_rate": 9.99997911760225e-06,
+      "loss": 0.1443,
+      "step": 174
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.261742691312454,
+      "learning_rate": 9.99996992935645e-06,
+      "loss": 0.1682,
+      "step": 175
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 24.032339010668686,
+      "learning_rate": 9.99995907052776e-06,
+      "loss": 0.1469,
+      "step": 176
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 58.31377133614243,
+      "learning_rate": 9.99994654111981e-06,
+      "loss": 0.2073,
+      "step": 177
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 23.70246515992443,
+      "learning_rate": 9.999932341136785e-06,
+      "loss": 0.1718,
+      "step": 178
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 26.891031514237994,
+      "learning_rate": 9.999916470583429e-06,
+      "loss": 0.1554,
+      "step": 179
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 57.563572942900166,
+      "learning_rate": 9.999898929465047e-06,
+      "loss": 0.1926,
+      "step": 180
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 36.34968696505108,
+      "learning_rate": 9.999879717787495e-06,
+      "loss": 0.1558,
+      "step": 181
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.794023955932686,
+      "learning_rate": 9.999858835557197e-06,
+      "loss": 0.145,
+      "step": 182
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 47.23120098027718,
+      "learning_rate": 9.999836282781128e-06,
+      "loss": 0.178,
+      "step": 183
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 40.114505438540625,
+      "learning_rate": 9.999812059466825e-06,
+      "loss": 0.2019,
+      "step": 184
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 25.380586865046148,
+      "learning_rate": 9.999786165622379e-06,
+      "loss": 0.1559,
+      "step": 185
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 29.25336121583889,
+      "learning_rate": 9.999758601256441e-06,
+      "loss": 0.1888,
+      "step": 186
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 64.84408243732132,
+      "learning_rate": 9.999729366378224e-06,
+      "loss": 0.1921,
+      "step": 187
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 22.31665172260842,
+      "learning_rate": 9.999698460997493e-06,
+      "loss": 0.1838,
+      "step": 188
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 18.476404539169845,
+      "learning_rate": 9.999665885124577e-06,
+      "loss": 0.1499,
+      "step": 189
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 41.418357640800615,
+      "learning_rate": 9.99963163877036e-06,
+      "loss": 0.2021,
+      "step": 190
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 31.53664585439554,
+      "learning_rate": 9.99959572194628e-06,
+      "loss": 0.2122,
+      "step": 191
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 21.575514536412193,
+      "learning_rate": 9.999558134664342e-06,
+      "loss": 0.1829,
+      "step": 192
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 39.02562270046022,
+      "learning_rate": 9.999518876937102e-06,
+      "loss": 0.1566,
+      "step": 193
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 33.24695587894434,
+      "learning_rate": 9.999477948777678e-06,
+      "loss": 0.1599,
+      "step": 194
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 45.78127128268334,
+      "learning_rate": 9.999435350199745e-06,
+      "loss": 0.1709,
+      "step": 195
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 42.42344263018681,
+      "learning_rate": 9.999391081217536e-06,
+      "loss": 0.1919,
+      "step": 196
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.73579636286334,
+      "learning_rate": 9.999345141845842e-06,
+      "loss": 0.1562,
+      "step": 197
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 42.85891427323893,
+      "learning_rate": 9.99929753210001e-06,
+      "loss": 0.1689,
+      "step": 198
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.259144765611043,
+      "learning_rate": 9.999248251995951e-06,
+      "loss": 0.1653,
+      "step": 199
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 24.27283819335801,
+      "learning_rate": 9.999197301550127e-06,
+      "loss": 0.1798,
+      "step": 200
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 27.899669266770783,
+      "learning_rate": 9.999144680779564e-06,
+      "loss": 0.1731,
+      "step": 201
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.179927364455224,
+      "learning_rate": 9.999090389701844e-06,
+      "loss": 0.1702,
+      "step": 202
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 19.847681296179072,
+      "learning_rate": 9.999034428335103e-06,
+      "loss": 0.1591,
+      "step": 203
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.752832888535757,
+      "learning_rate": 9.998976796698043e-06,
+      "loss": 0.1545,
+      "step": 204
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 20.029510235909346,
+      "learning_rate": 9.998917494809917e-06,
+      "loss": 0.1603,
+      "step": 205
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 15.214337375526133,
+      "learning_rate": 9.998856522690538e-06,
+      "loss": 0.1699,
+      "step": 206
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 12.852230970076132,
+      "learning_rate": 9.998793880360283e-06,
+      "loss": 0.1493,
+      "step": 207
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 42.16352485230185,
+      "learning_rate": 9.998729567840077e-06,
+      "loss": 0.2166,
+      "step": 208
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 43.29824172685643,
+      "learning_rate": 9.998663585151409e-06,
+      "loss": 0.1897,
+      "step": 209
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 42.33267438448777,
+      "learning_rate": 9.998595932316327e-06,
+      "loss": 0.1721,
+      "step": 210
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.107364224481227,
+      "learning_rate": 9.998526609357432e-06,
+      "loss": 0.1514,
+      "step": 211
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 40.03181460372598,
+      "learning_rate": 9.998455616297889e-06,
+      "loss": 0.1746,
+      "step": 212
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 49.846724341530525,
+      "learning_rate": 9.998382953161417e-06,
+      "loss": 0.2153,
+      "step": 213
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 23.55333318015724,
+      "learning_rate": 9.998308619972292e-06,
+      "loss": 0.1775,
+      "step": 214
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 49.65058025675812,
+      "learning_rate": 9.998232616755354e-06,
+      "loss": 0.1721,
+      "step": 215
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 30.06267959273612,
+      "learning_rate": 9.998154943535996e-06,
+      "loss": 0.1575,
+      "step": 216
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 31.9421224193554,
+      "learning_rate": 9.998075600340166e-06,
+      "loss": 0.179,
+      "step": 217
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 44.71443339581383,
+      "learning_rate": 9.997994587194381e-06,
+      "loss": 0.1708,
+      "step": 218
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.825036612048357,
+      "learning_rate": 9.997911904125704e-06,
+      "loss": 0.141,
+      "step": 219
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.105912112927012,
+      "learning_rate": 9.997827551161762e-06,
+      "loss": 0.1552,
+      "step": 220
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 30.337684796583403,
+      "learning_rate": 9.997741528330739e-06,
+      "loss": 0.1497,
+      "step": 221
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.256728997034248,
+      "learning_rate": 9.997653835661376e-06,
+      "loss": 0.1792,
+      "step": 222
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.710837732666267,
+      "learning_rate": 9.997564473182976e-06,
+      "loss": 0.1416,
+      "step": 223
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.138467739217358,
+      "learning_rate": 9.997473440925394e-06,
+      "loss": 0.1589,
+      "step": 224
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.011674393770035,
+      "learning_rate": 9.997380738919045e-06,
+      "loss": 0.1641,
+      "step": 225
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 45.558916966443206,
+      "learning_rate": 9.997286367194903e-06,
+      "loss": 0.1786,
+      "step": 226
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 22.828341098912837,
+      "learning_rate": 9.9971903257845e-06,
+      "loss": 0.1472,
+      "step": 227
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 32.44780904932061,
+      "learning_rate": 9.997092614719926e-06,
+      "loss": 0.1599,
+      "step": 228
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 37.937074029999266,
+      "learning_rate": 9.996993234033826e-06,
+      "loss": 0.1521,
+      "step": 229
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 22.707824607414384,
+      "learning_rate": 9.996892183759407e-06,
+      "loss": 0.1743,
+      "step": 230
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 46.45915188550997,
+      "learning_rate": 9.99678946393043e-06,
+      "loss": 0.1892,
+      "step": 231
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 24.319677361782876,
+      "learning_rate": 9.996685074581216e-06,
+      "loss": 0.1489,
+      "step": 232
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 18.661367787669693,
+      "learning_rate": 9.996579015746645e-06,
+      "loss": 0.1721,
+      "step": 233
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 57.942764629266655,
+      "learning_rate": 9.996471287462151e-06,
+      "loss": 0.1649,
+      "step": 234
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 13.048230837396387,
+      "learning_rate": 9.99636188976373e-06,
+      "loss": 0.1602,
+      "step": 235
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.709658285204258,
+      "learning_rate": 9.996250822687932e-06,
+      "loss": 0.1438,
+      "step": 236
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 31.617100347623733,
+      "learning_rate": 9.996138086271869e-06,
+      "loss": 0.1556,
+      "step": 237
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 35.27416734620715,
+      "learning_rate": 9.996023680553204e-06,
+      "loss": 0.1575,
+      "step": 238
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.153491233670788,
+      "learning_rate": 9.995907605570167e-06,
+      "loss": 0.1797,
+      "step": 239
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.987526278742785,
+      "learning_rate": 9.995789861361538e-06,
+      "loss": 0.2065,
+      "step": 240
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 64.32372496086278,
+      "learning_rate": 9.995670447966658e-06,
+      "loss": 0.1824,
+      "step": 241
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.645118470787404,
+      "learning_rate": 9.995549365425426e-06,
+      "loss": 0.1591,
+      "step": 242
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.7785990929647815,
+      "learning_rate": 9.995426613778297e-06,
+      "loss": 0.1615,
+      "step": 243
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 44.643088379285764,
+      "learning_rate": 9.995302193066286e-06,
+      "loss": 0.1528,
+      "step": 244
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 29.669669631908086,
+      "learning_rate": 9.995176103330962e-06,
+      "loss": 0.1643,
+      "step": 245
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.187868389208368,
+      "learning_rate": 9.995048344614455e-06,
+      "loss": 0.1855,
+      "step": 246
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 64.34845710145298,
+      "learning_rate": 9.994918916959453e-06,
+      "loss": 0.2019,
+      "step": 247
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 27.545718441791433,
+      "learning_rate": 9.994787820409198e-06,
+      "loss": 0.1816,
+      "step": 248
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 16.691064146642706,
+      "learning_rate": 9.994655055007491e-06,
+      "loss": 0.1548,
+      "step": 249
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 32.47918718068585,
+      "learning_rate": 9.994520620798696e-06,
+      "loss": 0.1423,
+      "step": 250
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 62.24279005895381,
+      "learning_rate": 9.994384517827726e-06,
+      "loss": 0.1908,
+      "step": 251
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.306470473367477,
+      "learning_rate": 9.994246746140057e-06,
+      "loss": 0.1655,
+      "step": 252
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 69.9581859530375,
+      "learning_rate": 9.99410730578172e-06,
+      "loss": 0.2007,
+      "step": 253
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 35.84402462268306,
+      "learning_rate": 9.993966196799304e-06,
+      "loss": 0.1494,
+      "step": 254
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.728361882638175,
+      "learning_rate": 9.993823419239959e-06,
+      "loss": 0.1531,
+      "step": 255
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 30.744672622096342,
+      "learning_rate": 9.993678973151388e-06,
+      "loss": 0.1378,
+      "step": 256
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 72.68699910576366,
+      "learning_rate": 9.993532858581853e-06,
+      "loss": 0.2354,
+      "step": 257
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.87452116844784,
+      "learning_rate": 9.993385075580173e-06,
+      "loss": 0.1495,
+      "step": 258
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 27.98767655605202,
+      "learning_rate": 9.993235624195728e-06,
+      "loss": 0.1709,
+      "step": 259
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 33.83809754783617,
+      "learning_rate": 9.993084504478448e-06,
+      "loss": 0.167,
+      "step": 260
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.335493128269599,
+      "learning_rate": 9.99293171647883e-06,
+      "loss": 0.1222,
+      "step": 261
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 13.010955660808273,
+      "learning_rate": 9.992777260247916e-06,
+      "loss": 0.1418,
+      "step": 262
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.84935039014272,
+      "learning_rate": 9.99262113583732e-06,
+      "loss": 0.1448,
+      "step": 263
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 15.275558317087714,
+      "learning_rate": 9.992463343299203e-06,
+      "loss": 0.1398,
+      "step": 264
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.577795034557614,
+      "learning_rate": 9.992303882686288e-06,
+      "loss": 0.1459,
+      "step": 265
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 21.96507681305609,
+      "learning_rate": 9.99214275405185e-06,
+      "loss": 0.1466,
+      "step": 266
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.113512224090927,
+      "learning_rate": 9.991979957449729e-06,
+      "loss": 0.149,
+      "step": 267
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.810162297246139,
+      "learning_rate": 9.991815492934318e-06,
+      "loss": 0.1479,
+      "step": 268
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 32.60107223287011,
+      "learning_rate": 9.991649360560565e-06,
+      "loss": 0.1899,
+      "step": 269
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 29.946520202273987,
+      "learning_rate": 9.99148156038398e-06,
+      "loss": 0.1804,
+      "step": 270
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.367035744107707,
+      "learning_rate": 9.991312092460626e-06,
+      "loss": 0.1296,
+      "step": 271
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 62.72671197950166,
+      "learning_rate": 9.991140956847128e-06,
+      "loss": 0.207,
+      "step": 272
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.62480621517589,
+      "learning_rate": 9.990968153600664e-06,
+      "loss": 0.1626,
+      "step": 273
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 19.419784824468074,
+      "learning_rate": 9.990793682778973e-06,
+      "loss": 0.139,
+      "step": 274
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 52.135212954885354,
+      "learning_rate": 9.990617544440346e-06,
+      "loss": 0.1566,
+      "step": 275
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 20.1662562272831,
+      "learning_rate": 9.990439738643635e-06,
+      "loss": 0.1516,
+      "step": 276
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 21.742228599596476,
+      "learning_rate": 9.99026026544825e-06,
+      "loss": 0.1516,
+      "step": 277
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 47.92258723772546,
+      "learning_rate": 9.990079124914156e-06,
+      "loss": 0.1448,
+      "step": 278
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 30.997197668797394,
+      "learning_rate": 9.989896317101873e-06,
+      "loss": 0.1375,
+      "step": 279
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 31.062475355277208,
+      "learning_rate": 9.989711842072482e-06,
+      "loss": 0.1689,
+      "step": 280
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 53.09032787145096,
+      "learning_rate": 9.989525699887619e-06,
+      "loss": 0.1543,
+      "step": 281
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 31.435016992194903,
+      "learning_rate": 9.989337890609478e-06,
+      "loss": 0.1792,
+      "step": 282
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 17.590958097222614,
+      "learning_rate": 9.98914841430081e-06,
+      "loss": 0.1298,
+      "step": 283
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.770018952416595,
+      "learning_rate": 9.988957271024922e-06,
+      "loss": 0.119,
+      "step": 284
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 67.99377764408035,
+      "learning_rate": 9.988764460845676e-06,
+      "loss": 0.2058,
+      "step": 285
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 37.91613195879691,
+      "learning_rate": 9.9885699838275e-06,
+      "loss": 0.1306,
+      "step": 286
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 26.261219214046495,
+      "learning_rate": 9.988373840035366e-06,
+      "loss": 0.1443,
+      "step": 287
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 42.836761940466495,
+      "learning_rate": 9.988176029534814e-06,
+      "loss": 0.161,
+      "step": 288
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 39.39654040488338,
+      "learning_rate": 9.987976552391933e-06,
+      "loss": 0.1567,
+      "step": 289
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.790620806136728,
+      "learning_rate": 9.987775408673373e-06,
+      "loss": 0.1569,
+      "step": 290
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 53.31062556612756,
+      "learning_rate": 9.987572598446337e-06,
+      "loss": 0.1945,
+      "step": 291
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 56.490600010392015,
+      "learning_rate": 9.987368121778594e-06,
+      "loss": 0.1736,
+      "step": 292
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.122995442967113,
+      "learning_rate": 9.98716197873846e-06,
+      "loss": 0.1388,
+      "step": 293
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 92.30833385479315,
+      "learning_rate": 9.98695416939481e-06,
+      "loss": 0.231,
+      "step": 294
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 35.983480070160525,
+      "learning_rate": 9.986744693817077e-06,
+      "loss": 0.1768,
+      "step": 295
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 16.767014422656075,
+      "learning_rate": 9.986533552075252e-06,
+      "loss": 0.1654,
+      "step": 296
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 34.16924846592324,
+      "learning_rate": 9.986320744239883e-06,
+      "loss": 0.1589,
+      "step": 297
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 51.06145849537146,
+      "learning_rate": 9.98610627038207e-06,
+      "loss": 0.1792,
+      "step": 298
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.179575987397332,
+      "learning_rate": 9.985890130573474e-06,
+      "loss": 0.1523,
+      "step": 299
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 32.87174916074473,
+      "learning_rate": 9.98567232488631e-06,
+      "loss": 0.1482,
+      "step": 300
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 20.716840238877516,
+      "learning_rate": 9.985452853393353e-06,
+      "loss": 0.1276,
+      "step": 301
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 39.212614000189724,
+      "learning_rate": 9.985231716167933e-06,
+      "loss": 0.15,
+      "step": 302
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 14.977736600228983,
+      "learning_rate": 9.985008913283933e-06,
+      "loss": 0.1562,
+      "step": 303
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 36.10332759742872,
+      "learning_rate": 9.984784444815799e-06,
+      "loss": 0.191,
+      "step": 304
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.909269570291388,
+      "learning_rate": 9.984558310838528e-06,
+      "loss": 0.1589,
+      "step": 305
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 28.923562213086704,
+      "learning_rate": 9.984330511427676e-06,
+      "loss": 0.1799,
+      "step": 306
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.490403752500722,
+      "learning_rate": 9.984101046659353e-06,
+      "loss": 0.1479,
+      "step": 307
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 30.913895702455253,
+      "learning_rate": 9.983869916610232e-06,
+      "loss": 0.1377,
+      "step": 308
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 53.35070194289101,
+      "learning_rate": 9.983637121357534e-06,
+      "loss": 0.1831,
+      "step": 309
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 19.191550606900922,
+      "learning_rate": 9.983402660979042e-06,
+      "loss": 0.1614,
+      "step": 310
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 46.927298296283055,
+      "learning_rate": 9.983166535553093e-06,
+      "loss": 0.1902,
+      "step": 311
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.604114451062268,
+      "learning_rate": 9.98292874515858e-06,
+      "loss": 0.1169,
+      "step": 312
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.658236921460949,
+      "learning_rate": 9.982689289874956e-06,
+      "loss": 0.1569,
+      "step": 313
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.620979122118501,
+      "learning_rate": 9.982448169782226e-06,
+      "loss": 0.1575,
+      "step": 314
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.20134743823668,
+      "learning_rate": 9.98220538496095e-06,
+      "loss": 0.1302,
+      "step": 315
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 21.83470224282846,
+      "learning_rate": 9.98196093549225e-06,
+      "loss": 0.1381,
+      "step": 316
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.019438949112746,
+      "learning_rate": 9.9817148214578e-06,
+      "loss": 0.1637,
+      "step": 317
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 18.531774403521805,
+      "learning_rate": 9.981467042939833e-06,
+      "loss": 0.1531,
+      "step": 318
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.772842664240888,
+      "learning_rate": 9.981217600021133e-06,
+      "loss": 0.1455,
+      "step": 319
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.688710490868121,
+      "learning_rate": 9.980966492785048e-06,
+      "loss": 0.1639,
+      "step": 320
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.259163694801227,
+      "learning_rate": 9.980713721315473e-06,
+      "loss": 0.1166,
+      "step": 321
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.099541659387623,
+      "learning_rate": 9.98045928569687e-06,
+      "loss": 0.1406,
+      "step": 322
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.661009836373966,
+      "learning_rate": 9.98020318601424e-06,
+      "loss": 0.1086,
+      "step": 323
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.038778597302965,
+      "learning_rate": 9.97994542235316e-06,
+      "loss": 0.1447,
+      "step": 324
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 21.75909194513975,
+      "learning_rate": 9.979685994799753e-06,
+      "loss": 0.1561,
+      "step": 325
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 35.89676316531792,
+      "learning_rate": 9.979424903440695e-06,
+      "loss": 0.1526,
+      "step": 326
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 26.47012119449328,
+      "learning_rate": 9.979162148363222e-06,
+      "loss": 0.1331,
+      "step": 327
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 41.47116585357957,
+      "learning_rate": 9.978897729655127e-06,
+      "loss": 0.1527,
+      "step": 328
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 68.50658423771155,
+      "learning_rate": 9.978631647404755e-06,
+      "loss": 0.2024,
+      "step": 329
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.536910352068493,
+      "learning_rate": 9.97836390170101e-06,
+      "loss": 0.1448,
+      "step": 330
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 62.646920155000565,
+      "learning_rate": 9.978094492633353e-06,
+      "loss": 0.1959,
+      "step": 331
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 74.9129334802522,
+      "learning_rate": 9.977823420291796e-06,
+      "loss": 0.2213,
+      "step": 332
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.926863527790253,
+      "learning_rate": 9.97755068476691e-06,
+      "loss": 0.1501,
+      "step": 333
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 29.900030068841634,
+      "learning_rate": 9.977276286149821e-06,
+      "loss": 0.1589,
+      "step": 334
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 64.98822420575073,
+      "learning_rate": 9.977000224532211e-06,
+      "loss": 0.1938,
+      "step": 335
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 43.849118052159554,
+      "learning_rate": 9.976722500006318e-06,
+      "loss": 0.1895,
+      "step": 336
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 16.878076037266382,
+      "learning_rate": 9.976443112664932e-06,
+      "loss": 0.1444,
+      "step": 337
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 58.54099348004542,
+      "learning_rate": 9.976162062601407e-06,
+      "loss": 0.1805,
+      "step": 338
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 37.890609096404795,
+      "learning_rate": 9.97587934990964e-06,
+      "loss": 0.1713,
+      "step": 339
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.701668469320554,
+      "learning_rate": 9.975594974684096e-06,
+      "loss": 0.1388,
+      "step": 340
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.179582917571072,
+      "learning_rate": 9.975308937019787e-06,
+      "loss": 0.1361,
+      "step": 341
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 38.612872861365716,
+      "learning_rate": 9.975021237012286e-06,
+      "loss": 0.1589,
+      "step": 342
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 28.653520911802577,
+      "learning_rate": 9.974731874757717e-06,
+      "loss": 0.1484,
+      "step": 343
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 28.38881703138808,
+      "learning_rate": 9.974440850352762e-06,
+      "loss": 0.1755,
+      "step": 344
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 48.04652377386355,
+      "learning_rate": 9.974148163894658e-06,
+      "loss": 0.1395,
+      "step": 345
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 28.398568336832877,
+      "learning_rate": 9.973853815481196e-06,
+      "loss": 0.1409,
+      "step": 346
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.919991106134674,
+      "learning_rate": 9.973557805210724e-06,
+      "loss": 0.1555,
+      "step": 347
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.837518594770932,
+      "learning_rate": 9.973260133182145e-06,
+      "loss": 0.1455,
+      "step": 348
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 27.31277900154392,
+      "learning_rate": 9.972960799494915e-06,
+      "loss": 0.1361,
+      "step": 349
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 12.083930952779552,
+      "learning_rate": 9.972659804249047e-06,
+      "loss": 0.1295,
+      "step": 350
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.435839166347306,
+      "learning_rate": 9.972357147545113e-06,
+      "loss": 0.1345,
+      "step": 351
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 28.605070082539534,
+      "learning_rate": 9.972052829484231e-06,
+      "loss": 0.1387,
+      "step": 352
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 13.819008184513397,
+      "learning_rate": 9.971746850168084e-06,
+      "loss": 0.1255,
+      "step": 353
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 15.558085688257078,
+      "learning_rate": 9.971439209698902e-06,
+      "loss": 0.1755,
+      "step": 354
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 14.892786074375907,
+      "learning_rate": 9.971129908179474e-06,
+      "loss": 0.1552,
+      "step": 355
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.814580259790407,
+      "learning_rate": 9.970818945713145e-06,
+      "loss": 0.1426,
+      "step": 356
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.246334439430278,
+      "learning_rate": 9.970506322403813e-06,
+      "loss": 0.1237,
+      "step": 357
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 36.7630773678966,
+      "learning_rate": 9.970192038355928e-06,
+      "loss": 0.1527,
+      "step": 358
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 18.698463504899674,
+      "learning_rate": 9.969876093674502e-06,
+      "loss": 0.1565,
+      "step": 359
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 34.419554162278324,
+      "learning_rate": 9.969558488465097e-06,
+      "loss": 0.1506,
+      "step": 360
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 36.93690244776614,
+      "learning_rate": 9.969239222833829e-06,
+      "loss": 0.1531,
+      "step": 361
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 33.49368454947981,
+      "learning_rate": 9.968918296887374e-06,
+      "loss": 0.1509,
+      "step": 362
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 25.28539142873406,
+      "learning_rate": 9.968595710732955e-06,
+      "loss": 0.1499,
+      "step": 363
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 23.43222923122621,
+      "learning_rate": 9.968271464478357e-06,
+      "loss": 0.1312,
+      "step": 364
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 19.51822930107335,
+      "learning_rate": 9.967945558231917e-06,
+      "loss": 0.144,
+      "step": 365
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 41.83722751889746,
+      "learning_rate": 9.967617992102526e-06,
+      "loss": 0.1533,
+      "step": 366
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 30.973390532695422,
+      "learning_rate": 9.967288766199628e-06,
+      "loss": 0.13,
+      "step": 367
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.81797488242334,
+      "learning_rate": 9.966957880633225e-06,
+      "loss": 0.1371,
+      "step": 368
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 18.097159112396604,
+      "learning_rate": 9.966625335513873e-06,
+      "loss": 0.1356,
+      "step": 369
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.8534602258826025,
+      "learning_rate": 9.96629113095268e-06,
+      "loss": 0.1406,
+      "step": 370
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 12.63522722168294,
+      "learning_rate": 9.965955267061309e-06,
+      "loss": 0.1616,
+      "step": 371
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 12.281840014620311,
+      "learning_rate": 9.965617743951982e-06,
+      "loss": 0.1528,
+      "step": 372
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.010291018078059,
+      "learning_rate": 9.965278561737466e-06,
+      "loss": 0.1039,
+      "step": 373
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 23.593676742558646,
+      "learning_rate": 9.964937720531094e-06,
+      "loss": 0.1334,
+      "step": 374
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.379963157519395,
+      "learning_rate": 9.964595220446744e-06,
+      "loss": 0.1658,
+      "step": 375
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.612895660033113,
+      "learning_rate": 9.964251061598853e-06,
+      "loss": 0.1724,
+      "step": 376
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.169065570952082,
+      "learning_rate": 9.96390524410241e-06,
+      "loss": 0.1475,
+      "step": 377
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 24.25698722300466,
+      "learning_rate": 9.96355776807296e-06,
+      "loss": 0.1232,
+      "step": 378
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 17.4894072996316,
+      "learning_rate": 9.9632086336266e-06,
+      "loss": 0.1412,
+      "step": 379
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 14.429277148705365,
+      "learning_rate": 9.962857840879983e-06,
+      "loss": 0.1322,
+      "step": 380
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 21.967245963743874,
+      "learning_rate": 9.962505389950317e-06,
+      "loss": 0.1565,
+      "step": 381
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 13.836611609961958,
+      "learning_rate": 9.962151280955359e-06,
+      "loss": 0.1473,
+      "step": 382
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.777348182737883,
+      "learning_rate": 9.961795514013424e-06,
+      "loss": 0.1934,
+      "step": 383
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 14.417467203134215,
+      "learning_rate": 9.961438089243384e-06,
+      "loss": 0.1414,
+      "step": 384
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.704688741271175,
+      "learning_rate": 9.961079006764659e-06,
+      "loss": 0.198,
+      "step": 385
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.379293993035889,
+      "learning_rate": 9.960718266697223e-06,
+      "loss": 0.115,
+      "step": 386
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 13.477558643143759,
+      "learning_rate": 9.960355869161609e-06,
+      "loss": 0.1284,
+      "step": 387
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.698091583062664,
+      "learning_rate": 9.959991814278898e-06,
+      "loss": 0.1287,
+      "step": 388
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 22.68559034944762,
+      "learning_rate": 9.95962610217073e-06,
+      "loss": 0.1202,
+      "step": 389
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.617129910996631,
+      "learning_rate": 9.959258732959296e-06,
+      "loss": 0.1614,
+      "step": 390
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 22.012768496639392,
+      "learning_rate": 9.958889706767341e-06,
+      "loss": 0.1481,
+      "step": 391
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 42.213262359287036,
+      "learning_rate": 9.95851902371816e-06,
+      "loss": 0.1747,
+      "step": 392
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.688717536140965,
+      "learning_rate": 9.95814668393561e-06,
+      "loss": 0.1383,
+      "step": 393
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 41.26203920787363,
+      "learning_rate": 9.957772687544094e-06,
+      "loss": 0.1768,
+      "step": 394
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.068392998011415,
+      "learning_rate": 9.95739703466857e-06,
+      "loss": 0.136,
+      "step": 395
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.825594261580451,
+      "learning_rate": 9.957019725434554e-06,
+      "loss": 0.1346,
+      "step": 396
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 14.922584492469332,
+      "learning_rate": 9.956640759968111e-06,
+      "loss": 0.1091,
+      "step": 397
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 15.362761377543631,
+      "learning_rate": 9.956260138395857e-06,
+      "loss": 0.1241,
+      "step": 398
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 14.478039428734272,
+      "learning_rate": 9.955877860844969e-06,
+      "loss": 0.1665,
+      "step": 399
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 15.342154188172143,
+      "learning_rate": 9.955493927443171e-06,
+      "loss": 0.1294,
+      "step": 400
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 16.72367814770503,
+      "learning_rate": 9.955108338318743e-06,
+      "loss": 0.1521,
+      "step": 401
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 12.948093169878645,
+      "learning_rate": 9.954721093600517e-06,
+      "loss": 0.1439,
+      "step": 402
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.973524690035784,
+      "learning_rate": 9.95433219341788e-06,
+      "loss": 0.1324,
+      "step": 403
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 15.47228950899974,
+      "learning_rate": 9.953941637900769e-06,
+      "loss": 0.1686,
+      "step": 404
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.160262840433568,
+      "learning_rate": 9.953549427179676e-06,
+      "loss": 0.1477,
+      "step": 405
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 25.931016059983655,
+      "learning_rate": 9.953155561385646e-06,
+      "loss": 0.1356,
+      "step": 406
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.62503524150755,
+      "learning_rate": 9.952760040650278e-06,
+      "loss": 0.1508,
+      "step": 407
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.370817286061716,
+      "learning_rate": 9.95236286510572e-06,
+      "loss": 0.1111,
+      "step": 408
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 17.969755002141245,
+      "learning_rate": 9.95196403488468e-06,
+      "loss": 0.1555,
+      "step": 409
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 15.602862431905702,
+      "learning_rate": 9.951563550120412e-06,
+      "loss": 0.1444,
+      "step": 410
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 24.892905576485063,
+      "learning_rate": 9.951161410946725e-06,
+      "loss": 0.142,
+      "step": 411
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 26.823371334885348,
+      "learning_rate": 9.950757617497983e-06,
+      "loss": 0.1376,
+      "step": 412
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 42.11109194980369,
+      "learning_rate": 9.950352169909101e-06,
+      "loss": 0.1213,
+      "step": 413
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.958217327562295,
+      "learning_rate": 9.949945068315544e-06,
+      "loss": 0.1626,
+      "step": 414
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 17.71107942350547,
+      "learning_rate": 9.949536312853334e-06,
+      "loss": 0.166,
+      "step": 415
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 33.36805721946892,
+      "learning_rate": 9.949125903659042e-06,
+      "loss": 0.1525,
+      "step": 416
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.084727395074816,
+      "learning_rate": 9.948713840869797e-06,
+      "loss": 0.1426,
+      "step": 417
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 29.756453469589164,
+      "learning_rate": 9.948300124623274e-06,
+      "loss": 0.1035,
+      "step": 418
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.824683629550345,
+      "learning_rate": 9.947884755057703e-06,
+      "loss": 0.14,
+      "step": 419
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.852603046561496,
+      "learning_rate": 9.947467732311868e-06,
+      "loss": 0.1642,
+      "step": 420
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 14.091641894723889,
+      "learning_rate": 9.947049056525104e-06,
+      "loss": 0.1337,
+      "step": 421
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.772867686006098,
+      "learning_rate": 9.9466287278373e-06,
+      "loss": 0.1405,
+      "step": 422
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 20.227967360512032,
+      "learning_rate": 9.946206746388892e-06,
+      "loss": 0.1464,
+      "step": 423
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 28.120732815939572,
+      "learning_rate": 9.94578311232087e-06,
+      "loss": 0.1587,
+      "step": 424
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.836326247915467,
+      "learning_rate": 9.945357825774786e-06,
+      "loss": 0.1632,
+      "step": 425
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 38.068508076771145,
+      "learning_rate": 9.944930886892731e-06,
+      "loss": 0.1635,
+      "step": 426
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 17.53039371409448,
+      "learning_rate": 9.944502295817353e-06,
+      "loss": 0.1393,
+      "step": 427
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.744624734916593,
+      "learning_rate": 9.944072052691853e-06,
+      "loss": 0.1299,
+      "step": 428
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 36.21266814484719,
+      "learning_rate": 9.943640157659984e-06,
+      "loss": 0.1241,
+      "step": 429
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 27.99729719398013,
+      "learning_rate": 9.94320661086605e-06,
+      "loss": 0.1172,
+      "step": 430
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 32.61398458459067,
+      "learning_rate": 9.942771412454906e-06,
+      "loss": 0.1688,
+      "step": 431
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 29.53139637962103,
+      "learning_rate": 9.942334562571961e-06,
+      "loss": 0.14,
+      "step": 432
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 70.15413954052487,
+      "learning_rate": 9.941896061363173e-06,
+      "loss": 0.1909,
+      "step": 433
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 41.57928064536469,
+      "learning_rate": 9.941455908975054e-06,
+      "loss": 0.1348,
+      "step": 434
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 38.37885664713142,
+      "learning_rate": 9.941014105554668e-06,
+      "loss": 0.1616,
+      "step": 435
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 49.537343514648235,
+      "learning_rate": 9.94057065124963e-06,
+      "loss": 0.1626,
+      "step": 436
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 54.24518982500366,
+      "learning_rate": 9.940125546208107e-06,
+      "loss": 0.1528,
+      "step": 437
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.439592562127047,
+      "learning_rate": 9.939678790578813e-06,
+      "loss": 0.1382,
+      "step": 438
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 50.401662662406025,
+      "learning_rate": 9.93923038451102e-06,
+      "loss": 0.1444,
+      "step": 439
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 27.565998455504264,
+      "learning_rate": 9.938780328154549e-06,
+      "loss": 0.1638,
+      "step": 440
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 29.80949294384711,
+      "learning_rate": 9.938328621659775e-06,
+      "loss": 0.177,
+      "step": 441
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 39.42630244801424,
+      "learning_rate": 9.937875265177615e-06,
+      "loss": 0.1831,
+      "step": 442
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.22622431582067,
+      "learning_rate": 9.937420258859547e-06,
+      "loss": 0.1394,
+      "step": 443
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 32.328178133199046,
+      "learning_rate": 9.9369636028576e-06,
+      "loss": 0.1349,
+      "step": 444
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.504231110559711,
+      "learning_rate": 9.936505297324346e-06,
+      "loss": 0.1211,
+      "step": 445
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 20.781251334768122,
+      "learning_rate": 9.936045342412917e-06,
+      "loss": 0.1482,
+      "step": 446
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 21.21248632778865,
+      "learning_rate": 9.93558373827699e-06,
+      "loss": 0.1327,
+      "step": 447
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 12.707782785702369,
+      "learning_rate": 9.935120485070799e-06,
+      "loss": 0.1494,
+      "step": 448
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.673210776385321,
+      "learning_rate": 9.934655582949123e-06,
+      "loss": 0.0964,
+      "step": 449
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.888011994441523,
+      "learning_rate": 9.934189032067296e-06,
+      "loss": 0.1555,
+      "step": 450
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 32.12516155863512,
+      "learning_rate": 9.933720832581197e-06,
+      "loss": 0.1355,
+      "step": 451
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 22.322386591262855,
+      "learning_rate": 9.933250984647266e-06,
+      "loss": 0.1368,
+      "step": 452
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 15.223180132034958,
+      "learning_rate": 9.932779488422484e-06,
+      "loss": 0.1383,
+      "step": 453
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 52.31779619507051,
+      "learning_rate": 9.93230634406439e-06,
+      "loss": 0.1522,
+      "step": 454
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 29.017829224877428,
+      "learning_rate": 9.931831551731067e-06,
+      "loss": 0.1445,
+      "step": 455
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 18.997292779869753,
+      "learning_rate": 9.931355111581154e-06,
+      "loss": 0.1554,
+      "step": 456
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 34.735932438290625,
+      "learning_rate": 9.930877023773837e-06,
+      "loss": 0.134,
+      "step": 457
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 45.54908316077474,
+      "learning_rate": 9.930397288468853e-06,
+      "loss": 0.1663,
+      "step": 458
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.033708570010118,
+      "learning_rate": 9.929915905826494e-06,
+      "loss": 0.1084,
+      "step": 459
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.323441132876292,
+      "learning_rate": 9.9294328760076e-06,
+      "loss": 0.1165,
+      "step": 460
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 47.919864734101694,
+      "learning_rate": 9.928948199173552e-06,
+      "loss": 0.1626,
+      "step": 461
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.461706134664676,
+      "learning_rate": 9.928461875486297e-06,
+      "loss": 0.1068,
+      "step": 462
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 15.03570267608453,
+      "learning_rate": 9.927973905108323e-06,
+      "loss": 0.1066,
+      "step": 463
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 29.02078498555869,
+      "learning_rate": 9.927484288202671e-06,
+      "loss": 0.1425,
+      "step": 464
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 25.201159295770108,
+      "learning_rate": 9.926993024932929e-06,
+      "loss": 0.1377,
+      "step": 465
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.395550286499605,
+      "learning_rate": 9.926500115463238e-06,
+      "loss": 0.1176,
+      "step": 466
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.158696809668983,
+      "learning_rate": 9.926005559958287e-06,
+      "loss": 0.1361,
+      "step": 467
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 25.515101194445457,
+      "learning_rate": 9.925509358583319e-06,
+      "loss": 0.1162,
+      "step": 468
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 18.28360522188909,
+      "learning_rate": 9.92501151150412e-06,
+      "loss": 0.1367,
+      "step": 469
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 31.778596429481162,
+      "learning_rate": 9.924512018887036e-06,
+      "loss": 0.1229,
+      "step": 470
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 13.711321378139074,
+      "learning_rate": 9.924010880898952e-06,
+      "loss": 0.1389,
+      "step": 471
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.3412870602780185,
+      "learning_rate": 9.923508097707306e-06,
+      "loss": 0.1394,
+      "step": 472
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 39.662327865375396,
+      "learning_rate": 9.923003669480094e-06,
+      "loss": 0.187,
+      "step": 473
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.585151391933716,
+      "learning_rate": 9.922497596385848e-06,
+      "loss": 0.1266,
+      "step": 474
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 25.903061542242025,
+      "learning_rate": 9.92198987859366e-06,
+      "loss": 0.0936,
+      "step": 475
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.982867596376823,
+      "learning_rate": 9.921480516273168e-06,
+      "loss": 0.1238,
+      "step": 476
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 13.564020531973936,
+      "learning_rate": 9.920969509594558e-06,
+      "loss": 0.126,
+      "step": 477
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.136096874644808,
+      "learning_rate": 9.920456858728567e-06,
+      "loss": 0.1329,
+      "step": 478
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.415457519145033,
+      "learning_rate": 9.919942563846482e-06,
+      "loss": 0.1044,
+      "step": 479
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.035743042678794,
+      "learning_rate": 9.919426625120137e-06,
+      "loss": 0.1689,
+      "step": 480
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.45092970947201,
+      "learning_rate": 9.918909042721918e-06,
+      "loss": 0.136,
+      "step": 481
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.68056597961224,
+      "learning_rate": 9.918389816824759e-06,
+      "loss": 0.1423,
+      "step": 482
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 20.55146614949313,
+      "learning_rate": 9.917868947602144e-06,
+      "loss": 0.1532,
+      "step": 483
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 34.77111147030742,
+      "learning_rate": 9.917346435228102e-06,
+      "loss": 0.1746,
+      "step": 484
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.0196084233144695,
+      "learning_rate": 9.916822279877217e-06,
+      "loss": 0.1279,
+      "step": 485
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.941526827167951,
+      "learning_rate": 9.91629648172462e-06,
+      "loss": 0.1146,
+      "step": 486
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.890997507039701,
+      "learning_rate": 9.915769040945984e-06,
+      "loss": 0.1028,
+      "step": 487
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 27.88496903893723,
+      "learning_rate": 9.915239957717542e-06,
+      "loss": 0.1274,
+      "step": 488
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.00778567508186,
+      "learning_rate": 9.91470923221607e-06,
+      "loss": 0.1411,
+      "step": 489
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 23.484883782999354,
+      "learning_rate": 9.914176864618891e-06,
+      "loss": 0.1384,
+      "step": 490
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.929422863937659,
+      "learning_rate": 9.913642855103881e-06,
+      "loss": 0.0734,
+      "step": 491
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.36947027897346,
+      "learning_rate": 9.913107203849464e-06,
+      "loss": 0.1497,
+      "step": 492
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 24.750798415291204,
+      "learning_rate": 9.912569911034607e-06,
+      "loss": 0.1672,
+      "step": 493
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.881614535806723,
+      "learning_rate": 9.912030976838832e-06,
+      "loss": 0.1456,
+      "step": 494
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 33.54932127569477,
+      "learning_rate": 9.911490401442205e-06,
+      "loss": 0.1489,
+      "step": 495
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 20.00784172694422,
+      "learning_rate": 9.910948185025345e-06,
+      "loss": 0.1536,
+      "step": 496
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.238510328306569,
+      "learning_rate": 9.910404327769414e-06,
+      "loss": 0.1207,
+      "step": 497
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 30.054761593942306,
+      "learning_rate": 9.909858829856127e-06,
+      "loss": 0.1364,
+      "step": 498
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.592134090075237,
+      "learning_rate": 9.909311691467744e-06,
+      "loss": 0.1006,
+      "step": 499
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.936467444451623,
+      "learning_rate": 9.908762912787073e-06,
+      "loss": 0.1512,
+      "step": 500
+    },
+    {
+      "epoch": 0.36,
+      "eval_avg_AUC": 0.8148136168882592,
+      "eval_avg_Accuracy": 0.7284897214854111,
+      "eval_avg_Accuracy-right": 0.8730272596843616,
+      "eval_avg_Accuracy-wrong": 0.4764612235615192,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6962655285457768,
+      "eval_last_AUC": 0.8285519781961206,
+      "eval_last_Accuracy": 0.7570457559681698,
+      "eval_last_Accuracy-right": 0.8302465110212599,
+      "eval_last_Accuracy-wrong": 0.6294064134637253,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.6999852156206944,
+      "eval_max_AUC": 0.7665118487150447,
+      "eval_max_Accuracy": 0.6521054376657824,
+      "eval_max_Accuracy-right": 0.9773053345506717,
+      "eval_max_Accuracy-wrong": 0.08505799408687742,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.6174539818202015,
+      "eval_min_AUC": 0.8243709198076318,
+      "eval_min_Accuracy": 0.7511604774535809,
+      "eval_min_Accuracy-right": 0.7600104343289422,
+      "eval_min_Accuracy-wrong": 0.735728906072322,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6964298045694274,
+      "eval_prod_AUC": 0.8209843050796547,
+      "eval_prod_Accuracy": 0.6548408488063661,
+      "eval_prod_Accuracy-right": 0.49582626842311206,
+      "eval_prod_Accuracy-wrong": 0.9321128041846714,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.6870361522901595,
+      "eval_runtime": 248.2297,
+      "eval_samples_per_second": 97.2,
+      "eval_steps_per_second": 3.038,
+      "eval_sum_AUC": 0.6857568992387502,
+      "eval_sum_Accuracy": 0.6382211538461539,
+      "eval_sum_Accuracy-right": 0.9962827703143342,
+      "eval_sum_Accuracy-wrong": 0.013873095292244713,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6648248225648956,
+      "step": 500
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.521136475767055,
+      "learning_rate": 9.908212493997473e-06,
+      "loss": 0.1351,
+      "step": 501
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 24.8215182959066,
+      "learning_rate": 9.90766043528285e-06,
+      "loss": 0.1265,
+      "step": 502
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 20.262220801212766,
+      "learning_rate": 9.907106736827654e-06,
+      "loss": 0.1382,
+      "step": 503
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.612067632756019,
+      "learning_rate": 9.906551398816886e-06,
+      "loss": 0.1176,
+      "step": 504
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.684225952657133,
+      "learning_rate": 9.9059944214361e-06,
+      "loss": 0.1636,
+      "step": 505
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.984066186445835,
+      "learning_rate": 9.905435804871387e-06,
+      "loss": 0.1366,
+      "step": 506
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 24.763096461281147,
+      "learning_rate": 9.904875549309391e-06,
+      "loss": 0.1584,
+      "step": 507
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.7311046700749975,
+      "learning_rate": 9.904313654937308e-06,
+      "loss": 0.1486,
+      "step": 508
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 51.33110066179473,
+      "learning_rate": 9.903750121942873e-06,
+      "loss": 0.1875,
+      "step": 509
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.408291940217502,
+      "learning_rate": 9.903184950514378e-06,
+      "loss": 0.1373,
+      "step": 510
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.255602958809023,
+      "learning_rate": 9.90261814084065e-06,
+      "loss": 0.1256,
+      "step": 511
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 17.357961967373914,
+      "learning_rate": 9.902049693111077e-06,
+      "loss": 0.1616,
+      "step": 512
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 28.352859541701022,
+      "learning_rate": 9.901479607515587e-06,
+      "loss": 0.1322,
+      "step": 513
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 18.349020721142626,
+      "learning_rate": 9.900907884244654e-06,
+      "loss": 0.1311,
+      "step": 514
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.109150918544185,
+      "learning_rate": 9.900334523489303e-06,
+      "loss": 0.1604,
+      "step": 515
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 29.842240993629268,
+      "learning_rate": 9.899759525441101e-06,
+      "loss": 0.1586,
+      "step": 516
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 32.448571178210294,
+      "learning_rate": 9.899182890292171e-06,
+      "loss": 0.1516,
+      "step": 517
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 37.541295217424135,
+      "learning_rate": 9.898604618235175e-06,
+      "loss": 0.1541,
+      "step": 518
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 59.74203765524549,
+      "learning_rate": 9.898024709463322e-06,
+      "loss": 0.1914,
+      "step": 519
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 51.553681345435216,
+      "learning_rate": 9.897443164170375e-06,
+      "loss": 0.1547,
+      "step": 520
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 17.65120220241073,
+      "learning_rate": 9.896859982550636e-06,
+      "loss": 0.1357,
+      "step": 521
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.5369157168625405,
+      "learning_rate": 9.89627516479896e-06,
+      "loss": 0.1232,
+      "step": 522
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 28.586892727146566,
+      "learning_rate": 9.895688711110739e-06,
+      "loss": 0.1234,
+      "step": 523
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.714799632443614,
+      "learning_rate": 9.895100621681923e-06,
+      "loss": 0.1794,
+      "step": 524
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.894818046361022,
+      "learning_rate": 9.894510896709003e-06,
+      "loss": 0.1145,
+      "step": 525
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.020267897450193,
+      "learning_rate": 9.893919536389017e-06,
+      "loss": 0.1331,
+      "step": 526
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.603140469816235,
+      "learning_rate": 9.89332654091955e-06,
+      "loss": 0.1317,
+      "step": 527
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.317710681474739,
+      "learning_rate": 9.892731910498731e-06,
+      "loss": 0.134,
+      "step": 528
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 18.48919116376951,
+      "learning_rate": 9.892135645325238e-06,
+      "loss": 0.127,
+      "step": 529
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 15.110852007766283,
+      "learning_rate": 9.891537745598293e-06,
+      "loss": 0.1333,
+      "step": 530
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 25.48093356108608,
+      "learning_rate": 9.89093821151767e-06,
+      "loss": 0.1425,
+      "step": 531
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.211704092158119,
+      "learning_rate": 9.89033704328368e-06,
+      "loss": 0.1261,
+      "step": 532
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.49842219441739,
+      "learning_rate": 9.889734241097186e-06,
+      "loss": 0.1227,
+      "step": 533
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.538899092333007,
+      "learning_rate": 9.889129805159595e-06,
+      "loss": 0.1333,
+      "step": 534
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.229548690859028,
+      "learning_rate": 9.888523735672861e-06,
+      "loss": 0.1207,
+      "step": 535
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.706186292987425,
+      "learning_rate": 9.887916032839482e-06,
+      "loss": 0.1376,
+      "step": 536
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 15.889794280637537,
+      "learning_rate": 9.887306696862504e-06,
+      "loss": 0.1122,
+      "step": 537
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.453535010562696,
+      "learning_rate": 9.886695727945515e-06,
+      "loss": 0.131,
+      "step": 538
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.612128252341902,
+      "learning_rate": 9.886083126292655e-06,
+      "loss": 0.1492,
+      "step": 539
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.075171549803063,
+      "learning_rate": 9.885468892108603e-06,
+      "loss": 0.1272,
+      "step": 540
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 26.914960079687482,
+      "learning_rate": 9.884853025598587e-06,
+      "loss": 0.1514,
+      "step": 541
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 43.779863637748186,
+      "learning_rate": 9.884235526968377e-06,
+      "loss": 0.1256,
+      "step": 542
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 37.73960304494291,
+      "learning_rate": 9.883616396424294e-06,
+      "loss": 0.1553,
+      "step": 543
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 12.454022843924934,
+      "learning_rate": 9.8829956341732e-06,
+      "loss": 0.1204,
+      "step": 544
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 43.19321310165527,
+      "learning_rate": 9.882373240422503e-06,
+      "loss": 0.1389,
+      "step": 545
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 48.79232289830731,
+      "learning_rate": 9.881749215380156e-06,
+      "loss": 0.13,
+      "step": 546
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 18.76530769424551,
+      "learning_rate": 9.881123559254658e-06,
+      "loss": 0.1158,
+      "step": 547
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 62.57220924147817,
+      "learning_rate": 9.880496272255053e-06,
+      "loss": 0.158,
+      "step": 548
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 12.500417488061052,
+      "learning_rate": 9.879867354590926e-06,
+      "loss": 0.1196,
+      "step": 549
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 15.875141788148147,
+      "learning_rate": 9.879236806472414e-06,
+      "loss": 0.1333,
+      "step": 550
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.301629680806376,
+      "learning_rate": 9.878604628110194e-06,
+      "loss": 0.1799,
+      "step": 551
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.264465519965361,
+      "learning_rate": 9.877970819715485e-06,
+      "loss": 0.1427,
+      "step": 552
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 30.949313988770307,
+      "learning_rate": 9.87733538150006e-06,
+      "loss": 0.142,
+      "step": 553
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.380549224768007,
+      "learning_rate": 9.876698313676225e-06,
+      "loss": 0.1191,
+      "step": 554
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.66859233407463,
+      "learning_rate": 9.876059616456842e-06,
+      "loss": 0.1405,
+      "step": 555
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 20.946002565023058,
+      "learning_rate": 9.875419290055305e-06,
+      "loss": 0.1102,
+      "step": 556
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 30.897928782214787,
+      "learning_rate": 9.874777334685565e-06,
+      "loss": 0.1711,
+      "step": 557
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 13.033106438202273,
+      "learning_rate": 9.874133750562108e-06,
+      "loss": 0.1622,
+      "step": 558
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 40.36685994127496,
+      "learning_rate": 9.873488537899967e-06,
+      "loss": 0.2061,
+      "step": 559
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 19.50274995202192,
+      "learning_rate": 9.872841696914721e-06,
+      "loss": 0.1294,
+      "step": 560
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.932353294948314,
+      "learning_rate": 9.872193227822492e-06,
+      "loss": 0.1265,
+      "step": 561
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 20.43294471193903,
+      "learning_rate": 9.871543130839944e-06,
+      "loss": 0.1456,
+      "step": 562
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 42.946806904632204,
+      "learning_rate": 9.870891406184288e-06,
+      "loss": 0.1626,
+      "step": 563
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 33.149765129877956,
+      "learning_rate": 9.870238054073275e-06,
+      "loss": 0.1593,
+      "step": 564
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.929277127548148,
+      "learning_rate": 9.869583074725206e-06,
+      "loss": 0.1477,
+      "step": 565
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 39.416395744901585,
+      "learning_rate": 9.868926468358919e-06,
+      "loss": 0.1432,
+      "step": 566
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 43.22729095180558,
+      "learning_rate": 9.868268235193796e-06,
+      "loss": 0.1406,
+      "step": 567
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 15.377542148018398,
+      "learning_rate": 9.867608375449772e-06,
+      "loss": 0.1235,
+      "step": 568
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 23.10191212472918,
+      "learning_rate": 9.866946889347311e-06,
+      "loss": 0.127,
+      "step": 569
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 32.3797801287966,
+      "learning_rate": 9.866283777107432e-06,
+      "loss": 0.1323,
+      "step": 570
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.095686964503419,
+      "learning_rate": 9.865619038951692e-06,
+      "loss": 0.1375,
+      "step": 571
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.775807625906909,
+      "learning_rate": 9.864952675102193e-06,
+      "loss": 0.1379,
+      "step": 572
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.286767820247743,
+      "learning_rate": 9.864284685781578e-06,
+      "loss": 0.1425,
+      "step": 573
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.147344431329742,
+      "learning_rate": 9.863615071213036e-06,
+      "loss": 0.1304,
+      "step": 574
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 28.412228671248563,
+      "learning_rate": 9.862943831620298e-06,
+      "loss": 0.1273,
+      "step": 575
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 30.96935582463328,
+      "learning_rate": 9.862270967227636e-06,
+      "loss": 0.1459,
+      "step": 576
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.987197903594031,
+      "learning_rate": 9.861596478259869e-06,
+      "loss": 0.139,
+      "step": 577
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 63.33976410883062,
+      "learning_rate": 9.860920364942353e-06,
+      "loss": 0.1904,
+      "step": 578
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.66543675696079,
+      "learning_rate": 9.860242627500994e-06,
+      "loss": 0.1125,
+      "step": 579
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.0806252525076,
+      "learning_rate": 9.859563266162231e-06,
+      "loss": 0.1321,
+      "step": 580
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.080418185339296,
+      "learning_rate": 9.858882281153058e-06,
+      "loss": 0.1157,
+      "step": 581
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 23.03682096474836,
+      "learning_rate": 9.858199672701e-06,
+      "loss": 0.1392,
+      "step": 582
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 31.28197071942006,
+      "learning_rate": 9.85751544103413e-06,
+      "loss": 0.1129,
+      "step": 583
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.68760436913685,
+      "learning_rate": 9.856829586381065e-06,
+      "loss": 0.1071,
+      "step": 584
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 73.28546422256532,
+      "learning_rate": 9.856142108970958e-06,
+      "loss": 0.1958,
+      "step": 585
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 18.720122616928183,
+      "learning_rate": 9.855453009033512e-06,
+      "loss": 0.1326,
+      "step": 586
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.463997560094482,
+      "learning_rate": 9.854762286798965e-06,
+      "loss": 0.1453,
+      "step": 587
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.391237164066708,
+      "learning_rate": 9.854069942498102e-06,
+      "loss": 0.1965,
+      "step": 588
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.599884051273923,
+      "learning_rate": 9.853375976362245e-06,
+      "loss": 0.1635,
+      "step": 589
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.441491117087743,
+      "learning_rate": 9.852680388623266e-06,
+      "loss": 0.1158,
+      "step": 590
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 34.51436948860912,
+      "learning_rate": 9.85198317951357e-06,
+      "loss": 0.1531,
+      "step": 591
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 17.99454842623697,
+      "learning_rate": 9.851284349266107e-06,
+      "loss": 0.1305,
+      "step": 592
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 16.123249191913267,
+      "learning_rate": 9.850583898114372e-06,
+      "loss": 0.1575,
+      "step": 593
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 48.69893321264437,
+      "learning_rate": 9.849881826292399e-06,
+      "loss": 0.1558,
+      "step": 594
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 34.46486038497999,
+      "learning_rate": 9.84917813403476e-06,
+      "loss": 0.1367,
+      "step": 595
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.597468969380602,
+      "learning_rate": 9.848472821576572e-06,
+      "loss": 0.1439,
+      "step": 596
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 52.9719490703774,
+      "learning_rate": 9.847765889153497e-06,
+      "loss": 0.1819,
+      "step": 597
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 36.233987937356446,
+      "learning_rate": 9.847057337001731e-06,
+      "loss": 0.1594,
+      "step": 598
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 20.295048431141648,
+      "learning_rate": 9.846347165358014e-06,
+      "loss": 0.1284,
+      "step": 599
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 42.84524320177862,
+      "learning_rate": 9.84563537445963e-06,
+      "loss": 0.1504,
+      "step": 600
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 62.1195121067531,
+      "learning_rate": 9.844921964544398e-06,
+      "loss": 0.1758,
+      "step": 601
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 47.2653880369296,
+      "learning_rate": 9.844206935850687e-06,
+      "loss": 0.1831,
+      "step": 602
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 29.4931297961525,
+      "learning_rate": 9.843490288617397e-06,
+      "loss": 0.1008,
+      "step": 603
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 78.26510483270128,
+      "learning_rate": 9.842772023083972e-06,
+      "loss": 0.1829,
+      "step": 604
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 72.53305356196587,
+      "learning_rate": 9.842052139490403e-06,
+      "loss": 0.2009,
+      "step": 605
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 23.74689339670262,
+      "learning_rate": 9.841330638077213e-06,
+      "loss": 0.139,
+      "step": 606
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 58.871630884771356,
+      "learning_rate": 9.840607519085467e-06,
+      "loss": 0.1533,
+      "step": 607
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 60.04727742225944,
+      "learning_rate": 9.839882782756778e-06,
+      "loss": 0.1533,
+      "step": 608
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 34.76813691419715,
+      "learning_rate": 9.839156429333291e-06,
+      "loss": 0.1517,
+      "step": 609
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 39.487093806576866,
+      "learning_rate": 9.838428459057694e-06,
+      "loss": 0.1306,
+      "step": 610
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 71.43206314131315,
+      "learning_rate": 9.837698872173214e-06,
+      "loss": 0.166,
+      "step": 611
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 49.09265681558195,
+      "learning_rate": 9.836967668923623e-06,
+      "loss": 0.1689,
+      "step": 612
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.462674294507499,
+      "learning_rate": 9.836234849553228e-06,
+      "loss": 0.1088,
+      "step": 613
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 70.19656687365818,
+      "learning_rate": 9.835500414306875e-06,
+      "loss": 0.2119,
+      "step": 614
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 62.173880695304916,
+      "learning_rate": 9.834764363429956e-06,
+      "loss": 0.1681,
+      "step": 615
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 43.48786940185642,
+      "learning_rate": 9.8340266971684e-06,
+      "loss": 0.172,
+      "step": 616
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 26.124867896320307,
+      "learning_rate": 9.83328741576867e-06,
+      "loss": 0.1342,
+      "step": 617
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 47.911183471370855,
+      "learning_rate": 9.832546519477778e-06,
+      "loss": 0.179,
+      "step": 618
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 89.25182761050898,
+      "learning_rate": 9.831804008543271e-06,
+      "loss": 0.2285,
+      "step": 619
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 23.042941448066955,
+      "learning_rate": 9.831059883213234e-06,
+      "loss": 0.1616,
+      "step": 620
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 58.09372311927322,
+      "learning_rate": 9.830314143736292e-06,
+      "loss": 0.1641,
+      "step": 621
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 28.97753294613849,
+      "learning_rate": 9.829566790361615e-06,
+      "loss": 0.1344,
+      "step": 622
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 49.28980917332516,
+      "learning_rate": 9.828817823338903e-06,
+      "loss": 0.1614,
+      "step": 623
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 32.04413626104762,
+      "learning_rate": 9.828067242918402e-06,
+      "loss": 0.1404,
+      "step": 624
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.78476590401764,
+      "learning_rate": 9.827315049350895e-06,
+      "loss": 0.1169,
+      "step": 625
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 81.63675673285175,
+      "learning_rate": 9.826561242887704e-06,
+      "loss": 0.1953,
+      "step": 626
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 48.11606730737452,
+      "learning_rate": 9.825805823780687e-06,
+      "loss": 0.1619,
+      "step": 627
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.907109159794875,
+      "learning_rate": 9.825048792282247e-06,
+      "loss": 0.131,
+      "step": 628
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.103553566829753,
+      "learning_rate": 9.824290148645322e-06,
+      "loss": 0.1139,
+      "step": 629
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 51.12278081641119,
+      "learning_rate": 9.823529893123384e-06,
+      "loss": 0.1527,
+      "step": 630
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 64.29064695958792,
+      "learning_rate": 9.822768025970456e-06,
+      "loss": 0.1838,
+      "step": 631
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 22.424660906876873,
+      "learning_rate": 9.822004547441088e-06,
+      "loss": 0.1398,
+      "step": 632
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.706055561297325,
+      "learning_rate": 9.821239457790373e-06,
+      "loss": 0.1428,
+      "step": 633
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 59.7429449871699,
+      "learning_rate": 9.82047275727394e-06,
+      "loss": 0.176,
+      "step": 634
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 67.77135894876831,
+      "learning_rate": 9.81970444614796e-06,
+      "loss": 0.1714,
+      "step": 635
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.668865071226387,
+      "learning_rate": 9.81893452466914e-06,
+      "loss": 0.1267,
+      "step": 636
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 44.76001772659544,
+      "learning_rate": 9.818162993094724e-06,
+      "loss": 0.1423,
+      "step": 637
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 53.39779453147049,
+      "learning_rate": 9.817389851682494e-06,
+      "loss": 0.1842,
+      "step": 638
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 24.28617527330543,
+      "learning_rate": 9.816615100690773e-06,
+      "loss": 0.1235,
+      "step": 639
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 19.154914114241624,
+      "learning_rate": 9.81583874037842e-06,
+      "loss": 0.1217,
+      "step": 640
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.613120002801404,
+      "learning_rate": 9.815060771004831e-06,
+      "loss": 0.1311,
+      "step": 641
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 27.240227333671324,
+      "learning_rate": 9.81428119282994e-06,
+      "loss": 0.1553,
+      "step": 642
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 21.0473186526304,
+      "learning_rate": 9.813500006114216e-06,
+      "loss": 0.1239,
+      "step": 643
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.104115765634024,
+      "learning_rate": 9.812717211118673e-06,
+      "loss": 0.1184,
+      "step": 644
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.249235905236901,
+      "learning_rate": 9.811932808104852e-06,
+      "loss": 0.1505,
+      "step": 645
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.4945638837559105,
+      "learning_rate": 9.811146797334838e-06,
+      "loss": 0.1547,
+      "step": 646
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.362754513773162,
+      "learning_rate": 9.810359179071255e-06,
+      "loss": 0.1442,
+      "step": 647
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 19.255632442668595,
+      "learning_rate": 9.809569953577258e-06,
+      "loss": 0.0961,
+      "step": 648
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.391685162385883,
+      "learning_rate": 9.808779121116542e-06,
+      "loss": 0.1421,
+      "step": 649
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.810940924831158,
+      "learning_rate": 9.807986681953341e-06,
+      "loss": 0.1198,
+      "step": 650
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.376506844728723,
+      "learning_rate": 9.807192636352422e-06,
+      "loss": 0.1383,
+      "step": 651
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 17.64775555566848,
+      "learning_rate": 9.80639698457909e-06,
+      "loss": 0.108,
+      "step": 652
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 19.559096284384445,
+      "learning_rate": 9.805599726899188e-06,
+      "loss": 0.1316,
+      "step": 653
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.461717264545886,
+      "learning_rate": 9.804800863579094e-06,
+      "loss": 0.1169,
+      "step": 654
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 20.446164874079532,
+      "learning_rate": 9.804000394885723e-06,
+      "loss": 0.13,
+      "step": 655
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 51.40477529762973,
+      "learning_rate": 9.803198321086527e-06,
+      "loss": 0.2056,
+      "step": 656
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.068665322892394,
+      "learning_rate": 9.802394642449494e-06,
+      "loss": 0.1346,
+      "step": 657
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.538054958638275,
+      "learning_rate": 9.801589359243147e-06,
+      "loss": 0.1362,
+      "step": 658
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 35.66887445030199,
+      "learning_rate": 9.800782471736547e-06,
+      "loss": 0.135,
+      "step": 659
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 31.297567953445206,
+      "learning_rate": 9.799973980199288e-06,
+      "loss": 0.1344,
+      "step": 660
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 22.996584888633382,
+      "learning_rate": 9.799163884901506e-06,
+      "loss": 0.1359,
+      "step": 661
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 38.59277267402653,
+      "learning_rate": 9.798352186113867e-06,
+      "loss": 0.1353,
+      "step": 662
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 43.973533584689015,
+      "learning_rate": 9.797538884107574e-06,
+      "loss": 0.1582,
+      "step": 663
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 32.03919442027451,
+      "learning_rate": 9.796723979154366e-06,
+      "loss": 0.1212,
+      "step": 664
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.076183766829495,
+      "learning_rate": 9.795907471526518e-06,
+      "loss": 0.1411,
+      "step": 665
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 53.124967231457354,
+      "learning_rate": 9.79508936149684e-06,
+      "loss": 0.1311,
+      "step": 666
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 39.74430808850389,
+      "learning_rate": 9.79426964933868e-06,
+      "loss": 0.1692,
+      "step": 667
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.484329880517071,
+      "learning_rate": 9.793448335325919e-06,
+      "loss": 0.1332,
+      "step": 668
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.266390261252008,
+      "learning_rate": 9.792625419732969e-06,
+      "loss": 0.121,
+      "step": 669
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 22.195451826375955,
+      "learning_rate": 9.791800902834787e-06,
+      "loss": 0.0991,
+      "step": 670
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 28.240346235800146,
+      "learning_rate": 9.790974784906855e-06,
+      "loss": 0.1233,
+      "step": 671
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.289549119474038,
+      "learning_rate": 9.790147066225198e-06,
+      "loss": 0.1588,
+      "step": 672
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 40.07083822351059,
+      "learning_rate": 9.789317747066369e-06,
+      "loss": 0.1315,
+      "step": 673
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 17.30108556214769,
+      "learning_rate": 9.788486827707462e-06,
+      "loss": 0.1672,
+      "step": 674
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 14.124723292613249,
+      "learning_rate": 9.7876543084261e-06,
+      "loss": 0.103,
+      "step": 675
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 27.255940377336177,
+      "learning_rate": 9.786820189500443e-06,
+      "loss": 0.1493,
+      "step": 676
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 32.00415533490613,
+      "learning_rate": 9.785984471209186e-06,
+      "loss": 0.1235,
+      "step": 677
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6327392186328873,
+      "learning_rate": 9.785147153831562e-06,
+      "loss": 0.1182,
+      "step": 678
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.6193887433702985,
+      "learning_rate": 9.784308237647329e-06,
+      "loss": 0.1451,
+      "step": 679
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.595692250193478,
+      "learning_rate": 9.783467722936786e-06,
+      "loss": 0.1777,
+      "step": 680
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 37.73689869043474,
+      "learning_rate": 9.782625609980767e-06,
+      "loss": 0.1315,
+      "step": 681
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 52.16154314059323,
+      "learning_rate": 9.781781899060635e-06,
+      "loss": 0.1628,
+      "step": 682
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.752406905653005,
+      "learning_rate": 9.78093659045829e-06,
+      "loss": 0.1372,
+      "step": 683
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 32.17012777708828,
+      "learning_rate": 9.780089684456164e-06,
+      "loss": 0.1354,
+      "step": 684
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 30.835260778351316,
+      "learning_rate": 9.779241181337228e-06,
+      "loss": 0.1133,
+      "step": 685
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 32.630866103987174,
+      "learning_rate": 9.778391081384979e-06,
+      "loss": 0.1271,
+      "step": 686
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 23.691240269860604,
+      "learning_rate": 9.777539384883453e-06,
+      "loss": 0.1061,
+      "step": 687
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.821205705393695,
+      "learning_rate": 9.776686092117216e-06,
+      "loss": 0.1611,
+      "step": 688
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 28.128381007348157,
+      "learning_rate": 9.775831203371371e-06,
+      "loss": 0.1252,
+      "step": 689
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 62.024712894378,
+      "learning_rate": 9.774974718931551e-06,
+      "loss": 0.2048,
+      "step": 690
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 30.29155638874213,
+      "learning_rate": 9.774116639083923e-06,
+      "loss": 0.1371,
+      "step": 691
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 17.014556677775154,
+      "learning_rate": 9.773256964115189e-06,
+      "loss": 0.0955,
+      "step": 692
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 55.463045257964794,
+      "learning_rate": 9.772395694312583e-06,
+      "loss": 0.1831,
+      "step": 693
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 24.022624545256317,
+      "learning_rate": 9.771532829963865e-06,
+      "loss": 0.1633,
+      "step": 694
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 20.571501004321036,
+      "learning_rate": 9.770668371357344e-06,
+      "loss": 0.1271,
+      "step": 695
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 30.701880427913235,
+      "learning_rate": 9.769802318781842e-06,
+      "loss": 0.1296,
+      "step": 696
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 25.625148935174835,
+      "learning_rate": 9.76893467252673e-06,
+      "loss": 0.1271,
+      "step": 697
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.240286750024976,
+      "learning_rate": 9.768065432881903e-06,
+      "loss": 0.1227,
+      "step": 698
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 19.49975018445689,
+      "learning_rate": 9.767194600137789e-06,
+      "loss": 0.1124,
+      "step": 699
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.416082012842279,
+      "learning_rate": 9.766322174585347e-06,
+      "loss": 0.1313,
+      "step": 700
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 15.26643688822226,
+      "learning_rate": 9.765448156516077e-06,
+      "loss": 0.1049,
+      "step": 701
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 23.46621543879915,
+      "learning_rate": 9.764572546222e-06,
+      "loss": 0.1229,
+      "step": 702
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 20.60346076669723,
+      "learning_rate": 9.763695343995674e-06,
+      "loss": 0.1364,
+      "step": 703
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 14.594095858724636,
+      "learning_rate": 9.762816550130192e-06,
+      "loss": 0.0992,
+      "step": 704
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 21.631405632050786,
+      "learning_rate": 9.76193616491917e-06,
+      "loss": 0.1521,
+      "step": 705
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.107810318183903,
+      "learning_rate": 9.761054188656766e-06,
+      "loss": 0.1497,
+      "step": 706
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.5375760091303,
+      "learning_rate": 9.760170621637661e-06,
+      "loss": 0.1255,
+      "step": 707
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.731462678449901,
+      "learning_rate": 9.759285464157073e-06,
+      "loss": 0.1245,
+      "step": 708
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.6574842453813687,
+      "learning_rate": 9.758398716510751e-06,
+      "loss": 0.1086,
+      "step": 709
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 30.727691788456823,
+      "learning_rate": 9.75751037899497e-06,
+      "loss": 0.1281,
+      "step": 710
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 27.202375481436885,
+      "learning_rate": 9.756620451906543e-06,
+      "loss": 0.1276,
+      "step": 711
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 18.595561761908904,
+      "learning_rate": 9.75572893554281e-06,
+      "loss": 0.1384,
+      "step": 712
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 14.144175158912619,
+      "learning_rate": 9.754835830201645e-06,
+      "loss": 0.1586,
+      "step": 713
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 37.849196456539325,
+      "learning_rate": 9.753941136181448e-06,
+      "loss": 0.145,
+      "step": 714
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 35.69633971283892,
+      "learning_rate": 9.753044853781155e-06,
+      "loss": 0.1268,
+      "step": 715
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.152193957134935,
+      "learning_rate": 9.75214698330023e-06,
+      "loss": 0.1831,
+      "step": 716
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 27.653182844039453,
+      "learning_rate": 9.751247525038669e-06,
+      "loss": 0.1306,
+      "step": 717
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 27.593230490233363,
+      "learning_rate": 9.750346479296998e-06,
+      "loss": 0.1471,
+      "step": 718
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.305679091229352,
+      "learning_rate": 9.74944384637627e-06,
+      "loss": 0.1173,
+      "step": 719
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 27.29753578820685,
+      "learning_rate": 9.748539626578076e-06,
+      "loss": 0.1168,
+      "step": 720
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.465815777296448,
+      "learning_rate": 9.747633820204527e-06,
+      "loss": 0.1176,
+      "step": 721
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 12.977436523645396,
+      "learning_rate": 9.746726427558276e-06,
+      "loss": 0.1294,
+      "step": 722
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.50325075384258,
+      "learning_rate": 9.745817448942496e-06,
+      "loss": 0.1541,
+      "step": 723
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 25.299637458129375,
+      "learning_rate": 9.744906884660894e-06,
+      "loss": 0.146,
+      "step": 724
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 29.392294563570957,
+      "learning_rate": 9.743994735017708e-06,
+      "loss": 0.1144,
+      "step": 725
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.051587116954162,
+      "learning_rate": 9.743081000317703e-06,
+      "loss": 0.1433,
+      "step": 726
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 38.587545048429,
+      "learning_rate": 9.742165680866173e-06,
+      "loss": 0.1388,
+      "step": 727
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 23.864218562404943,
+      "learning_rate": 9.741248776968947e-06,
+      "loss": 0.1458,
+      "step": 728
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 17.858968253120967,
+      "learning_rate": 9.740330288932379e-06,
+      "loss": 0.1136,
+      "step": 729
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.168166530378955,
+      "learning_rate": 9.73941021706335e-06,
+      "loss": 0.1274,
+      "step": 730
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 20.348246500543265,
+      "learning_rate": 9.738488561669272e-06,
+      "loss": 0.1316,
+      "step": 731
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 31.053717210440414,
+      "learning_rate": 9.737565323058094e-06,
+      "loss": 0.1594,
+      "step": 732
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 21.28339041051843,
+      "learning_rate": 9.736640501538281e-06,
+      "loss": 0.1228,
+      "step": 733
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 15.286309070528286,
+      "learning_rate": 9.735714097418835e-06,
+      "loss": 0.1377,
+      "step": 734
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 27.796625488655682,
+      "learning_rate": 9.734786111009287e-06,
+      "loss": 0.1254,
+      "step": 735
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.5056897186412606,
+      "learning_rate": 9.73385654261969e-06,
+      "loss": 0.0974,
+      "step": 736
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.791484316668564,
+      "learning_rate": 9.732925392560634e-06,
+      "loss": 0.131,
+      "step": 737
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 14.411536482139912,
+      "learning_rate": 9.731992661143233e-06,
+      "loss": 0.1107,
+      "step": 738
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 15.29447975137697,
+      "learning_rate": 9.731058348679128e-06,
+      "loss": 0.1244,
+      "step": 739
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 26.753002422549987,
+      "learning_rate": 9.73012245548049e-06,
+      "loss": 0.1729,
+      "step": 740
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 15.269880803641675,
+      "learning_rate": 9.729184981860023e-06,
+      "loss": 0.1367,
+      "step": 741
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 23.286568445497416,
+      "learning_rate": 9.728245928130949e-06,
+      "loss": 0.1204,
+      "step": 742
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 12.124695582884476,
+      "learning_rate": 9.727305294607024e-06,
+      "loss": 0.1174,
+      "step": 743
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.340039718599179,
+      "learning_rate": 9.726363081602532e-06,
+      "loss": 0.1272,
+      "step": 744
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.03978216102685,
+      "learning_rate": 9.725419289432287e-06,
+      "loss": 0.1349,
+      "step": 745
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 12.900890675498434,
+      "learning_rate": 9.724473918411624e-06,
+      "loss": 0.1295,
+      "step": 746
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 12.021350602091163,
+      "learning_rate": 9.723526968856408e-06,
+      "loss": 0.1057,
+      "step": 747
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.989225561921891,
+      "learning_rate": 9.722578441083035e-06,
+      "loss": 0.0867,
+      "step": 748
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 17.269981813172723,
+      "learning_rate": 9.721628335408423e-06,
+      "loss": 0.1116,
+      "step": 749
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 30.291178256649264,
+      "learning_rate": 9.720676652150025e-06,
+      "loss": 0.1224,
+      "step": 750
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.188545883869058,
+      "learning_rate": 9.719723391625813e-06,
+      "loss": 0.1566,
+      "step": 751
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.183215999542815,
+      "learning_rate": 9.718768554154287e-06,
+      "loss": 0.1754,
+      "step": 752
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.300039794356254,
+      "learning_rate": 9.717812140054479e-06,
+      "loss": 0.1091,
+      "step": 753
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 43.092405088706,
+      "learning_rate": 9.716854149645945e-06,
+      "loss": 0.1327,
+      "step": 754
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.989995368482804,
+      "learning_rate": 9.715894583248764e-06,
+      "loss": 0.1329,
+      "step": 755
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.450637838006689,
+      "learning_rate": 9.714933441183549e-06,
+      "loss": 0.1046,
+      "step": 756
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.370431682700567,
+      "learning_rate": 9.713970723771432e-06,
+      "loss": 0.094,
+      "step": 757
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.439106294822273,
+      "learning_rate": 9.713006431334076e-06,
+      "loss": 0.1075,
+      "step": 758
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.568446270079447,
+      "learning_rate": 9.71204056419367e-06,
+      "loss": 0.1646,
+      "step": 759
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.719463921354748,
+      "learning_rate": 9.711073122672928e-06,
+      "loss": 0.1296,
+      "step": 760
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 32.803026706716224,
+      "learning_rate": 9.71010410709509e-06,
+      "loss": 0.1307,
+      "step": 761
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.96453423011384,
+      "learning_rate": 9.70913351778392e-06,
+      "loss": 0.1228,
+      "step": 762
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 20.833929487014423,
+      "learning_rate": 9.708161355063714e-06,
+      "loss": 0.1479,
+      "step": 763
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 39.33943975858833,
+      "learning_rate": 9.707187619259286e-06,
+      "loss": 0.123,
+      "step": 764
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.376268195351014,
+      "learning_rate": 9.706212310695981e-06,
+      "loss": 0.125,
+      "step": 765
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.440271953630097,
+      "learning_rate": 9.705235429699666e-06,
+      "loss": 0.1115,
+      "step": 766
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 20.249022741934898,
+      "learning_rate": 9.704256976596737e-06,
+      "loss": 0.1263,
+      "step": 767
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 26.41981337531156,
+      "learning_rate": 9.703276951714114e-06,
+      "loss": 0.1115,
+      "step": 768
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 36.217352696254,
+      "learning_rate": 9.70229535537924e-06,
+      "loss": 0.0952,
+      "step": 769
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 14.397083314980746,
+      "learning_rate": 9.701312187920084e-06,
+      "loss": 0.1769,
+      "step": 770
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 26.0325827243847,
+      "learning_rate": 9.700327449665143e-06,
+      "loss": 0.1141,
+      "step": 771
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 36.351045967305154,
+      "learning_rate": 9.699341140943434e-06,
+      "loss": 0.1384,
+      "step": 772
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 23.25267147893353,
+      "learning_rate": 9.698353262084501e-06,
+      "loss": 0.1324,
+      "step": 773
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.482877435007008,
+      "learning_rate": 9.697363813418414e-06,
+      "loss": 0.1206,
+      "step": 774
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 12.560594535886262,
+      "learning_rate": 9.696372795275766e-06,
+      "loss": 0.1587,
+      "step": 775
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 14.141509443672007,
+      "learning_rate": 9.695380207987675e-06,
+      "loss": 0.0968,
+      "step": 776
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 23.18138524562932,
+      "learning_rate": 9.69438605188578e-06,
+      "loss": 0.1436,
+      "step": 777
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 18.79428105301295,
+      "learning_rate": 9.69339032730225e-06,
+      "loss": 0.1552,
+      "step": 778
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 30.44553819976487,
+      "learning_rate": 9.692393034569776e-06,
+      "loss": 0.1146,
+      "step": 779
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.399140231533663,
+      "learning_rate": 9.69139417402157e-06,
+      "loss": 0.1072,
+      "step": 780
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 19.39038831015829,
+      "learning_rate": 9.690393745991368e-06,
+      "loss": 0.1361,
+      "step": 781
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 16.807652464179398,
+      "learning_rate": 9.689391750813436e-06,
+      "loss": 0.1516,
+      "step": 782
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 12.595719018779906,
+      "learning_rate": 9.688388188822556e-06,
+      "loss": 0.1456,
+      "step": 783
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 78.16077311228811,
+      "learning_rate": 9.687383060354038e-06,
+      "loss": 0.2327,
+      "step": 784
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 12.822682377103986,
+      "learning_rate": 9.686376365743714e-06,
+      "loss": 0.1251,
+      "step": 785
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 35.69822535004326,
+      "learning_rate": 9.685368105327938e-06,
+      "loss": 0.1688,
+      "step": 786
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 27.540323241637438,
+      "learning_rate": 9.684358279443593e-06,
+      "loss": 0.1223,
+      "step": 787
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 52.99880117052681,
+      "learning_rate": 9.683346888428074e-06,
+      "loss": 0.1387,
+      "step": 788
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 38.310273349733706,
+      "learning_rate": 9.68233393261931e-06,
+      "loss": 0.156,
+      "step": 789
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 34.726834784529174,
+      "learning_rate": 9.681319412355748e-06,
+      "loss": 0.0992,
+      "step": 790
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 68.29196974086027,
+      "learning_rate": 9.680303327976356e-06,
+      "loss": 0.1697,
+      "step": 791
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 44.244078950175414,
+      "learning_rate": 9.679285679820628e-06,
+      "loss": 0.1471,
+      "step": 792
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.851951807112226,
+      "learning_rate": 9.67826646822858e-06,
+      "loss": 0.1339,
+      "step": 793
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 49.15989066226097,
+      "learning_rate": 9.677245693540749e-06,
+      "loss": 0.1746,
+      "step": 794
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 29.52416076684413,
+      "learning_rate": 9.676223356098194e-06,
+      "loss": 0.1154,
+      "step": 795
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.38163332741748,
+      "learning_rate": 9.675199456242499e-06,
+      "loss": 0.1305,
+      "step": 796
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.925248900398602,
+      "learning_rate": 9.674173994315764e-06,
+      "loss": 0.1724,
+      "step": 797
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 17.954529179430356,
+      "learning_rate": 9.67314697066062e-06,
+      "loss": 0.1324,
+      "step": 798
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 19.573255779518618,
+      "learning_rate": 9.672118385620209e-06,
+      "loss": 0.1199,
+      "step": 799
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 45.10260866334031,
+      "learning_rate": 9.671088239538204e-06,
+      "loss": 0.168,
+      "step": 800
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.096128376621521,
+      "learning_rate": 9.670056532758798e-06,
+      "loss": 0.113,
+      "step": 801
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 41.288709677166764,
+      "learning_rate": 9.669023265626698e-06,
+      "loss": 0.1699,
+      "step": 802
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 50.206876725438384,
+      "learning_rate": 9.66798843848714e-06,
+      "loss": 0.1154,
+      "step": 803
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 18.766537968172486,
+      "learning_rate": 9.666952051685882e-06,
+      "loss": 0.1078,
+      "step": 804
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 27.654720064496317,
+      "learning_rate": 9.665914105569196e-06,
+      "loss": 0.1472,
+      "step": 805
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 30.465348594344608,
+      "learning_rate": 9.664874600483883e-06,
+      "loss": 0.1125,
+      "step": 806
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 35.45942332075495,
+      "learning_rate": 9.663833536777256e-06,
+      "loss": 0.1239,
+      "step": 807
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.15194949775956,
+      "learning_rate": 9.662790914797158e-06,
+      "loss": 0.1382,
+      "step": 808
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.445521315197055,
+      "learning_rate": 9.661746734891947e-06,
+      "loss": 0.1438,
+      "step": 809
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 18.926741667241487,
+      "learning_rate": 9.6607009974105e-06,
+      "loss": 0.1327,
+      "step": 810
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 34.07683918961295,
+      "learning_rate": 9.659653702702223e-06,
+      "loss": 0.1337,
+      "step": 811
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 31.666273539929072,
+      "learning_rate": 9.658604851117032e-06,
+      "loss": 0.1421,
+      "step": 812
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.290269985502701,
+      "learning_rate": 9.65755444300537e-06,
+      "loss": 0.1259,
+      "step": 813
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 20.748229179136143,
+      "learning_rate": 9.656502478718197e-06,
+      "loss": 0.1207,
+      "step": 814
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 18.76253063384707,
+      "learning_rate": 9.655448958606994e-06,
+      "loss": 0.1289,
+      "step": 815
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 47.068157538168485,
+      "learning_rate": 9.654393883023763e-06,
+      "loss": 0.1449,
+      "step": 816
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 16.508111889804976,
+      "learning_rate": 9.653337252321023e-06,
+      "loss": 0.137,
+      "step": 817
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.161976884102493,
+      "learning_rate": 9.652279066851811e-06,
+      "loss": 0.126,
+      "step": 818
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 69.48129884052561,
+      "learning_rate": 9.651219326969694e-06,
+      "loss": 0.179,
+      "step": 819
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 25.548064760384744,
+      "learning_rate": 9.650158033028743e-06,
+      "loss": 0.1292,
+      "step": 820
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 22.848051753734424,
+      "learning_rate": 9.64909518538356e-06,
+      "loss": 0.1185,
+      "step": 821
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 38.528184137095906,
+      "learning_rate": 9.648030784389264e-06,
+      "loss": 0.1333,
+      "step": 822
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 40.289153016277766,
+      "learning_rate": 9.646964830401487e-06,
+      "loss": 0.1868,
+      "step": 823
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 28.63000932230696,
+      "learning_rate": 9.645897323776386e-06,
+      "loss": 0.1309,
+      "step": 824
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 29.390175324133896,
+      "learning_rate": 9.644828264870634e-06,
+      "loss": 0.1494,
+      "step": 825
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 29.520098412687467,
+      "learning_rate": 9.643757654041423e-06,
+      "loss": 0.1147,
+      "step": 826
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 30.864194139048347,
+      "learning_rate": 9.642685491646467e-06,
+      "loss": 0.1078,
+      "step": 827
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 42.21262734452298,
+      "learning_rate": 9.641611778043992e-06,
+      "loss": 0.1384,
+      "step": 828
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.587684898717356,
+      "learning_rate": 9.64053651359275e-06,
+      "loss": 0.1387,
+      "step": 829
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 36.28856657814182,
+      "learning_rate": 9.639459698652e-06,
+      "loss": 0.1475,
+      "step": 830
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 40.3795735401778,
+      "learning_rate": 9.63838133358153e-06,
+      "loss": 0.1454,
+      "step": 831
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 56.58394981425765,
+      "learning_rate": 9.637301418741643e-06,
+      "loss": 0.1733,
+      "step": 832
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.372522640119524,
+      "learning_rate": 9.636219954493157e-06,
+      "loss": 0.1172,
+      "step": 833
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.10785762912234,
+      "learning_rate": 9.635136941197409e-06,
+      "loss": 0.1244,
+      "step": 834
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 47.76352747100944,
+      "learning_rate": 9.634052379216256e-06,
+      "loss": 0.1663,
+      "step": 835
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 27.543918826169143,
+      "learning_rate": 9.632966268912067e-06,
+      "loss": 0.1227,
+      "step": 836
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 27.936130596657357,
+      "learning_rate": 9.631878610647734e-06,
+      "loss": 0.14,
+      "step": 837
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 35.17792432965117,
+      "learning_rate": 9.630789404786664e-06,
+      "loss": 0.156,
+      "step": 838
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 28.829265986491016,
+      "learning_rate": 9.629698651692779e-06,
+      "loss": 0.1443,
+      "step": 839
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 24.10923790760327,
+      "learning_rate": 9.62860635173052e-06,
+      "loss": 0.1364,
+      "step": 840
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.678265188180967,
+      "learning_rate": 9.627512505264847e-06,
+      "loss": 0.1251,
+      "step": 841
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 30.518872167549794,
+      "learning_rate": 9.626417112661233e-06,
+      "loss": 0.1193,
+      "step": 842
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.30248553766188,
+      "learning_rate": 9.62532017428567e-06,
+      "loss": 0.1294,
+      "step": 843
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.95928112006393,
+      "learning_rate": 9.624221690504663e-06,
+      "loss": 0.1318,
+      "step": 844
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.975229344522406,
+      "learning_rate": 9.623121661685239e-06,
+      "loss": 0.1351,
+      "step": 845
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.673008949146638,
+      "learning_rate": 9.622020088194934e-06,
+      "loss": 0.1339,
+      "step": 846
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.7187307082075485,
+      "learning_rate": 9.62091697040181e-06,
+      "loss": 0.109,
+      "step": 847
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 19.086793156437807,
+      "learning_rate": 9.619812308674434e-06,
+      "loss": 0.1528,
+      "step": 848
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 15.448285601263763,
+      "learning_rate": 9.618706103381896e-06,
+      "loss": 0.137,
+      "step": 849
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.43458261032492,
+      "learning_rate": 9.6175983548938e-06,
+      "loss": 0.1368,
+      "step": 850
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.695331340739746,
+      "learning_rate": 9.616489063580265e-06,
+      "loss": 0.1494,
+      "step": 851
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 16.36884007886458,
+      "learning_rate": 9.615378229811927e-06,
+      "loss": 0.098,
+      "step": 852
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.709638349793122,
+      "learning_rate": 9.614265853959935e-06,
+      "loss": 0.1013,
+      "step": 853
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 13.96540985996891,
+      "learning_rate": 9.613151936395952e-06,
+      "loss": 0.1692,
+      "step": 854
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.1809794075417095,
+      "learning_rate": 9.612036477492163e-06,
+      "loss": 0.1151,
+      "step": 855
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 31.864538660439695,
+      "learning_rate": 9.610919477621262e-06,
+      "loss": 0.094,
+      "step": 856
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.028949180354989,
+      "learning_rate": 9.609800937156459e-06,
+      "loss": 0.1671,
+      "step": 857
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 13.125900295808705,
+      "learning_rate": 9.60868085647148e-06,
+      "loss": 0.1697,
+      "step": 858
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 26.03409453086144,
+      "learning_rate": 9.607559235940562e-06,
+      "loss": 0.1094,
+      "step": 859
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 14.356106276553628,
+      "learning_rate": 9.60643607593846e-06,
+      "loss": 0.1176,
+      "step": 860
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.48182805841283,
+      "learning_rate": 9.605311376840446e-06,
+      "loss": 0.1034,
+      "step": 861
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.737169065727766,
+      "learning_rate": 9.604185139022302e-06,
+      "loss": 0.1119,
+      "step": 862
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.543502918761671,
+      "learning_rate": 9.603057362860323e-06,
+      "loss": 0.1512,
+      "step": 863
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 13.137753210423396,
+      "learning_rate": 9.60192804873132e-06,
+      "loss": 0.0869,
+      "step": 864
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.238531279627713,
+      "learning_rate": 9.60079719701262e-06,
+      "loss": 0.1251,
+      "step": 865
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.442828257491122,
+      "learning_rate": 9.599664808082058e-06,
+      "loss": 0.1073,
+      "step": 866
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 45.80139113041787,
+      "learning_rate": 9.598530882317992e-06,
+      "loss": 0.1849,
+      "step": 867
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.013274249831894,
+      "learning_rate": 9.59739542009928e-06,
+      "loss": 0.1652,
+      "step": 868
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 20.94648720784535,
+      "learning_rate": 9.596258421805306e-06,
+      "loss": 0.1349,
+      "step": 869
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 20.308753874252467,
+      "learning_rate": 9.595119887815962e-06,
+      "loss": 0.1345,
+      "step": 870
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 30.38587830318456,
+      "learning_rate": 9.593979818511655e-06,
+      "loss": 0.1326,
+      "step": 871
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.113951857489782,
+      "learning_rate": 9.592838214273298e-06,
+      "loss": 0.1516,
+      "step": 872
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.08267552970721,
+      "learning_rate": 9.591695075482326e-06,
+      "loss": 0.1118,
+      "step": 873
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 21.290502900984546,
+      "learning_rate": 9.590550402520683e-06,
+      "loss": 0.1262,
+      "step": 874
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 40.747754501062765,
+      "learning_rate": 9.589404195770821e-06,
+      "loss": 0.145,
+      "step": 875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.363040561381,
+      "learning_rate": 9.588256455615716e-06,
+      "loss": 0.1309,
+      "step": 876
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.164698662773475,
+      "learning_rate": 9.587107182438846e-06,
+      "loss": 0.1272,
+      "step": 877
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 46.7965238441854,
+      "learning_rate": 9.585956376624204e-06,
+      "loss": 0.1318,
+      "step": 878
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.889294742395169,
+      "learning_rate": 9.584804038556297e-06,
+      "loss": 0.1427,
+      "step": 879
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.060406285399305,
+      "learning_rate": 9.58365016862014e-06,
+      "loss": 0.1256,
+      "step": 880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 40.00127209719933,
+      "learning_rate": 9.582494767201265e-06,
+      "loss": 0.1309,
+      "step": 881
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.84245575000966,
+      "learning_rate": 9.581337834685713e-06,
+      "loss": 0.104,
+      "step": 882
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 23.402757032695074,
+      "learning_rate": 9.580179371460034e-06,
+      "loss": 0.1289,
+      "step": 883
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 24.43621959954114,
+      "learning_rate": 9.579019377911296e-06,
+      "loss": 0.1401,
+      "step": 884
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.753446651852203,
+      "learning_rate": 9.57785785442707e-06,
+      "loss": 0.1206,
+      "step": 885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.301431458818122,
+      "learning_rate": 9.576694801395447e-06,
+      "loss": 0.1083,
+      "step": 886
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 15.99251554227728,
+      "learning_rate": 9.57553021920502e-06,
+      "loss": 0.1708,
+      "step": 887
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 27.354550811963918,
+      "learning_rate": 9.574364108244903e-06,
+      "loss": 0.1095,
+      "step": 888
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 14.453620875280626,
+      "learning_rate": 9.573196468904711e-06,
+      "loss": 0.1439,
+      "step": 889
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 23.906822227072045,
+      "learning_rate": 9.572027301574576e-06,
+      "loss": 0.177,
+      "step": 890
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 40.74283625046122,
+      "learning_rate": 9.570856606645139e-06,
+      "loss": 0.1349,
+      "step": 891
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 32.77881193498861,
+      "learning_rate": 9.569684384507547e-06,
+      "loss": 0.1088,
+      "step": 892
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.788598652893294,
+      "learning_rate": 9.568510635553466e-06,
+      "loss": 0.1145,
+      "step": 893
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.486909948066521,
+      "learning_rate": 9.567335360175065e-06,
+      "loss": 0.131,
+      "step": 894
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.250156911609917,
+      "learning_rate": 9.566158558765026e-06,
+      "loss": 0.1603,
+      "step": 895
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 23.236140333378174,
+      "learning_rate": 9.564980231716541e-06,
+      "loss": 0.1176,
+      "step": 896
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 37.16329810519581,
+      "learning_rate": 9.56380037942331e-06,
+      "loss": 0.1315,
+      "step": 897
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.545747912275482,
+      "learning_rate": 9.562619002279541e-06,
+      "loss": 0.1576,
+      "step": 898
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.7428484359296625,
+      "learning_rate": 9.561436100679959e-06,
+      "loss": 0.1204,
+      "step": 899
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 34.822828382532826,
+      "learning_rate": 9.56025167501979e-06,
+      "loss": 0.1407,
+      "step": 900
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.274669158416529,
+      "learning_rate": 9.559065725694775e-06,
+      "loss": 0.1305,
+      "step": 901
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.2730681558009245,
+      "learning_rate": 9.55787825310116e-06,
+      "loss": 0.1417,
+      "step": 902
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.919028484956142,
+      "learning_rate": 9.5566892576357e-06,
+      "loss": 0.0978,
+      "step": 903
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.4710721096915,
+      "learning_rate": 9.555498739695665e-06,
+      "loss": 0.1519,
+      "step": 904
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.993053552415446,
+      "learning_rate": 9.554306699678827e-06,
+      "loss": 0.1193,
+      "step": 905
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 29.810998416607553,
+      "learning_rate": 9.553113137983467e-06,
+      "loss": 0.1377,
+      "step": 906
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.703199156948056,
+      "learning_rate": 9.551918055008378e-06,
+      "loss": 0.125,
+      "step": 907
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.792361535663362,
+      "learning_rate": 9.55072145115286e-06,
+      "loss": 0.1246,
+      "step": 908
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 38.43165453677999,
+      "learning_rate": 9.54952332681672e-06,
+      "loss": 0.1543,
+      "step": 909
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 20.810549271538743,
+      "learning_rate": 9.54832368240027e-06,
+      "loss": 0.1169,
+      "step": 910
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 14.216738798712827,
+      "learning_rate": 9.54712251830434e-06,
+      "loss": 0.1195,
+      "step": 911
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 37.7812086135415,
+      "learning_rate": 9.545919834930257e-06,
+      "loss": 0.1229,
+      "step": 912
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 16.088887693282086,
+      "learning_rate": 9.54471563267986e-06,
+      "loss": 0.1521,
+      "step": 913
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.086877374953741,
+      "learning_rate": 9.543509911955497e-06,
+      "loss": 0.1245,
+      "step": 914
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.884647790510014,
+      "learning_rate": 9.542302673160021e-06,
+      "loss": 0.1477,
+      "step": 915
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 13.471052689354279,
+      "learning_rate": 9.541093916696793e-06,
+      "loss": 0.1655,
+      "step": 916
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.860831191922153,
+      "learning_rate": 9.539883642969681e-06,
+      "loss": 0.0962,
+      "step": 917
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.950951525429048,
+      "learning_rate": 9.53867185238306e-06,
+      "loss": 0.1212,
+      "step": 918
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 29.397346489598785,
+      "learning_rate": 9.53745854534181e-06,
+      "loss": 0.1531,
+      "step": 919
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 14.700115456334249,
+      "learning_rate": 9.536243722251321e-06,
+      "loss": 0.1633,
+      "step": 920
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 31.465682765345253,
+      "learning_rate": 9.53502738351749e-06,
+      "loss": 0.1273,
+      "step": 921
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 13.342265225292849,
+      "learning_rate": 9.533809529546716e-06,
+      "loss": 0.0986,
+      "step": 922
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.06223070740461,
+      "learning_rate": 9.532590160745906e-06,
+      "loss": 0.1138,
+      "step": 923
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.891965791948293,
+      "learning_rate": 9.531369277522475e-06,
+      "loss": 0.1008,
+      "step": 924
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 14.385809821534846,
+      "learning_rate": 9.530146880284343e-06,
+      "loss": 0.1107,
+      "step": 925
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.594656018466403,
+      "learning_rate": 9.528922969439935e-06,
+      "loss": 0.1097,
+      "step": 926
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 23.884572371985932,
+      "learning_rate": 9.527697545398183e-06,
+      "loss": 0.1483,
+      "step": 927
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.235374187689162,
+      "learning_rate": 9.526470608568521e-06,
+      "loss": 0.1179,
+      "step": 928
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.498044658058308,
+      "learning_rate": 9.525242159360897e-06,
+      "loss": 0.1262,
+      "step": 929
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.326025536830267,
+      "learning_rate": 9.524012198185755e-06,
+      "loss": 0.1978,
+      "step": 930
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 15.950579961915821,
+      "learning_rate": 9.522780725454048e-06,
+      "loss": 0.1472,
+      "step": 931
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 16.911939075772374,
+      "learning_rate": 9.521547741577232e-06,
+      "loss": 0.1405,
+      "step": 932
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.840723626936537,
+      "learning_rate": 9.520313246967277e-06,
+      "loss": 0.1378,
+      "step": 933
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.194781971670784,
+      "learning_rate": 9.519077242036643e-06,
+      "loss": 0.1351,
+      "step": 934
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 26.571594739816774,
+      "learning_rate": 9.517839727198306e-06,
+      "loss": 0.1461,
+      "step": 935
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 25.488817918137755,
+      "learning_rate": 9.516600702865742e-06,
+      "loss": 0.1245,
+      "step": 936
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.428264723234935,
+      "learning_rate": 9.51536016945293e-06,
+      "loss": 0.1256,
+      "step": 937
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.398832656295712,
+      "learning_rate": 9.514118127374358e-06,
+      "loss": 0.118,
+      "step": 938
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 28.08087002312531,
+      "learning_rate": 9.512874577045016e-06,
+      "loss": 0.1302,
+      "step": 939
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 20.624474099676195,
+      "learning_rate": 9.511629518880394e-06,
+      "loss": 0.1273,
+      "step": 940
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.613846962609959,
+      "learning_rate": 9.510382953296492e-06,
+      "loss": 0.0986,
+      "step": 941
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 35.918107603359275,
+      "learning_rate": 9.50913488070981e-06,
+      "loss": 0.1638,
+      "step": 942
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 26.49689819650897,
+      "learning_rate": 9.50788530153735e-06,
+      "loss": 0.1265,
+      "step": 943
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 16.996031473631124,
+      "learning_rate": 9.506634216196621e-06,
+      "loss": 0.1117,
+      "step": 944
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 29.06015069003101,
+      "learning_rate": 9.505381625105636e-06,
+      "loss": 0.1605,
+      "step": 945
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 30.453905417068526,
+      "learning_rate": 9.504127528682907e-06,
+      "loss": 0.1602,
+      "step": 946
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.628167915371439,
+      "learning_rate": 9.502871927347452e-06,
+      "loss": 0.101,
+      "step": 947
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.321842222130279,
+      "learning_rate": 9.501614821518789e-06,
+      "loss": 0.1279,
+      "step": 948
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.367606237846145,
+      "learning_rate": 9.500356211616941e-06,
+      "loss": 0.1444,
+      "step": 949
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 52.609164817487056,
+      "learning_rate": 9.499096098062435e-06,
+      "loss": 0.1614,
+      "step": 950
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.849007077901183,
+      "learning_rate": 9.497834481276293e-06,
+      "loss": 0.1133,
+      "step": 951
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.628839587514296,
+      "learning_rate": 9.496571361680052e-06,
+      "loss": 0.1362,
+      "step": 952
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 22.70492961795664,
+      "learning_rate": 9.495306739695738e-06,
+      "loss": 0.1348,
+      "step": 953
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 23.438670549885423,
+      "learning_rate": 9.494040615745887e-06,
+      "loss": 0.1139,
+      "step": 954
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.777828080048308,
+      "learning_rate": 9.492772990253535e-06,
+      "loss": 0.1149,
+      "step": 955
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 17.01356783506953,
+      "learning_rate": 9.49150386364222e-06,
+      "loss": 0.1173,
+      "step": 956
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 20.86435129792081,
+      "learning_rate": 9.490233236335977e-06,
+      "loss": 0.1227,
+      "step": 957
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 21.168010427297975,
+      "learning_rate": 9.488961108759349e-06,
+      "loss": 0.0913,
+      "step": 958
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 12.351204194124849,
+      "learning_rate": 9.487687481337377e-06,
+      "loss": 0.1135,
+      "step": 959
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.269203529612833,
+      "learning_rate": 9.486412354495605e-06,
+      "loss": 0.1016,
+      "step": 960
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.879287993430144,
+      "learning_rate": 9.485135728660073e-06,
+      "loss": 0.1355,
+      "step": 961
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 19.29718867284448,
+      "learning_rate": 9.48385760425733e-06,
+      "loss": 0.1233,
+      "step": 962
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.109327624294433,
+      "learning_rate": 9.482577981714417e-06,
+      "loss": 0.1221,
+      "step": 963
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.843077711835422,
+      "learning_rate": 9.481296861458881e-06,
+      "loss": 0.1158,
+      "step": 964
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.528475309427824,
+      "learning_rate": 9.480014243918769e-06,
+      "loss": 0.168,
+      "step": 965
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 15.294376985491054,
+      "learning_rate": 9.478730129522627e-06,
+      "loss": 0.1021,
+      "step": 966
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 23.53373087484837,
+      "learning_rate": 9.477444518699501e-06,
+      "loss": 0.156,
+      "step": 967
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.345322636300342,
+      "learning_rate": 9.476157411878937e-06,
+      "loss": 0.1395,
+      "step": 968
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 29.858200916606965,
+      "learning_rate": 9.474868809490984e-06,
+      "loss": 0.1115,
+      "step": 969
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.955084258234279,
+      "learning_rate": 9.473578711966185e-06,
+      "loss": 0.1597,
+      "step": 970
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.600246069094517,
+      "learning_rate": 9.472287119735588e-06,
+      "loss": 0.1439,
+      "step": 971
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.893917106674843,
+      "learning_rate": 9.470994033230735e-06,
+      "loss": 0.1113,
+      "step": 972
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.074697713863284,
+      "learning_rate": 9.469699452883672e-06,
+      "loss": 0.1354,
+      "step": 973
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.174208973947078,
+      "learning_rate": 9.468403379126943e-06,
+      "loss": 0.1121,
+      "step": 974
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 18.846130115274406,
+      "learning_rate": 9.46710581239359e-06,
+      "loss": 0.163,
+      "step": 975
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.592869078244344,
+      "learning_rate": 9.465806753117153e-06,
+      "loss": 0.1381,
+      "step": 976
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.7149921395013608,
+      "learning_rate": 9.464506201731674e-06,
+      "loss": 0.0979,
+      "step": 977
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.5795453248401605,
+      "learning_rate": 9.463204158671687e-06,
+      "loss": 0.1036,
+      "step": 978
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.827533012507683,
+      "learning_rate": 9.461900624372233e-06,
+      "loss": 0.1218,
+      "step": 979
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.285235509914184,
+      "learning_rate": 9.460595599268848e-06,
+      "loss": 0.1506,
+      "step": 980
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.469405449404304,
+      "learning_rate": 9.45928908379756e-06,
+      "loss": 0.0923,
+      "step": 981
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 13.088900389735285,
+      "learning_rate": 9.457981078394905e-06,
+      "loss": 0.1519,
+      "step": 982
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.416521356456194,
+      "learning_rate": 9.45667158349791e-06,
+      "loss": 0.1038,
+      "step": 983
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.000029309311584,
+      "learning_rate": 9.4553605995441e-06,
+      "loss": 0.1067,
+      "step": 984
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.179225960011602,
+      "learning_rate": 9.4540481269715e-06,
+      "loss": 0.1224,
+      "step": 985
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.010485857421125,
+      "learning_rate": 9.452734166218635e-06,
+      "loss": 0.1298,
+      "step": 986
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 28.699516768299603,
+      "learning_rate": 9.451418717724518e-06,
+      "loss": 0.1122,
+      "step": 987
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 17.47564053882029,
+      "learning_rate": 9.45010178192867e-06,
+      "loss": 0.105,
+      "step": 988
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 22.22446380391295,
+      "learning_rate": 9.448783359271102e-06,
+      "loss": 0.1224,
+      "step": 989
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.196080255624146,
+      "learning_rate": 9.44746345019232e-06,
+      "loss": 0.1234,
+      "step": 990
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 23.353760929453095,
+      "learning_rate": 9.446142055133333e-06,
+      "loss": 0.1925,
+      "step": 991
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.684837863530596,
+      "learning_rate": 9.444819174535647e-06,
+      "loss": 0.1752,
+      "step": 992
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 41.91167980648941,
+      "learning_rate": 9.443494808841255e-06,
+      "loss": 0.1741,
+      "step": 993
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 25.021333305393867,
+      "learning_rate": 9.442168958492657e-06,
+      "loss": 0.1377,
+      "step": 994
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 17.303187659047722,
+      "learning_rate": 9.44084162393284e-06,
+      "loss": 0.1587,
+      "step": 995
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.881013346273633,
+      "learning_rate": 9.439512805605294e-06,
+      "loss": 0.1152,
+      "step": 996
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.679523342858859,
+      "learning_rate": 9.438182503954002e-06,
+      "loss": 0.1433,
+      "step": 997
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.809722363046939,
+      "learning_rate": 9.43685071942344e-06,
+      "loss": 0.1212,
+      "step": 998
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 16.300528267719702,
+      "learning_rate": 9.435517452458584e-06,
+      "loss": 0.1003,
+      "step": 999
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.142200136073567,
+      "learning_rate": 9.434182703504904e-06,
+      "loss": 0.1167,
+      "step": 1000
+    },
+    {
+      "epoch": 0.71,
+      "eval_avg_AUC": 0.8306426631221906,
+      "eval_avg_Accuracy": 0.697198275862069,
+      "eval_avg_Accuracy-right": 0.9357636624494587,
+      "eval_avg_Accuracy-wrong": 0.2812144644075506,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.7007286233364818,
+      "eval_last_AUC": 0.8499919642378029,
+      "eval_last_Accuracy": 0.7737483421750663,
+      "eval_last_Accuracy-right": 0.8576366244945872,
+      "eval_last_Accuracy-wrong": 0.6274732772344781,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.7122224635862333,
+      "eval_max_AUC": 0.7712377534346357,
+      "eval_max_Accuracy": 0.6421170424403183,
+      "eval_max_Accuracy-right": 0.9876744489370027,
+      "eval_max_Accuracy-wrong": 0.03957243575164885,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.6283448927147767,
+      "eval_min_AUC": 0.8459536118200042,
+      "eval_min_Accuracy": 0.7756962864721485,
+      "eval_min_Accuracy-right": 0.8238554845441503,
+      "eval_min_Accuracy-wrong": 0.6917216283829885,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.7067690279119097,
+      "eval_prod_AUC": 0.8340437720336431,
+      "eval_prod_Accuracy": 0.7153929045092838,
+      "eval_prod_Accuracy-right": 0.5985391939480892,
+      "eval_prod_Accuracy-wrong": 0.9191494200591313,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.6925029632488553,
+      "eval_runtime": 246.6967,
+      "eval_samples_per_second": 97.804,
+      "eval_steps_per_second": 3.056,
+      "eval_sum_AUC": 0.6696619705759399,
+      "eval_sum_Accuracy": 0.6358173076923077,
+      "eval_sum_Accuracy-right": 0.9999347854441111,
+      "eval_sum_Accuracy-wrong": 0.0009097111667045713,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6722441685783858,
+      "step": 1000
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 26.90301939765578,
+      "learning_rate": 9.432846473008363e-06,
+      "loss": 0.1073,
+      "step": 1001
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 22.473750563787483,
+      "learning_rate": 9.431508761415422e-06,
+      "loss": 0.1453,
+      "step": 1002
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.035685660671625,
+      "learning_rate": 9.430169569173034e-06,
+      "loss": 0.1307,
+      "step": 1003
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.372775149242225,
+      "learning_rate": 9.428828896728645e-06,
+      "loss": 0.1189,
+      "step": 1004
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 14.367906249168973,
+      "learning_rate": 9.427486744530205e-06,
+      "loss": 0.1089,
+      "step": 1005
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 21.80296312308416,
+      "learning_rate": 9.426143113026147e-06,
+      "loss": 0.1641,
+      "step": 1006
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 12.689136775090986,
+      "learning_rate": 9.424798002665405e-06,
+      "loss": 0.1456,
+      "step": 1007
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 19.914315829482117,
+      "learning_rate": 9.423451413897406e-06,
+      "loss": 0.1381,
+      "step": 1008
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 17.487013924724515,
+      "learning_rate": 9.42210334717207e-06,
+      "loss": 0.1842,
+      "step": 1009
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 41.591369148501364,
+      "learning_rate": 9.42075380293981e-06,
+      "loss": 0.1582,
+      "step": 1010
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 21.325164316590822,
+      "learning_rate": 9.419402781651537e-06,
+      "loss": 0.1091,
+      "step": 1011
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 27.610236895046608,
+      "learning_rate": 9.418050283758647e-06,
+      "loss": 0.1791,
+      "step": 1012
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 48.91961383487278,
+      "learning_rate": 9.416696309713038e-06,
+      "loss": 0.1597,
+      "step": 1013
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.314404735669536,
+      "learning_rate": 9.415340859967099e-06,
+      "loss": 0.1124,
+      "step": 1014
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.467559964579186,
+      "learning_rate": 9.413983934973709e-06,
+      "loss": 0.1421,
+      "step": 1015
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 39.483948098381305,
+      "learning_rate": 9.412625535186242e-06,
+      "loss": 0.1479,
+      "step": 1016
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 25.24978912021175,
+      "learning_rate": 9.411265661058565e-06,
+      "loss": 0.1482,
+      "step": 1017
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 13.0958518186262,
+      "learning_rate": 9.409904313045038e-06,
+      "loss": 0.1525,
+      "step": 1018
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 15.169402604256561,
+      "learning_rate": 9.408541491600511e-06,
+      "loss": 0.1279,
+      "step": 1019
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 65.13621930141626,
+      "learning_rate": 9.407177197180328e-06,
+      "loss": 0.1594,
+      "step": 1020
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 47.31199672873694,
+      "learning_rate": 9.405811430240329e-06,
+      "loss": 0.1588,
+      "step": 1021
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 15.929281558822629,
+      "learning_rate": 9.404444191236837e-06,
+      "loss": 0.1355,
+      "step": 1022
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 43.31429781114566,
+      "learning_rate": 9.403075480626674e-06,
+      "loss": 0.1611,
+      "step": 1023
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 50.56085481452662,
+      "learning_rate": 9.401705298867151e-06,
+      "loss": 0.1643,
+      "step": 1024
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 36.11556641873227,
+      "learning_rate": 9.400333646416073e-06,
+      "loss": 0.1428,
+      "step": 1025
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.0767580228440075,
+      "learning_rate": 9.398960523731735e-06,
+      "loss": 0.1222,
+      "step": 1026
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 48.47371243487692,
+      "learning_rate": 9.397585931272919e-06,
+      "loss": 0.1434,
+      "step": 1027
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 47.87986732416734,
+      "learning_rate": 9.396209869498905e-06,
+      "loss": 0.1599,
+      "step": 1028
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 39.188166800682154,
+      "learning_rate": 9.39483233886946e-06,
+      "loss": 0.1154,
+      "step": 1029
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.175502788260841,
+      "learning_rate": 9.393453339844842e-06,
+      "loss": 0.1465,
+      "step": 1030
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 15.843982783240834,
+      "learning_rate": 9.392072872885802e-06,
+      "loss": 0.1418,
+      "step": 1031
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 51.04924891166709,
+      "learning_rate": 9.39069093845358e-06,
+      "loss": 0.1541,
+      "step": 1032
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 28.092328125663045,
+      "learning_rate": 9.389307537009902e-06,
+      "loss": 0.1537,
+      "step": 1033
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 30.374904328612487,
+      "learning_rate": 9.387922669016992e-06,
+      "loss": 0.1475,
+      "step": 1034
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 16.81604856905053,
+      "learning_rate": 9.386536334937557e-06,
+      "loss": 0.1382,
+      "step": 1035
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 37.54277803011268,
+      "learning_rate": 9.385148535234799e-06,
+      "loss": 0.1389,
+      "step": 1036
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 37.991731569487285,
+      "learning_rate": 9.383759270372408e-06,
+      "loss": 0.1583,
+      "step": 1037
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 43.38772303094576,
+      "learning_rate": 9.382368540814563e-06,
+      "loss": 0.1724,
+      "step": 1038
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 17.689671523576397,
+      "learning_rate": 9.380976347025932e-06,
+      "loss": 0.1157,
+      "step": 1039
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 44.41572273988692,
+      "learning_rate": 9.379582689471671e-06,
+      "loss": 0.1479,
+      "step": 1040
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 24.499163894929794,
+      "learning_rate": 9.378187568617431e-06,
+      "loss": 0.1245,
+      "step": 1041
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 34.745047025595184,
+      "learning_rate": 9.376790984929348e-06,
+      "loss": 0.1395,
+      "step": 1042
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 18.610150565591297,
+      "learning_rate": 9.37539293887404e-06,
+      "loss": 0.1469,
+      "step": 1043
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 29.510494891674,
+      "learning_rate": 9.373993430918626e-06,
+      "loss": 0.1155,
+      "step": 1044
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 26.89748864189952,
+      "learning_rate": 9.372592461530708e-06,
+      "loss": 0.1505,
+      "step": 1045
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.407826158103616,
+      "learning_rate": 9.371190031178372e-06,
+      "loss": 0.1257,
+      "step": 1046
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 20.146410557268332,
+      "learning_rate": 9.369786140330198e-06,
+      "loss": 0.1201,
+      "step": 1047
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.649809459235866,
+      "learning_rate": 9.368380789455251e-06,
+      "loss": 0.1188,
+      "step": 1048
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 20.935495774130015,
+      "learning_rate": 9.36697397902309e-06,
+      "loss": 0.1434,
+      "step": 1049
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 25.501227504620445,
+      "learning_rate": 9.365565709503748e-06,
+      "loss": 0.1395,
+      "step": 1050
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.080004066515803,
+      "learning_rate": 9.364155981367761e-06,
+      "loss": 0.0883,
+      "step": 1051
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.423934524437326,
+      "learning_rate": 9.36274479508614e-06,
+      "loss": 0.1187,
+      "step": 1052
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 34.962167370390645,
+      "learning_rate": 9.361332151130396e-06,
+      "loss": 0.12,
+      "step": 1053
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 21.079610427912726,
+      "learning_rate": 9.359918049972512e-06,
+      "loss": 0.1295,
+      "step": 1054
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 28.618833107684583,
+      "learning_rate": 9.358502492084969e-06,
+      "loss": 0.1395,
+      "step": 1055
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.8759001732284855,
+      "learning_rate": 9.35708547794073e-06,
+      "loss": 0.1281,
+      "step": 1056
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 37.51265565785163,
+      "learning_rate": 9.355667008013249e-06,
+      "loss": 0.1451,
+      "step": 1057
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 63.60202101689444,
+      "learning_rate": 9.354247082776459e-06,
+      "loss": 0.1753,
+      "step": 1058
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 23.159854031756225,
+      "learning_rate": 9.352825702704784e-06,
+      "loss": 0.132,
+      "step": 1059
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 46.848693056166454,
+      "learning_rate": 9.351402868273136e-06,
+      "loss": 0.1409,
+      "step": 1060
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 35.31714858575622,
+      "learning_rate": 9.349978579956908e-06,
+      "loss": 0.1536,
+      "step": 1061
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 39.11227629802165,
+      "learning_rate": 9.348552838231983e-06,
+      "loss": 0.1378,
+      "step": 1062
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 14.341752586653023,
+      "learning_rate": 9.347125643574726e-06,
+      "loss": 0.1119,
+      "step": 1063
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 31.512222693182753,
+      "learning_rate": 9.345696996461992e-06,
+      "loss": 0.1544,
+      "step": 1064
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 44.78047175611448,
+      "learning_rate": 9.344266897371114e-06,
+      "loss": 0.1526,
+      "step": 1065
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 13.860702506152542,
+      "learning_rate": 9.34283534677992e-06,
+      "loss": 0.132,
+      "step": 1066
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.914053857460699,
+      "learning_rate": 9.341402345166714e-06,
+      "loss": 0.146,
+      "step": 1067
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.244521794867931,
+      "learning_rate": 9.33996789301029e-06,
+      "loss": 0.1675,
+      "step": 1068
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 30.55515143164621,
+      "learning_rate": 9.338531990789926e-06,
+      "loss": 0.1207,
+      "step": 1069
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.578710866057671,
+      "learning_rate": 9.33709463898538e-06,
+      "loss": 0.1359,
+      "step": 1070
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.390861205292069,
+      "learning_rate": 9.335655838076902e-06,
+      "loss": 0.0799,
+      "step": 1071
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.0312126352617454,
+      "learning_rate": 9.33421558854522e-06,
+      "loss": 0.1311,
+      "step": 1072
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 16.82866199649571,
+      "learning_rate": 9.332773890871548e-06,
+      "loss": 0.1306,
+      "step": 1073
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 23.34045634748492,
+      "learning_rate": 9.331330745537586e-06,
+      "loss": 0.1274,
+      "step": 1074
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 21.331301913161205,
+      "learning_rate": 9.329886153025513e-06,
+      "loss": 0.1211,
+      "step": 1075
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.711724010388526,
+      "learning_rate": 9.328440113817995e-06,
+      "loss": 0.1337,
+      "step": 1076
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 22.866056035390898,
+      "learning_rate": 9.326992628398182e-06,
+      "loss": 0.1196,
+      "step": 1077
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.234289721717951,
+      "learning_rate": 9.325543697249706e-06,
+      "loss": 0.1526,
+      "step": 1078
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 29.859928319090212,
+      "learning_rate": 9.324093320856679e-06,
+      "loss": 0.137,
+      "step": 1079
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 15.250096335579789,
+      "learning_rate": 9.3226414997037e-06,
+      "loss": 0.116,
+      "step": 1080
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 13.223197640483088,
+      "learning_rate": 9.32118823427585e-06,
+      "loss": 0.1439,
+      "step": 1081
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.3596389314819675,
+      "learning_rate": 9.319733525058694e-06,
+      "loss": 0.1238,
+      "step": 1082
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.286656373449186,
+      "learning_rate": 9.318277372538274e-06,
+      "loss": 0.1317,
+      "step": 1083
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 14.109356967510104,
+      "learning_rate": 9.316819777201119e-06,
+      "loss": 0.1257,
+      "step": 1084
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.223027875056157,
+      "learning_rate": 9.315360739534235e-06,
+      "loss": 0.1648,
+      "step": 1085
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.052529390869049,
+      "learning_rate": 9.313900260025121e-06,
+      "loss": 0.1232,
+      "step": 1086
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 14.816538587879734,
+      "learning_rate": 9.312438339161746e-06,
+      "loss": 0.1198,
+      "step": 1087
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.283898279259918,
+      "learning_rate": 9.310974977432565e-06,
+      "loss": 0.1261,
+      "step": 1088
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 19.36116992418242,
+      "learning_rate": 9.309510175326515e-06,
+      "loss": 0.1118,
+      "step": 1089
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.860358397763082,
+      "learning_rate": 9.308043933333012e-06,
+      "loss": 0.1211,
+      "step": 1090
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.600425083069407,
+      "learning_rate": 9.306576251941957e-06,
+      "loss": 0.1387,
+      "step": 1091
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 49.923681618239925,
+      "learning_rate": 9.305107131643729e-06,
+      "loss": 0.1647,
+      "step": 1092
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.41603756614823,
+      "learning_rate": 9.303636572929188e-06,
+      "loss": 0.1338,
+      "step": 1093
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 15.05350452195723,
+      "learning_rate": 9.302164576289674e-06,
+      "loss": 0.1469,
+      "step": 1094
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 28.33566516662762,
+      "learning_rate": 9.30069114221701e-06,
+      "loss": 0.153,
+      "step": 1095
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 23.243076526598593,
+      "learning_rate": 9.299216271203498e-06,
+      "loss": 0.1603,
+      "step": 1096
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.582545901960945,
+      "learning_rate": 9.297739963741918e-06,
+      "loss": 0.1294,
+      "step": 1097
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.667560754149541,
+      "learning_rate": 9.296262220325535e-06,
+      "loss": 0.11,
+      "step": 1098
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.168707735076364,
+      "learning_rate": 9.294783041448088e-06,
+      "loss": 0.127,
+      "step": 1099
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.57179862351945,
+      "learning_rate": 9.293302427603796e-06,
+      "loss": 0.1414,
+      "step": 1100
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 26.656118865639158,
+      "learning_rate": 9.291820379287364e-06,
+      "loss": 0.1296,
+      "step": 1101
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.172589483208834,
+      "learning_rate": 9.29033689699397e-06,
+      "loss": 0.1492,
+      "step": 1102
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 14.01791519086498,
+      "learning_rate": 9.288851981219273e-06,
+      "loss": 0.1327,
+      "step": 1103
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 33.859914601995875,
+      "learning_rate": 9.28736563245941e-06,
+      "loss": 0.1571,
+      "step": 1104
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 21.956222104782483,
+      "learning_rate": 9.285877851210999e-06,
+      "loss": 0.1,
+      "step": 1105
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 18.10448242341278,
+      "learning_rate": 9.284388637971136e-06,
+      "loss": 0.1234,
+      "step": 1106
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 30.84897554982973,
+      "learning_rate": 9.282897993237392e-06,
+      "loss": 0.1714,
+      "step": 1107
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 37.9856161444381,
+      "learning_rate": 9.281405917507824e-06,
+      "loss": 0.1244,
+      "step": 1108
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 15.563367998567477,
+      "learning_rate": 9.279912411280958e-06,
+      "loss": 0.1194,
+      "step": 1109
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.245185274327808,
+      "learning_rate": 9.278417475055803e-06,
+      "loss": 0.1119,
+      "step": 1110
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.455063691988519,
+      "learning_rate": 9.276921109331845e-06,
+      "loss": 0.1165,
+      "step": 1111
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 25.8213778871432,
+      "learning_rate": 9.275423314609049e-06,
+      "loss": 0.152,
+      "step": 1112
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 13.869919626594394,
+      "learning_rate": 9.273924091387855e-06,
+      "loss": 0.115,
+      "step": 1113
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.074864485432894,
+      "learning_rate": 9.272423440169181e-06,
+      "loss": 0.1091,
+      "step": 1114
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.656458241700619,
+      "learning_rate": 9.270921361454424e-06,
+      "loss": 0.0948,
+      "step": 1115
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 37.09019385402894,
+      "learning_rate": 9.269417855745453e-06,
+      "loss": 0.1337,
+      "step": 1116
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 33.08524762310668,
+      "learning_rate": 9.267912923544621e-06,
+      "loss": 0.1455,
+      "step": 1117
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.26175581668237,
+      "learning_rate": 9.266406565354753e-06,
+      "loss": 0.1221,
+      "step": 1118
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 23.067596741868993,
+      "learning_rate": 9.26489878167915e-06,
+      "loss": 0.1331,
+      "step": 1119
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 28.961300839321403,
+      "learning_rate": 9.263389573021592e-06,
+      "loss": 0.1156,
+      "step": 1120
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 36.58260181012153,
+      "learning_rate": 9.261878939886332e-06,
+      "loss": 0.1499,
+      "step": 1121
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 15.19087787168311,
+      "learning_rate": 9.2603668827781e-06,
+      "loss": 0.1367,
+      "step": 1122
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 20.099032625961726,
+      "learning_rate": 9.258853402202106e-06,
+      "loss": 0.1182,
+      "step": 1123
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 33.55489255269871,
+      "learning_rate": 9.25733849866403e-06,
+      "loss": 0.1436,
+      "step": 1124
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 22.799341496962658,
+      "learning_rate": 9.255822172670028e-06,
+      "loss": 0.119,
+      "step": 1125
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.031052291448315,
+      "learning_rate": 9.254304424726734e-06,
+      "loss": 0.1013,
+      "step": 1126
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.390605015029899,
+      "learning_rate": 9.252785255341256e-06,
+      "loss": 0.1116,
+      "step": 1127
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.188962911089087,
+      "learning_rate": 9.251264665021178e-06,
+      "loss": 0.1255,
+      "step": 1128
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 28.64616641930042,
+      "learning_rate": 9.249742654274554e-06,
+      "loss": 0.1237,
+      "step": 1129
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.921072334184103,
+      "learning_rate": 9.24821922360992e-06,
+      "loss": 0.1051,
+      "step": 1130
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 22.18891754085567,
+      "learning_rate": 9.246694373536277e-06,
+      "loss": 0.1155,
+      "step": 1131
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 22.768049432903226,
+      "learning_rate": 9.245168104563112e-06,
+      "loss": 0.1561,
+      "step": 1132
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.694267206038536,
+      "learning_rate": 9.243640417200376e-06,
+      "loss": 0.1177,
+      "step": 1133
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 19.472208799074128,
+      "learning_rate": 9.242111311958502e-06,
+      "loss": 0.1261,
+      "step": 1134
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.137828446129272,
+      "learning_rate": 9.240580789348385e-06,
+      "loss": 0.1493,
+      "step": 1135
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 17.02804572717195,
+      "learning_rate": 9.23904884988141e-06,
+      "loss": 0.1364,
+      "step": 1136
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.782199051410428,
+      "learning_rate": 9.237515494069417e-06,
+      "loss": 0.1243,
+      "step": 1137
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.708252427463937,
+      "learning_rate": 9.235980722424737e-06,
+      "loss": 0.1117,
+      "step": 1138
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.366900740290782,
+      "learning_rate": 9.234444535460161e-06,
+      "loss": 0.1058,
+      "step": 1139
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.735023185477729,
+      "learning_rate": 9.232906933688959e-06,
+      "loss": 0.1042,
+      "step": 1140
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 17.878356014797415,
+      "learning_rate": 9.231367917624872e-06,
+      "loss": 0.1143,
+      "step": 1141
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 13.500635215475153,
+      "learning_rate": 9.229827487782115e-06,
+      "loss": 0.1472,
+      "step": 1142
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 28.914775597777798,
+      "learning_rate": 9.228285644675372e-06,
+      "loss": 0.1571,
+      "step": 1143
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.679656527001652,
+      "learning_rate": 9.226742388819804e-06,
+      "loss": 0.1243,
+      "step": 1144
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 42.56589275402226,
+      "learning_rate": 9.225197720731039e-06,
+      "loss": 0.1748,
+      "step": 1145
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 33.46124830306206,
+      "learning_rate": 9.223651640925181e-06,
+      "loss": 0.1602,
+      "step": 1146
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.354447656553011,
+      "learning_rate": 9.222104149918804e-06,
+      "loss": 0.1125,
+      "step": 1147
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.943626062124622,
+      "learning_rate": 9.220555248228954e-06,
+      "loss": 0.1055,
+      "step": 1148
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 22.06199413082995,
+      "learning_rate": 9.219004936373146e-06,
+      "loss": 0.1202,
+      "step": 1149
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 78.80684998307608,
+      "learning_rate": 9.217453214869368e-06,
+      "loss": 0.2026,
+      "step": 1150
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 33.723692179365536,
+      "learning_rate": 9.21590008423608e-06,
+      "loss": 0.1237,
+      "step": 1151
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 20.8665304253155,
+      "learning_rate": 9.214345544992214e-06,
+      "loss": 0.136,
+      "step": 1152
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 66.63596872046561,
+      "learning_rate": 9.212789597657167e-06,
+      "loss": 0.1742,
+      "step": 1153
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 39.21810406123508,
+      "learning_rate": 9.21123224275081e-06,
+      "loss": 0.1675,
+      "step": 1154
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 23.06081056803674,
+      "learning_rate": 9.209673480793486e-06,
+      "loss": 0.1375,
+      "step": 1155
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 25.325897571173787,
+      "learning_rate": 9.208113312306006e-06,
+      "loss": 0.1434,
+      "step": 1156
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 52.68852183497776,
+      "learning_rate": 9.206551737809653e-06,
+      "loss": 0.1707,
+      "step": 1157
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.831358394803084,
+      "learning_rate": 9.204988757826173e-06,
+      "loss": 0.1233,
+      "step": 1158
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 16.788606151895603,
+      "learning_rate": 9.203424372877791e-06,
+      "loss": 0.116,
+      "step": 1159
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.314075233780184,
+      "learning_rate": 9.201858583487195e-06,
+      "loss": 0.1305,
+      "step": 1160
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 24.77340581749608,
+      "learning_rate": 9.200291390177546e-06,
+      "loss": 0.1249,
+      "step": 1161
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.838543735694156,
+      "learning_rate": 9.198722793472471e-06,
+      "loss": 0.0964,
+      "step": 1162
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.489793512918954,
+      "learning_rate": 9.197152793896068e-06,
+      "loss": 0.1027,
+      "step": 1163
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 17.370296763714304,
+      "learning_rate": 9.195581391972903e-06,
+      "loss": 0.1373,
+      "step": 1164
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.574790093828888,
+      "learning_rate": 9.194008588228011e-06,
+      "loss": 0.1179,
+      "step": 1165
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.243110764589119,
+      "learning_rate": 9.192434383186894e-06,
+      "loss": 0.1274,
+      "step": 1166
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.676977628339888,
+      "learning_rate": 9.190858777375523e-06,
+      "loss": 0.1256,
+      "step": 1167
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 19.690069281828507,
+      "learning_rate": 9.18928177132034e-06,
+      "loss": 0.1165,
+      "step": 1168
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.038428559464359,
+      "learning_rate": 9.187703365548248e-06,
+      "loss": 0.1094,
+      "step": 1169
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 26.547275502331654,
+      "learning_rate": 9.186123560586623e-06,
+      "loss": 0.2321,
+      "step": 1170
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.193929639465265,
+      "learning_rate": 9.18454235696331e-06,
+      "loss": 0.1295,
+      "step": 1171
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 26.734089366689073,
+      "learning_rate": 9.182959755206613e-06,
+      "loss": 0.1381,
+      "step": 1172
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.039923169051226,
+      "learning_rate": 9.181375755845314e-06,
+      "loss": 0.1196,
+      "step": 1173
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.543614416837632,
+      "learning_rate": 9.179790359408655e-06,
+      "loss": 0.1206,
+      "step": 1174
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.105282043598741,
+      "learning_rate": 9.178203566426344e-06,
+      "loss": 0.1093,
+      "step": 1175
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 20.186482641410777,
+      "learning_rate": 9.176615377428563e-06,
+      "loss": 0.1453,
+      "step": 1176
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.872272725289261,
+      "learning_rate": 9.175025792945951e-06,
+      "loss": 0.0986,
+      "step": 1177
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.738372660616186,
+      "learning_rate": 9.173434813509618e-06,
+      "loss": 0.0973,
+      "step": 1178
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.83532907369495,
+      "learning_rate": 9.171842439651143e-06,
+      "loss": 0.1072,
+      "step": 1179
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.269615641040455,
+      "learning_rate": 9.170248671902565e-06,
+      "loss": 0.106,
+      "step": 1180
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 19.59295084898656,
+      "learning_rate": 9.168653510796392e-06,
+      "loss": 0.1466,
+      "step": 1181
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.073455582971626,
+      "learning_rate": 9.167056956865596e-06,
+      "loss": 0.1572,
+      "step": 1182
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.573875087408759,
+      "learning_rate": 9.165459010643618e-06,
+      "loss": 0.1233,
+      "step": 1183
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.690591395107708,
+      "learning_rate": 9.16385967266436e-06,
+      "loss": 0.111,
+      "step": 1184
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.771275364512612,
+      "learning_rate": 9.16225894346219e-06,
+      "loss": 0.1333,
+      "step": 1185
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 24.721033576277165,
+      "learning_rate": 9.160656823571942e-06,
+      "loss": 0.1638,
+      "step": 1186
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 15.57598172877004,
+      "learning_rate": 9.159053313528913e-06,
+      "loss": 0.1168,
+      "step": 1187
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 26.636854607871165,
+      "learning_rate": 9.15744841386887e-06,
+      "loss": 0.1198,
+      "step": 1188
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.98440431085169,
+      "learning_rate": 9.155842125128033e-06,
+      "loss": 0.1067,
+      "step": 1189
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 31.917676167530633,
+      "learning_rate": 9.154234447843098e-06,
+      "loss": 0.1306,
+      "step": 1190
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.019593781975672,
+      "learning_rate": 9.152625382551217e-06,
+      "loss": 0.0909,
+      "step": 1191
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.806640969390639,
+      "learning_rate": 9.15101492979001e-06,
+      "loss": 0.1046,
+      "step": 1192
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 28.841021287717343,
+      "learning_rate": 9.149403090097557e-06,
+      "loss": 0.1171,
+      "step": 1193
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 16.465965710527545,
+      "learning_rate": 9.147789864012408e-06,
+      "loss": 0.1447,
+      "step": 1194
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 24.666832031038798,
+      "learning_rate": 9.146175252073568e-06,
+      "loss": 0.1456,
+      "step": 1195
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 19.683015165025946,
+      "learning_rate": 9.144559254820511e-06,
+      "loss": 0.1213,
+      "step": 1196
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.122237884315906,
+      "learning_rate": 9.14294187279317e-06,
+      "loss": 0.1357,
+      "step": 1197
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.419084999748459,
+      "learning_rate": 9.141323106531943e-06,
+      "loss": 0.1353,
+      "step": 1198
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 33.4219861921213,
+      "learning_rate": 9.139702956577693e-06,
+      "loss": 0.1405,
+      "step": 1199
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 23.625021627700484,
+      "learning_rate": 9.138081423471736e-06,
+      "loss": 0.1046,
+      "step": 1200
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.839297028784905,
+      "learning_rate": 9.136458507755862e-06,
+      "loss": 0.1106,
+      "step": 1201
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.004864533298464,
+      "learning_rate": 9.134834209972314e-06,
+      "loss": 0.104,
+      "step": 1202
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 34.08869419046535,
+      "learning_rate": 9.133208530663801e-06,
+      "loss": 0.1288,
+      "step": 1203
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 26.979912078853985,
+      "learning_rate": 9.131581470373495e-06,
+      "loss": 0.1449,
+      "step": 1204
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 26.374270596425987,
+      "learning_rate": 9.129953029645022e-06,
+      "loss": 0.1167,
+      "step": 1205
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.895140545361581,
+      "learning_rate": 9.128323209022478e-06,
+      "loss": 0.1267,
+      "step": 1206
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 51.67643816360633,
+      "learning_rate": 9.126692009050415e-06,
+      "loss": 0.1334,
+      "step": 1207
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.7195350284212845,
+      "learning_rate": 9.125059430273848e-06,
+      "loss": 0.1033,
+      "step": 1208
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.68146047091333,
+      "learning_rate": 9.123425473238253e-06,
+      "loss": 0.1194,
+      "step": 1209
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.479614467784797,
+      "learning_rate": 9.121790138489564e-06,
+      "loss": 0.1483,
+      "step": 1210
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.88151888755877,
+      "learning_rate": 9.120153426574177e-06,
+      "loss": 0.1454,
+      "step": 1211
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 21.761167483206385,
+      "learning_rate": 9.118515338038947e-06,
+      "loss": 0.1039,
+      "step": 1212
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 15.054311022200379,
+      "learning_rate": 9.11687587343119e-06,
+      "loss": 0.127,
+      "step": 1213
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.715530741583724,
+      "learning_rate": 9.115235033298682e-06,
+      "loss": 0.1182,
+      "step": 1214
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.72131217945236,
+      "learning_rate": 9.113592818189661e-06,
+      "loss": 0.1331,
+      "step": 1215
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.593220467655973,
+      "learning_rate": 9.111949228652816e-06,
+      "loss": 0.1128,
+      "step": 1216
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.570725755965095,
+      "learning_rate": 9.110304265237304e-06,
+      "loss": 0.1183,
+      "step": 1217
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 17.867725431027747,
+      "learning_rate": 9.10865792849274e-06,
+      "loss": 0.1313,
+      "step": 1218
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 41.74139211185614,
+      "learning_rate": 9.107010218969191e-06,
+      "loss": 0.1473,
+      "step": 1219
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 13.698548574340123,
+      "learning_rate": 9.10536113721719e-06,
+      "loss": 0.1401,
+      "step": 1220
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 56.22024108773497,
+      "learning_rate": 9.103710683787728e-06,
+      "loss": 0.129,
+      "step": 1221
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.143381456660101,
+      "learning_rate": 9.102058859232247e-06,
+      "loss": 0.1459,
+      "step": 1222
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.58414656315145,
+      "learning_rate": 9.100405664102656e-06,
+      "loss": 0.1403,
+      "step": 1223
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 33.73692819416245,
+      "learning_rate": 9.098751098951317e-06,
+      "loss": 0.1604,
+      "step": 1224
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 17.199990598047965,
+      "learning_rate": 9.09709516433105e-06,
+      "loss": 0.1375,
+      "step": 1225
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 15.42145961364144,
+      "learning_rate": 9.095437860795138e-06,
+      "loss": 0.1305,
+      "step": 1226
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 36.82138829966084,
+      "learning_rate": 9.09377918889731e-06,
+      "loss": 0.1387,
+      "step": 1227
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.248193826270128,
+      "learning_rate": 9.092119149191765e-06,
+      "loss": 0.1298,
+      "step": 1228
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.168241999958529,
+      "learning_rate": 9.090457742233152e-06,
+      "loss": 0.0919,
+      "step": 1229
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.413448398731609,
+      "learning_rate": 9.088794968576575e-06,
+      "loss": 0.1368,
+      "step": 1230
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.297298848033256,
+      "learning_rate": 9.087130828777598e-06,
+      "loss": 0.1672,
+      "step": 1231
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.995001961847258,
+      "learning_rate": 9.085465323392243e-06,
+      "loss": 0.1167,
+      "step": 1232
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.999725996925589,
+      "learning_rate": 9.083798452976988e-06,
+      "loss": 0.1295,
+      "step": 1233
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 20.375505744587326,
+      "learning_rate": 9.082130218088762e-06,
+      "loss": 0.1641,
+      "step": 1234
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.932899862057339,
+      "learning_rate": 9.080460619284954e-06,
+      "loss": 0.1132,
+      "step": 1235
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 27.645263230888336,
+      "learning_rate": 9.07878965712341e-06,
+      "loss": 0.1514,
+      "step": 1236
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.21802883875594,
+      "learning_rate": 9.077117332162427e-06,
+      "loss": 0.1427,
+      "step": 1237
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 18.322663661950923,
+      "learning_rate": 9.075443644960761e-06,
+      "loss": 0.1166,
+      "step": 1238
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.316841214679732,
+      "learning_rate": 9.07376859607762e-06,
+      "loss": 0.0997,
+      "step": 1239
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.244511662257982,
+      "learning_rate": 9.072092186072675e-06,
+      "loss": 0.1416,
+      "step": 1240
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 19.860679127095636,
+      "learning_rate": 9.070414415506038e-06,
+      "loss": 0.1356,
+      "step": 1241
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.605522811610166,
+      "learning_rate": 9.068735284938288e-06,
+      "loss": 0.1052,
+      "step": 1242
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 17.842863169949126,
+      "learning_rate": 9.067054794930452e-06,
+      "loss": 0.1169,
+      "step": 1243
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.718865016595677,
+      "learning_rate": 9.065372946044014e-06,
+      "loss": 0.1428,
+      "step": 1244
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 20.576896857972773,
+      "learning_rate": 9.063689738840911e-06,
+      "loss": 0.1407,
+      "step": 1245
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.3907765645746935,
+      "learning_rate": 9.06200517388353e-06,
+      "loss": 0.1586,
+      "step": 1246
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.412682061788436,
+      "learning_rate": 9.060319251734723e-06,
+      "loss": 0.1168,
+      "step": 1247
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.216748606683582,
+      "learning_rate": 9.058631972957783e-06,
+      "loss": 0.179,
+      "step": 1248
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 13.251080346982283,
+      "learning_rate": 9.056943338116461e-06,
+      "loss": 0.1057,
+      "step": 1249
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 13.556345489969091,
+      "learning_rate": 9.055253347774961e-06,
+      "loss": 0.1252,
+      "step": 1250
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.70070366178762,
+      "learning_rate": 9.053562002497943e-06,
+      "loss": 0.121,
+      "step": 1251
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.542057911871553,
+      "learning_rate": 9.051869302850515e-06,
+      "loss": 0.0962,
+      "step": 1252
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 16.718582711323283,
+      "learning_rate": 9.05017524939824e-06,
+      "loss": 0.1232,
+      "step": 1253
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 13.719584747865103,
+      "learning_rate": 9.048479842707132e-06,
+      "loss": 0.1292,
+      "step": 1254
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.570019714838347,
+      "learning_rate": 9.046783083343657e-06,
+      "loss": 0.0828,
+      "step": 1255
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.629289413960565,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.1039,
+      "step": 1256
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.7497642996638705,
+      "learning_rate": 9.043385508867741e-06,
+      "loss": 0.0898,
+      "step": 1257
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 20.039832970250274,
+      "learning_rate": 9.041684694890492e-06,
+      "loss": 0.1107,
+      "step": 1258
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 15.178772636850542,
+      "learning_rate": 9.03998253051126e-06,
+      "loss": 0.1654,
+      "step": 1259
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 27.037718152114405,
+      "learning_rate": 9.038279016298773e-06,
+      "loss": 0.1035,
+      "step": 1260
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 28.49390797177503,
+      "learning_rate": 9.036574152822206e-06,
+      "loss": 0.1362,
+      "step": 1261
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.695671686306866,
+      "learning_rate": 9.034867940651186e-06,
+      "loss": 0.1486,
+      "step": 1262
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.332210434139537,
+      "learning_rate": 9.033160380355789e-06,
+      "loss": 0.1077,
+      "step": 1263
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.95118182352966,
+      "learning_rate": 9.031451472506544e-06,
+      "loss": 0.1095,
+      "step": 1264
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.219397289040563,
+      "learning_rate": 9.029741217674428e-06,
+      "loss": 0.1373,
+      "step": 1265
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 28.958428037574063,
+      "learning_rate": 9.02802961643087e-06,
+      "loss": 0.1364,
+      "step": 1266
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 27.25383513627309,
+      "learning_rate": 9.026316669347747e-06,
+      "loss": 0.1401,
+      "step": 1267
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.0655709620800975,
+      "learning_rate": 9.024602376997387e-06,
+      "loss": 0.1539,
+      "step": 1268
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 45.88384945997455,
+      "learning_rate": 9.022886739952565e-06,
+      "loss": 0.1285,
+      "step": 1269
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 26.66446887230687,
+      "learning_rate": 9.02116975878651e-06,
+      "loss": 0.1138,
+      "step": 1270
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.862633449690026,
+      "learning_rate": 9.019451434072894e-06,
+      "loss": 0.1599,
+      "step": 1271
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.445395355203088,
+      "learning_rate": 9.017731766385844e-06,
+      "loss": 0.1327,
+      "step": 1272
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.717224738562036,
+      "learning_rate": 9.016010756299934e-06,
+      "loss": 0.1062,
+      "step": 1273
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 33.84566034930815,
+      "learning_rate": 9.014288404390182e-06,
+      "loss": 0.1376,
+      "step": 1274
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 17.609479593872077,
+      "learning_rate": 9.012564711232059e-06,
+      "loss": 0.1116,
+      "step": 1275
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.705629267795132,
+      "learning_rate": 9.010839677401484e-06,
+      "loss": 0.1307,
+      "step": 1276
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4202600275574486,
+      "learning_rate": 9.009113303474822e-06,
+      "loss": 0.0847,
+      "step": 1277
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 27.46790309202585,
+      "learning_rate": 9.007385590028887e-06,
+      "loss": 0.1199,
+      "step": 1278
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 19.134070635982145,
+      "learning_rate": 9.005656537640942e-06,
+      "loss": 0.1385,
+      "step": 1279
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.867626401898946,
+      "learning_rate": 9.003926146888691e-06,
+      "loss": 0.098,
+      "step": 1280
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.301604393023416,
+      "learning_rate": 9.002194418350291e-06,
+      "loss": 0.1766,
+      "step": 1281
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.47168882178934,
+      "learning_rate": 9.000461352604349e-06,
+      "loss": 0.1528,
+      "step": 1282
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 20.94266276844876,
+      "learning_rate": 8.99872695022991e-06,
+      "loss": 0.13,
+      "step": 1283
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.568146188671397,
+      "learning_rate": 8.996991211806471e-06,
+      "loss": 0.1041,
+      "step": 1284
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 19.767793790726582,
+      "learning_rate": 8.995254137913977e-06,
+      "loss": 0.1321,
+      "step": 1285
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.925767904518081,
+      "learning_rate": 8.99351572913281e-06,
+      "loss": 0.1204,
+      "step": 1286
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 17.665286809508974,
+      "learning_rate": 8.991775986043814e-06,
+      "loss": 0.0851,
+      "step": 1287
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 15.691280452120386,
+      "learning_rate": 8.990034909228262e-06,
+      "loss": 0.1456,
+      "step": 1288
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.619640637451463,
+      "learning_rate": 8.988292499267885e-06,
+      "loss": 0.0916,
+      "step": 1289
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 39.48387258198995,
+      "learning_rate": 8.986548756744852e-06,
+      "loss": 0.1195,
+      "step": 1290
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.63610318578367,
+      "learning_rate": 8.98480368224178e-06,
+      "loss": 0.1492,
+      "step": 1291
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 20.03898408603859,
+      "learning_rate": 8.98305727634173e-06,
+      "loss": 0.1219,
+      "step": 1292
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.857773381391331,
+      "learning_rate": 8.981309539628212e-06,
+      "loss": 0.132,
+      "step": 1293
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.80086758659809,
+      "learning_rate": 8.979560472685174e-06,
+      "loss": 0.1019,
+      "step": 1294
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.467420571204822,
+      "learning_rate": 8.977810076097013e-06,
+      "loss": 0.0927,
+      "step": 1295
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.798830013093253,
+      "learning_rate": 8.97605835044857e-06,
+      "loss": 0.1462,
+      "step": 1296
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3782928073977945,
+      "learning_rate": 8.974305296325125e-06,
+      "loss": 0.0842,
+      "step": 1297
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.272978733091367,
+      "learning_rate": 8.97255091431241e-06,
+      "loss": 0.1271,
+      "step": 1298
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 16.81966484700589,
+      "learning_rate": 8.970795204996597e-06,
+      "loss": 0.1229,
+      "step": 1299
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.739832847910312,
+      "learning_rate": 8.969038168964298e-06,
+      "loss": 0.1266,
+      "step": 1300
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.173928247267657,
+      "learning_rate": 8.967279806802576e-06,
+      "loss": 0.1152,
+      "step": 1301
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 22.14994655091031,
+      "learning_rate": 8.965520119098926e-06,
+      "loss": 0.1097,
+      "step": 1302
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 54.5742266393221,
+      "learning_rate": 8.9637591064413e-06,
+      "loss": 0.1621,
+      "step": 1303
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.6238048135336,
+      "learning_rate": 8.961996769418077e-06,
+      "loss": 0.0958,
+      "step": 1304
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 29.088793120301332,
+      "learning_rate": 8.960233108618092e-06,
+      "loss": 0.1401,
+      "step": 1305
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 51.52087793146708,
+      "learning_rate": 8.958468124630617e-06,
+      "loss": 0.1482,
+      "step": 1306
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 26.746196577981273,
+      "learning_rate": 8.956701818045363e-06,
+      "loss": 0.1227,
+      "step": 1307
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 35.98626766193302,
+      "learning_rate": 8.954934189452489e-06,
+      "loss": 0.1271,
+      "step": 1308
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.706810349285421,
+      "learning_rate": 8.953165239442589e-06,
+      "loss": 0.1436,
+      "step": 1309
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 17.689232861773036,
+      "learning_rate": 8.951394968606704e-06,
+      "loss": 0.1155,
+      "step": 1310
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.4873562096415744,
+      "learning_rate": 8.949623377536314e-06,
+      "loss": 0.1095,
+      "step": 1311
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.977404966444077,
+      "learning_rate": 8.947850466823343e-06,
+      "loss": 0.0917,
+      "step": 1312
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.150139134114927,
+      "learning_rate": 8.946076237060148e-06,
+      "loss": 0.1312,
+      "step": 1313
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.547066209117817,
+      "learning_rate": 8.944300688839538e-06,
+      "loss": 0.1211,
+      "step": 1314
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.574430114106702,
+      "learning_rate": 8.942523822754751e-06,
+      "loss": 0.1184,
+      "step": 1315
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.767237626071879,
+      "learning_rate": 8.940745639399477e-06,
+      "loss": 0.1439,
+      "step": 1316
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 18.217321006016,
+      "learning_rate": 8.938966139367837e-06,
+      "loss": 0.124,
+      "step": 1317
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.12123329875439,
+      "learning_rate": 8.937185323254395e-06,
+      "loss": 0.0945,
+      "step": 1318
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 20.992000622397523,
+      "learning_rate": 8.935403191654155e-06,
+      "loss": 0.0821,
+      "step": 1319
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 20.121514370471925,
+      "learning_rate": 8.933619745162559e-06,
+      "loss": 0.1375,
+      "step": 1320
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 19.672716362914535,
+      "learning_rate": 8.931834984375492e-06,
+      "loss": 0.1133,
+      "step": 1321
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.854511996956678,
+      "learning_rate": 8.930048909889272e-06,
+      "loss": 0.1428,
+      "step": 1322
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.981569840262182,
+      "learning_rate": 8.928261522300665e-06,
+      "loss": 0.0844,
+      "step": 1323
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 20.95789494137274,
+      "learning_rate": 8.926472822206869e-06,
+      "loss": 0.1074,
+      "step": 1324
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.085896564148001,
+      "learning_rate": 8.924682810205519e-06,
+      "loss": 0.1525,
+      "step": 1325
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.93521143384227,
+      "learning_rate": 8.922891486894692e-06,
+      "loss": 0.1149,
+      "step": 1326
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.576240348621916,
+      "learning_rate": 8.921098852872904e-06,
+      "loss": 0.123,
+      "step": 1327
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.371219220183408,
+      "learning_rate": 8.919304908739106e-06,
+      "loss": 0.1293,
+      "step": 1328
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.67584580080675,
+      "learning_rate": 8.917509655092691e-06,
+      "loss": 0.1666,
+      "step": 1329
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.493405850966914,
+      "learning_rate": 8.915713092533483e-06,
+      "loss": 0.1056,
+      "step": 1330
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 13.499498266925373,
+      "learning_rate": 8.913915221661748e-06,
+      "loss": 0.1012,
+      "step": 1331
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 30.9937070736886,
+      "learning_rate": 8.912116043078188e-06,
+      "loss": 0.1466,
+      "step": 1332
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 14.4905788898255,
+      "learning_rate": 8.910315557383944e-06,
+      "loss": 0.1387,
+      "step": 1333
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.442340648313488,
+      "learning_rate": 8.90851376518059e-06,
+      "loss": 0.1321,
+      "step": 1334
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 19.692706543472003,
+      "learning_rate": 8.906710667070136e-06,
+      "loss": 0.1663,
+      "step": 1335
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.236348097631177,
+      "learning_rate": 8.904906263655036e-06,
+      "loss": 0.1521,
+      "step": 1336
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.189125432366566,
+      "learning_rate": 8.903100555538169e-06,
+      "loss": 0.1282,
+      "step": 1337
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.711104406960347,
+      "learning_rate": 8.90129354332286e-06,
+      "loss": 0.146,
+      "step": 1338
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 33.786095997393936,
+      "learning_rate": 8.899485227612865e-06,
+      "loss": 0.1194,
+      "step": 1339
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.881201085589049,
+      "learning_rate": 8.897675609012372e-06,
+      "loss": 0.1199,
+      "step": 1340
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.300747945263061,
+      "learning_rate": 8.895864688126013e-06,
+      "loss": 0.0984,
+      "step": 1341
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 19.177339508083687,
+      "learning_rate": 8.894052465558846e-06,
+      "loss": 0.0996,
+      "step": 1342
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.531710013911395,
+      "learning_rate": 8.892238941916372e-06,
+      "loss": 0.1389,
+      "step": 1343
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.131493938719796,
+      "learning_rate": 8.890424117804522e-06,
+      "loss": 0.1129,
+      "step": 1344
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 12.104920527712904,
+      "learning_rate": 8.88860799382966e-06,
+      "loss": 0.1772,
+      "step": 1345
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 22.119517879117527,
+      "learning_rate": 8.88679057059859e-06,
+      "loss": 0.1145,
+      "step": 1346
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.741445859808426,
+      "learning_rate": 8.884971848718544e-06,
+      "loss": 0.1284,
+      "step": 1347
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.598276011199355,
+      "learning_rate": 8.883151828797194e-06,
+      "loss": 0.1213,
+      "step": 1348
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 19.888338261433105,
+      "learning_rate": 8.88133051144264e-06,
+      "loss": 0.1331,
+      "step": 1349
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 26.720524373083247,
+      "learning_rate": 8.87950789726342e-06,
+      "loss": 0.1427,
+      "step": 1350
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.296969295372449,
+      "learning_rate": 8.8776839868685e-06,
+      "loss": 0.1285,
+      "step": 1351
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.519639013034927,
+      "learning_rate": 8.875858780867286e-06,
+      "loss": 0.1084,
+      "step": 1352
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.140053963785225,
+      "learning_rate": 8.87403227986961e-06,
+      "loss": 0.1273,
+      "step": 1353
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.666680845030289,
+      "learning_rate": 8.872204484485743e-06,
+      "loss": 0.1301,
+      "step": 1354
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 32.518052200093564,
+      "learning_rate": 8.870375395326384e-06,
+      "loss": 0.1344,
+      "step": 1355
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 24.980724053437154,
+      "learning_rate": 8.868545013002665e-06,
+      "loss": 0.1077,
+      "step": 1356
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 15.549275682764785,
+      "learning_rate": 8.866713338126152e-06,
+      "loss": 0.13,
+      "step": 1357
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.581867589291187,
+      "learning_rate": 8.86488037130884e-06,
+      "loss": 0.13,
+      "step": 1358
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 13.682883003986115,
+      "learning_rate": 8.863046113163158e-06,
+      "loss": 0.0698,
+      "step": 1359
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 67.34606591529005,
+      "learning_rate": 8.861210564301967e-06,
+      "loss": 0.2075,
+      "step": 1360
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 16.000577042347064,
+      "learning_rate": 8.859373725338558e-06,
+      "loss": 0.1465,
+      "step": 1361
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.734984778309666,
+      "learning_rate": 8.857535596886652e-06,
+      "loss": 0.1035,
+      "step": 1362
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 19.14624542885837,
+      "learning_rate": 8.855696179560402e-06,
+      "loss": 0.1437,
+      "step": 1363
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 72.76158246694781,
+      "learning_rate": 8.85385547397439e-06,
+      "loss": 0.1649,
+      "step": 1364
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.2260099295276845,
+      "learning_rate": 8.852013480743632e-06,
+      "loss": 0.1321,
+      "step": 1365
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.131312974812547,
+      "learning_rate": 8.850170200483573e-06,
+      "loss": 0.1362,
+      "step": 1366
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.780427057540104,
+      "learning_rate": 8.848325633810083e-06,
+      "loss": 0.0975,
+      "step": 1367
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 38.41930656679251,
+      "learning_rate": 8.84647978133947e-06,
+      "loss": 0.1438,
+      "step": 1368
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 31.457034993977672,
+      "learning_rate": 8.844632643688467e-06,
+      "loss": 0.1525,
+      "step": 1369
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.017217356990798,
+      "learning_rate": 8.842784221474237e-06,
+      "loss": 0.1115,
+      "step": 1370
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 15.961687118243042,
+      "learning_rate": 8.840934515314372e-06,
+      "loss": 0.173,
+      "step": 1371
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 35.13831841169741,
+      "learning_rate": 8.839083525826893e-06,
+      "loss": 0.1478,
+      "step": 1372
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 56.57243122107551,
+      "learning_rate": 8.837231253630247e-06,
+      "loss": 0.1528,
+      "step": 1373
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.078243066957646,
+      "learning_rate": 8.835377699343318e-06,
+      "loss": 0.1129,
+      "step": 1374
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.098116188214989,
+      "learning_rate": 8.83352286358541e-06,
+      "loss": 0.1077,
+      "step": 1375
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 27.118006043141744,
+      "learning_rate": 8.83166674697626e-06,
+      "loss": 0.1112,
+      "step": 1376
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 35.911096298188546,
+      "learning_rate": 8.829809350136027e-06,
+      "loss": 0.1365,
+      "step": 1377
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 25.339967860287878,
+      "learning_rate": 8.827950673685306e-06,
+      "loss": 0.1319,
+      "step": 1378
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.559736283305232,
+      "learning_rate": 8.826090718245112e-06,
+      "loss": 0.1271,
+      "step": 1379
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.481495463900032,
+      "learning_rate": 8.824229484436894e-06,
+      "loss": 0.1123,
+      "step": 1380
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.710361054334273,
+      "learning_rate": 8.822366972882523e-06,
+      "loss": 0.1602,
+      "step": 1381
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 46.045622161560836,
+      "learning_rate": 8.820503184204299e-06,
+      "loss": 0.1102,
+      "step": 1382
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 57.381890629821335,
+      "learning_rate": 8.818638119024949e-06,
+      "loss": 0.1418,
+      "step": 1383
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 32.56557318410978,
+      "learning_rate": 8.816771777967623e-06,
+      "loss": 0.1338,
+      "step": 1384
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 59.30261725642934,
+      "learning_rate": 8.814904161655904e-06,
+      "loss": 0.1843,
+      "step": 1385
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 55.152942998068376,
+      "learning_rate": 8.813035270713796e-06,
+      "loss": 0.144,
+      "step": 1386
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 54.60066261642631,
+      "learning_rate": 8.811165105765732e-06,
+      "loss": 0.14,
+      "step": 1387
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.534246320273303,
+      "learning_rate": 8.809293667436565e-06,
+      "loss": 0.124,
+      "step": 1388
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 25.51275813460598,
+      "learning_rate": 8.80742095635158e-06,
+      "loss": 0.1329,
+      "step": 1389
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 41.23937073867847,
+      "learning_rate": 8.805546973136481e-06,
+      "loss": 0.139,
+      "step": 1390
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 44.99322436803483,
+      "learning_rate": 8.803671718417407e-06,
+      "loss": 0.1451,
+      "step": 1391
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 19.302534314720663,
+      "learning_rate": 8.80179519282091e-06,
+      "loss": 0.1173,
+      "step": 1392
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.692753538243162,
+      "learning_rate": 8.799917396973976e-06,
+      "loss": 0.0931,
+      "step": 1393
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 55.478799678782806,
+      "learning_rate": 8.798038331504008e-06,
+      "loss": 0.1469,
+      "step": 1394
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 51.060808138084866,
+      "learning_rate": 8.79615799703884e-06,
+      "loss": 0.1503,
+      "step": 1395
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 18.71241704458466,
+      "learning_rate": 8.794276394206722e-06,
+      "loss": 0.1154,
+      "step": 1396
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 20.47077129185746,
+      "learning_rate": 8.792393523636337e-06,
+      "loss": 0.1069,
+      "step": 1397
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 50.377448768013004,
+      "learning_rate": 8.790509385956784e-06,
+      "loss": 0.1875,
+      "step": 1398
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 24.38834863041375,
+      "learning_rate": 8.788623981797592e-06,
+      "loss": 0.145,
+      "step": 1399
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 18.459905174452402,
+      "learning_rate": 8.786737311788708e-06,
+      "loss": 0.1196,
+      "step": 1400
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.4253124318053634,
+      "learning_rate": 8.784849376560503e-06,
+      "loss": 0.0878,
+      "step": 1401
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 32.20871321310459,
+      "learning_rate": 8.78296017674377e-06,
+      "loss": 0.0911,
+      "step": 1402
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 29.868559521129765,
+      "learning_rate": 8.781069712969726e-06,
+      "loss": 0.0909,
+      "step": 1403
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 28.47627393710747,
+      "learning_rate": 8.779177985870012e-06,
+      "loss": 0.0869,
+      "step": 1404
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.353515032854893,
+      "learning_rate": 8.77728499607669e-06,
+      "loss": 0.0629,
+      "step": 1405
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 23.98867948652184,
+      "learning_rate": 8.775390744222238e-06,
+      "loss": 0.1105,
+      "step": 1406
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 20.72490090675031,
+      "learning_rate": 8.773495230939567e-06,
+      "loss": 0.0758,
+      "step": 1407
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 28.331372328309556,
+      "learning_rate": 8.771598456861998e-06,
+      "loss": 0.0789,
+      "step": 1408
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.026063050859235,
+      "learning_rate": 8.769700422623283e-06,
+      "loss": 0.0461,
+      "step": 1409
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 26.292659193113025,
+      "learning_rate": 8.767801128857588e-06,
+      "loss": 0.1025,
+      "step": 1410
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 7.794861855473636,
+      "learning_rate": 8.765900576199502e-06,
+      "loss": 0.0879,
+      "step": 1411
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 18.682338368358316,
+      "learning_rate": 8.763998765284036e-06,
+      "loss": 0.0944,
+      "step": 1412
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 11.412070873783431,
+      "learning_rate": 8.76209569674662e-06,
+      "loss": 0.0685,
+      "step": 1413
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 10.884099960069042,
+      "learning_rate": 8.760191371223104e-06,
+      "loss": 0.0643,
+      "step": 1414
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 9.034014941112474,
+      "learning_rate": 8.758285789349759e-06,
+      "loss": 0.1039,
+      "step": 1415
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 7.709293674256186,
+      "learning_rate": 8.756378951763277e-06,
+      "loss": 0.093,
+      "step": 1416
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.1454163736142275,
+      "learning_rate": 8.754470859100765e-06,
+      "loss": 0.0861,
+      "step": 1417
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 17.634562346783152,
+      "learning_rate": 8.752561511999754e-06,
+      "loss": 0.0907,
+      "step": 1418
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 5.054823596253129,
+      "learning_rate": 8.750650911098193e-06,
+      "loss": 0.0757,
+      "step": 1419
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 10.241226207922915,
+      "learning_rate": 8.748739057034447e-06,
+      "loss": 0.081,
+      "step": 1420
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.495722997870084,
+      "learning_rate": 8.746825950447302e-06,
+      "loss": 0.0734,
+      "step": 1421
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 17.82904323610015,
+      "learning_rate": 8.744911591975967e-06,
+      "loss": 0.0651,
+      "step": 1422
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 17.484241379617817,
+      "learning_rate": 8.742995982260059e-06,
+      "loss": 0.0819,
+      "step": 1423
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 18.69316674886684,
+      "learning_rate": 8.741079121939621e-06,
+      "loss": 0.0981,
+      "step": 1424
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.329418925647879,
+      "learning_rate": 8.739161011655113e-06,
+      "loss": 0.087,
+      "step": 1425
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.751184813489952,
+      "learning_rate": 8.737241652047408e-06,
+      "loss": 0.086,
+      "step": 1426
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 6.326549122631243,
+      "learning_rate": 8.735321043757805e-06,
+      "loss": 0.0751,
+      "step": 1427
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 10.632712593121198,
+      "learning_rate": 8.73339918742801e-06,
+      "loss": 0.0657,
+      "step": 1428
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 15.207265708492715,
+      "learning_rate": 8.731476083700154e-06,
+      "loss": 0.117,
+      "step": 1429
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 11.850694276149138,
+      "learning_rate": 8.729551733216779e-06,
+      "loss": 0.0742,
+      "step": 1430
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 12.467249641043043,
+      "learning_rate": 8.727626136620848e-06,
+      "loss": 0.0967,
+      "step": 1431
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 16.67115883838518,
+      "learning_rate": 8.725699294555739e-06,
+      "loss": 0.0746,
+      "step": 1432
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 21.221164427537694,
+      "learning_rate": 8.723771207665245e-06,
+      "loss": 0.0936,
+      "step": 1433
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.9808876084963707,
+      "learning_rate": 8.721841876593576e-06,
+      "loss": 0.0742,
+      "step": 1434
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 13.177543099730642,
+      "learning_rate": 8.719911301985355e-06,
+      "loss": 0.0726,
+      "step": 1435
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 5.383309194526871,
+      "learning_rate": 8.717979484485628e-06,
+      "loss": 0.0803,
+      "step": 1436
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 8.67535542961095,
+      "learning_rate": 8.716046424739845e-06,
+      "loss": 0.0964,
+      "step": 1437
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 23.546000226981334,
+      "learning_rate": 8.714112123393882e-06,
+      "loss": 0.0906,
+      "step": 1438
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 5.478616913562366,
+      "learning_rate": 8.712176581094025e-06,
+      "loss": 0.1,
+      "step": 1439
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 12.072295746434312,
+      "learning_rate": 8.710239798486972e-06,
+      "loss": 0.1013,
+      "step": 1440
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 6.156588542394663,
+      "learning_rate": 8.708301776219838e-06,
+      "loss": 0.0717,
+      "step": 1441
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 7.03749840440953,
+      "learning_rate": 8.706362514940153e-06,
+      "loss": 0.071,
+      "step": 1442
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 21.908593717607186,
+      "learning_rate": 8.704422015295861e-06,
+      "loss": 0.0908,
+      "step": 1443
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 12.743381450833912,
+      "learning_rate": 8.702480277935319e-06,
+      "loss": 0.093,
+      "step": 1444
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 12.237240562039307,
+      "learning_rate": 8.700537303507298e-06,
+      "loss": 0.0611,
+      "step": 1445
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 31.205512369091654,
+      "learning_rate": 8.69859309266098e-06,
+      "loss": 0.1085,
+      "step": 1446
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 6.508224635069914,
+      "learning_rate": 8.696647646045962e-06,
+      "loss": 0.0961,
+      "step": 1447
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 6.020337116432563,
+      "learning_rate": 8.694700964312257e-06,
+      "loss": 0.1194,
+      "step": 1448
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 11.85315501709236,
+      "learning_rate": 8.692753048110283e-06,
+      "loss": 0.1057,
+      "step": 1449
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 16.76312648328666,
+      "learning_rate": 8.690803898090878e-06,
+      "loss": 0.0859,
+      "step": 1450
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 18.181912184082726,
+      "learning_rate": 8.68885351490529e-06,
+      "loss": 0.0883,
+      "step": 1451
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 15.561501662544323,
+      "learning_rate": 8.686901899205177e-06,
+      "loss": 0.0615,
+      "step": 1452
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 10.141158740015651,
+      "learning_rate": 8.684949051642609e-06,
+      "loss": 0.0827,
+      "step": 1453
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.131395886113488,
+      "learning_rate": 8.68299497287007e-06,
+      "loss": 0.0664,
+      "step": 1454
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 35.18287088578419,
+      "learning_rate": 8.681039663540454e-06,
+      "loss": 0.0862,
+      "step": 1455
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 6.000912763873758,
+      "learning_rate": 8.679083124307064e-06,
+      "loss": 0.0785,
+      "step": 1456
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.954545314588621,
+      "learning_rate": 8.67712535582362e-06,
+      "loss": 0.0663,
+      "step": 1457
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 16.363053312350825,
+      "learning_rate": 8.675166358744247e-06,
+      "loss": 0.0945,
+      "step": 1458
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 32.23364016473821,
+      "learning_rate": 8.67320613372348e-06,
+      "loss": 0.1357,
+      "step": 1459
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 19.06812434731879,
+      "learning_rate": 8.67124468141627e-06,
+      "loss": 0.0738,
+      "step": 1460
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.336167974534424,
+      "learning_rate": 8.669282002477975e-06,
+      "loss": 0.116,
+      "step": 1461
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 30.216375689397136,
+      "learning_rate": 8.66731809756436e-06,
+      "loss": 0.0871,
+      "step": 1462
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 9.454651839939022,
+      "learning_rate": 8.665352967331604e-06,
+      "loss": 0.0981,
+      "step": 1463
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.880572461820449,
+      "learning_rate": 8.66338661243629e-06,
+      "loss": 0.0784,
+      "step": 1464
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 48.047477471559795,
+      "learning_rate": 8.661419033535419e-06,
+      "loss": 0.1255,
+      "step": 1465
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.296796150731295,
+      "learning_rate": 8.659450231286392e-06,
+      "loss": 0.0761,
+      "step": 1466
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 23.57114189695642,
+      "learning_rate": 8.657480206347024e-06,
+      "loss": 0.0864,
+      "step": 1467
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 25.506822238745904,
+      "learning_rate": 8.655508959375536e-06,
+      "loss": 0.0881,
+      "step": 1468
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 8.686210935110786,
+      "learning_rate": 8.653536491030559e-06,
+      "loss": 0.0715,
+      "step": 1469
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 7.926723814269391,
+      "learning_rate": 8.651562801971131e-06,
+      "loss": 0.0723,
+      "step": 1470
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 11.287008911087467,
+      "learning_rate": 8.649587892856698e-06,
+      "loss": 0.0819,
+      "step": 1471
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 17.18659129208043,
+      "learning_rate": 8.647611764347114e-06,
+      "loss": 0.0782,
+      "step": 1472
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 7.819540078505,
+      "learning_rate": 8.64563441710264e-06,
+      "loss": 0.1422,
+      "step": 1473
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 17.558856807362183,
+      "learning_rate": 8.643655851783947e-06,
+      "loss": 0.0733,
+      "step": 1474
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 16.51338622173928,
+      "learning_rate": 8.641676069052104e-06,
+      "loss": 0.0606,
+      "step": 1475
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 7.477740828151368,
+      "learning_rate": 8.639695069568602e-06,
+      "loss": 0.0682,
+      "step": 1476
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 14.973985830755504,
+      "learning_rate": 8.637712853995324e-06,
+      "loss": 0.105,
+      "step": 1477
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 15.603054833838772,
+      "learning_rate": 8.635729422994566e-06,
+      "loss": 0.087,
+      "step": 1478
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.290026085929057,
+      "learning_rate": 8.633744777229029e-06,
+      "loss": 0.0738,
+      "step": 1479
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.7390029716415563,
+      "learning_rate": 8.63175891736182e-06,
+      "loss": 0.0673,
+      "step": 1480
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 5.473545590164696,
+      "learning_rate": 8.629771844056452e-06,
+      "loss": 0.0686,
+      "step": 1481
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 14.272289047677544,
+      "learning_rate": 8.627783557976846e-06,
+      "loss": 0.0627,
+      "step": 1482
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 12.905709838650964,
+      "learning_rate": 8.62579405978732e-06,
+      "loss": 0.0811,
+      "step": 1483
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 13.0467547255787,
+      "learning_rate": 8.623803350152606e-06,
+      "loss": 0.0734,
+      "step": 1484
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.523481106503748,
+      "learning_rate": 8.621811429737837e-06,
+      "loss": 0.0906,
+      "step": 1485
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 26.36033723371387,
+      "learning_rate": 8.619818299208548e-06,
+      "loss": 0.0687,
+      "step": 1486
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 13.671703283276997,
+      "learning_rate": 8.617823959230683e-06,
+      "loss": 0.0829,
+      "step": 1487
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 13.44761997098398,
+      "learning_rate": 8.615828410470589e-06,
+      "loss": 0.0809,
+      "step": 1488
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 12.38686214081961,
+      "learning_rate": 8.613831653595013e-06,
+      "loss": 0.1403,
+      "step": 1489
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 29.543403014919598,
+      "learning_rate": 8.61183368927111e-06,
+      "loss": 0.0684,
+      "step": 1490
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 9.143382134106462,
+      "learning_rate": 8.609834518166439e-06,
+      "loss": 0.0809,
+      "step": 1491
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 14.121857063675124,
+      "learning_rate": 8.607834140948958e-06,
+      "loss": 0.0721,
+      "step": 1492
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 14.681563500765165,
+      "learning_rate": 8.60583255828703e-06,
+      "loss": 0.0598,
+      "step": 1493
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 11.402342240873446,
+      "learning_rate": 8.603829770849421e-06,
+      "loss": 0.0837,
+      "step": 1494
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 34.80956350140246,
+      "learning_rate": 8.601825779305302e-06,
+      "loss": 0.0812,
+      "step": 1495
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 25.664482199439714,
+      "learning_rate": 8.59982058432424e-06,
+      "loss": 0.0872,
+      "step": 1496
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.942475728354198,
+      "learning_rate": 8.597814186576212e-06,
+      "loss": 0.0752,
+      "step": 1497
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 21.889841226015804,
+      "learning_rate": 8.595806586731589e-06,
+      "loss": 0.0869,
+      "step": 1498
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 47.39501041858811,
+      "learning_rate": 8.59379778546115e-06,
+      "loss": 0.1328,
+      "step": 1499
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 23.843981585654273,
+      "learning_rate": 8.591787783436073e-06,
+      "loss": 0.0833,
+      "step": 1500
+    },
+    {
+      "epoch": 1.07,
+      "eval_avg_AUC": 0.8187337540188822,
+      "eval_avg_Accuracy": 0.7427470159151194,
+      "eval_avg_Accuracy-right": 0.8404199817399244,
+      "eval_avg_Accuracy-wrong": 0.5724357516488515,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6880919420787215,
+      "eval_last_AUC": 0.8358783930516435,
+      "eval_last_Accuracy": 0.7628066976127321,
+      "eval_last_Accuracy-right": 0.8261379940002609,
+      "eval_last_Accuracy-wrong": 0.6523766204230157,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.7015684504151366,
+      "eval_max_AUC": 0.7842336113536985,
+      "eval_max_Accuracy": 0.658363726790451,
+      "eval_max_Accuracy-right": 0.9557845311073432,
+      "eval_max_Accuracy-wrong": 0.13975437798498977,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.6534520933484,
+      "eval_min_AUC": 0.822015164482916,
+      "eval_min_Accuracy": 0.7524452917771883,
+      "eval_min_Accuracy-right": 0.7425981479066127,
+      "eval_min_Accuracy-wrong": 0.7696156470320673,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6912215159082531,
+      "eval_prod_AUC": 0.819309531656854,
+      "eval_prod_Accuracy": 0.5770888594164456,
+      "eval_prod_Accuracy-right": 0.35946263205947565,
+      "eval_prod_Accuracy-wrong": 0.9565612917898567,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.6874273712080468,
+      "eval_runtime": 248.3319,
+      "eval_samples_per_second": 97.16,
+      "eval_steps_per_second": 3.036,
+      "eval_sum_AUC": 0.7137200687510031,
+      "eval_sum_Accuracy": 0.64253149867374,
+      "eval_sum_Accuracy-right": 0.9941959045258901,
+      "eval_sum_Accuracy-wrong": 0.029338185126222424,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6771763816044019,
+      "step": 1500
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 22.34310628383116,
+      "learning_rate": 8.589776581327936e-06,
+      "loss": 0.1263,
+      "step": 1501
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 22.26638647085685,
+      "learning_rate": 8.587764179808716e-06,
+      "loss": 0.0922,
+      "step": 1502
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 34.88623829684697,
+      "learning_rate": 8.5857505795508e-06,
+      "loss": 0.0751,
+      "step": 1503
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 19.362303625609805,
+      "learning_rate": 8.583735781226964e-06,
+      "loss": 0.0897,
+      "step": 1504
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 15.607349323647277,
+      "learning_rate": 8.581719785510391e-06,
+      "loss": 0.0867,
+      "step": 1505
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 11.194617495715056,
+      "learning_rate": 8.579702593074666e-06,
+      "loss": 0.0877,
+      "step": 1506
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 19.404643890863266,
+      "learning_rate": 8.577684204593767e-06,
+      "loss": 0.0745,
+      "step": 1507
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 12.482102245943857,
+      "learning_rate": 8.575664620742073e-06,
+      "loss": 0.0892,
+      "step": 1508
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 17.587811648576377,
+      "learning_rate": 8.57364384219437e-06,
+      "loss": 0.0742,
+      "step": 1509
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 19.655816441505596,
+      "learning_rate": 8.571621869625835e-06,
+      "loss": 0.093,
+      "step": 1510
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 30.03043984591091,
+      "learning_rate": 8.569598703712045e-06,
+      "loss": 0.0831,
+      "step": 1511
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 21.842875103984415,
+      "learning_rate": 8.56757434512898e-06,
+      "loss": 0.0834,
+      "step": 1512
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 12.39324221817927,
+      "learning_rate": 8.565548794553016e-06,
+      "loss": 0.0818,
+      "step": 1513
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 43.49885915979903,
+      "learning_rate": 8.563522052660925e-06,
+      "loss": 0.1049,
+      "step": 1514
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.834088679375381,
+      "learning_rate": 8.561494120129878e-06,
+      "loss": 0.1014,
+      "step": 1515
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.466927403243657,
+      "learning_rate": 8.55946499763745e-06,
+      "loss": 0.0749,
+      "step": 1516
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 10.911200002806018,
+      "learning_rate": 8.557434685861604e-06,
+      "loss": 0.0834,
+      "step": 1517
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 33.65082017177862,
+      "learning_rate": 8.555403185480706e-06,
+      "loss": 0.0867,
+      "step": 1518
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 12.496700495977283,
+      "learning_rate": 8.553370497173518e-06,
+      "loss": 0.0687,
+      "step": 1519
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 13.198970575902443,
+      "learning_rate": 8.551336621619202e-06,
+      "loss": 0.1044,
+      "step": 1520
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 6.252606767967176,
+      "learning_rate": 8.549301559497309e-06,
+      "loss": 0.0756,
+      "step": 1521
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 19.71387444708082,
+      "learning_rate": 8.547265311487794e-06,
+      "loss": 0.0796,
+      "step": 1522
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 13.494983831348298,
+      "learning_rate": 8.545227878271004e-06,
+      "loss": 0.0994,
+      "step": 1523
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 31.544098364929408,
+      "learning_rate": 8.543189260527685e-06,
+      "loss": 0.0847,
+      "step": 1524
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 7.203150409865368,
+      "learning_rate": 8.541149458938972e-06,
+      "loss": 0.0718,
+      "step": 1525
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 6.754617592326723,
+      "learning_rate": 8.539108474186408e-06,
+      "loss": 0.0932,
+      "step": 1526
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 28.549577345821078,
+      "learning_rate": 8.53706630695192e-06,
+      "loss": 0.0979,
+      "step": 1527
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 22.622866020467605,
+      "learning_rate": 8.535022957917833e-06,
+      "loss": 0.0777,
+      "step": 1528
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 12.70378295713489,
+      "learning_rate": 8.53297842776687e-06,
+      "loss": 0.1194,
+      "step": 1529
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 9.841315864137869,
+      "learning_rate": 8.530932717182148e-06,
+      "loss": 0.1196,
+      "step": 1530
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 38.579351262870325,
+      "learning_rate": 8.528885826847173e-06,
+      "loss": 0.0726,
+      "step": 1531
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 29.95101519238097,
+      "learning_rate": 8.52683775744585e-06,
+      "loss": 0.0677,
+      "step": 1532
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 18.422090044359035,
+      "learning_rate": 8.524788509662478e-06,
+      "loss": 0.069,
+      "step": 1533
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 18.93769711941636,
+      "learning_rate": 8.522738084181749e-06,
+      "loss": 0.0963,
+      "step": 1534
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 22.851640489831993,
+      "learning_rate": 8.52068648168875e-06,
+      "loss": 0.0773,
+      "step": 1535
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 51.74572885384836,
+      "learning_rate": 8.518633702868955e-06,
+      "loss": 0.1013,
+      "step": 1536
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 12.047738013956764,
+      "learning_rate": 8.516579748408237e-06,
+      "loss": 0.0697,
+      "step": 1537
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.15886762195816,
+      "learning_rate": 8.514524618992864e-06,
+      "loss": 0.0896,
+      "step": 1538
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 47.185535045680865,
+      "learning_rate": 8.51246831530949e-06,
+      "loss": 0.1111,
+      "step": 1539
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 56.966239713258744,
+      "learning_rate": 8.510410838045165e-06,
+      "loss": 0.1395,
+      "step": 1540
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 13.443718359824823,
+      "learning_rate": 8.508352187887329e-06,
+      "loss": 0.0826,
+      "step": 1541
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 21.80564065604766,
+      "learning_rate": 8.506292365523816e-06,
+      "loss": 0.0814,
+      "step": 1542
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 24.864346346179897,
+      "learning_rate": 8.504231371642852e-06,
+      "loss": 0.1256,
+      "step": 1543
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 54.020150916277935,
+      "learning_rate": 8.502169206933053e-06,
+      "loss": 0.1006,
+      "step": 1544
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 25.153860020715037,
+      "learning_rate": 8.500105872083424e-06,
+      "loss": 0.0862,
+      "step": 1545
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 8.72803160180219,
+      "learning_rate": 8.498041367783367e-06,
+      "loss": 0.0757,
+      "step": 1546
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 44.857111526137345,
+      "learning_rate": 8.49597569472267e-06,
+      "loss": 0.1027,
+      "step": 1547
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 39.333400063483836,
+      "learning_rate": 8.493908853591515e-06,
+      "loss": 0.0902,
+      "step": 1548
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 18.48890358440609,
+      "learning_rate": 8.491840845080467e-06,
+      "loss": 0.0837,
+      "step": 1549
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 5.104802433608296,
+      "learning_rate": 8.489771669880489e-06,
+      "loss": 0.0821,
+      "step": 1550
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 17.233922736828326,
+      "learning_rate": 8.487701328682932e-06,
+      "loss": 0.0866,
+      "step": 1551
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 31.87602778903619,
+      "learning_rate": 8.485629822179533e-06,
+      "loss": 0.084,
+      "step": 1552
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 34.96613218097158,
+      "learning_rate": 8.483557151062423e-06,
+      "loss": 0.0948,
+      "step": 1553
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 5.749975742639682,
+      "learning_rate": 8.481483316024117e-06,
+      "loss": 0.0853,
+      "step": 1554
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 12.11538066802083,
+      "learning_rate": 8.479408317757525e-06,
+      "loss": 0.11,
+      "step": 1555
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 35.93167259883927,
+      "learning_rate": 8.477332156955942e-06,
+      "loss": 0.0968,
+      "step": 1556
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 45.13111946912449,
+      "learning_rate": 8.475254834313051e-06,
+      "loss": 0.1176,
+      "step": 1557
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 8.099342716236617,
+      "learning_rate": 8.473176350522925e-06,
+      "loss": 0.0784,
+      "step": 1558
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 6.454912943269717,
+      "learning_rate": 8.471096706280022e-06,
+      "loss": 0.1095,
+      "step": 1559
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 27.021317013221832,
+      "learning_rate": 8.469015902279191e-06,
+      "loss": 0.0663,
+      "step": 1560
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 34.78660476793632,
+      "learning_rate": 8.466933939215669e-06,
+      "loss": 0.1093,
+      "step": 1561
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 21.9565790223612,
+      "learning_rate": 8.464850817785075e-06,
+      "loss": 0.0702,
+      "step": 1562
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 11.065785710336767,
+      "learning_rate": 8.462766538683422e-06,
+      "loss": 0.0822,
+      "step": 1563
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 27.319906992202778,
+      "learning_rate": 8.460681102607106e-06,
+      "loss": 0.0803,
+      "step": 1564
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 12.384719292234578,
+      "learning_rate": 8.45859451025291e-06,
+      "loss": 0.0873,
+      "step": 1565
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.592013770399037,
+      "learning_rate": 8.456506762317998e-06,
+      "loss": 0.1086,
+      "step": 1566
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 15.454204756646234,
+      "learning_rate": 8.454417859499932e-06,
+      "loss": 0.1068,
+      "step": 1567
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 13.97423432264866,
+      "learning_rate": 8.45232780249665e-06,
+      "loss": 0.0629,
+      "step": 1568
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.11051784415923,
+      "learning_rate": 8.450236592006481e-06,
+      "loss": 0.0693,
+      "step": 1569
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 10.52568011914339,
+      "learning_rate": 8.448144228728135e-06,
+      "loss": 0.0842,
+      "step": 1570
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 13.957701644591483,
+      "learning_rate": 8.446050713360711e-06,
+      "loss": 0.0625,
+      "step": 1571
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 8.651605410737629,
+      "learning_rate": 8.443956046603692e-06,
+      "loss": 0.0853,
+      "step": 1572
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 11.215401871180575,
+      "learning_rate": 8.441860229156944e-06,
+      "loss": 0.0886,
+      "step": 1573
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 16.2620319467978,
+      "learning_rate": 8.439763261720716e-06,
+      "loss": 0.1127,
+      "step": 1574
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 19.65795515380451,
+      "learning_rate": 8.43766514499565e-06,
+      "loss": 0.0867,
+      "step": 1575
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 15.569766631054197,
+      "learning_rate": 8.435565879682759e-06,
+      "loss": 0.0986,
+      "step": 1576
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.803571295203503,
+      "learning_rate": 8.433465466483452e-06,
+      "loss": 0.0811,
+      "step": 1577
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 8.29721737997988,
+      "learning_rate": 8.431363906099513e-06,
+      "loss": 0.0776,
+      "step": 1578
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 15.709180796487498,
+      "learning_rate": 8.429261199233114e-06,
+      "loss": 0.0936,
+      "step": 1579
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.65559442729195,
+      "learning_rate": 8.427157346586807e-06,
+      "loss": 0.0811,
+      "step": 1580
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 9.08245662725313,
+      "learning_rate": 8.42505234886353e-06,
+      "loss": 0.1066,
+      "step": 1581
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 15.963511956163712,
+      "learning_rate": 8.422946206766598e-06,
+      "loss": 0.0867,
+      "step": 1582
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 16.28827683392989,
+      "learning_rate": 8.420838920999718e-06,
+      "loss": 0.0611,
+      "step": 1583
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 28.81251875839677,
+      "learning_rate": 8.418730492266968e-06,
+      "loss": 0.086,
+      "step": 1584
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 10.509715363597547,
+      "learning_rate": 8.416620921272818e-06,
+      "loss": 0.074,
+      "step": 1585
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 21.04769017279374,
+      "learning_rate": 8.414510208722111e-06,
+      "loss": 0.0928,
+      "step": 1586
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 29.427287736707218,
+      "learning_rate": 8.412398355320078e-06,
+      "loss": 0.0986,
+      "step": 1587
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 12.45398042864309,
+      "learning_rate": 8.410285361772328e-06,
+      "loss": 0.0876,
+      "step": 1588
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 14.68115002988725,
+      "learning_rate": 8.408171228784847e-06,
+      "loss": 0.0646,
+      "step": 1589
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 23.379480357894426,
+      "learning_rate": 8.406055957064014e-06,
+      "loss": 0.0698,
+      "step": 1590
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 33.42786888545678,
+      "learning_rate": 8.403939547316576e-06,
+      "loss": 0.1045,
+      "step": 1591
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 18.63204059915201,
+      "learning_rate": 8.401822000249661e-06,
+      "loss": 0.0681,
+      "step": 1592
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 34.22901610610889,
+      "learning_rate": 8.399703316570788e-06,
+      "loss": 0.0841,
+      "step": 1593
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 7.427652450081075,
+      "learning_rate": 8.397583496987846e-06,
+      "loss": 0.0715,
+      "step": 1594
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 10.923600709443539,
+      "learning_rate": 8.395462542209106e-06,
+      "loss": 0.0789,
+      "step": 1595
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.337513283657069,
+      "learning_rate": 8.393340452943219e-06,
+      "loss": 0.0902,
+      "step": 1596
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 27.489704010210236,
+      "learning_rate": 8.391217229899211e-06,
+      "loss": 0.1047,
+      "step": 1597
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 21.02825485654168,
+      "learning_rate": 8.389092873786495e-06,
+      "loss": 0.0721,
+      "step": 1598
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 7.16992364078436,
+      "learning_rate": 8.386967385314857e-06,
+      "loss": 0.0811,
+      "step": 1599
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 12.091998028936846,
+      "learning_rate": 8.384840765194458e-06,
+      "loss": 0.0624,
+      "step": 1600
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 51.589968347407535,
+      "learning_rate": 8.382713014135846e-06,
+      "loss": 0.1481,
+      "step": 1601
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 23.84886638204971,
+      "learning_rate": 8.38058413284994e-06,
+      "loss": 0.0939,
+      "step": 1602
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 21.203344409082234,
+      "learning_rate": 8.37845412204804e-06,
+      "loss": 0.0793,
+      "step": 1603
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 24.550507200743613,
+      "learning_rate": 8.376322982441821e-06,
+      "loss": 0.0908,
+      "step": 1604
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 19.77879329537523,
+      "learning_rate": 8.374190714743338e-06,
+      "loss": 0.0679,
+      "step": 1605
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 29.6562862247157,
+      "learning_rate": 8.37205731966502e-06,
+      "loss": 0.0901,
+      "step": 1606
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.814058600851218,
+      "learning_rate": 8.369922797919672e-06,
+      "loss": 0.0811,
+      "step": 1607
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 39.8862320984172,
+      "learning_rate": 8.367787150220481e-06,
+      "loss": 0.0906,
+      "step": 1608
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 28.666925523395125,
+      "learning_rate": 8.365650377281004e-06,
+      "loss": 0.0891,
+      "step": 1609
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 21.206415015810865,
+      "learning_rate": 8.36351247981518e-06,
+      "loss": 0.0583,
+      "step": 1610
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.871052189203208,
+      "learning_rate": 8.361373458537316e-06,
+      "loss": 0.0898,
+      "step": 1611
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 32.08612248377674,
+      "learning_rate": 8.359233314162102e-06,
+      "loss": 0.1099,
+      "step": 1612
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 28.62885152393508,
+      "learning_rate": 8.357092047404598e-06,
+      "loss": 0.0684,
+      "step": 1613
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 12.122663530621193,
+      "learning_rate": 8.354949658980243e-06,
+      "loss": 0.0867,
+      "step": 1614
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 10.300802431022989,
+      "learning_rate": 8.352806149604847e-06,
+      "loss": 0.0674,
+      "step": 1615
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 14.64308369502829,
+      "learning_rate": 8.350661519994596e-06,
+      "loss": 0.1304,
+      "step": 1616
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 14.480196587169626,
+      "learning_rate": 8.348515770866051e-06,
+      "loss": 0.1102,
+      "step": 1617
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 25.760410586184832,
+      "learning_rate": 8.346368902936149e-06,
+      "loss": 0.1083,
+      "step": 1618
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 5.429983401294056,
+      "learning_rate": 8.344220916922195e-06,
+      "loss": 0.0852,
+      "step": 1619
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 13.695217928576449,
+      "learning_rate": 8.342071813541873e-06,
+      "loss": 0.0719,
+      "step": 1620
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 18.550814397047013,
+      "learning_rate": 8.339921593513239e-06,
+      "loss": 0.1259,
+      "step": 1621
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 5.205918109284469,
+      "learning_rate": 8.337770257554721e-06,
+      "loss": 0.0732,
+      "step": 1622
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.675401155226301,
+      "learning_rate": 8.335617806385119e-06,
+      "loss": 0.0649,
+      "step": 1623
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 5.349145931438254,
+      "learning_rate": 8.333464240723608e-06,
+      "loss": 0.0719,
+      "step": 1624
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 21.629273939026348,
+      "learning_rate": 8.331309561289734e-06,
+      "loss": 0.089,
+      "step": 1625
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 13.33043717619918,
+      "learning_rate": 8.329153768803415e-06,
+      "loss": 0.0852,
+      "step": 1626
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 22.20088070759164,
+      "learning_rate": 8.326996863984942e-06,
+      "loss": 0.1255,
+      "step": 1627
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 14.843059951898912,
+      "learning_rate": 8.324838847554976e-06,
+      "loss": 0.1042,
+      "step": 1628
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 28.36135952406115,
+      "learning_rate": 8.322679720234553e-06,
+      "loss": 0.0717,
+      "step": 1629
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 21.469435660824352,
+      "learning_rate": 8.320519482745076e-06,
+      "loss": 0.0778,
+      "step": 1630
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 6.953286758260457,
+      "learning_rate": 8.31835813580832e-06,
+      "loss": 0.0997,
+      "step": 1631
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 45.73029621305888,
+      "learning_rate": 8.316195680146431e-06,
+      "loss": 0.1168,
+      "step": 1632
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 27.95351189790114,
+      "learning_rate": 8.314032116481927e-06,
+      "loss": 0.1123,
+      "step": 1633
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 19.796349764504168,
+      "learning_rate": 8.311867445537694e-06,
+      "loss": 0.0734,
+      "step": 1634
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 30.2548744067372,
+      "learning_rate": 8.30970166803699e-06,
+      "loss": 0.1013,
+      "step": 1635
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 32.50299572337184,
+      "learning_rate": 8.307534784703438e-06,
+      "loss": 0.0886,
+      "step": 1636
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 23.322609551734118,
+      "learning_rate": 8.305366796261036e-06,
+      "loss": 0.0909,
+      "step": 1637
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 18.271103305323894,
+      "learning_rate": 8.303197703434151e-06,
+      "loss": 0.111,
+      "step": 1638
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 37.2120104735074,
+      "learning_rate": 8.301027506947516e-06,
+      "loss": 0.1141,
+      "step": 1639
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 26.100003029295397,
+      "learning_rate": 8.298856207526234e-06,
+      "loss": 0.0931,
+      "step": 1640
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 13.681595047192792,
+      "learning_rate": 8.296683805895777e-06,
+      "loss": 0.0827,
+      "step": 1641
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.15509856357399,
+      "learning_rate": 8.294510302781984e-06,
+      "loss": 0.0536,
+      "step": 1642
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 26.6273672687637,
+      "learning_rate": 8.29233569891106e-06,
+      "loss": 0.082,
+      "step": 1643
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 24.756619530194843,
+      "learning_rate": 8.290159995009586e-06,
+      "loss": 0.1047,
+      "step": 1644
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 23.963631520040913,
+      "learning_rate": 8.2879831918045e-06,
+      "loss": 0.0957,
+      "step": 1645
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 14.18439751234935,
+      "learning_rate": 8.285805290023119e-06,
+      "loss": 0.1011,
+      "step": 1646
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 42.43033780709877,
+      "learning_rate": 8.283626290393112e-06,
+      "loss": 0.1035,
+      "step": 1647
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 50.888225111603596,
+      "learning_rate": 8.28144619364253e-06,
+      "loss": 0.0992,
+      "step": 1648
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 6.112792402279644,
+      "learning_rate": 8.279265000499783e-06,
+      "loss": 0.076,
+      "step": 1649
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 22.063075048957163,
+      "learning_rate": 8.277082711693645e-06,
+      "loss": 0.0963,
+      "step": 1650
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 15.773608935822898,
+      "learning_rate": 8.274899327953261e-06,
+      "loss": 0.1035,
+      "step": 1651
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 11.5617974343692,
+      "learning_rate": 8.272714850008142e-06,
+      "loss": 0.1187,
+      "step": 1652
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 15.775555813001663,
+      "learning_rate": 8.270529278588158e-06,
+      "loss": 0.1015,
+      "step": 1653
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 14.11119372001958,
+      "learning_rate": 8.268342614423553e-06,
+      "loss": 0.0741,
+      "step": 1654
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 22.852005917074994,
+      "learning_rate": 8.26615485824493e-06,
+      "loss": 0.0905,
+      "step": 1655
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 10.383090597471917,
+      "learning_rate": 8.263966010783259e-06,
+      "loss": 0.0772,
+      "step": 1656
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 20.778806982750876,
+      "learning_rate": 8.261776072769878e-06,
+      "loss": 0.0751,
+      "step": 1657
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 32.605084582643634,
+      "learning_rate": 8.259585044936484e-06,
+      "loss": 0.0916,
+      "step": 1658
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 16.061701345538165,
+      "learning_rate": 8.257392928015138e-06,
+      "loss": 0.0689,
+      "step": 1659
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.880742039534995,
+      "learning_rate": 8.25519972273827e-06,
+      "loss": 0.0938,
+      "step": 1660
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 38.374981943860945,
+      "learning_rate": 8.253005429838667e-06,
+      "loss": 0.0822,
+      "step": 1661
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 31.098642690273483,
+      "learning_rate": 8.250810050049488e-06,
+      "loss": 0.0938,
+      "step": 1662
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 19.79092470836297,
+      "learning_rate": 8.248613584104245e-06,
+      "loss": 0.073,
+      "step": 1663
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 10.879540641520684,
+      "learning_rate": 8.246416032736824e-06,
+      "loss": 0.0814,
+      "step": 1664
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 35.35301070268229,
+      "learning_rate": 8.244217396681461e-06,
+      "loss": 0.0746,
+      "step": 1665
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 32.156642571249634,
+      "learning_rate": 8.242017676672766e-06,
+      "loss": 0.1055,
+      "step": 1666
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.761163190326076,
+      "learning_rate": 8.239816873445705e-06,
+      "loss": 0.0907,
+      "step": 1667
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 7.04210178955987,
+      "learning_rate": 8.237614987735607e-06,
+      "loss": 0.0601,
+      "step": 1668
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 18.6708744832657,
+      "learning_rate": 8.235412020278164e-06,
+      "loss": 0.0577,
+      "step": 1669
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 18.924197204027514,
+      "learning_rate": 8.233207971809427e-06,
+      "loss": 0.0748,
+      "step": 1670
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.057484238524624,
+      "learning_rate": 8.23100284306581e-06,
+      "loss": 0.0565,
+      "step": 1671
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.302716030736626,
+      "learning_rate": 8.228796634784086e-06,
+      "loss": 0.0578,
+      "step": 1672
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 7.599559113192868,
+      "learning_rate": 8.226589347701396e-06,
+      "loss": 0.0682,
+      "step": 1673
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 7.509062555162927,
+      "learning_rate": 8.224380982555226e-06,
+      "loss": 0.0937,
+      "step": 1674
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 15.548280441790565,
+      "learning_rate": 8.222171540083442e-06,
+      "loss": 0.1221,
+      "step": 1675
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 19.248751048696665,
+      "learning_rate": 8.219961021024251e-06,
+      "loss": 0.0949,
+      "step": 1676
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 12.733356442809999,
+      "learning_rate": 8.217749426116238e-06,
+      "loss": 0.0925,
+      "step": 1677
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 19.53707190829396,
+      "learning_rate": 8.215536756098327e-06,
+      "loss": 0.0745,
+      "step": 1678
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.487045075606509,
+      "learning_rate": 8.21332301170982e-06,
+      "loss": 0.0776,
+      "step": 1679
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 7.500739390496608,
+      "learning_rate": 8.211108193690369e-06,
+      "loss": 0.1046,
+      "step": 1680
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 7.048956581103097,
+      "learning_rate": 8.208892302779982e-06,
+      "loss": 0.0927,
+      "step": 1681
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 13.805405676241328,
+      "learning_rate": 8.206675339719034e-06,
+      "loss": 0.0771,
+      "step": 1682
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 15.463386393474316,
+      "learning_rate": 8.204457305248253e-06,
+      "loss": 0.0728,
+      "step": 1683
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 15.88977529722202,
+      "learning_rate": 8.202238200108721e-06,
+      "loss": 0.0798,
+      "step": 1684
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 17.967914215750977,
+      "learning_rate": 8.200018025041887e-06,
+      "loss": 0.1217,
+      "step": 1685
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 28.123029815824335,
+      "learning_rate": 8.19779678078955e-06,
+      "loss": 0.084,
+      "step": 1686
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 34.442573623512985,
+      "learning_rate": 8.195574468093872e-06,
+      "loss": 0.1146,
+      "step": 1687
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 13.908280626151955,
+      "learning_rate": 8.193351087697366e-06,
+      "loss": 0.0895,
+      "step": 1688
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 17.926660993388946,
+      "learning_rate": 8.191126640342906e-06,
+      "loss": 0.0702,
+      "step": 1689
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 18.323048704617886,
+      "learning_rate": 8.18890112677372e-06,
+      "loss": 0.083,
+      "step": 1690
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 15.46406615699352,
+      "learning_rate": 8.186674547733398e-06,
+      "loss": 0.0956,
+      "step": 1691
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 12.626608111415358,
+      "learning_rate": 8.184446903965875e-06,
+      "loss": 0.1058,
+      "step": 1692
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 14.08504431889377,
+      "learning_rate": 8.182218196215452e-06,
+      "loss": 0.1021,
+      "step": 1693
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 18.58510270452204,
+      "learning_rate": 8.17998842522678e-06,
+      "loss": 0.0598,
+      "step": 1694
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 10.819894273976141,
+      "learning_rate": 8.17775759174487e-06,
+      "loss": 0.1018,
+      "step": 1695
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 26.40052974707958,
+      "learning_rate": 8.17552569651508e-06,
+      "loss": 0.0984,
+      "step": 1696
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 6.349053416270252,
+      "learning_rate": 8.173292740283135e-06,
+      "loss": 0.0953,
+      "step": 1697
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 19.342874409960686,
+      "learning_rate": 8.171058723795097e-06,
+      "loss": 0.0953,
+      "step": 1698
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 27.925603947809098,
+      "learning_rate": 8.168823647797401e-06,
+      "loss": 0.1146,
+      "step": 1699
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 77.29392142046062,
+      "learning_rate": 8.166587513036826e-06,
+      "loss": 0.1232,
+      "step": 1700
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 16.95185229227817,
+      "learning_rate": 8.164350320260502e-06,
+      "loss": 0.0662,
+      "step": 1701
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 21.974085417487288,
+      "learning_rate": 8.16211207021592e-06,
+      "loss": 0.0947,
+      "step": 1702
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.92675942999132,
+      "learning_rate": 8.15987276365092e-06,
+      "loss": 0.1031,
+      "step": 1703
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.05360437061264,
+      "learning_rate": 8.157632401313696e-06,
+      "loss": 0.1014,
+      "step": 1704
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.27119929232412,
+      "learning_rate": 8.155390983952795e-06,
+      "loss": 0.0781,
+      "step": 1705
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 4.3328835898750695,
+      "learning_rate": 8.153148512317117e-06,
+      "loss": 0.0669,
+      "step": 1706
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 11.753037962222542,
+      "learning_rate": 8.150904987155911e-06,
+      "loss": 0.0864,
+      "step": 1707
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 16.976159095803407,
+      "learning_rate": 8.148660409218786e-06,
+      "loss": 0.1355,
+      "step": 1708
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 5.360819721353881,
+      "learning_rate": 8.146414779255689e-06,
+      "loss": 0.1117,
+      "step": 1709
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 17.94580268964115,
+      "learning_rate": 8.144168098016933e-06,
+      "loss": 0.071,
+      "step": 1710
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 22.882975417855626,
+      "learning_rate": 8.141920366253173e-06,
+      "loss": 0.089,
+      "step": 1711
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 8.388813141318186,
+      "learning_rate": 8.139671584715419e-06,
+      "loss": 0.088,
+      "step": 1712
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 19.66518390061167,
+      "learning_rate": 8.137421754155031e-06,
+      "loss": 0.1162,
+      "step": 1713
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 5.009113369388294,
+      "learning_rate": 8.13517087532372e-06,
+      "loss": 0.0576,
+      "step": 1714
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 14.062847580323941,
+      "learning_rate": 8.132918948973543e-06,
+      "loss": 0.0834,
+      "step": 1715
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 23.85559125219598,
+      "learning_rate": 8.130665975856913e-06,
+      "loss": 0.0881,
+      "step": 1716
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 25.11264401301826,
+      "learning_rate": 8.128411956726592e-06,
+      "loss": 0.1072,
+      "step": 1717
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 9.641948331236978,
+      "learning_rate": 8.126156892335686e-06,
+      "loss": 0.0957,
+      "step": 1718
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.36919292282907,
+      "learning_rate": 8.123900783437655e-06,
+      "loss": 0.1229,
+      "step": 1719
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 7.6768793485009965,
+      "learning_rate": 8.121643630786308e-06,
+      "loss": 0.1084,
+      "step": 1720
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 11.056594907308932,
+      "learning_rate": 8.1193854351358e-06,
+      "loss": 0.08,
+      "step": 1721
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 13.02410509965871,
+      "learning_rate": 8.11712619724064e-06,
+      "loss": 0.0721,
+      "step": 1722
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 8.593130940430317,
+      "learning_rate": 8.114865917855676e-06,
+      "loss": 0.0872,
+      "step": 1723
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 5.9240949988331035,
+      "learning_rate": 8.112604597736113e-06,
+      "loss": 0.0928,
+      "step": 1724
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 22.20784644695516,
+      "learning_rate": 8.110342237637501e-06,
+      "loss": 0.0628,
+      "step": 1725
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 10.921566844245627,
+      "learning_rate": 8.108078838315732e-06,
+      "loss": 0.0618,
+      "step": 1726
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 8.111574295790442,
+      "learning_rate": 8.105814400527052e-06,
+      "loss": 0.0785,
+      "step": 1727
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 8.289683871686902,
+      "learning_rate": 8.103548925028054e-06,
+      "loss": 0.1143,
+      "step": 1728
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 14.203313022838092,
+      "learning_rate": 8.101282412575673e-06,
+      "loss": 0.098,
+      "step": 1729
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 36.75773831067622,
+      "learning_rate": 8.099014863927192e-06,
+      "loss": 0.0892,
+      "step": 1730
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 11.032969099373302,
+      "learning_rate": 8.096746279840245e-06,
+      "loss": 0.0819,
+      "step": 1731
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 27.902605396219545,
+      "learning_rate": 8.094476661072806e-06,
+      "loss": 0.0928,
+      "step": 1732
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 30.50428948295198,
+      "learning_rate": 8.092206008383195e-06,
+      "loss": 0.0852,
+      "step": 1733
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 26.430089450059356,
+      "learning_rate": 8.089934322530082e-06,
+      "loss": 0.1184,
+      "step": 1734
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 21.899501892048658,
+      "learning_rate": 8.087661604272477e-06,
+      "loss": 0.0836,
+      "step": 1735
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 20.60542568709364,
+      "learning_rate": 8.08538785436974e-06,
+      "loss": 0.097,
+      "step": 1736
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 22.941207777328003,
+      "learning_rate": 8.08311307358157e-06,
+      "loss": 0.1018,
+      "step": 1737
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 6.3384979167538775,
+      "learning_rate": 8.080837262668017e-06,
+      "loss": 0.0703,
+      "step": 1738
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 6.570398124978654,
+      "learning_rate": 8.078560422389472e-06,
+      "loss": 0.0669,
+      "step": 1739
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 21.414339602584793,
+      "learning_rate": 8.076282553506664e-06,
+      "loss": 0.0767,
+      "step": 1740
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 18.95372832984155,
+      "learning_rate": 8.074003656780678e-06,
+      "loss": 0.0938,
+      "step": 1741
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 10.549353249338639,
+      "learning_rate": 8.071723732972933e-06,
+      "loss": 0.0778,
+      "step": 1742
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 16.938439415836026,
+      "learning_rate": 8.069442782845191e-06,
+      "loss": 0.1041,
+      "step": 1743
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 27.528040440849797,
+      "learning_rate": 8.067160807159566e-06,
+      "loss": 0.1102,
+      "step": 1744
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 34.857238335191894,
+      "learning_rate": 8.064877806678504e-06,
+      "loss": 0.1146,
+      "step": 1745
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 19.588923727022934,
+      "learning_rate": 8.062593782164798e-06,
+      "loss": 0.1074,
+      "step": 1746
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.262300732118138,
+      "learning_rate": 8.060308734381585e-06,
+      "loss": 0.0928,
+      "step": 1747
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 21.63069752918576,
+      "learning_rate": 8.05802266409234e-06,
+      "loss": 0.0915,
+      "step": 1748
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 17.730732468468556,
+      "learning_rate": 8.055735572060883e-06,
+      "loss": 0.0682,
+      "step": 1749
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 15.174261533303634,
+      "learning_rate": 8.053447459051374e-06,
+      "loss": 0.0715,
+      "step": 1750
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 14.744003542025283,
+      "learning_rate": 8.051158325828315e-06,
+      "loss": 0.0828,
+      "step": 1751
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 16.141517917244126,
+      "learning_rate": 8.048868173156546e-06,
+      "loss": 0.0897,
+      "step": 1752
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 27.216388324679485,
+      "learning_rate": 8.046577001801248e-06,
+      "loss": 0.1003,
+      "step": 1753
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 12.025194670403664,
+      "learning_rate": 8.044284812527949e-06,
+      "loss": 0.1011,
+      "step": 1754
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 17.72776253943835,
+      "learning_rate": 8.041991606102507e-06,
+      "loss": 0.1263,
+      "step": 1755
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 6.379107531262259,
+      "learning_rate": 8.039697383291127e-06,
+      "loss": 0.068,
+      "step": 1756
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.098384373784489,
+      "learning_rate": 8.037402144860353e-06,
+      "loss": 0.0898,
+      "step": 1757
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 11.389445025532513,
+      "learning_rate": 8.035105891577064e-06,
+      "loss": 0.0896,
+      "step": 1758
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.922481057387875,
+      "learning_rate": 8.032808624208485e-06,
+      "loss": 0.1005,
+      "step": 1759
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 25.126444637378192,
+      "learning_rate": 8.030510343522172e-06,
+      "loss": 0.0884,
+      "step": 1760
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 12.30642758311427,
+      "learning_rate": 8.02821105028602e-06,
+      "loss": 0.076,
+      "step": 1761
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 6.33267110220372,
+      "learning_rate": 8.025910745268276e-06,
+      "loss": 0.1035,
+      "step": 1762
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 19.02072960195617,
+      "learning_rate": 8.023609429237504e-06,
+      "loss": 0.0708,
+      "step": 1763
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 6.3968909866767545,
+      "learning_rate": 8.021307102962623e-06,
+      "loss": 0.0759,
+      "step": 1764
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 23.20723618473892,
+      "learning_rate": 8.019003767212881e-06,
+      "loss": 0.0717,
+      "step": 1765
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 32.55425111351312,
+      "learning_rate": 8.016699422757865e-06,
+      "loss": 0.126,
+      "step": 1766
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 8.562249690813024,
+      "learning_rate": 8.014394070367499e-06,
+      "loss": 0.1367,
+      "step": 1767
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 36.530412481845815,
+      "learning_rate": 8.012087710812047e-06,
+      "loss": 0.1062,
+      "step": 1768
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 38.24443582138145,
+      "learning_rate": 8.009780344862101e-06,
+      "loss": 0.0983,
+      "step": 1769
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 19.804828163884277,
+      "learning_rate": 8.0074719732886e-06,
+      "loss": 0.0815,
+      "step": 1770
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 19.841573958043558,
+      "learning_rate": 8.005162596862812e-06,
+      "loss": 0.0888,
+      "step": 1771
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 36.698744929178375,
+      "learning_rate": 8.002852216356343e-06,
+      "loss": 0.1433,
+      "step": 1772
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 41.58220436938131,
+      "learning_rate": 8.000540832541132e-06,
+      "loss": 0.1224,
+      "step": 1773
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 16.701232099627646,
+      "learning_rate": 7.99822844618946e-06,
+      "loss": 0.1053,
+      "step": 1774
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 23.90585403265861,
+      "learning_rate": 7.995915058073933e-06,
+      "loss": 0.1041,
+      "step": 1775
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 22.792947908502864,
+      "learning_rate": 7.9936006689675e-06,
+      "loss": 0.0842,
+      "step": 1776
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 28.537044176142224,
+      "learning_rate": 7.99128527964344e-06,
+      "loss": 0.1215,
+      "step": 1777
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 11.80089352447319,
+      "learning_rate": 7.988968890875368e-06,
+      "loss": 0.0922,
+      "step": 1778
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.311869719605001,
+      "learning_rate": 7.986651503437233e-06,
+      "loss": 0.0958,
+      "step": 1779
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 21.233417174050086,
+      "learning_rate": 7.984333118103318e-06,
+      "loss": 0.1084,
+      "step": 1780
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 8.340066676153791,
+      "learning_rate": 7.982013735648235e-06,
+      "loss": 0.0981,
+      "step": 1781
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 24.679907926446546,
+      "learning_rate": 7.979693356846937e-06,
+      "loss": 0.1503,
+      "step": 1782
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 25.530690055700195,
+      "learning_rate": 7.977371982474705e-06,
+      "loss": 0.1339,
+      "step": 1783
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 7.597467209345345,
+      "learning_rate": 7.975049613307151e-06,
+      "loss": 0.1124,
+      "step": 1784
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 10.392568087679368,
+      "learning_rate": 7.972726250120225e-06,
+      "loss": 0.1146,
+      "step": 1785
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 7.338556178755361,
+      "learning_rate": 7.970401893690202e-06,
+      "loss": 0.1012,
+      "step": 1786
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 22.062178634771666,
+      "learning_rate": 7.968076544793696e-06,
+      "loss": 0.0973,
+      "step": 1787
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 6.614753659360179,
+      "learning_rate": 7.965750204207647e-06,
+      "loss": 0.0793,
+      "step": 1788
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 6.37429769749807,
+      "learning_rate": 7.96342287270933e-06,
+      "loss": 0.0891,
+      "step": 1789
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 6.518825899555654,
+      "learning_rate": 7.96109455107635e-06,
+      "loss": 0.0751,
+      "step": 1790
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 39.72404138798098,
+      "learning_rate": 7.958765240086639e-06,
+      "loss": 0.1064,
+      "step": 1791
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 13.961553572881497,
+      "learning_rate": 7.956434940518468e-06,
+      "loss": 0.0696,
+      "step": 1792
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 13.524395349313767,
+      "learning_rate": 7.954103653150432e-06,
+      "loss": 0.1025,
+      "step": 1793
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 12.512822700266833,
+      "learning_rate": 7.951771378761455e-06,
+      "loss": 0.0912,
+      "step": 1794
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 31.031749476523494,
+      "learning_rate": 7.949438118130797e-06,
+      "loss": 0.1554,
+      "step": 1795
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 21.655240692631608,
+      "learning_rate": 7.94710387203804e-06,
+      "loss": 0.1049,
+      "step": 1796
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 23.573349737519887,
+      "learning_rate": 7.944768641263101e-06,
+      "loss": 0.0951,
+      "step": 1797
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 6.423633965336433,
+      "learning_rate": 7.942432426586224e-06,
+      "loss": 0.0883,
+      "step": 1798
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 33.7924509461897,
+      "learning_rate": 7.94009522878798e-06,
+      "loss": 0.1217,
+      "step": 1799
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 24.608886378571242,
+      "learning_rate": 7.937757048649274e-06,
+      "loss": 0.1155,
+      "step": 1800
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 31.5944802631455,
+      "learning_rate": 7.935417886951332e-06,
+      "loss": 0.1301,
+      "step": 1801
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 24.56355408273472,
+      "learning_rate": 7.933077744475713e-06,
+      "loss": 0.0983,
+      "step": 1802
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 12.600335223994328,
+      "learning_rate": 7.930736622004301e-06,
+      "loss": 0.0992,
+      "step": 1803
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 17.365260585180668,
+      "learning_rate": 7.928394520319311e-06,
+      "loss": 0.1122,
+      "step": 1804
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 5.327789011223336,
+      "learning_rate": 7.926051440203278e-06,
+      "loss": 0.0663,
+      "step": 1805
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.382747080442627,
+      "learning_rate": 7.923707382439073e-06,
+      "loss": 0.082,
+      "step": 1806
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 9.124410943625675,
+      "learning_rate": 7.921362347809888e-06,
+      "loss": 0.1038,
+      "step": 1807
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 19.289726689987205,
+      "learning_rate": 7.919016337099242e-06,
+      "loss": 0.1105,
+      "step": 1808
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 7.964546128423192,
+      "learning_rate": 7.916669351090981e-06,
+      "loss": 0.0864,
+      "step": 1809
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 4.824052539383046,
+      "learning_rate": 7.914321390569278e-06,
+      "loss": 0.0694,
+      "step": 1810
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 8.33312536205575,
+      "learning_rate": 7.911972456318629e-06,
+      "loss": 0.1254,
+      "step": 1811
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 6.507945976935379,
+      "learning_rate": 7.909622549123855e-06,
+      "loss": 0.0936,
+      "step": 1812
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 13.143972556032457,
+      "learning_rate": 7.907271669770107e-06,
+      "loss": 0.0985,
+      "step": 1813
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 12.491831143764811,
+      "learning_rate": 7.904919819042855e-06,
+      "loss": 0.1169,
+      "step": 1814
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 13.22535591359354,
+      "learning_rate": 7.902566997727896e-06,
+      "loss": 0.1021,
+      "step": 1815
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.970708252245554,
+      "learning_rate": 7.900213206611353e-06,
+      "loss": 0.1017,
+      "step": 1816
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 26.696684320453635,
+      "learning_rate": 7.897858446479672e-06,
+      "loss": 0.1003,
+      "step": 1817
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 31.522235837524857,
+      "learning_rate": 7.895502718119618e-06,
+      "loss": 0.1056,
+      "step": 1818
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 11.01303572938718,
+      "learning_rate": 7.89314602231829e-06,
+      "loss": 0.0793,
+      "step": 1819
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 34.25204366156014,
+      "learning_rate": 7.8907883598631e-06,
+      "loss": 0.1014,
+      "step": 1820
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 40.92644841373539,
+      "learning_rate": 7.888429731541784e-06,
+      "loss": 0.1143,
+      "step": 1821
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 5.3982661444680335,
+      "learning_rate": 7.886070138142407e-06,
+      "loss": 0.071,
+      "step": 1822
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 32.32811463120643,
+      "learning_rate": 7.883709580453354e-06,
+      "loss": 0.1158,
+      "step": 1823
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 19.031316596008647,
+      "learning_rate": 7.88134805926333e-06,
+      "loss": 0.0977,
+      "step": 1824
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 15.045104813096087,
+      "learning_rate": 7.878985575361362e-06,
+      "loss": 0.0897,
+      "step": 1825
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 21.035459837713777,
+      "learning_rate": 7.876622129536801e-06,
+      "loss": 0.1024,
+      "step": 1826
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 13.660668150169178,
+      "learning_rate": 7.874257722579319e-06,
+      "loss": 0.0901,
+      "step": 1827
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 13.858105852883831,
+      "learning_rate": 7.871892355278906e-06,
+      "loss": 0.0676,
+      "step": 1828
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 30.715956025025108,
+      "learning_rate": 7.869526028425878e-06,
+      "loss": 0.1143,
+      "step": 1829
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 20.527869229321144,
+      "learning_rate": 7.867158742810866e-06,
+      "loss": 0.0834,
+      "step": 1830
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 7.21553644321985,
+      "learning_rate": 7.864790499224825e-06,
+      "loss": 0.069,
+      "step": 1831
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 15.834412029043872,
+      "learning_rate": 7.86242129845903e-06,
+      "loss": 0.0689,
+      "step": 1832
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 16.28611478117609,
+      "learning_rate": 7.860051141305074e-06,
+      "loss": 0.0878,
+      "step": 1833
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 18.470539205773807,
+      "learning_rate": 7.857680028554873e-06,
+      "loss": 0.0988,
+      "step": 1834
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 14.517688298308093,
+      "learning_rate": 7.855307961000656e-06,
+      "loss": 0.0839,
+      "step": 1835
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 8.127093098505616,
+      "learning_rate": 7.852934939434977e-06,
+      "loss": 0.0699,
+      "step": 1836
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 15.700240348518008,
+      "learning_rate": 7.850560964650707e-06,
+      "loss": 0.1311,
+      "step": 1837
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 33.191617585225714,
+      "learning_rate": 7.848186037441035e-06,
+      "loss": 0.1548,
+      "step": 1838
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 38.7790193219392,
+      "learning_rate": 7.845810158599467e-06,
+      "loss": 0.0975,
+      "step": 1839
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 16.850557590964687,
+      "learning_rate": 7.84343332891983e-06,
+      "loss": 0.0818,
+      "step": 1840
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 28.72144236111981,
+      "learning_rate": 7.841055549196267e-06,
+      "loss": 0.1003,
+      "step": 1841
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 21.710573467867395,
+      "learning_rate": 7.838676820223234e-06,
+      "loss": 0.0881,
+      "step": 1842
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 35.03115815820891,
+      "learning_rate": 7.836297142795515e-06,
+      "loss": 0.0961,
+      "step": 1843
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 12.832217299261412,
+      "learning_rate": 7.833916517708203e-06,
+      "loss": 0.0826,
+      "step": 1844
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 35.32560567226581,
+      "learning_rate": 7.831534945756703e-06,
+      "loss": 0.1127,
+      "step": 1845
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 27.26210477910772,
+      "learning_rate": 7.82915242773675e-06,
+      "loss": 0.0912,
+      "step": 1846
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 12.005708358639822,
+      "learning_rate": 7.826768964444384e-06,
+      "loss": 0.0798,
+      "step": 1847
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 21.5695948891362,
+      "learning_rate": 7.824384556675966e-06,
+      "loss": 0.0976,
+      "step": 1848
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 32.38044189186778,
+      "learning_rate": 7.821999205228168e-06,
+      "loss": 0.13,
+      "step": 1849
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 78.64683053933734,
+      "learning_rate": 7.819612910897985e-06,
+      "loss": 0.2098,
+      "step": 1850
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 18.63082381659104,
+      "learning_rate": 7.817225674482717e-06,
+      "loss": 0.0945,
+      "step": 1851
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 36.11550459481589,
+      "learning_rate": 7.814837496779988e-06,
+      "loss": 0.0802,
+      "step": 1852
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 57.52467349010068,
+      "learning_rate": 7.812448378587731e-06,
+      "loss": 0.1255,
+      "step": 1853
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 38.526328381998304,
+      "learning_rate": 7.810058320704194e-06,
+      "loss": 0.11,
+      "step": 1854
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 7.853053246927483,
+      "learning_rate": 7.807667323927941e-06,
+      "loss": 0.0726,
+      "step": 1855
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 32.05414416545823,
+      "learning_rate": 7.80527538905785e-06,
+      "loss": 0.0781,
+      "step": 1856
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 50.865898236236205,
+      "learning_rate": 7.802882516893106e-06,
+      "loss": 0.1028,
+      "step": 1857
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 31.687847029880817,
+      "learning_rate": 7.800488708233219e-06,
+      "loss": 0.0911,
+      "step": 1858
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.587216812760387,
+      "learning_rate": 7.798093963877998e-06,
+      "loss": 0.0892,
+      "step": 1859
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 32.49642886667288,
+      "learning_rate": 7.795698284627575e-06,
+      "loss": 0.1219,
+      "step": 1860
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 32.145261643797554,
+      "learning_rate": 7.793301671282391e-06,
+      "loss": 0.1083,
+      "step": 1861
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 24.419766007483094,
+      "learning_rate": 7.7909041246432e-06,
+      "loss": 0.0844,
+      "step": 1862
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.398067835525364,
+      "learning_rate": 7.788505645511065e-06,
+      "loss": 0.1044,
+      "step": 1863
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 30.039981713715175,
+      "learning_rate": 7.786106234687362e-06,
+      "loss": 0.0833,
+      "step": 1864
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 47.34351073345086,
+      "learning_rate": 7.783705892973782e-06,
+      "loss": 0.1405,
+      "step": 1865
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 12.349109025106152,
+      "learning_rate": 7.78130462117232e-06,
+      "loss": 0.0813,
+      "step": 1866
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 9.908058765643409,
+      "learning_rate": 7.778902420085289e-06,
+      "loss": 0.0681,
+      "step": 1867
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.3494219707093515,
+      "learning_rate": 7.776499290515304e-06,
+      "loss": 0.0879,
+      "step": 1868
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 11.040424391158401,
+      "learning_rate": 7.7740952332653e-06,
+      "loss": 0.0918,
+      "step": 1869
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 24.22287383306475,
+      "learning_rate": 7.771690249138517e-06,
+      "loss": 0.0968,
+      "step": 1870
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 11.109944144915485,
+      "learning_rate": 7.769284338938502e-06,
+      "loss": 0.1071,
+      "step": 1871
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 9.840567242675737,
+      "learning_rate": 7.766877503469117e-06,
+      "loss": 0.1022,
+      "step": 1872
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.497594965854643,
+      "learning_rate": 7.764469743534529e-06,
+      "loss": 0.0892,
+      "step": 1873
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 34.64644715009004,
+      "learning_rate": 7.762061059939214e-06,
+      "loss": 0.0901,
+      "step": 1874
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 17.020914763066287,
+      "learning_rate": 7.759651453487963e-06,
+      "loss": 0.0775,
+      "step": 1875
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 41.443228475209594,
+      "learning_rate": 7.757240924985866e-06,
+      "loss": 0.1339,
+      "step": 1876
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 22.513277159467158,
+      "learning_rate": 7.754829475238323e-06,
+      "loss": 0.105,
+      "step": 1877
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 21.265724907023937,
+      "learning_rate": 7.752417105051051e-06,
+      "loss": 0.1527,
+      "step": 1878
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 48.24610738893118,
+      "learning_rate": 7.750003815230062e-06,
+      "loss": 0.0941,
+      "step": 1879
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 6.8333978374856015,
+      "learning_rate": 7.747589606581686e-06,
+      "loss": 0.0563,
+      "step": 1880
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 17.172826245285943,
+      "learning_rate": 7.745174479912551e-06,
+      "loss": 0.1003,
+      "step": 1881
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 10.693710087910775,
+      "learning_rate": 7.742758436029596e-06,
+      "loss": 0.1112,
+      "step": 1882
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 18.75748741216541,
+      "learning_rate": 7.740341475740068e-06,
+      "loss": 0.0944,
+      "step": 1883
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 29.562990385496306,
+      "learning_rate": 7.737923599851519e-06,
+      "loss": 0.0795,
+      "step": 1884
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 6.523359606962466,
+      "learning_rate": 7.735504809171801e-06,
+      "loss": 0.0846,
+      "step": 1885
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 6.171598232912832,
+      "learning_rate": 7.733085104509084e-06,
+      "loss": 0.0924,
+      "step": 1886
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 29.73271224436981,
+      "learning_rate": 7.730664486671831e-06,
+      "loss": 0.1106,
+      "step": 1887
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 30.655108989409246,
+      "learning_rate": 7.72824295646882e-06,
+      "loss": 0.0946,
+      "step": 1888
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 11.90587683853757,
+      "learning_rate": 7.725820514709124e-06,
+      "loss": 0.0774,
+      "step": 1889
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 23.959766879108276,
+      "learning_rate": 7.723397162202128e-06,
+      "loss": 0.1327,
+      "step": 1890
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 18.9085622070494,
+      "learning_rate": 7.720972899757522e-06,
+      "loss": 0.0879,
+      "step": 1891
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 23.68721606790896,
+      "learning_rate": 7.718547728185293e-06,
+      "loss": 0.1257,
+      "step": 1892
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.866401919651024,
+      "learning_rate": 7.716121648295738e-06,
+      "loss": 0.0869,
+      "step": 1893
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 12.203988683415128,
+      "learning_rate": 7.713694660899455e-06,
+      "loss": 0.087,
+      "step": 1894
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 47.327087832020034,
+      "learning_rate": 7.711266766807345e-06,
+      "loss": 0.1295,
+      "step": 1895
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 37.47630197076899,
+      "learning_rate": 7.708837966830615e-06,
+      "loss": 0.1118,
+      "step": 1896
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 15.991272091833782,
+      "learning_rate": 7.706408261780769e-06,
+      "loss": 0.075,
+      "step": 1897
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 35.91340982420784,
+      "learning_rate": 7.703977652469618e-06,
+      "loss": 0.1104,
+      "step": 1898
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 31.129407279590517,
+      "learning_rate": 7.701546139709272e-06,
+      "loss": 0.0825,
+      "step": 1899
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 6.731671111833922,
+      "learning_rate": 7.69911372431215e-06,
+      "loss": 0.1019,
+      "step": 1900
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.665645173275369,
+      "learning_rate": 7.696680407090962e-06,
+      "loss": 0.1041,
+      "step": 1901
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 14.15220002595494,
+      "learning_rate": 7.694246188858726e-06,
+      "loss": 0.087,
+      "step": 1902
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 11.320413195892144,
+      "learning_rate": 7.691811070428758e-06,
+      "loss": 0.1177,
+      "step": 1903
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 39.10964804895109,
+      "learning_rate": 7.689375052614681e-06,
+      "loss": 0.1274,
+      "step": 1904
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 12.690397207526837,
+      "learning_rate": 7.686938136230408e-06,
+      "loss": 0.1031,
+      "step": 1905
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 15.07798986443616,
+      "learning_rate": 7.684500322090162e-06,
+      "loss": 0.1309,
+      "step": 1906
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.8068024583617093,
+      "learning_rate": 7.68206161100846e-06,
+      "loss": 0.0712,
+      "step": 1907
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 7.903182113883539,
+      "learning_rate": 7.679622003800122e-06,
+      "loss": 0.0999,
+      "step": 1908
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 41.30605521807284,
+      "learning_rate": 7.677181501280266e-06,
+      "loss": 0.0956,
+      "step": 1909
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 13.709690616051338,
+      "learning_rate": 7.674740104264308e-06,
+      "loss": 0.0869,
+      "step": 1910
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 13.260512003188966,
+      "learning_rate": 7.672297813567968e-06,
+      "loss": 0.1622,
+      "step": 1911
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 21.84565087029452,
+      "learning_rate": 7.669854630007257e-06,
+      "loss": 0.1017,
+      "step": 1912
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 33.420476181478065,
+      "learning_rate": 7.667410554398486e-06,
+      "loss": 0.098,
+      "step": 1913
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 31.57040112746546,
+      "learning_rate": 7.664965587558271e-06,
+      "loss": 0.1077,
+      "step": 1914
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 4.308046009484161,
+      "learning_rate": 7.662519730303517e-06,
+      "loss": 0.0771,
+      "step": 1915
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 60.79999436143695,
+      "learning_rate": 7.660072983451433e-06,
+      "loss": 0.1399,
+      "step": 1916
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 18.3987711534509,
+      "learning_rate": 7.657625347819522e-06,
+      "loss": 0.1149,
+      "step": 1917
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 6.265288838813149,
+      "learning_rate": 7.655176824225582e-06,
+      "loss": 0.0807,
+      "step": 1918
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 6.300914605261491,
+      "learning_rate": 7.652727413487716e-06,
+      "loss": 0.0993,
+      "step": 1919
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 22.124179459804967,
+      "learning_rate": 7.650277116424313e-06,
+      "loss": 0.0684,
+      "step": 1920
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 25.625468309309074,
+      "learning_rate": 7.647825933854063e-06,
+      "loss": 0.1117,
+      "step": 1921
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 7.624552453874793,
+      "learning_rate": 7.645373866595953e-06,
+      "loss": 0.1179,
+      "step": 1922
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 6.194508526818871,
+      "learning_rate": 7.642920915469265e-06,
+      "loss": 0.0785,
+      "step": 1923
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 24.148000648762252,
+      "learning_rate": 7.640467081293573e-06,
+      "loss": 0.1417,
+      "step": 1924
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 25.24596065135585,
+      "learning_rate": 7.638012364888751e-06,
+      "loss": 0.1062,
+      "step": 1925
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 11.663735609649464,
+      "learning_rate": 7.635556767074965e-06,
+      "loss": 0.0919,
+      "step": 1926
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 19.983238031321108,
+      "learning_rate": 7.633100288672674e-06,
+      "loss": 0.0861,
+      "step": 1927
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 24.516987429952046,
+      "learning_rate": 7.630642930502634e-06,
+      "loss": 0.1084,
+      "step": 1928
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 23.7079785548782,
+      "learning_rate": 7.628184693385896e-06,
+      "loss": 0.0987,
+      "step": 1929
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.632779942934862,
+      "learning_rate": 7.625725578143801e-06,
+      "loss": 0.0803,
+      "step": 1930
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 26.75562561931583,
+      "learning_rate": 7.6232655855979844e-06,
+      "loss": 0.12,
+      "step": 1931
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.220766214367428,
+      "learning_rate": 7.620804716570376e-06,
+      "loss": 0.1036,
+      "step": 1932
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 35.404659677885,
+      "learning_rate": 7.618342971883199e-06,
+      "loss": 0.1191,
+      "step": 1933
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 5.539001124567162,
+      "learning_rate": 7.615880352358967e-06,
+      "loss": 0.0887,
+      "step": 1934
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 33.35991052555321,
+      "learning_rate": 7.613416858820486e-06,
+      "loss": 0.0751,
+      "step": 1935
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 43.77244661048231,
+      "learning_rate": 7.6109524920908575e-06,
+      "loss": 0.105,
+      "step": 1936
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.095324647325247,
+      "learning_rate": 7.608487252993471e-06,
+      "loss": 0.1018,
+      "step": 1937
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.482483420441799,
+      "learning_rate": 7.6060211423520095e-06,
+      "loss": 0.0607,
+      "step": 1938
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 39.33723284296271,
+      "learning_rate": 7.6035541609904425e-06,
+      "loss": 0.1287,
+      "step": 1939
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 16.80992526081807,
+      "learning_rate": 7.60108630973304e-06,
+      "loss": 0.0977,
+      "step": 1940
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 14.940668855422615,
+      "learning_rate": 7.598617589404354e-06,
+      "loss": 0.0879,
+      "step": 1941
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 7.542100439749129,
+      "learning_rate": 7.596148000829229e-06,
+      "loss": 0.1262,
+      "step": 1942
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 17.47476515065765,
+      "learning_rate": 7.593677544832802e-06,
+      "loss": 0.1219,
+      "step": 1943
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 7.605399845369206,
+      "learning_rate": 7.5912062222404965e-06,
+      "loss": 0.064,
+      "step": 1944
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 30.299165014514305,
+      "learning_rate": 7.588734033878031e-06,
+      "loss": 0.1134,
+      "step": 1945
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 12.53081257535877,
+      "learning_rate": 7.586260980571407e-06,
+      "loss": 0.0906,
+      "step": 1946
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 28.76809322028387,
+      "learning_rate": 7.5837870631469165e-06,
+      "loss": 0.1503,
+      "step": 1947
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 21.855556241959924,
+      "learning_rate": 7.581312282431143e-06,
+      "loss": 0.0753,
+      "step": 1948
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 29.170849942865587,
+      "learning_rate": 7.578836639250958e-06,
+      "loss": 0.0962,
+      "step": 1949
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 11.690758795177931,
+      "learning_rate": 7.576360134433517e-06,
+      "loss": 0.0917,
+      "step": 1950
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 12.408711643478298,
+      "learning_rate": 7.5738827688062676e-06,
+      "loss": 0.0862,
+      "step": 1951
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 13.868396523210292,
+      "learning_rate": 7.571404543196943e-06,
+      "loss": 0.1176,
+      "step": 1952
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 26.57491798793351,
+      "learning_rate": 7.568925458433567e-06,
+      "loss": 0.1008,
+      "step": 1953
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 5.974416690063627,
+      "learning_rate": 7.566445515344445e-06,
+      "loss": 0.0807,
+      "step": 1954
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 14.981211780495261,
+      "learning_rate": 7.563964714758172e-06,
+      "loss": 0.0732,
+      "step": 1955
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 30.306827143553917,
+      "learning_rate": 7.561483057503632e-06,
+      "loss": 0.1293,
+      "step": 1956
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 4.761468592139051,
+      "learning_rate": 7.559000544409991e-06,
+      "loss": 0.0891,
+      "step": 1957
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.534845276627891,
+      "learning_rate": 7.556517176306704e-06,
+      "loss": 0.0975,
+      "step": 1958
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 18.880359885969494,
+      "learning_rate": 7.554032954023508e-06,
+      "loss": 0.0979,
+      "step": 1959
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 15.143403477025016,
+      "learning_rate": 7.55154787839043e-06,
+      "loss": 0.1144,
+      "step": 1960
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 15.521438570335969,
+      "learning_rate": 7.5490619502377805e-06,
+      "loss": 0.1129,
+      "step": 1961
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 18.5242277495622,
+      "learning_rate": 7.546575170396153e-06,
+      "loss": 0.1074,
+      "step": 1962
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 24.927697963385498,
+      "learning_rate": 7.544087539696427e-06,
+      "loss": 0.1021,
+      "step": 1963
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.703714901569482,
+      "learning_rate": 7.541599058969766e-06,
+      "loss": 0.0956,
+      "step": 1964
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 34.83739720762142,
+      "learning_rate": 7.539109729047619e-06,
+      "loss": 0.121,
+      "step": 1965
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 10.37898184340536,
+      "learning_rate": 7.5366195507617155e-06,
+      "loss": 0.0914,
+      "step": 1966
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 32.17544750214366,
+      "learning_rate": 7.534128524944071e-06,
+      "loss": 0.0848,
+      "step": 1967
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 47.63891616778719,
+      "learning_rate": 7.531636652426985e-06,
+      "loss": 0.144,
+      "step": 1968
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 12.949628774093883,
+      "learning_rate": 7.529143934043036e-06,
+      "loss": 0.1061,
+      "step": 1969
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 32.96795634535262,
+      "learning_rate": 7.526650370625088e-06,
+      "loss": 0.1438,
+      "step": 1970
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 55.11333254799953,
+      "learning_rate": 7.5241559630062896e-06,
+      "loss": 0.1469,
+      "step": 1971
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 18.925573978897116,
+      "learning_rate": 7.5216607120200655e-06,
+      "loss": 0.0874,
+      "step": 1972
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 21.262962891421868,
+      "learning_rate": 7.519164618500127e-06,
+      "loss": 0.0812,
+      "step": 1973
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 41.7596554934021,
+      "learning_rate": 7.5166676832804655e-06,
+      "loss": 0.1268,
+      "step": 1974
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 11.588135089329397,
+      "learning_rate": 7.514169907195352e-06,
+      "loss": 0.0864,
+      "step": 1975
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 5.342692979719796,
+      "learning_rate": 7.511671291079342e-06,
+      "loss": 0.0705,
+      "step": 1976
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 5.712686440388007,
+      "learning_rate": 7.509171835767268e-06,
+      "loss": 0.0907,
+      "step": 1977
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 36.70804565028893,
+      "learning_rate": 7.506671542094246e-06,
+      "loss": 0.1252,
+      "step": 1978
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 10.489002626915171,
+      "learning_rate": 7.504170410895668e-06,
+      "loss": 0.0996,
+      "step": 1979
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 6.3444733677159295,
+      "learning_rate": 7.501668443007212e-06,
+      "loss": 0.0735,
+      "step": 1980
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 11.093855896464442,
+      "learning_rate": 7.499165639264828e-06,
+      "loss": 0.115,
+      "step": 1981
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 33.7558992291076,
+      "learning_rate": 7.496662000504752e-06,
+      "loss": 0.165,
+      "step": 1982
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 10.907611777929603,
+      "learning_rate": 7.4941575275634945e-06,
+      "loss": 0.11,
+      "step": 1983
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 28.17923318389303,
+      "learning_rate": 7.49165222127785e-06,
+      "loss": 0.0878,
+      "step": 1984
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 15.100103791027154,
+      "learning_rate": 7.489146082484882e-06,
+      "loss": 0.0727,
+      "step": 1985
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 12.562326160266043,
+      "learning_rate": 7.486639112021944e-06,
+      "loss": 0.0812,
+      "step": 1986
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 31.277970594845105,
+      "learning_rate": 7.484131310726658e-06,
+      "loss": 0.0872,
+      "step": 1987
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 14.473359713155952,
+      "learning_rate": 7.481622679436929e-06,
+      "loss": 0.0759,
+      "step": 1988
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.391891563984436,
+      "learning_rate": 7.479113218990934e-06,
+      "loss": 0.0832,
+      "step": 1989
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 6.8735457218956455,
+      "learning_rate": 7.4766029302271335e-06,
+      "loss": 0.108,
+      "step": 1990
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 17.97877074755628,
+      "learning_rate": 7.474091813984261e-06,
+      "loss": 0.0837,
+      "step": 1991
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 15.971760147819229,
+      "learning_rate": 7.471579871101326e-06,
+      "loss": 0.1023,
+      "step": 1992
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.376728168558531,
+      "learning_rate": 7.4690671024176165e-06,
+      "loss": 0.1162,
+      "step": 1993
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 9.360966837136843,
+      "learning_rate": 7.466553508772695e-06,
+      "loss": 0.0769,
+      "step": 1994
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 8.398769699308323,
+      "learning_rate": 7.4640390910064e-06,
+      "loss": 0.093,
+      "step": 1995
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 19.282073477802033,
+      "learning_rate": 7.461523849958845e-06,
+      "loss": 0.1223,
+      "step": 1996
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 9.488837006383456,
+      "learning_rate": 7.459007786470418e-06,
+      "loss": 0.0822,
+      "step": 1997
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 16.12399085356851,
+      "learning_rate": 7.4564909013817845e-06,
+      "loss": 0.1228,
+      "step": 1998
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 7.449212798239846,
+      "learning_rate": 7.45397319553388e-06,
+      "loss": 0.0666,
+      "step": 1999
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 30.988058304482948,
+      "learning_rate": 7.451454669767919e-06,
+      "loss": 0.1119,
+      "step": 2000
+    },
+    {
+      "epoch": 1.43,
+      "eval_avg_AUC": 0.7993414523799219,
+      "eval_avg_Accuracy": 0.6962864721485411,
+      "eval_avg_Accuracy-right": 0.895069779574801,
+      "eval_avg_Accuracy-wrong": 0.34967022970206957,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6534253440240803,
+      "eval_last_AUC": 0.796335909721104,
+      "eval_last_Accuracy": 0.7389340185676393,
+      "eval_last_Accuracy-right": 0.8014868918742664,
+      "eval_last_Accuracy-wrong": 0.6298612690470775,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.6609678828685568,
+      "eval_max_AUC": 0.731932816756531,
+      "eval_max_Accuracy": 0.6443136604774535,
+      "eval_max_Accuracy-right": 0.9846745793661145,
+      "eval_max_Accuracy-wrong": 0.05083011143961792,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.597156835188284,
+      "eval_min_AUC": 0.8068730624550771,
+      "eval_min_Accuracy": 0.7409234084880637,
+      "eval_min_Accuracy-right": 0.7598800052171645,
+      "eval_min_Accuracy-wrong": 0.7078690015919945,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.662910126847143,
+      "eval_prod_AUC": 0.8081306191935946,
+      "eval_prod_Accuracy": 0.7138594164456233,
+      "eval_prod_Accuracy-right": 0.6302334681100822,
+      "eval_prod_Accuracy-wrong": 0.8596770525358198,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.653658027762353,
+      "eval_runtime": 246.3352,
+      "eval_samples_per_second": 97.948,
+      "eval_steps_per_second": 3.061,
+      "eval_sum_AUC": 0.6609696244629366,
+      "eval_sum_Accuracy": 0.6379310344827587,
+      "eval_sum_Accuracy-right": 0.9988261379940002,
+      "eval_sum_Accuracy-wrong": 0.008642256083693428,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6421408268547636,
+      "step": 2000
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 13.839120903654035,
+      "learning_rate": 7.448935324925386e-06,
+      "loss": 0.0573,
+      "step": 2001
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 6.253220256521355,
+      "learning_rate": 7.446415161848043e-06,
+      "loss": 0.0869,
+      "step": 2002
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.932247825123236,
+      "learning_rate": 7.443894181377921e-06,
+      "loss": 0.0869,
+      "step": 2003
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 14.774479223409482,
+      "learning_rate": 7.441372384357328e-06,
+      "loss": 0.0901,
+      "step": 2004
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 28.556495798503843,
+      "learning_rate": 7.438849771628844e-06,
+      "loss": 0.1105,
+      "step": 2005
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.991255023773615,
+      "learning_rate": 7.43632634403532e-06,
+      "loss": 0.0887,
+      "step": 2006
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.019756556058779,
+      "learning_rate": 7.433802102419878e-06,
+      "loss": 0.1019,
+      "step": 2007
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 26.827197018870162,
+      "learning_rate": 7.431277047625918e-06,
+      "loss": 0.1045,
+      "step": 2008
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 40.7772525438193,
+      "learning_rate": 7.428751180497104e-06,
+      "loss": 0.1014,
+      "step": 2009
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 10.794657275533215,
+      "learning_rate": 7.426224501877376e-06,
+      "loss": 0.1172,
+      "step": 2010
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 4.9017737328348465,
+      "learning_rate": 7.423697012610947e-06,
+      "loss": 0.0688,
+      "step": 2011
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 32.11030414551294,
+      "learning_rate": 7.421168713542294e-06,
+      "loss": 0.1136,
+      "step": 2012
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 23.50894816197612,
+      "learning_rate": 7.418639605516172e-06,
+      "loss": 0.103,
+      "step": 2013
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 7.586007382354752,
+      "learning_rate": 7.416109689377603e-06,
+      "loss": 0.121,
+      "step": 2014
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.23229798707242,
+      "learning_rate": 7.413578965971876e-06,
+      "loss": 0.0738,
+      "step": 2015
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 5.177266007361741,
+      "learning_rate": 7.411047436144556e-06,
+      "loss": 0.0636,
+      "step": 2016
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 7.750691106855549,
+      "learning_rate": 7.408515100741471e-06,
+      "loss": 0.1152,
+      "step": 2017
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 33.35658858310154,
+      "learning_rate": 7.405981960608725e-06,
+      "loss": 0.0986,
+      "step": 2018
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.949061067234727,
+      "learning_rate": 7.403448016592685e-06,
+      "loss": 0.0984,
+      "step": 2019
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 30.906573797223782,
+      "learning_rate": 7.400913269539988e-06,
+      "loss": 0.119,
+      "step": 2020
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 14.698468315096022,
+      "learning_rate": 7.398377720297541e-06,
+      "loss": 0.0934,
+      "step": 2021
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 38.759538949910294,
+      "learning_rate": 7.39584136971252e-06,
+      "loss": 0.1078,
+      "step": 2022
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 12.61303665133732,
+      "learning_rate": 7.393304218632364e-06,
+      "loss": 0.0614,
+      "step": 2023
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 33.99016387504461,
+      "learning_rate": 7.390766267904783e-06,
+      "loss": 0.1074,
+      "step": 2024
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 25.92261989754995,
+      "learning_rate": 7.3882275183777554e-06,
+      "loss": 0.1655,
+      "step": 2025
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 40.11705918481798,
+      "learning_rate": 7.385687970899523e-06,
+      "loss": 0.1057,
+      "step": 2026
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.286685919964656,
+      "learning_rate": 7.3831476263185965e-06,
+      "loss": 0.0775,
+      "step": 2027
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 15.0192688550375,
+      "learning_rate": 7.380606485483751e-06,
+      "loss": 0.0768,
+      "step": 2028
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 55.38449400925222,
+      "learning_rate": 7.378064549244031e-06,
+      "loss": 0.1389,
+      "step": 2029
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 23.97034166915613,
+      "learning_rate": 7.375521818448741e-06,
+      "loss": 0.1333,
+      "step": 2030
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 13.121072563372268,
+      "learning_rate": 7.372978293947459e-06,
+      "loss": 0.0995,
+      "step": 2031
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 45.71363021283654,
+      "learning_rate": 7.3704339765900205e-06,
+      "loss": 0.1544,
+      "step": 2032
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 29.422717589921906,
+      "learning_rate": 7.367888867226531e-06,
+      "loss": 0.1324,
+      "step": 2033
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 38.96958949899909,
+      "learning_rate": 7.365342966707359e-06,
+      "loss": 0.1,
+      "step": 2034
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 6.289954542505453,
+      "learning_rate": 7.362796275883135e-06,
+      "loss": 0.0754,
+      "step": 2035
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 10.57359176786369,
+      "learning_rate": 7.360248795604758e-06,
+      "loss": 0.0861,
+      "step": 2036
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 24.090251255820395,
+      "learning_rate": 7.3577005267233885e-06,
+      "loss": 0.0708,
+      "step": 2037
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 37.93968118245042,
+      "learning_rate": 7.355151470090449e-06,
+      "loss": 0.1194,
+      "step": 2038
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 13.449078150315941,
+      "learning_rate": 7.352601626557628e-06,
+      "loss": 0.0925,
+      "step": 2039
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 18.95997171971615,
+      "learning_rate": 7.350050996976875e-06,
+      "loss": 0.1005,
+      "step": 2040
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 26.92934782855485,
+      "learning_rate": 7.347499582200404e-06,
+      "loss": 0.0961,
+      "step": 2041
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 22.73730072835748,
+      "learning_rate": 7.344947383080687e-06,
+      "loss": 0.1058,
+      "step": 2042
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 22.503461102204398,
+      "learning_rate": 7.342394400470463e-06,
+      "loss": 0.0862,
+      "step": 2043
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 20.649364677252795,
+      "learning_rate": 7.339840635222732e-06,
+      "loss": 0.0664,
+      "step": 2044
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 11.339548594219169,
+      "learning_rate": 7.337286088190754e-06,
+      "loss": 0.0889,
+      "step": 2045
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 11.400084265838423,
+      "learning_rate": 7.334730760228049e-06,
+      "loss": 0.0951,
+      "step": 2046
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 14.486076375014688,
+      "learning_rate": 7.332174652188401e-06,
+      "loss": 0.1224,
+      "step": 2047
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 34.7417013692277,
+      "learning_rate": 7.329617764925853e-06,
+      "loss": 0.1178,
+      "step": 2048
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 30.777860474477603,
+      "learning_rate": 7.32706009929471e-06,
+      "loss": 0.1587,
+      "step": 2049
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 18.592979828725138,
+      "learning_rate": 7.324501656149532e-06,
+      "loss": 0.1007,
+      "step": 2050
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 42.347294863149365,
+      "learning_rate": 7.321942436345146e-06,
+      "loss": 0.1045,
+      "step": 2051
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 21.001661806227776,
+      "learning_rate": 7.319382440736632e-06,
+      "loss": 0.1101,
+      "step": 2052
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 19.577035199548465,
+      "learning_rate": 7.316821670179335e-06,
+      "loss": 0.1144,
+      "step": 2053
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.230363742790843,
+      "learning_rate": 7.314260125528854e-06,
+      "loss": 0.1404,
+      "step": 2054
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 17.499694248271904,
+      "learning_rate": 7.311697807641048e-06,
+      "loss": 0.0714,
+      "step": 2055
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 23.806058024752186,
+      "learning_rate": 7.3091347173720386e-06,
+      "loss": 0.1033,
+      "step": 2056
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 6.085617803027079,
+      "learning_rate": 7.3065708555781986e-06,
+      "loss": 0.0939,
+      "step": 2057
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 34.50468476483311,
+      "learning_rate": 7.304006223116162e-06,
+      "loss": 0.1208,
+      "step": 2058
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 13.365971501697732,
+      "learning_rate": 7.301440820842822e-06,
+      "loss": 0.09,
+      "step": 2059
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 10.044928264199827,
+      "learning_rate": 7.298874649615327e-06,
+      "loss": 0.153,
+      "step": 2060
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 28.362345328223054,
+      "learning_rate": 7.29630771029108e-06,
+      "loss": 0.0945,
+      "step": 2061
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 6.658453340675622,
+      "learning_rate": 7.293740003727745e-06,
+      "loss": 0.0809,
+      "step": 2062
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 15.121775553304609,
+      "learning_rate": 7.291171530783241e-06,
+      "loss": 0.1129,
+      "step": 2063
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 11.416830583602653,
+      "learning_rate": 7.288602292315742e-06,
+      "loss": 0.1333,
+      "step": 2064
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 5.084441019797188,
+      "learning_rate": 7.286032289183679e-06,
+      "loss": 0.1097,
+      "step": 2065
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 15.642621559627102,
+      "learning_rate": 7.283461522245736e-06,
+      "loss": 0.1122,
+      "step": 2066
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 41.35476137894392,
+      "learning_rate": 7.280889992360856e-06,
+      "loss": 0.1206,
+      "step": 2067
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 21.18129028339432,
+      "learning_rate": 7.278317700388232e-06,
+      "loss": 0.1133,
+      "step": 2068
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 19.902882472946708,
+      "learning_rate": 7.275744647187318e-06,
+      "loss": 0.1512,
+      "step": 2069
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 15.939602818419818,
+      "learning_rate": 7.273170833617818e-06,
+      "loss": 0.1044,
+      "step": 2070
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 31.533034131028536,
+      "learning_rate": 7.2705962605396895e-06,
+      "loss": 0.1394,
+      "step": 2071
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 9.111943201249398,
+      "learning_rate": 7.268020928813147e-06,
+      "loss": 0.0912,
+      "step": 2072
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 13.514191499285351,
+      "learning_rate": 7.265444839298656e-06,
+      "loss": 0.0923,
+      "step": 2073
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 6.330749877489256,
+      "learning_rate": 7.262867992856934e-06,
+      "loss": 0.0975,
+      "step": 2074
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 30.935621396553604,
+      "learning_rate": 7.260290390348956e-06,
+      "loss": 0.1388,
+      "step": 2075
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 6.688613763402811,
+      "learning_rate": 7.257712032635946e-06,
+      "loss": 0.1047,
+      "step": 2076
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 13.33335451161966,
+      "learning_rate": 7.255132920579382e-06,
+      "loss": 0.1041,
+      "step": 2077
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.619130876094033,
+      "learning_rate": 7.252553055040991e-06,
+      "loss": 0.0897,
+      "step": 2078
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 12.168562646783498,
+      "learning_rate": 7.249972436882756e-06,
+      "loss": 0.1006,
+      "step": 2079
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 14.278860858221417,
+      "learning_rate": 7.247391066966909e-06,
+      "loss": 0.0814,
+      "step": 2080
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 13.260089786437403,
+      "learning_rate": 7.244808946155933e-06,
+      "loss": 0.0886,
+      "step": 2081
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 5.521768030802181,
+      "learning_rate": 7.242226075312564e-06,
+      "loss": 0.0858,
+      "step": 2082
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 13.198034640001179,
+      "learning_rate": 7.239642455299787e-06,
+      "loss": 0.0963,
+      "step": 2083
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 35.149002150189034,
+      "learning_rate": 7.237058086980835e-06,
+      "loss": 0.1558,
+      "step": 2084
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 10.810236800782775,
+      "learning_rate": 7.234472971219197e-06,
+      "loss": 0.08,
+      "step": 2085
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 14.523629121141207,
+      "learning_rate": 7.231887108878606e-06,
+      "loss": 0.1177,
+      "step": 2086
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.493707671164756,
+      "learning_rate": 7.229300500823047e-06,
+      "loss": 0.1161,
+      "step": 2087
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 7.760754279339331,
+      "learning_rate": 7.226713147916754e-06,
+      "loss": 0.0781,
+      "step": 2088
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 10.837060450720395,
+      "learning_rate": 7.22412505102421e-06,
+      "loss": 0.0732,
+      "step": 2089
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 21.583304857531907,
+      "learning_rate": 7.221536211010147e-06,
+      "loss": 0.0729,
+      "step": 2090
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 12.445127735446684,
+      "learning_rate": 7.2189466287395425e-06,
+      "loss": 0.0918,
+      "step": 2091
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 8.94159436880388,
+      "learning_rate": 7.216356305077625e-06,
+      "loss": 0.1051,
+      "step": 2092
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 18.138755911165397,
+      "learning_rate": 7.21376524088987e-06,
+      "loss": 0.1355,
+      "step": 2093
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 15.689142765328521,
+      "learning_rate": 7.211173437042001e-06,
+      "loss": 0.0781,
+      "step": 2094
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 37.30887264410398,
+      "learning_rate": 7.208580894399986e-06,
+      "loss": 0.1034,
+      "step": 2095
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 27.998517100165653,
+      "learning_rate": 7.205987613830043e-06,
+      "loss": 0.1226,
+      "step": 2096
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 27.401516777324407,
+      "learning_rate": 7.203393596198635e-06,
+      "loss": 0.1133,
+      "step": 2097
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 23.244970427895833,
+      "learning_rate": 7.200798842372472e-06,
+      "loss": 0.1119,
+      "step": 2098
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 24.537751698472384,
+      "learning_rate": 7.198203353218508e-06,
+      "loss": 0.1145,
+      "step": 2099
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 6.5385945727391395,
+      "learning_rate": 7.195607129603946e-06,
+      "loss": 0.0945,
+      "step": 2100
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 17.09406318610685,
+      "learning_rate": 7.19301017239623e-06,
+      "loss": 0.0776,
+      "step": 2101
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.953522177542275,
+      "learning_rate": 7.190412482463054e-06,
+      "loss": 0.1013,
+      "step": 2102
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 31.262974276965277,
+      "learning_rate": 7.187814060672354e-06,
+      "loss": 0.1171,
+      "step": 2103
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 10.270590362702018,
+      "learning_rate": 7.1852149078923105e-06,
+      "loss": 0.1014,
+      "step": 2104
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 12.499466735424543,
+      "learning_rate": 7.1826150249913495e-06,
+      "loss": 0.131,
+      "step": 2105
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 26.68680284528321,
+      "learning_rate": 7.18001441283814e-06,
+      "loss": 0.0918,
+      "step": 2106
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 42.91007158573037,
+      "learning_rate": 7.1774130723015955e-06,
+      "loss": 0.1384,
+      "step": 2107
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 8.298129546397226,
+      "learning_rate": 7.17481100425087e-06,
+      "loss": 0.0711,
+      "step": 2108
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 22.30804777349356,
+      "learning_rate": 7.172208209555365e-06,
+      "loss": 0.077,
+      "step": 2109
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 21.45735632031128,
+      "learning_rate": 7.1696046890847206e-06,
+      "loss": 0.1058,
+      "step": 2110
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 21.40607494365893,
+      "learning_rate": 7.167000443708823e-06,
+      "loss": 0.1253,
+      "step": 2111
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 7.216056021799717,
+      "learning_rate": 7.164395474297798e-06,
+      "loss": 0.067,
+      "step": 2112
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 22.5190941505317,
+      "learning_rate": 7.161789781722016e-06,
+      "loss": 0.1038,
+      "step": 2113
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 24.586064614156214,
+      "learning_rate": 7.159183366852085e-06,
+      "loss": 0.1046,
+      "step": 2114
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 11.51113614776396,
+      "learning_rate": 7.156576230558859e-06,
+      "loss": 0.1046,
+      "step": 2115
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 70.95164168879344,
+      "learning_rate": 7.153968373713429e-06,
+      "loss": 0.1827,
+      "step": 2116
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 6.699926065717833,
+      "learning_rate": 7.1513597971871295e-06,
+      "loss": 0.0992,
+      "step": 2117
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 11.794229177328235,
+      "learning_rate": 7.148750501851532e-06,
+      "loss": 0.0793,
+      "step": 2118
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 18.355183637296523,
+      "learning_rate": 7.1461404885784545e-06,
+      "loss": 0.1051,
+      "step": 2119
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 32.86488907052558,
+      "learning_rate": 7.1435297582399475e-06,
+      "loss": 0.1007,
+      "step": 2120
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 21.387125829223244,
+      "learning_rate": 7.140918311708306e-06,
+      "loss": 0.0792,
+      "step": 2121
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 13.590993567540462,
+      "learning_rate": 7.138306149856062e-06,
+      "loss": 0.075,
+      "step": 2122
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 6.447054862968271,
+      "learning_rate": 7.1356932735559905e-06,
+      "loss": 0.085,
+      "step": 2123
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 25.482506147936167,
+      "learning_rate": 7.133079683681099e-06,
+      "loss": 0.1274,
+      "step": 2124
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 9.387675092660638,
+      "learning_rate": 7.130465381104635e-06,
+      "loss": 0.0701,
+      "step": 2125
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.715433430188718,
+      "learning_rate": 7.1278503667000885e-06,
+      "loss": 0.0797,
+      "step": 2126
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 8.423200658182854,
+      "learning_rate": 7.125234641341185e-06,
+      "loss": 0.0934,
+      "step": 2127
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 6.710068469682492,
+      "learning_rate": 7.1226182059018835e-06,
+      "loss": 0.0822,
+      "step": 2128
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 13.329201988437877,
+      "learning_rate": 7.120001061256387e-06,
+      "loss": 0.0792,
+      "step": 2129
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 22.059536956159867,
+      "learning_rate": 7.1173832082791294e-06,
+      "loss": 0.1038,
+      "step": 2130
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 13.826617022708637,
+      "learning_rate": 7.114764647844788e-06,
+      "loss": 0.1299,
+      "step": 2131
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 17.18931427687854,
+      "learning_rate": 7.112145380828267e-06,
+      "loss": 0.0983,
+      "step": 2132
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 12.456228718757368,
+      "learning_rate": 7.109525408104717e-06,
+      "loss": 0.125,
+      "step": 2133
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 12.457600501785418,
+      "learning_rate": 7.106904730549517e-06,
+      "loss": 0.0661,
+      "step": 2134
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 19.169864989415867,
+      "learning_rate": 7.104283349038285e-06,
+      "loss": 0.0616,
+      "step": 2135
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 28.451596210521583,
+      "learning_rate": 7.101661264446875e-06,
+      "loss": 0.1067,
+      "step": 2136
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 11.557672547776232,
+      "learning_rate": 7.099038477651371e-06,
+      "loss": 0.0716,
+      "step": 2137
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 13.915155495212185,
+      "learning_rate": 7.096414989528095e-06,
+      "loss": 0.0714,
+      "step": 2138
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 28.111132386929246,
+      "learning_rate": 7.093790800953606e-06,
+      "loss": 0.0987,
+      "step": 2139
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 13.342771869341991,
+      "learning_rate": 7.091165912804693e-06,
+      "loss": 0.1157,
+      "step": 2140
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.294032473258458,
+      "learning_rate": 7.088540325958379e-06,
+      "loss": 0.1033,
+      "step": 2141
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 6.93465060964492,
+      "learning_rate": 7.085914041291921e-06,
+      "loss": 0.0911,
+      "step": 2142
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 17.880292154170586,
+      "learning_rate": 7.08328705968281e-06,
+      "loss": 0.1404,
+      "step": 2143
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 22.929359757090396,
+      "learning_rate": 7.080659382008772e-06,
+      "loss": 0.1053,
+      "step": 2144
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 7.958105578684251,
+      "learning_rate": 7.078031009147759e-06,
+      "loss": 0.1392,
+      "step": 2145
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 14.942761890735525,
+      "learning_rate": 7.075401941977961e-06,
+      "loss": 0.0994,
+      "step": 2146
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 6.000079482176476,
+      "learning_rate": 7.072772181377798e-06,
+      "loss": 0.0935,
+      "step": 2147
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 9.21102239643325,
+      "learning_rate": 7.070141728225922e-06,
+      "loss": 0.0652,
+      "step": 2148
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 10.92020872783142,
+      "learning_rate": 7.067510583401217e-06,
+      "loss": 0.0845,
+      "step": 2149
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 4.251963896243572,
+      "learning_rate": 7.0648787477827965e-06,
+      "loss": 0.0601,
+      "step": 2150
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 7.8790995364041905,
+      "learning_rate": 7.062246222250005e-06,
+      "loss": 0.1079,
+      "step": 2151
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 16.979064512752753,
+      "learning_rate": 7.05961300768242e-06,
+      "loss": 0.1107,
+      "step": 2152
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 39.1816186306808,
+      "learning_rate": 7.056979104959847e-06,
+      "loss": 0.1908,
+      "step": 2153
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 15.074164527605571,
+      "learning_rate": 7.054344514962319e-06,
+      "loss": 0.0779,
+      "step": 2154
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 19.191667670584966,
+      "learning_rate": 7.051709238570106e-06,
+      "loss": 0.1405,
+      "step": 2155
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 27.248491502820745,
+      "learning_rate": 7.0490732766637e-06,
+      "loss": 0.1021,
+      "step": 2156
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 9.506985241043234,
+      "learning_rate": 7.046436630123826e-06,
+      "loss": 0.1384,
+      "step": 2157
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 6.615642105259546,
+      "learning_rate": 7.043799299831438e-06,
+      "loss": 0.1138,
+      "step": 2158
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 8.07347512799611,
+      "learning_rate": 7.041161286667713e-06,
+      "loss": 0.1147,
+      "step": 2159
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 24.75914750945603,
+      "learning_rate": 7.038522591514061e-06,
+      "loss": 0.0977,
+      "step": 2160
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 20.871761477687055,
+      "learning_rate": 7.035883215252123e-06,
+      "loss": 0.0725,
+      "step": 2161
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 7.2177050883819165,
+      "learning_rate": 7.03324315876376e-06,
+      "loss": 0.0817,
+      "step": 2162
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 10.886438037155624,
+      "learning_rate": 7.030602422931065e-06,
+      "loss": 0.0947,
+      "step": 2163
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 32.18182550206542,
+      "learning_rate": 7.027961008636359e-06,
+      "loss": 0.1033,
+      "step": 2164
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 10.8123263405742,
+      "learning_rate": 7.025318916762185e-06,
+      "loss": 0.0902,
+      "step": 2165
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 5.74661964617569,
+      "learning_rate": 7.022676148191315e-06,
+      "loss": 0.1548,
+      "step": 2166
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.66124484632883,
+      "learning_rate": 7.020032703806748e-06,
+      "loss": 0.1166,
+      "step": 2167
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 5.129462459098309,
+      "learning_rate": 7.017388584491709e-06,
+      "loss": 0.0863,
+      "step": 2168
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 19.70040390124027,
+      "learning_rate": 7.014743791129644e-06,
+      "loss": 0.1052,
+      "step": 2169
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 4.783745164454452,
+      "learning_rate": 7.012098324604231e-06,
+      "loss": 0.0804,
+      "step": 2170
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 5.83062015200628,
+      "learning_rate": 7.009452185799368e-06,
+      "loss": 0.0876,
+      "step": 2171
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 8.96404912676571,
+      "learning_rate": 7.00680537559918e-06,
+      "loss": 0.0917,
+      "step": 2172
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 26.88319898284892,
+      "learning_rate": 7.0041578948880155e-06,
+      "loss": 0.0736,
+      "step": 2173
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 11.385232172270138,
+      "learning_rate": 7.001509744550446e-06,
+      "loss": 0.1053,
+      "step": 2174
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 7.771562344479279,
+      "learning_rate": 6.998860925471267e-06,
+      "loss": 0.0853,
+      "step": 2175
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 18.270461704058164,
+      "learning_rate": 6.9962114385355e-06,
+      "loss": 0.0884,
+      "step": 2176
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 43.811495080381825,
+      "learning_rate": 6.993561284628388e-06,
+      "loss": 0.1189,
+      "step": 2177
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 9.153350063653045,
+      "learning_rate": 6.990910464635395e-06,
+      "loss": 0.075,
+      "step": 2178
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 16.548387260500697,
+      "learning_rate": 6.9882589794422105e-06,
+      "loss": 0.0931,
+      "step": 2179
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 25.197771744639976,
+      "learning_rate": 6.9856068299347455e-06,
+      "loss": 0.1284,
+      "step": 2180
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 30.717886675806074,
+      "learning_rate": 6.98295401699913e-06,
+      "loss": 0.0989,
+      "step": 2181
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 22.815922730000807,
+      "learning_rate": 6.980300541521721e-06,
+      "loss": 0.1202,
+      "step": 2182
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 8.881826965201908,
+      "learning_rate": 6.977646404389092e-06,
+      "loss": 0.0764,
+      "step": 2183
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 74.805756583807,
+      "learning_rate": 6.9749916064880404e-06,
+      "loss": 0.1982,
+      "step": 2184
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 15.273061210962908,
+      "learning_rate": 6.972336148705583e-06,
+      "loss": 0.0898,
+      "step": 2185
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 26.839917671298732,
+      "learning_rate": 6.969680031928959e-06,
+      "loss": 0.1118,
+      "step": 2186
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 18.784018802631863,
+      "learning_rate": 6.967023257045624e-06,
+      "loss": 0.1224,
+      "step": 2187
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 42.06255919354335,
+      "learning_rate": 6.96436582494326e-06,
+      "loss": 0.0928,
+      "step": 2188
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 38.1069266876009,
+      "learning_rate": 6.961707736509759e-06,
+      "loss": 0.1373,
+      "step": 2189
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 29.757725569858078,
+      "learning_rate": 6.959048992633241e-06,
+      "loss": 0.0897,
+      "step": 2190
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 4.726770945951311,
+      "learning_rate": 6.956389594202041e-06,
+      "loss": 0.0917,
+      "step": 2191
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 36.79881925491694,
+      "learning_rate": 6.953729542104713e-06,
+      "loss": 0.1097,
+      "step": 2192
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 44.56686035867685,
+      "learning_rate": 6.951068837230032e-06,
+      "loss": 0.1515,
+      "step": 2193
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 27.897014911286362,
+      "learning_rate": 6.9484074804669865e-06,
+      "loss": 0.1221,
+      "step": 2194
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 17.338158874149745,
+      "learning_rate": 6.945745472704786e-06,
+      "loss": 0.1188,
+      "step": 2195
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 38.83775106356088,
+      "learning_rate": 6.943082814832858e-06,
+      "loss": 0.0985,
+      "step": 2196
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 44.22872017990922,
+      "learning_rate": 6.940419507740843e-06,
+      "loss": 0.1453,
+      "step": 2197
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 23.616953797426014,
+      "learning_rate": 6.937755552318606e-06,
+      "loss": 0.0958,
+      "step": 2198
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 16.228217397227283,
+      "learning_rate": 6.935090949456219e-06,
+      "loss": 0.0956,
+      "step": 2199
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 16.949120884184445,
+      "learning_rate": 6.93242570004398e-06,
+      "loss": 0.1097,
+      "step": 2200
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 45.65816773721843,
+      "learning_rate": 6.929759804972394e-06,
+      "loss": 0.1361,
+      "step": 2201
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 28.0710224178119,
+      "learning_rate": 6.92709326513219e-06,
+      "loss": 0.11,
+      "step": 2202
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 13.390078383095538,
+      "learning_rate": 6.924426081414305e-06,
+      "loss": 0.088,
+      "step": 2203
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 5.427364387155767,
+      "learning_rate": 6.921758254709897e-06,
+      "loss": 0.0829,
+      "step": 2204
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 26.214900361222632,
+      "learning_rate": 6.919089785910336e-06,
+      "loss": 0.0849,
+      "step": 2205
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 31.71611812858684,
+      "learning_rate": 6.916420675907207e-06,
+      "loss": 0.0957,
+      "step": 2206
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 21.55026185223569,
+      "learning_rate": 6.9137509255923085e-06,
+      "loss": 0.1045,
+      "step": 2207
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 5.845490265416999,
+      "learning_rate": 6.911080535857655e-06,
+      "loss": 0.0778,
+      "step": 2208
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 18.755449635394918,
+      "learning_rate": 6.908409507595472e-06,
+      "loss": 0.114,
+      "step": 2209
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 27.69838224283327,
+      "learning_rate": 6.905737841698201e-06,
+      "loss": 0.1097,
+      "step": 2210
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 13.295405659497893,
+      "learning_rate": 6.903065539058496e-06,
+      "loss": 0.0918,
+      "step": 2211
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.442426195870846,
+      "learning_rate": 6.900392600569219e-06,
+      "loss": 0.0647,
+      "step": 2212
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 8.137208442510627,
+      "learning_rate": 6.897719027123451e-06,
+      "loss": 0.1304,
+      "step": 2213
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 29.345677962603688,
+      "learning_rate": 6.895044819614484e-06,
+      "loss": 0.1,
+      "step": 2214
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 17.490421643869904,
+      "learning_rate": 6.8923699789358185e-06,
+      "loss": 0.0791,
+      "step": 2215
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 14.661343717167206,
+      "learning_rate": 6.88969450598117e-06,
+      "loss": 0.068,
+      "step": 2216
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 18.33228470361208,
+      "learning_rate": 6.887018401644463e-06,
+      "loss": 0.1045,
+      "step": 2217
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.382139849610716,
+      "learning_rate": 6.884341666819832e-06,
+      "loss": 0.0778,
+      "step": 2218
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 9.342522202577735,
+      "learning_rate": 6.881664302401626e-06,
+      "loss": 0.0865,
+      "step": 2219
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.14544309202564,
+      "learning_rate": 6.878986309284401e-06,
+      "loss": 0.0933,
+      "step": 2220
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 6.725068137788035,
+      "learning_rate": 6.876307688362925e-06,
+      "loss": 0.0895,
+      "step": 2221
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 5.054484543343092,
+      "learning_rate": 6.873628440532175e-06,
+      "loss": 0.0834,
+      "step": 2222
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 10.231322227289729,
+      "learning_rate": 6.8709485666873375e-06,
+      "loss": 0.104,
+      "step": 2223
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 5.956539872126496,
+      "learning_rate": 6.868268067723808e-06,
+      "loss": 0.1002,
+      "step": 2224
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 17.94119750705428,
+      "learning_rate": 6.86558694453719e-06,
+      "loss": 0.1049,
+      "step": 2225
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 7.710067333196714,
+      "learning_rate": 6.8629051980233e-06,
+      "loss": 0.0728,
+      "step": 2226
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 13.507965465766109,
+      "learning_rate": 6.860222829078156e-06,
+      "loss": 0.0928,
+      "step": 2227
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 15.267717469976887,
+      "learning_rate": 6.857539838597987e-06,
+      "loss": 0.0722,
+      "step": 2228
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 6.716234484479027,
+      "learning_rate": 6.8548562274792325e-06,
+      "loss": 0.0989,
+      "step": 2229
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 32.705783237669955,
+      "learning_rate": 6.8521719966185355e-06,
+      "loss": 0.1067,
+      "step": 2230
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 9.096469392127245,
+      "learning_rate": 6.8494871469127474e-06,
+      "loss": 0.1183,
+      "step": 2231
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 7.547206226988792,
+      "learning_rate": 6.846801679258926e-06,
+      "loss": 0.1047,
+      "step": 2232
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 13.029631509644526,
+      "learning_rate": 6.844115594554338e-06,
+      "loss": 0.0861,
+      "step": 2233
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 43.40421796011947,
+      "learning_rate": 6.841428893696453e-06,
+      "loss": 0.1598,
+      "step": 2234
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 18.30412375137895,
+      "learning_rate": 6.838741577582946e-06,
+      "loss": 0.1125,
+      "step": 2235
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 4.815879130885394,
+      "learning_rate": 6.836053647111701e-06,
+      "loss": 0.0886,
+      "step": 2236
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 20.612409133400302,
+      "learning_rate": 6.833365103180806e-06,
+      "loss": 0.1157,
+      "step": 2237
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 23.966186458842202,
+      "learning_rate": 6.830675946688552e-06,
+      "loss": 0.0892,
+      "step": 2238
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 7.940623113295723,
+      "learning_rate": 6.827986178533437e-06,
+      "loss": 0.1117,
+      "step": 2239
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 7.44672935549102,
+      "learning_rate": 6.825295799614163e-06,
+      "loss": 0.1162,
+      "step": 2240
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 14.891138354626657,
+      "learning_rate": 6.822604810829634e-06,
+      "loss": 0.0913,
+      "step": 2241
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 17.653225783254584,
+      "learning_rate": 6.819913213078961e-06,
+      "loss": 0.0951,
+      "step": 2242
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 5.9242045134536525,
+      "learning_rate": 6.817221007261456e-06,
+      "loss": 0.1039,
+      "step": 2243
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 14.31038138115803,
+      "learning_rate": 6.814528194276636e-06,
+      "loss": 0.0687,
+      "step": 2244
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 8.40265784360439,
+      "learning_rate": 6.811834775024219e-06,
+      "loss": 0.1125,
+      "step": 2245
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 20.076086550456807,
+      "learning_rate": 6.809140750404127e-06,
+      "loss": 0.0917,
+      "step": 2246
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 23.9239943476738,
+      "learning_rate": 6.8064461213164825e-06,
+      "loss": 0.1105,
+      "step": 2247
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 7.421604483486793,
+      "learning_rate": 6.803750888661611e-06,
+      "loss": 0.0867,
+      "step": 2248
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 26.261644106440436,
+      "learning_rate": 6.8010550533400425e-06,
+      "loss": 0.1093,
+      "step": 2249
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 24.854024525208718,
+      "learning_rate": 6.798358616252503e-06,
+      "loss": 0.09,
+      "step": 2250
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 28.96708219096894,
+      "learning_rate": 6.795661578299924e-06,
+      "loss": 0.0673,
+      "step": 2251
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 11.997934589159817,
+      "learning_rate": 6.792963940383436e-06,
+      "loss": 0.132,
+      "step": 2252
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 51.52301595812061,
+      "learning_rate": 6.790265703404368e-06,
+      "loss": 0.1309,
+      "step": 2253
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 33.198458196121244,
+      "learning_rate": 6.787566868264253e-06,
+      "loss": 0.1149,
+      "step": 2254
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 39.63747128767255,
+      "learning_rate": 6.7848674358648195e-06,
+      "loss": 0.1301,
+      "step": 2255
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 16.270808913561332,
+      "learning_rate": 6.782167407108001e-06,
+      "loss": 0.1317,
+      "step": 2256
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 32.72395746940256,
+      "learning_rate": 6.779466782895926e-06,
+      "loss": 0.087,
+      "step": 2257
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 48.91482478473609,
+      "learning_rate": 6.7767655641309234e-06,
+      "loss": 0.1608,
+      "step": 2258
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 34.08193473685804,
+      "learning_rate": 6.7740637517155205e-06,
+      "loss": 0.1154,
+      "step": 2259
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 12.943831638411528,
+      "learning_rate": 6.771361346552445e-06,
+      "loss": 0.1069,
+      "step": 2260
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 7.135693851021515,
+      "learning_rate": 6.7686583495446164e-06,
+      "loss": 0.0879,
+      "step": 2261
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 37.11558243569602,
+      "learning_rate": 6.765954761595161e-06,
+      "loss": 0.1265,
+      "step": 2262
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 51.426289905430444,
+      "learning_rate": 6.763250583607392e-06,
+      "loss": 0.1113,
+      "step": 2263
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 24.538812354381417,
+      "learning_rate": 6.7605458164848316e-06,
+      "loss": 0.1005,
+      "step": 2264
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 13.374751197386294,
+      "learning_rate": 6.75784046113119e-06,
+      "loss": 0.1232,
+      "step": 2265
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 19.870883488045923,
+      "learning_rate": 6.755134518450377e-06,
+      "loss": 0.1033,
+      "step": 2266
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.625372295659911,
+      "learning_rate": 6.752427989346497e-06,
+      "loss": 0.1268,
+      "step": 2267
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 10.575802908582206,
+      "learning_rate": 6.749720874723854e-06,
+      "loss": 0.0843,
+      "step": 2268
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 25.526322460359744,
+      "learning_rate": 6.747013175486944e-06,
+      "loss": 0.1187,
+      "step": 2269
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.939467806031972,
+      "learning_rate": 6.74430489254046e-06,
+      "loss": 0.088,
+      "step": 2270
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 4.729990582136349,
+      "learning_rate": 6.741596026789288e-06,
+      "loss": 0.1049,
+      "step": 2271
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 6.467119491306653,
+      "learning_rate": 6.7388865791385124e-06,
+      "loss": 0.0942,
+      "step": 2272
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 9.876980913769884,
+      "learning_rate": 6.736176550493411e-06,
+      "loss": 0.1053,
+      "step": 2273
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 20.755840347838348,
+      "learning_rate": 6.7334659417594514e-06,
+      "loss": 0.116,
+      "step": 2274
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 17.06569796812762,
+      "learning_rate": 6.730754753842303e-06,
+      "loss": 0.1023,
+      "step": 2275
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 7.8737047418848904,
+      "learning_rate": 6.728042987647818e-06,
+      "loss": 0.0779,
+      "step": 2276
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 31.73870174511549,
+      "learning_rate": 6.725330644082054e-06,
+      "loss": 0.1567,
+      "step": 2277
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 27.556880854348176,
+      "learning_rate": 6.7226177240512516e-06,
+      "loss": 0.085,
+      "step": 2278
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 11.788183835441814,
+      "learning_rate": 6.7199042284618484e-06,
+      "loss": 0.093,
+      "step": 2279
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 31.31737563867473,
+      "learning_rate": 6.717190158220475e-06,
+      "loss": 0.1143,
+      "step": 2280
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.564357936167133,
+      "learning_rate": 6.714475514233951e-06,
+      "loss": 0.0759,
+      "step": 2281
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 46.40291596398642,
+      "learning_rate": 6.71176029740929e-06,
+      "loss": 0.1226,
+      "step": 2282
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 24.154317765086333,
+      "learning_rate": 6.709044508653697e-06,
+      "loss": 0.1367,
+      "step": 2283
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 6.925533803113575,
+      "learning_rate": 6.706328148874568e-06,
+      "loss": 0.1111,
+      "step": 2284
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 17.52523515570555,
+      "learning_rate": 6.703611218979488e-06,
+      "loss": 0.113,
+      "step": 2285
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 32.63171057290881,
+      "learning_rate": 6.700893719876234e-06,
+      "loss": 0.1052,
+      "step": 2286
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 35.04462192513845,
+      "learning_rate": 6.698175652472774e-06,
+      "loss": 0.0858,
+      "step": 2287
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 18.72923605107616,
+      "learning_rate": 6.695457017677263e-06,
+      "loss": 0.1011,
+      "step": 2288
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 14.768416928691146,
+      "learning_rate": 6.692737816398048e-06,
+      "loss": 0.1317,
+      "step": 2289
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 21.649718208972313,
+      "learning_rate": 6.6900180495436664e-06,
+      "loss": 0.1016,
+      "step": 2290
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 16.92349184199488,
+      "learning_rate": 6.68729771802284e-06,
+      "loss": 0.1027,
+      "step": 2291
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 18.73784013126941,
+      "learning_rate": 6.6845768227444855e-06,
+      "loss": 0.0793,
+      "step": 2292
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.89593946985034,
+      "learning_rate": 6.681855364617702e-06,
+      "loss": 0.0908,
+      "step": 2293
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 12.205255707572324,
+      "learning_rate": 6.67913334455178e-06,
+      "loss": 0.0995,
+      "step": 2294
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 16.757751110332908,
+      "learning_rate": 6.676410763456197e-06,
+      "loss": 0.1001,
+      "step": 2295
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 7.269571127624379,
+      "learning_rate": 6.673687622240619e-06,
+      "loss": 0.088,
+      "step": 2296
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 9.609560802494752,
+      "learning_rate": 6.670963921814896e-06,
+      "loss": 0.1106,
+      "step": 2297
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 8.246135299481141,
+      "learning_rate": 6.668239663089069e-06,
+      "loss": 0.1433,
+      "step": 2298
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 18.24562720984829,
+      "learning_rate": 6.665514846973361e-06,
+      "loss": 0.0901,
+      "step": 2299
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 13.37161564678813,
+      "learning_rate": 6.662789474378186e-06,
+      "loss": 0.1213,
+      "step": 2300
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 11.844017421917352,
+      "learning_rate": 6.6600635462141415e-06,
+      "loss": 0.0905,
+      "step": 2301
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 13.504340961380517,
+      "learning_rate": 6.657337063392011e-06,
+      "loss": 0.1224,
+      "step": 2302
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 5.623690274333707,
+      "learning_rate": 6.654610026822761e-06,
+      "loss": 0.0778,
+      "step": 2303
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 10.443915001651478,
+      "learning_rate": 6.651882437417546e-06,
+      "loss": 0.1146,
+      "step": 2304
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.261849012510766,
+      "learning_rate": 6.649154296087705e-06,
+      "loss": 0.1248,
+      "step": 2305
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.712674765149288,
+      "learning_rate": 6.646425603744759e-06,
+      "loss": 0.1034,
+      "step": 2306
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 8.076830070701273,
+      "learning_rate": 6.643696361300418e-06,
+      "loss": 0.1305,
+      "step": 2307
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 16.23453792461202,
+      "learning_rate": 6.6409665696665715e-06,
+      "loss": 0.0792,
+      "step": 2308
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 6.928315712867063,
+      "learning_rate": 6.638236229755292e-06,
+      "loss": 0.1429,
+      "step": 2309
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 5.653617940327973,
+      "learning_rate": 6.635505342478838e-06,
+      "loss": 0.0847,
+      "step": 2310
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.207528691085546,
+      "learning_rate": 6.632773908749649e-06,
+      "loss": 0.0923,
+      "step": 2311
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 26.119937550135738,
+      "learning_rate": 6.630041929480349e-06,
+      "loss": 0.1027,
+      "step": 2312
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 24.180449478429487,
+      "learning_rate": 6.627309405583741e-06,
+      "loss": 0.1044,
+      "step": 2313
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 20.448643687176528,
+      "learning_rate": 6.624576337972815e-06,
+      "loss": 0.0922,
+      "step": 2314
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 11.591680766177754,
+      "learning_rate": 6.621842727560737e-06,
+      "loss": 0.0912,
+      "step": 2315
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 4.828986274241329,
+      "learning_rate": 6.6191085752608575e-06,
+      "loss": 0.0772,
+      "step": 2316
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.954083488247607,
+      "learning_rate": 6.616373881986708e-06,
+      "loss": 0.1047,
+      "step": 2317
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 5.6263248883934125,
+      "learning_rate": 6.613638648652002e-06,
+      "loss": 0.0961,
+      "step": 2318
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.087116155084636,
+      "learning_rate": 6.610902876170631e-06,
+      "loss": 0.0953,
+      "step": 2319
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 18.559528074367204,
+      "learning_rate": 6.608166565456666e-06,
+      "loss": 0.1322,
+      "step": 2320
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 36.09006097390221,
+      "learning_rate": 6.605429717424359e-06,
+      "loss": 0.0972,
+      "step": 2321
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.741647863961911,
+      "learning_rate": 6.602692332988143e-06,
+      "loss": 0.0908,
+      "step": 2322
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 30.835808916611533,
+      "learning_rate": 6.5999544130626305e-06,
+      "loss": 0.0942,
+      "step": 2323
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 14.628025775072933,
+      "learning_rate": 6.597215958562608e-06,
+      "loss": 0.1154,
+      "step": 2324
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 19.30215107454651,
+      "learning_rate": 6.5944769704030465e-06,
+      "loss": 0.0925,
+      "step": 2325
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 16.28763762345787,
+      "learning_rate": 6.591737449499092e-06,
+      "loss": 0.1129,
+      "step": 2326
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.830045219544676,
+      "learning_rate": 6.58899739676607e-06,
+      "loss": 0.0909,
+      "step": 2327
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 8.33628287139074,
+      "learning_rate": 6.586256813119482e-06,
+      "loss": 0.0975,
+      "step": 2328
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 16.449451637998475,
+      "learning_rate": 6.583515699475009e-06,
+      "loss": 0.129,
+      "step": 2329
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 6.413192128563511,
+      "learning_rate": 6.580774056748508e-06,
+      "loss": 0.0936,
+      "step": 2330
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 24.55379335050217,
+      "learning_rate": 6.578031885856011e-06,
+      "loss": 0.0896,
+      "step": 2331
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 16.570585530302036,
+      "learning_rate": 6.575289187713731e-06,
+      "loss": 0.0884,
+      "step": 2332
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 16.171822796583307,
+      "learning_rate": 6.572545963238053e-06,
+      "loss": 0.0837,
+      "step": 2333
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 7.644151539511672,
+      "learning_rate": 6.569802213345537e-06,
+      "loss": 0.1268,
+      "step": 2334
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 14.08472143552305,
+      "learning_rate": 6.5670579389529255e-06,
+      "loss": 0.0915,
+      "step": 2335
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.674980055295068,
+      "learning_rate": 6.56431314097713e-06,
+      "loss": 0.0601,
+      "step": 2336
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 20.618948771998678,
+      "learning_rate": 6.561567820335236e-06,
+      "loss": 0.0776,
+      "step": 2337
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 7.837958868858976,
+      "learning_rate": 6.558821977944508e-06,
+      "loss": 0.0777,
+      "step": 2338
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 10.081314355572308,
+      "learning_rate": 6.556075614722383e-06,
+      "loss": 0.097,
+      "step": 2339
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 12.979423171549898,
+      "learning_rate": 6.553328731586473e-06,
+      "loss": 0.1097,
+      "step": 2340
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 31.475378687444337,
+      "learning_rate": 6.550581329454561e-06,
+      "loss": 0.1274,
+      "step": 2341
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 19.29402235295992,
+      "learning_rate": 6.547833409244606e-06,
+      "loss": 0.0825,
+      "step": 2342
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 17.39491613562142,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.1018,
+      "step": 2343
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 6.751666154792371,
+      "learning_rate": 6.542336018263262e-06,
+      "loss": 0.0856,
+      "step": 2344
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 28.598101139931043,
+      "learning_rate": 6.539586549328656e-06,
+      "loss": 0.1866,
+      "step": 2345
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.960017160676509,
+      "learning_rate": 6.536836565989565e-06,
+      "loss": 0.0786,
+      "step": 2346
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 12.28562563434325,
+      "learning_rate": 6.534086069164813e-06,
+      "loss": 0.1039,
+      "step": 2347
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 11.822129310031302,
+      "learning_rate": 6.531335059773392e-06,
+      "loss": 0.0911,
+      "step": 2348
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 33.14817099069893,
+      "learning_rate": 6.528583538734463e-06,
+      "loss": 0.115,
+      "step": 2349
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 9.926580522705304,
+      "learning_rate": 6.525831506967361e-06,
+      "loss": 0.1115,
+      "step": 2350
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 25.639367084236095,
+      "learning_rate": 6.523078965391592e-06,
+      "loss": 0.1034,
+      "step": 2351
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 6.074092820606506,
+      "learning_rate": 6.520325914926831e-06,
+      "loss": 0.0667,
+      "step": 2352
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 16.162347392614848,
+      "learning_rate": 6.517572356492922e-06,
+      "loss": 0.0854,
+      "step": 2353
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 4.991411546020857,
+      "learning_rate": 6.514818291009881e-06,
+      "loss": 0.1069,
+      "step": 2354
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 15.6323688780031,
+      "learning_rate": 6.512063719397894e-06,
+      "loss": 0.0802,
+      "step": 2355
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.816757264191924,
+      "learning_rate": 6.5093086425773126e-06,
+      "loss": 0.1061,
+      "step": 2356
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 11.042693972596268,
+      "learning_rate": 6.506553061468659e-06,
+      "loss": 0.0801,
+      "step": 2357
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 7.60615774529474,
+      "learning_rate": 6.5037969769926256e-06,
+      "loss": 0.1073,
+      "step": 2358
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 7.938469871759891,
+      "learning_rate": 6.501040390070071e-06,
+      "loss": 0.1047,
+      "step": 2359
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.407806278240889,
+      "learning_rate": 6.498283301622022e-06,
+      "loss": 0.0829,
+      "step": 2360
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 21.855833457945497,
+      "learning_rate": 6.495525712569673e-06,
+      "loss": 0.0896,
+      "step": 2361
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 24.55266138680911,
+      "learning_rate": 6.492767623834385e-06,
+      "loss": 0.0958,
+      "step": 2362
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 7.844762029942168,
+      "learning_rate": 6.490009036337687e-06,
+      "loss": 0.1097,
+      "step": 2363
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 14.157061696422106,
+      "learning_rate": 6.487249951001276e-06,
+      "loss": 0.0968,
+      "step": 2364
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 23.749244203785135,
+      "learning_rate": 6.484490368747012e-06,
+      "loss": 0.1128,
+      "step": 2365
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 6.384281292680775,
+      "learning_rate": 6.4817302904969226e-06,
+      "loss": 0.1133,
+      "step": 2366
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 12.27197718764886,
+      "learning_rate": 6.4789697171732024e-06,
+      "loss": 0.1234,
+      "step": 2367
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 6.1705409111581115,
+      "learning_rate": 6.476208649698209e-06,
+      "loss": 0.1206,
+      "step": 2368
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 9.676461914963904,
+      "learning_rate": 6.473447088994467e-06,
+      "loss": 0.0778,
+      "step": 2369
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 26.76018722096519,
+      "learning_rate": 6.470685035984667e-06,
+      "loss": 0.1274,
+      "step": 2370
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 12.24487360044837,
+      "learning_rate": 6.467922491591658e-06,
+      "loss": 0.0955,
+      "step": 2371
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 27.638340021369807,
+      "learning_rate": 6.465159456738461e-06,
+      "loss": 0.0926,
+      "step": 2372
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 4.785845976111559,
+      "learning_rate": 6.462395932348257e-06,
+      "loss": 0.0978,
+      "step": 2373
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.312844920909349,
+      "learning_rate": 6.459631919344389e-06,
+      "loss": 0.067,
+      "step": 2374
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.685594864488179,
+      "learning_rate": 6.456867418650366e-06,
+      "loss": 0.0913,
+      "step": 2375
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 9.9810523595601,
+      "learning_rate": 6.454102431189859e-06,
+      "loss": 0.1007,
+      "step": 2376
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 6.265965225581489,
+      "learning_rate": 6.4513369578867026e-06,
+      "loss": 0.104,
+      "step": 2377
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 16.893745136728803,
+      "learning_rate": 6.448570999664894e-06,
+      "loss": 0.1005,
+      "step": 2378
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 28.908957709905614,
+      "learning_rate": 6.4458045574485875e-06,
+      "loss": 0.1255,
+      "step": 2379
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 24.22139522152318,
+      "learning_rate": 6.443037632162104e-06,
+      "loss": 0.0996,
+      "step": 2380
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 26.572515462610134,
+      "learning_rate": 6.440270224729927e-06,
+      "loss": 0.0901,
+      "step": 2381
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 12.207827481160598,
+      "learning_rate": 6.437502336076695e-06,
+      "loss": 0.1221,
+      "step": 2382
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 19.41999789803127,
+      "learning_rate": 6.4347339671272155e-06,
+      "loss": 0.0783,
+      "step": 2383
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 16.873084876197144,
+      "learning_rate": 6.431965118806449e-06,
+      "loss": 0.094,
+      "step": 2384
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 16.372299469008183,
+      "learning_rate": 6.42919579203952e-06,
+      "loss": 0.1016,
+      "step": 2385
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 27.497086153318044,
+      "learning_rate": 6.4264259877517124e-06,
+      "loss": 0.1262,
+      "step": 2386
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 24.053964065962585,
+      "learning_rate": 6.423655706868468e-06,
+      "loss": 0.0745,
+      "step": 2387
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 22.25200523393893,
+      "learning_rate": 6.4208849503153915e-06,
+      "loss": 0.0837,
+      "step": 2388
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 11.764255564259074,
+      "learning_rate": 6.418113719018242e-06,
+      "loss": 0.1034,
+      "step": 2389
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 30.75721694150146,
+      "learning_rate": 6.415342013902939e-06,
+      "loss": 0.1151,
+      "step": 2390
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 5.427156665392917,
+      "learning_rate": 6.412569835895562e-06,
+      "loss": 0.0956,
+      "step": 2391
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 18.046997378438945,
+      "learning_rate": 6.409797185922349e-06,
+      "loss": 0.1169,
+      "step": 2392
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 7.538487314232114,
+      "learning_rate": 6.40702406490969e-06,
+      "loss": 0.1074,
+      "step": 2393
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 14.150795863197285,
+      "learning_rate": 6.404250473784138e-06,
+      "loss": 0.0994,
+      "step": 2394
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 24.75329036726323,
+      "learning_rate": 6.401476413472404e-06,
+      "loss": 0.1069,
+      "step": 2395
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 12.43088365316308,
+      "learning_rate": 6.398701884901348e-06,
+      "loss": 0.1107,
+      "step": 2396
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 8.74404078917541,
+      "learning_rate": 6.3959268889979956e-06,
+      "loss": 0.0984,
+      "step": 2397
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.577324694920657,
+      "learning_rate": 6.393151426689522e-06,
+      "loss": 0.1191,
+      "step": 2398
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 13.63480133602996,
+      "learning_rate": 6.390375498903263e-06,
+      "loss": 0.0992,
+      "step": 2399
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 14.152036274866392,
+      "learning_rate": 6.387599106566705e-06,
+      "loss": 0.0937,
+      "step": 2400
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.121175600708925,
+      "learning_rate": 6.384822250607495e-06,
+      "loss": 0.1039,
+      "step": 2401
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.47698057810036,
+      "learning_rate": 6.382044931953431e-06,
+      "loss": 0.0828,
+      "step": 2402
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 10.08821825459922,
+      "learning_rate": 6.379267151532467e-06,
+      "loss": 0.135,
+      "step": 2403
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.9574752880861226,
+      "learning_rate": 6.376488910272709e-06,
+      "loss": 0.0742,
+      "step": 2404
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 5.2069319299192065,
+      "learning_rate": 6.373710209102423e-06,
+      "loss": 0.1099,
+      "step": 2405
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 26.51909679991199,
+      "learning_rate": 6.370931048950022e-06,
+      "loss": 0.0972,
+      "step": 2406
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 18.736652018749133,
+      "learning_rate": 6.368151430744075e-06,
+      "loss": 0.1042,
+      "step": 2407
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 15.498335426749598,
+      "learning_rate": 6.365371355413306e-06,
+      "loss": 0.1053,
+      "step": 2408
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 35.0436335302218,
+      "learning_rate": 6.362590823886588e-06,
+      "loss": 0.094,
+      "step": 2409
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 35.01179148908803,
+      "learning_rate": 6.359809837092947e-06,
+      "loss": 0.1014,
+      "step": 2410
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 8.194118078049293,
+      "learning_rate": 6.357028395961566e-06,
+      "loss": 0.0864,
+      "step": 2411
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 6.697601824559924,
+      "learning_rate": 6.354246501421777e-06,
+      "loss": 0.12,
+      "step": 2412
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 15.721750841959683,
+      "learning_rate": 6.3514641544030575e-06,
+      "loss": 0.1021,
+      "step": 2413
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 23.028362842896957,
+      "learning_rate": 6.348681355835043e-06,
+      "loss": 0.106,
+      "step": 2414
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 23.07284177900744,
+      "learning_rate": 6.345898106647521e-06,
+      "loss": 0.1036,
+      "step": 2415
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 14.759239504580846,
+      "learning_rate": 6.3431144077704245e-06,
+      "loss": 0.1005,
+      "step": 2416
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 7.849400235997493,
+      "learning_rate": 6.340330260133839e-06,
+      "loss": 0.1296,
+      "step": 2417
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 13.264559348008135,
+      "learning_rate": 6.337545664668001e-06,
+      "loss": 0.1018,
+      "step": 2418
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 54.345532921290065,
+      "learning_rate": 6.334760622303294e-06,
+      "loss": 0.1444,
+      "step": 2419
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 11.648785289643577,
+      "learning_rate": 6.331975133970255e-06,
+      "loss": 0.0713,
+      "step": 2420
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 20.282763003405808,
+      "learning_rate": 6.329189200599566e-06,
+      "loss": 0.0861,
+      "step": 2421
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 19.267431318750052,
+      "learning_rate": 6.326402823122059e-06,
+      "loss": 0.061,
+      "step": 2422
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 16.487214621707068,
+      "learning_rate": 6.3236160024687134e-06,
+      "loss": 0.0946,
+      "step": 2423
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 8.987316243770564,
+      "learning_rate": 6.3208287395706595e-06,
+      "loss": 0.1047,
+      "step": 2424
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 23.44498703021866,
+      "learning_rate": 6.3180410353591735e-06,
+      "loss": 0.1006,
+      "step": 2425
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 11.465144785371189,
+      "learning_rate": 6.315252890765678e-06,
+      "loss": 0.0963,
+      "step": 2426
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 5.166961299823076,
+      "learning_rate": 6.312464306721745e-06,
+      "loss": 0.0905,
+      "step": 2427
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 9.634991916602496,
+      "learning_rate": 6.309675284159093e-06,
+      "loss": 0.0927,
+      "step": 2428
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 11.247409453302282,
+      "learning_rate": 6.306885824009585e-06,
+      "loss": 0.0801,
+      "step": 2429
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 6.056553339693292,
+      "learning_rate": 6.3040959272052315e-06,
+      "loss": 0.0787,
+      "step": 2430
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 13.238200555357215,
+      "learning_rate": 6.301305594678189e-06,
+      "loss": 0.0916,
+      "step": 2431
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.245856258170428,
+      "learning_rate": 6.2985148273607586e-06,
+      "loss": 0.0818,
+      "step": 2432
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 16.58705389948258,
+      "learning_rate": 6.29572362618539e-06,
+      "loss": 0.0864,
+      "step": 2433
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 24.4358198678026,
+      "learning_rate": 6.292931992084672e-06,
+      "loss": 0.1365,
+      "step": 2434
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 14.195560279980855,
+      "learning_rate": 6.290139925991345e-06,
+      "loss": 0.1036,
+      "step": 2435
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 14.566630680835873,
+      "learning_rate": 6.287347428838289e-06,
+      "loss": 0.067,
+      "step": 2436
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 10.846086245631618,
+      "learning_rate": 6.2845545015585275e-06,
+      "loss": 0.126,
+      "step": 2437
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 6.710131680304093,
+      "learning_rate": 6.281761145085232e-06,
+      "loss": 0.0868,
+      "step": 2438
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 9.77685196724837,
+      "learning_rate": 6.278967360351712e-06,
+      "loss": 0.0619,
+      "step": 2439
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 22.26451062145235,
+      "learning_rate": 6.276173148291425e-06,
+      "loss": 0.1013,
+      "step": 2440
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 4.273890399178918,
+      "learning_rate": 6.273378509837969e-06,
+      "loss": 0.0831,
+      "step": 2441
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 22.997496293271507,
+      "learning_rate": 6.2705834459250825e-06,
+      "loss": 0.1004,
+      "step": 2442
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 11.57569324607047,
+      "learning_rate": 6.2677879574866515e-06,
+      "loss": 0.0858,
+      "step": 2443
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 33.78323186801771,
+      "learning_rate": 6.264992045456699e-06,
+      "loss": 0.111,
+      "step": 2444
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.134971347540446,
+      "learning_rate": 6.262195710769391e-06,
+      "loss": 0.089,
+      "step": 2445
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 26.861775030048683,
+      "learning_rate": 6.259398954359037e-06,
+      "loss": 0.1116,
+      "step": 2446
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 7.188038510887187,
+      "learning_rate": 6.256601777160082e-06,
+      "loss": 0.1488,
+      "step": 2447
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 23.487091229583946,
+      "learning_rate": 6.253804180107116e-06,
+      "loss": 0.1021,
+      "step": 2448
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.589395392272444,
+      "learning_rate": 6.2510061641348695e-06,
+      "loss": 0.1018,
+      "step": 2449
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 13.31139980832625,
+      "learning_rate": 6.248207730178211e-06,
+      "loss": 0.0819,
+      "step": 2450
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.211913937617608,
+      "learning_rate": 6.245408879172148e-06,
+      "loss": 0.1106,
+      "step": 2451
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.641616344245171,
+      "learning_rate": 6.24260961205183e-06,
+      "loss": 0.0833,
+      "step": 2452
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 21.256785415127464,
+      "learning_rate": 6.239809929752544e-06,
+      "loss": 0.0796,
+      "step": 2453
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 10.627318442838185,
+      "learning_rate": 6.237009833209715e-06,
+      "loss": 0.1066,
+      "step": 2454
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 12.179171841920033,
+      "learning_rate": 6.2342093233589095e-06,
+      "loss": 0.1462,
+      "step": 2455
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 14.930949343627839,
+      "learning_rate": 6.231408401135828e-06,
+      "loss": 0.1081,
+      "step": 2456
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 27.796422103824614,
+      "learning_rate": 6.228607067476311e-06,
+      "loss": 0.1116,
+      "step": 2457
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 19.282120894394577,
+      "learning_rate": 6.225805323316336e-06,
+      "loss": 0.1353,
+      "step": 2458
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 24.664287072388348,
+      "learning_rate": 6.223003169592018e-06,
+      "loss": 0.1069,
+      "step": 2459
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 11.014341949865857,
+      "learning_rate": 6.220200607239609e-06,
+      "loss": 0.1069,
+      "step": 2460
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.682585859806537,
+      "learning_rate": 6.217397637195497e-06,
+      "loss": 0.1176,
+      "step": 2461
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.254486314903181,
+      "learning_rate": 6.214594260396206e-06,
+      "loss": 0.078,
+      "step": 2462
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 6.715958544218641,
+      "learning_rate": 6.211790477778399e-06,
+      "loss": 0.0914,
+      "step": 2463
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 9.16222865335177,
+      "learning_rate": 6.208986290278866e-06,
+      "loss": 0.0664,
+      "step": 2464
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 17.220240072763353,
+      "learning_rate": 6.206181698834544e-06,
+      "loss": 0.1018,
+      "step": 2465
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 16.278402617080406,
+      "learning_rate": 6.2033767043824955e-06,
+      "loss": 0.0988,
+      "step": 2466
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 16.789027698060593,
+      "learning_rate": 6.200571307859923e-06,
+      "loss": 0.1199,
+      "step": 2467
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 16.858190409234904,
+      "learning_rate": 6.197765510204161e-06,
+      "loss": 0.1025,
+      "step": 2468
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 20.311780934271166,
+      "learning_rate": 6.19495931235268e-06,
+      "loss": 0.0986,
+      "step": 2469
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 8.24075864852246,
+      "learning_rate": 6.19215271524308e-06,
+      "loss": 0.0897,
+      "step": 2470
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.537619256321832,
+      "learning_rate": 6.189345719813099e-06,
+      "loss": 0.1053,
+      "step": 2471
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 19.985538588188874,
+      "learning_rate": 6.186538327000609e-06,
+      "loss": 0.0933,
+      "step": 2472
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 10.951691550287652,
+      "learning_rate": 6.183730537743607e-06,
+      "loss": 0.0876,
+      "step": 2473
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 14.796383785272676,
+      "learning_rate": 6.18092235298023e-06,
+      "loss": 0.1133,
+      "step": 2474
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 8.222463592646822,
+      "learning_rate": 6.178113773648745e-06,
+      "loss": 0.0845,
+      "step": 2475
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.583268825766998,
+      "learning_rate": 6.175304800687551e-06,
+      "loss": 0.0778,
+      "step": 2476
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.546628402576362,
+      "learning_rate": 6.172495435035176e-06,
+      "loss": 0.1199,
+      "step": 2477
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.515726741394918,
+      "learning_rate": 6.169685677630284e-06,
+      "loss": 0.0916,
+      "step": 2478
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.737944386905074,
+      "learning_rate": 6.1668755294116655e-06,
+      "loss": 0.0611,
+      "step": 2479
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.071570166337342,
+      "learning_rate": 6.1640649913182436e-06,
+      "loss": 0.0988,
+      "step": 2480
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 5.529604937790912,
+      "learning_rate": 6.161254064289072e-06,
+      "loss": 0.084,
+      "step": 2481
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 13.061035895196843,
+      "learning_rate": 6.158442749263332e-06,
+      "loss": 0.0971,
+      "step": 2482
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 17.682527665030868,
+      "learning_rate": 6.155631047180337e-06,
+      "loss": 0.0814,
+      "step": 2483
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 14.400108863552052,
+      "learning_rate": 6.152818958979529e-06,
+      "loss": 0.0972,
+      "step": 2484
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 4.875457146337422,
+      "learning_rate": 6.1500064856004796e-06,
+      "loss": 0.0699,
+      "step": 2485
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 15.446178764057269,
+      "learning_rate": 6.147193627982887e-06,
+      "loss": 0.099,
+      "step": 2486
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 7.245784964129334,
+      "learning_rate": 6.144380387066581e-06,
+      "loss": 0.1026,
+      "step": 2487
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 24.36888800556823,
+      "learning_rate": 6.141566763791518e-06,
+      "loss": 0.1219,
+      "step": 2488
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 11.458177901064403,
+      "learning_rate": 6.138752759097778e-06,
+      "loss": 0.0728,
+      "step": 2489
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.774247738507147,
+      "learning_rate": 6.135938373925576e-06,
+      "loss": 0.0879,
+      "step": 2490
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.476440380457411,
+      "learning_rate": 6.133123609215249e-06,
+      "loss": 0.0761,
+      "step": 2491
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.095496325152822,
+      "learning_rate": 6.130308465907263e-06,
+      "loss": 0.0991,
+      "step": 2492
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.029873546120703,
+      "learning_rate": 6.127492944942209e-06,
+      "loss": 0.0861,
+      "step": 2493
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 19.09411821013496,
+      "learning_rate": 6.124677047260805e-06,
+      "loss": 0.0801,
+      "step": 2494
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 18.044110197346168,
+      "learning_rate": 6.121860773803895e-06,
+      "loss": 0.099,
+      "step": 2495
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 9.097311899728787,
+      "learning_rate": 6.119044125512447e-06,
+      "loss": 0.082,
+      "step": 2496
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 10.196462877637837,
+      "learning_rate": 6.116227103327559e-06,
+      "loss": 0.1102,
+      "step": 2497
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 17.411823414220972,
+      "learning_rate": 6.113409708190447e-06,
+      "loss": 0.1019,
+      "step": 2498
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 18.155557304557057,
+      "learning_rate": 6.1105919410424566e-06,
+      "loss": 0.1013,
+      "step": 2499
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 16.792200022527275,
+      "learning_rate": 6.107773802825055e-06,
+      "loss": 0.1157,
+      "step": 2500
+    },
+    {
+      "epoch": 1.78,
+      "eval_avg_AUC": 0.8162853827527864,
+      "eval_avg_Accuracy": 0.7180039787798409,
+      "eval_avg_Accuracy-right": 0.8944176340159123,
+      "eval_avg_Accuracy-wrong": 0.4103934500795997,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6879813648500559,
+      "eval_last_AUC": 0.8267063817923214,
+      "eval_last_Accuracy": 0.7504559018567639,
+      "eval_last_Accuracy-right": 0.8129646537107083,
+      "eval_last_Accuracy-wrong": 0.6414600864225608,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.6987671721798714,
+      "eval_max_AUC": 0.7501564696977459,
+      "eval_max_Accuracy": 0.6474635278514589,
+      "eval_max_Accuracy-right": 0.9788704838920047,
+      "eval_max_Accuracy-wrong": 0.0695929042528997,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.6173876633213748,
+      "eval_min_AUC": 0.824010385799939,
+      "eval_min_Accuracy": 0.7522795092838196,
+      "eval_min_Accuracy-right": 0.7631407330116082,
+      "eval_min_Accuracy-wrong": 0.7333409142597226,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6985406564564841,
+      "eval_prod_AUC": 0.8230216036527745,
+      "eval_prod_Accuracy": 0.712367374005305,
+      "eval_prod_Accuracy-right": 0.617320986044085,
+      "eval_prod_Accuracy-wrong": 0.8780987036615875,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.6893611435134097,
+      "eval_runtime": 247.0285,
+      "eval_samples_per_second": 97.673,
+      "eval_steps_per_second": 3.052,
+      "eval_sum_AUC": 0.6795322314303073,
+      "eval_sum_Accuracy": 0.6392572944297082,
+      "eval_sum_Accuracy-right": 0.9985000652145559,
+      "eval_sum_Accuracy-wrong": 0.01284967022970207,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6636898117156945,
+      "step": 2500
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 27.66785481931168,
+      "learning_rate": 6.1049552944798355e-06,
+      "loss": 0.1031,
+      "step": 2501
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 13.877238096501856,
+      "learning_rate": 6.102136416948513e-06,
+      "loss": 0.0922,
+      "step": 2502
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 14.832991307429838,
+      "learning_rate": 6.099317171172929e-06,
+      "loss": 0.1277,
+      "step": 2503
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 13.151505015978923,
+      "learning_rate": 6.0964975580950445e-06,
+      "loss": 0.088,
+      "step": 2504
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 14.194545354333274,
+      "learning_rate": 6.093677578656946e-06,
+      "loss": 0.1083,
+      "step": 2505
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.548368155403958,
+      "learning_rate": 6.090857233800839e-06,
+      "loss": 0.0778,
+      "step": 2506
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 29.162206973113896,
+      "learning_rate": 6.0880365244690546e-06,
+      "loss": 0.1195,
+      "step": 2507
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 8.958049048309908,
+      "learning_rate": 6.085215451604044e-06,
+      "loss": 0.0724,
+      "step": 2508
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 29.68574871023764,
+      "learning_rate": 6.082394016148379e-06,
+      "loss": 0.0842,
+      "step": 2509
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 34.56039583857383,
+      "learning_rate": 6.079572219044755e-06,
+      "loss": 0.1259,
+      "step": 2510
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 12.146084312643838,
+      "learning_rate": 6.076750061235985e-06,
+      "loss": 0.0802,
+      "step": 2511
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 17.542205198619733,
+      "learning_rate": 6.073927543665008e-06,
+      "loss": 0.1012,
+      "step": 2512
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 28.234603954607753,
+      "learning_rate": 6.071104667274875e-06,
+      "loss": 0.1429,
+      "step": 2513
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 22.631420233858943,
+      "learning_rate": 6.068281433008765e-06,
+      "loss": 0.1251,
+      "step": 2514
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.105380110647712,
+      "learning_rate": 6.0654578418099715e-06,
+      "loss": 0.0927,
+      "step": 2515
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 13.931020901993193,
+      "learning_rate": 6.062633894621909e-06,
+      "loss": 0.0729,
+      "step": 2516
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 26.71305807576688,
+      "learning_rate": 6.0598095923881105e-06,
+      "loss": 0.1173,
+      "step": 2517
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 23.253308243155512,
+      "learning_rate": 6.056984936052229e-06,
+      "loss": 0.092,
+      "step": 2518
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 11.399153561239252,
+      "learning_rate": 6.054159926558033e-06,
+      "loss": 0.1056,
+      "step": 2519
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 32.15751111995552,
+      "learning_rate": 6.051334564849413e-06,
+      "loss": 0.1127,
+      "step": 2520
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 7.043809237704044,
+      "learning_rate": 6.048508851870372e-06,
+      "loss": 0.0915,
+      "step": 2521
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 17.66369609169051,
+      "learning_rate": 6.045682788565036e-06,
+      "loss": 0.0673,
+      "step": 2522
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 11.479436418010804,
+      "learning_rate": 6.042856375877644e-06,
+      "loss": 0.1146,
+      "step": 2523
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 12.596499035596006,
+      "learning_rate": 6.040029614752551e-06,
+      "loss": 0.1011,
+      "step": 2524
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 27.997863595749756,
+      "learning_rate": 6.037202506134234e-06,
+      "loss": 0.0824,
+      "step": 2525
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 21.9014456038331,
+      "learning_rate": 6.03437505096728e-06,
+      "loss": 0.0857,
+      "step": 2526
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 11.901316851446808,
+      "learning_rate": 6.0315472501963955e-06,
+      "loss": 0.1244,
+      "step": 2527
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.351044648698084,
+      "learning_rate": 6.028719104766402e-06,
+      "loss": 0.0792,
+      "step": 2528
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 21.352307377026413,
+      "learning_rate": 6.025890615622233e-06,
+      "loss": 0.1433,
+      "step": 2529
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 26.395611117406183,
+      "learning_rate": 6.023061783708941e-06,
+      "loss": 0.0824,
+      "step": 2530
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.387064879765818,
+      "learning_rate": 6.020232609971694e-06,
+      "loss": 0.093,
+      "step": 2531
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 7.527848046304538,
+      "learning_rate": 6.017403095355766e-06,
+      "loss": 0.106,
+      "step": 2532
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 30.56609985998804,
+      "learning_rate": 6.014573240806553e-06,
+      "loss": 0.1035,
+      "step": 2533
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 18.801880253810776,
+      "learning_rate": 6.011743047269563e-06,
+      "loss": 0.1088,
+      "step": 2534
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 9.532091172107885,
+      "learning_rate": 6.008912515690415e-06,
+      "loss": 0.0753,
+      "step": 2535
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 12.658859923242405,
+      "learning_rate": 6.006081647014842e-06,
+      "loss": 0.0805,
+      "step": 2536
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 29.195819838210074,
+      "learning_rate": 6.00325044218869e-06,
+      "loss": 0.1105,
+      "step": 2537
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 18.51162600271363,
+      "learning_rate": 6.000418902157919e-06,
+      "loss": 0.1062,
+      "step": 2538
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 44.1732405351717,
+      "learning_rate": 5.997587027868598e-06,
+      "loss": 0.1221,
+      "step": 2539
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 29.309897683209353,
+      "learning_rate": 5.994754820266908e-06,
+      "loss": 0.0969,
+      "step": 2540
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.87554200493296,
+      "learning_rate": 5.991922280299143e-06,
+      "loss": 0.0918,
+      "step": 2541
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 7.609783728113712,
+      "learning_rate": 5.989089408911706e-06,
+      "loss": 0.0911,
+      "step": 2542
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 14.418526098296795,
+      "learning_rate": 5.986256207051113e-06,
+      "loss": 0.1036,
+      "step": 2543
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 58.75440816966157,
+      "learning_rate": 5.98342267566399e-06,
+      "loss": 0.1527,
+      "step": 2544
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 32.69007877459673,
+      "learning_rate": 5.9805888156970714e-06,
+      "loss": 0.1378,
+      "step": 2545
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 28.46074577433084,
+      "learning_rate": 5.977754628097203e-06,
+      "loss": 0.0883,
+      "step": 2546
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 36.58388507387772,
+      "learning_rate": 5.97492011381134e-06,
+      "loss": 0.088,
+      "step": 2547
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 40.783240549154485,
+      "learning_rate": 5.972085273786547e-06,
+      "loss": 0.0876,
+      "step": 2548
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 27.81800708078887,
+      "learning_rate": 5.969250108969995e-06,
+      "loss": 0.1101,
+      "step": 2549
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.59147754384461,
+      "learning_rate": 5.966414620308965e-06,
+      "loss": 0.092,
+      "step": 2550
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 28.42033100421827,
+      "learning_rate": 5.9635788087508474e-06,
+      "loss": 0.1016,
+      "step": 2551
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 34.74266526203714,
+      "learning_rate": 5.960742675243139e-06,
+      "loss": 0.0794,
+      "step": 2552
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 21.553513330683376,
+      "learning_rate": 5.957906220733447e-06,
+      "loss": 0.0999,
+      "step": 2553
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 5.114093682906385,
+      "learning_rate": 5.9550694461694806e-06,
+      "loss": 0.0754,
+      "step": 2554
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 7.8221315937987494,
+      "learning_rate": 5.95223235249906e-06,
+      "loss": 0.099,
+      "step": 2555
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 16.717018622807625,
+      "learning_rate": 5.949394940670112e-06,
+      "loss": 0.0839,
+      "step": 2556
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.214246667729665,
+      "learning_rate": 5.946557211630667e-06,
+      "loss": 0.0936,
+      "step": 2557
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.657749645898761,
+      "learning_rate": 5.943719166328864e-06,
+      "loss": 0.0894,
+      "step": 2558
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 5.149493730592742,
+      "learning_rate": 5.940880805712945e-06,
+      "loss": 0.0777,
+      "step": 2559
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 6.2592533001332225,
+      "learning_rate": 5.938042130731262e-06,
+      "loss": 0.1044,
+      "step": 2560
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 32.71064285721712,
+      "learning_rate": 5.935203142332267e-06,
+      "loss": 0.1262,
+      "step": 2561
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 6.8348479431905105,
+      "learning_rate": 5.932363841464519e-06,
+      "loss": 0.0815,
+      "step": 2562
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 30.961681239814183,
+      "learning_rate": 5.9295242290766805e-06,
+      "loss": 0.1127,
+      "step": 2563
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 18.416776056408292,
+      "learning_rate": 5.9266843061175216e-06,
+      "loss": 0.0973,
+      "step": 2564
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 10.93811275476671,
+      "learning_rate": 5.92384407353591e-06,
+      "loss": 0.1251,
+      "step": 2565
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 5.447480213973523,
+      "learning_rate": 5.921003532280822e-06,
+      "loss": 0.0801,
+      "step": 2566
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 7.813801842459522,
+      "learning_rate": 5.918162683301336e-06,
+      "loss": 0.1125,
+      "step": 2567
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 28.216557976157063,
+      "learning_rate": 5.91532152754663e-06,
+      "loss": 0.1176,
+      "step": 2568
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 19.011803166843467,
+      "learning_rate": 5.91248006596599e-06,
+      "loss": 0.1005,
+      "step": 2569
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 22.129204827206834,
+      "learning_rate": 5.909638299508798e-06,
+      "loss": 0.0659,
+      "step": 2570
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 8.030892883456888,
+      "learning_rate": 5.906796229124543e-06,
+      "loss": 0.101,
+      "step": 2571
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 16.213774164615035,
+      "learning_rate": 5.903953855762812e-06,
+      "loss": 0.0795,
+      "step": 2572
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.430953096593536,
+      "learning_rate": 5.901111180373298e-06,
+      "loss": 0.1147,
+      "step": 2573
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 43.58218646987675,
+      "learning_rate": 5.898268203905788e-06,
+      "loss": 0.1244,
+      "step": 2574
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 36.70832267287658,
+      "learning_rate": 5.895424927310174e-06,
+      "loss": 0.1086,
+      "step": 2575
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 23.88936445567082,
+      "learning_rate": 5.89258135153645e-06,
+      "loss": 0.1085,
+      "step": 2576
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 43.53612470901018,
+      "learning_rate": 5.889737477534704e-06,
+      "loss": 0.103,
+      "step": 2577
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 34.14388783488702,
+      "learning_rate": 5.886893306255129e-06,
+      "loss": 0.1014,
+      "step": 2578
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 21.552187716115085,
+      "learning_rate": 5.884048838648017e-06,
+      "loss": 0.1003,
+      "step": 2579
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 6.854450569018791,
+      "learning_rate": 5.881204075663755e-06,
+      "loss": 0.1073,
+      "step": 2580
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 21.51464575703357,
+      "learning_rate": 5.878359018252831e-06,
+      "loss": 0.1342,
+      "step": 2581
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 41.8483916842586,
+      "learning_rate": 5.8755136673658365e-06,
+      "loss": 0.1082,
+      "step": 2582
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 37.636416874159046,
+      "learning_rate": 5.872668023953449e-06,
+      "loss": 0.0976,
+      "step": 2583
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.552098998633584,
+      "learning_rate": 5.869822088966455e-06,
+      "loss": 0.0833,
+      "step": 2584
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 10.713494577494489,
+      "learning_rate": 5.866975863355734e-06,
+      "loss": 0.0893,
+      "step": 2585
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 23.2091103383476,
+      "learning_rate": 5.864129348072261e-06,
+      "loss": 0.1205,
+      "step": 2586
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 31.823573317574535,
+      "learning_rate": 5.861282544067112e-06,
+      "loss": 0.1232,
+      "step": 2587
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 22.21437732970618,
+      "learning_rate": 5.8584354522914555e-06,
+      "loss": 0.1088,
+      "step": 2588
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 11.487315525706093,
+      "learning_rate": 5.855588073696559e-06,
+      "loss": 0.0837,
+      "step": 2589
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 22.281223727113854,
+      "learning_rate": 5.852740409233785e-06,
+      "loss": 0.0732,
+      "step": 2590
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 24.22884461950451,
+      "learning_rate": 5.849892459854588e-06,
+      "loss": 0.0881,
+      "step": 2591
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 25.959868893514688,
+      "learning_rate": 5.847044226510524e-06,
+      "loss": 0.0851,
+      "step": 2592
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 17.515409904214152,
+      "learning_rate": 5.84419571015324e-06,
+      "loss": 0.1182,
+      "step": 2593
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 14.3143330651017,
+      "learning_rate": 5.8413469117344766e-06,
+      "loss": 0.1311,
+      "step": 2594
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 13.137514401761287,
+      "learning_rate": 5.838497832206074e-06,
+      "loss": 0.0912,
+      "step": 2595
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 6.325028784584776,
+      "learning_rate": 5.835648472519958e-06,
+      "loss": 0.0896,
+      "step": 2596
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 24.617506819797416,
+      "learning_rate": 5.832798833628156e-06,
+      "loss": 0.0967,
+      "step": 2597
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.056349053356649,
+      "learning_rate": 5.829948916482784e-06,
+      "loss": 0.0851,
+      "step": 2598
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.059342030707879,
+      "learning_rate": 5.827098722036053e-06,
+      "loss": 0.0838,
+      "step": 2599
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 21.709897870658843,
+      "learning_rate": 5.824248251240265e-06,
+      "loss": 0.1256,
+      "step": 2600
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 21.868681173826214,
+      "learning_rate": 5.8213975050478155e-06,
+      "loss": 0.0929,
+      "step": 2601
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.614721781840734,
+      "learning_rate": 5.818546484411191e-06,
+      "loss": 0.1243,
+      "step": 2602
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 10.034829947869662,
+      "learning_rate": 5.815695190282974e-06,
+      "loss": 0.1393,
+      "step": 2603
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 13.607835352795393,
+      "learning_rate": 5.81284362361583e-06,
+      "loss": 0.1298,
+      "step": 2604
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 10.740168450428905,
+      "learning_rate": 5.809991785362525e-06,
+      "loss": 0.0995,
+      "step": 2605
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 10.42600482469918,
+      "learning_rate": 5.8071396764759065e-06,
+      "loss": 0.1045,
+      "step": 2606
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 14.34092985354534,
+      "learning_rate": 5.804287297908923e-06,
+      "loss": 0.099,
+      "step": 2607
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 12.665461324372817,
+      "learning_rate": 5.801434650614601e-06,
+      "loss": 0.1346,
+      "step": 2608
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 6.536614697320999,
+      "learning_rate": 5.798581735546066e-06,
+      "loss": 0.1066,
+      "step": 2609
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.547830795525716,
+      "learning_rate": 5.79572855365653e-06,
+      "loss": 0.0737,
+      "step": 2610
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 22.917140065123103,
+      "learning_rate": 5.792875105899294e-06,
+      "loss": 0.1536,
+      "step": 2611
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 14.280858621166734,
+      "learning_rate": 5.790021393227747e-06,
+      "loss": 0.1257,
+      "step": 2612
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.601450815452466,
+      "learning_rate": 5.787167416595369e-06,
+      "loss": 0.0939,
+      "step": 2613
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 23.25278219450862,
+      "learning_rate": 5.784313176955726e-06,
+      "loss": 0.1099,
+      "step": 2614
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 13.52299424961681,
+      "learning_rate": 5.781458675262472e-06,
+      "loss": 0.0918,
+      "step": 2615
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 13.140181284280587,
+      "learning_rate": 5.778603912469349e-06,
+      "loss": 0.1211,
+      "step": 2616
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 17.987639399325246,
+      "learning_rate": 5.775748889530187e-06,
+      "loss": 0.1158,
+      "step": 2617
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.029185285872469,
+      "learning_rate": 5.772893607398901e-06,
+      "loss": 0.0793,
+      "step": 2618
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 10.243106638432673,
+      "learning_rate": 5.770038067029496e-06,
+      "loss": 0.0837,
+      "step": 2619
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 7.120185441810166,
+      "learning_rate": 5.76718226937606e-06,
+      "loss": 0.071,
+      "step": 2620
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 17.79704560468372,
+      "learning_rate": 5.764326215392768e-06,
+      "loss": 0.0668,
+      "step": 2621
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 23.092612411054002,
+      "learning_rate": 5.761469906033879e-06,
+      "loss": 0.0911,
+      "step": 2622
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.923215155786091,
+      "learning_rate": 5.758613342253743e-06,
+      "loss": 0.0652,
+      "step": 2623
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 12.23566762430646,
+      "learning_rate": 5.7557565250067896e-06,
+      "loss": 0.093,
+      "step": 2624
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 29.68819365852448,
+      "learning_rate": 5.752899455247532e-06,
+      "loss": 0.0942,
+      "step": 2625
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 25.11407988203367,
+      "learning_rate": 5.750042133930571e-06,
+      "loss": 0.1067,
+      "step": 2626
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 13.029913042547735,
+      "learning_rate": 5.7471845620105925e-06,
+      "loss": 0.095,
+      "step": 2627
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.765283553153818,
+      "learning_rate": 5.744326740442364e-06,
+      "loss": 0.1107,
+      "step": 2628
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 14.7909670270782,
+      "learning_rate": 5.741468670180737e-06,
+      "loss": 0.1256,
+      "step": 2629
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 16.345896107400005,
+      "learning_rate": 5.738610352180645e-06,
+      "loss": 0.1219,
+      "step": 2630
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 6.525593347954734,
+      "learning_rate": 5.735751787397106e-06,
+      "loss": 0.0771,
+      "step": 2631
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 29.13519994204438,
+      "learning_rate": 5.732892976785218e-06,
+      "loss": 0.1133,
+      "step": 2632
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.037077891178535,
+      "learning_rate": 5.730033921300166e-06,
+      "loss": 0.0765,
+      "step": 2633
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 11.01310603937196,
+      "learning_rate": 5.7271746218972105e-06,
+      "loss": 0.0965,
+      "step": 2634
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 10.383120380608432,
+      "learning_rate": 5.724315079531697e-06,
+      "loss": 0.0765,
+      "step": 2635
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 14.053650424576215,
+      "learning_rate": 5.721455295159053e-06,
+      "loss": 0.1095,
+      "step": 2636
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 13.49823698629184,
+      "learning_rate": 5.7185952697347844e-06,
+      "loss": 0.1095,
+      "step": 2637
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 15.100988227402953,
+      "learning_rate": 5.71573500421448e-06,
+      "loss": 0.0826,
+      "step": 2638
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 18.69455842462284,
+      "learning_rate": 5.712874499553807e-06,
+      "loss": 0.1101,
+      "step": 2639
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 8.131975095287098,
+      "learning_rate": 5.710013756708513e-06,
+      "loss": 0.1218,
+      "step": 2640
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 18.309767202091365,
+      "learning_rate": 5.707152776634427e-06,
+      "loss": 0.0981,
+      "step": 2641
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 17.760090104193882,
+      "learning_rate": 5.704291560287454e-06,
+      "loss": 0.1068,
+      "step": 2642
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 17.356458991602008,
+      "learning_rate": 5.701430108623578e-06,
+      "loss": 0.0968,
+      "step": 2643
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 8.346261757755903,
+      "learning_rate": 5.698568422598867e-06,
+      "loss": 0.0783,
+      "step": 2644
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 10.936135220918278,
+      "learning_rate": 5.69570650316946e-06,
+      "loss": 0.0816,
+      "step": 2645
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 12.619615725003502,
+      "learning_rate": 5.69284435129158e-06,
+      "loss": 0.105,
+      "step": 2646
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.963461261946324,
+      "learning_rate": 5.689981967921523e-06,
+      "loss": 0.0933,
+      "step": 2647
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.236246645246913,
+      "learning_rate": 5.6871193540156666e-06,
+      "loss": 0.0918,
+      "step": 2648
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 12.81684610998191,
+      "learning_rate": 5.684256510530461e-06,
+      "loss": 0.1476,
+      "step": 2649
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 11.153293940717711,
+      "learning_rate": 5.68139343842244e-06,
+      "loss": 0.0962,
+      "step": 2650
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 14.765702957258346,
+      "learning_rate": 5.678530138648204e-06,
+      "loss": 0.0981,
+      "step": 2651
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.541403950790154,
+      "learning_rate": 5.675666612164436e-06,
+      "loss": 0.1184,
+      "step": 2652
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 27.66496804411558,
+      "learning_rate": 5.672802859927895e-06,
+      "loss": 0.093,
+      "step": 2653
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.315476185301423,
+      "learning_rate": 5.669938882895412e-06,
+      "loss": 0.0898,
+      "step": 2654
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.178791444866505,
+      "learning_rate": 5.667074682023896e-06,
+      "loss": 0.0924,
+      "step": 2655
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 13.533905636009482,
+      "learning_rate": 5.664210258270331e-06,
+      "loss": 0.1217,
+      "step": 2656
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 18.36169164512371,
+      "learning_rate": 5.661345612591771e-06,
+      "loss": 0.0782,
+      "step": 2657
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 7.538301490553328,
+      "learning_rate": 5.6584807459453515e-06,
+      "loss": 0.0942,
+      "step": 2658
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 10.264391649732847,
+      "learning_rate": 5.655615659288274e-06,
+      "loss": 0.1205,
+      "step": 2659
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.794709350968707,
+      "learning_rate": 5.652750353577818e-06,
+      "loss": 0.1161,
+      "step": 2660
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 13.99871974442599,
+      "learning_rate": 5.649884829771337e-06,
+      "loss": 0.1078,
+      "step": 2661
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 7.867770861552083,
+      "learning_rate": 5.6470190888262545e-06,
+      "loss": 0.0922,
+      "step": 2662
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 17.41387225248798,
+      "learning_rate": 5.644153131700067e-06,
+      "loss": 0.1163,
+      "step": 2663
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 39.2761918554314,
+      "learning_rate": 5.6412869593503476e-06,
+      "loss": 0.1678,
+      "step": 2664
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.993724717189073,
+      "learning_rate": 5.638420572734733e-06,
+      "loss": 0.0949,
+      "step": 2665
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 10.59354372045396,
+      "learning_rate": 5.63555397281094e-06,
+      "loss": 0.0689,
+      "step": 2666
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 24.937857294603422,
+      "learning_rate": 5.632687160536751e-06,
+      "loss": 0.0983,
+      "step": 2667
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 41.37417208875139,
+      "learning_rate": 5.629820136870022e-06,
+      "loss": 0.1111,
+      "step": 2668
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 7.437831972634287,
+      "learning_rate": 5.626952902768678e-06,
+      "loss": 0.1069,
+      "step": 2669
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 14.905422981253997,
+      "learning_rate": 5.624085459190717e-06,
+      "loss": 0.0795,
+      "step": 2670
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 18.705922511860052,
+      "learning_rate": 5.621217807094202e-06,
+      "loss": 0.0938,
+      "step": 2671
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 31.833596818269967,
+      "learning_rate": 5.618349947437272e-06,
+      "loss": 0.1108,
+      "step": 2672
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 18.633295676824503,
+      "learning_rate": 5.615481881178132e-06,
+      "loss": 0.1321,
+      "step": 2673
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.9713241603062395,
+      "learning_rate": 5.612613609275054e-06,
+      "loss": 0.0859,
+      "step": 2674
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 11.386906666975248,
+      "learning_rate": 5.609745132686383e-06,
+      "loss": 0.1326,
+      "step": 2675
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 33.73569341071956,
+      "learning_rate": 5.60687645237053e-06,
+      "loss": 0.108,
+      "step": 2676
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 37.4374522679626,
+      "learning_rate": 5.604007569285973e-06,
+      "loss": 0.135,
+      "step": 2677
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 15.482062569547526,
+      "learning_rate": 5.6011384843912605e-06,
+      "loss": 0.098,
+      "step": 2678
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 23.405290661541976,
+      "learning_rate": 5.598269198645008e-06,
+      "loss": 0.0983,
+      "step": 2679
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 12.43065289610563,
+      "learning_rate": 5.5953997130058945e-06,
+      "loss": 0.1143,
+      "step": 2680
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 25.630102473252506,
+      "learning_rate": 5.5925300284326715e-06,
+      "loss": 0.1412,
+      "step": 2681
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 45.50344170828186,
+      "learning_rate": 5.5896601458841505e-06,
+      "loss": 0.1259,
+      "step": 2682
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 30.06550381414141,
+      "learning_rate": 5.586790066319217e-06,
+      "loss": 0.0741,
+      "step": 2683
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 7.271702734925442,
+      "learning_rate": 5.583919790696814e-06,
+      "loss": 0.0865,
+      "step": 2684
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 24.493565366620036,
+      "learning_rate": 5.581049319975957e-06,
+      "loss": 0.0911,
+      "step": 2685
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 10.872141679797997,
+      "learning_rate": 5.57817865511572e-06,
+      "loss": 0.1267,
+      "step": 2686
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 31.622766470405587,
+      "learning_rate": 5.575307797075249e-06,
+      "loss": 0.1034,
+      "step": 2687
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 33.281994290944546,
+      "learning_rate": 5.572436746813748e-06,
+      "loss": 0.1101,
+      "step": 2688
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 9.921027242320786,
+      "learning_rate": 5.5695655052904905e-06,
+      "loss": 0.1333,
+      "step": 2689
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.842511723248251,
+      "learning_rate": 5.566694073464812e-06,
+      "loss": 0.0715,
+      "step": 2690
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 31.392960437519093,
+      "learning_rate": 5.56382245229611e-06,
+      "loss": 0.1133,
+      "step": 2691
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 33.161077306234404,
+      "learning_rate": 5.560950642743847e-06,
+      "loss": 0.1158,
+      "step": 2692
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 7.72182359708755,
+      "learning_rate": 5.558078645767547e-06,
+      "loss": 0.082,
+      "step": 2693
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 33.264247772214716,
+      "learning_rate": 5.5552064623267986e-06,
+      "loss": 0.1349,
+      "step": 2694
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 18.089787658256412,
+      "learning_rate": 5.5523340933812505e-06,
+      "loss": 0.1017,
+      "step": 2695
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 15.398888267717105,
+      "learning_rate": 5.549461539890616e-06,
+      "loss": 0.0844,
+      "step": 2696
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 24.67559393234702,
+      "learning_rate": 5.546588802814669e-06,
+      "loss": 0.0819,
+      "step": 2697
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 18.90602279218803,
+      "learning_rate": 5.543715883113241e-06,
+      "loss": 0.0938,
+      "step": 2698
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 26.617935820127663,
+      "learning_rate": 5.540842781746231e-06,
+      "loss": 0.0928,
+      "step": 2699
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.410245020771018,
+      "learning_rate": 5.537969499673598e-06,
+      "loss": 0.1094,
+      "step": 2700
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 12.46263647968909,
+      "learning_rate": 5.535096037855353e-06,
+      "loss": 0.0858,
+      "step": 2701
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 24.402438809378204,
+      "learning_rate": 5.532222397251576e-06,
+      "loss": 0.0955,
+      "step": 2702
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 18.28253890357763,
+      "learning_rate": 5.529348578822403e-06,
+      "loss": 0.132,
+      "step": 2703
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 18.717809032260746,
+      "learning_rate": 5.526474583528032e-06,
+      "loss": 0.137,
+      "step": 2704
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 24.233609934720455,
+      "learning_rate": 5.523600412328716e-06,
+      "loss": 0.1124,
+      "step": 2705
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 12.27153406116996,
+      "learning_rate": 5.520726066184769e-06,
+      "loss": 0.1357,
+      "step": 2706
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 36.491820592631434,
+      "learning_rate": 5.517851546056566e-06,
+      "loss": 0.0982,
+      "step": 2707
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 24.71547500695131,
+      "learning_rate": 5.5149768529045355e-06,
+      "loss": 0.0795,
+      "step": 2708
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 14.636866072311467,
+      "learning_rate": 5.512101987689168e-06,
+      "loss": 0.0864,
+      "step": 2709
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 7.1436136261471574,
+      "learning_rate": 5.509226951371006e-06,
+      "loss": 0.0834,
+      "step": 2710
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 20.77019856777048,
+      "learning_rate": 5.506351744910654e-06,
+      "loss": 0.0908,
+      "step": 2711
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 20.18000054845474,
+      "learning_rate": 5.503476369268773e-06,
+      "loss": 0.0805,
+      "step": 2712
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.97702212144314,
+      "learning_rate": 5.50060082540608e-06,
+      "loss": 0.0823,
+      "step": 2713
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 18.93799132198841,
+      "learning_rate": 5.4977251142833445e-06,
+      "loss": 0.1274,
+      "step": 2714
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 18.781493822249985,
+      "learning_rate": 5.494849236861397e-06,
+      "loss": 0.093,
+      "step": 2715
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 25.586693957909443,
+      "learning_rate": 5.491973194101122e-06,
+      "loss": 0.1316,
+      "step": 2716
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 33.83182887402268,
+      "learning_rate": 5.4890969869634606e-06,
+      "loss": 0.1128,
+      "step": 2717
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 39.09734754951871,
+      "learning_rate": 5.486220616409403e-06,
+      "loss": 0.1139,
+      "step": 2718
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 17.296007817801954,
+      "learning_rate": 5.4833440834e-06,
+      "loss": 0.0948,
+      "step": 2719
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 36.970414189809944,
+      "learning_rate": 5.480467388896353e-06,
+      "loss": 0.1217,
+      "step": 2720
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 41.98479812243881,
+      "learning_rate": 5.477590533859623e-06,
+      "loss": 0.1089,
+      "step": 2721
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 39.39789221503169,
+      "learning_rate": 5.474713519251018e-06,
+      "loss": 0.1365,
+      "step": 2722
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 10.022368374310624,
+      "learning_rate": 5.471836346031802e-06,
+      "loss": 0.0945,
+      "step": 2723
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 11.411241170144551,
+      "learning_rate": 5.468959015163293e-06,
+      "loss": 0.0825,
+      "step": 2724
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 17.404101960272982,
+      "learning_rate": 5.46608152760686e-06,
+      "loss": 0.0732,
+      "step": 2725
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 39.62197496277501,
+      "learning_rate": 5.463203884323926e-06,
+      "loss": 0.1486,
+      "step": 2726
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 27.49778841473844,
+      "learning_rate": 5.460326086275964e-06,
+      "loss": 0.1128,
+      "step": 2727
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 6.379646370483621,
+      "learning_rate": 5.4574481344245015e-06,
+      "loss": 0.1173,
+      "step": 2728
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 21.947214598255393,
+      "learning_rate": 5.454570029731115e-06,
+      "loss": 0.1093,
+      "step": 2729
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 18.335934862211666,
+      "learning_rate": 5.451691773157431e-06,
+      "loss": 0.1183,
+      "step": 2730
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 51.38289661985466,
+      "learning_rate": 5.448813365665129e-06,
+      "loss": 0.1513,
+      "step": 2731
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 17.977264185528178,
+      "learning_rate": 5.44593480821594e-06,
+      "loss": 0.1239,
+      "step": 2732
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 6.769399150938259,
+      "learning_rate": 5.443056101771643e-06,
+      "loss": 0.0939,
+      "step": 2733
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 49.50787284234708,
+      "learning_rate": 5.44017724729407e-06,
+      "loss": 0.1066,
+      "step": 2734
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 41.47491103746017,
+      "learning_rate": 5.437298245745093e-06,
+      "loss": 0.1199,
+      "step": 2735
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 31.21920187501198,
+      "learning_rate": 5.434419098086645e-06,
+      "loss": 0.1379,
+      "step": 2736
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 4.019317654949999,
+      "learning_rate": 5.431539805280702e-06,
+      "loss": 0.0836,
+      "step": 2737
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 16.908346573560465,
+      "learning_rate": 5.428660368289289e-06,
+      "loss": 0.1027,
+      "step": 2738
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 54.3446520419247,
+      "learning_rate": 5.42578078807448e-06,
+      "loss": 0.1442,
+      "step": 2739
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 29.604460610263637,
+      "learning_rate": 5.422901065598395e-06,
+      "loss": 0.0917,
+      "step": 2740
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 18.052870526381152,
+      "learning_rate": 5.4200212018232024e-06,
+      "loss": 0.0837,
+      "step": 2741
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.485821680212534,
+      "learning_rate": 5.41714119771112e-06,
+      "loss": 0.0809,
+      "step": 2742
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 23.671337108776125,
+      "learning_rate": 5.414261054224412e-06,
+      "loss": 0.0701,
+      "step": 2743
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 13.628921621805322,
+      "learning_rate": 5.411380772325383e-06,
+      "loss": 0.132,
+      "step": 2744
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 58.95361204923983,
+      "learning_rate": 5.408500352976392e-06,
+      "loss": 0.177,
+      "step": 2745
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 13.789171355412767,
+      "learning_rate": 5.40561979713984e-06,
+      "loss": 0.0959,
+      "step": 2746
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.334704571609194,
+      "learning_rate": 5.402739105778175e-06,
+      "loss": 0.1243,
+      "step": 2747
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 46.74494923664595,
+      "learning_rate": 5.399858279853889e-06,
+      "loss": 0.1034,
+      "step": 2748
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 37.41227408696971,
+      "learning_rate": 5.39697732032952e-06,
+      "loss": 0.0923,
+      "step": 2749
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 41.32562404134608,
+      "learning_rate": 5.394096228167648e-06,
+      "loss": 0.1169,
+      "step": 2750
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 11.165428505092693,
+      "learning_rate": 5.391215004330903e-06,
+      "loss": 0.0894,
+      "step": 2751
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 25.320083374694267,
+      "learning_rate": 5.388333649781951e-06,
+      "loss": 0.121,
+      "step": 2752
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 34.05814695395899,
+      "learning_rate": 5.3854521654835105e-06,
+      "loss": 0.1019,
+      "step": 2753
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 37.729778469348126,
+      "learning_rate": 5.3825705523983366e-06,
+      "loss": 0.1252,
+      "step": 2754
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.7768601785361735,
+      "learning_rate": 5.37968881148923e-06,
+      "loss": 0.1022,
+      "step": 2755
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 21.434040827191126,
+      "learning_rate": 5.376806943719033e-06,
+      "loss": 0.106,
+      "step": 2756
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 16.80181493949827,
+      "learning_rate": 5.373924950050633e-06,
+      "loss": 0.093,
+      "step": 2757
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 48.37549473256851,
+      "learning_rate": 5.371042831446957e-06,
+      "loss": 0.172,
+      "step": 2758
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.651227352253567,
+      "learning_rate": 5.3681605888709755e-06,
+      "loss": 0.0944,
+      "step": 2759
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 8.240276586363413,
+      "learning_rate": 5.365278223285698e-06,
+      "loss": 0.1755,
+      "step": 2760
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 13.781340769294333,
+      "learning_rate": 5.362395735654175e-06,
+      "loss": 0.1061,
+      "step": 2761
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 36.77153746292403,
+      "learning_rate": 5.3595131269395015e-06,
+      "loss": 0.1104,
+      "step": 2762
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 26.923003378387772,
+      "learning_rate": 5.356630398104811e-06,
+      "loss": 0.1163,
+      "step": 2763
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 19.27470596602233,
+      "learning_rate": 5.353747550113274e-06,
+      "loss": 0.0679,
+      "step": 2764
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 27.07027957372197,
+      "learning_rate": 5.350864583928106e-06,
+      "loss": 0.1111,
+      "step": 2765
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 7.765524049007483,
+      "learning_rate": 5.347981500512558e-06,
+      "loss": 0.085,
+      "step": 2766
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 41.86517901009098,
+      "learning_rate": 5.345098300829924e-06,
+      "loss": 0.1456,
+      "step": 2767
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.349243010931892,
+      "learning_rate": 5.342214985843534e-06,
+      "loss": 0.0994,
+      "step": 2768
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 16.401974170097617,
+      "learning_rate": 5.339331556516755e-06,
+      "loss": 0.0795,
+      "step": 2769
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 16.03990289754173,
+      "learning_rate": 5.336448013812996e-06,
+      "loss": 0.1182,
+      "step": 2770
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 17.74040652818955,
+      "learning_rate": 5.333564358695701e-06,
+      "loss": 0.0903,
+      "step": 2771
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 41.62125267709523,
+      "learning_rate": 5.330680592128355e-06,
+      "loss": 0.1292,
+      "step": 2772
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 11.822329634056095,
+      "learning_rate": 5.3277967150744755e-06,
+      "loss": 0.0974,
+      "step": 2773
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 17.12557284468244,
+      "learning_rate": 5.324912728497621e-06,
+      "loss": 0.104,
+      "step": 2774
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 51.98493528591735,
+      "learning_rate": 5.322028633361386e-06,
+      "loss": 0.1464,
+      "step": 2775
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 38.343198477712235,
+      "learning_rate": 5.319144430629397e-06,
+      "loss": 0.1155,
+      "step": 2776
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 13.395904291966062,
+      "learning_rate": 5.316260121265323e-06,
+      "loss": 0.1008,
+      "step": 2777
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 12.23983059469622,
+      "learning_rate": 5.313375706232864e-06,
+      "loss": 0.1072,
+      "step": 2778
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.743029425887316,
+      "learning_rate": 5.310491186495757e-06,
+      "loss": 0.0966,
+      "step": 2779
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 28.054949717811493,
+      "learning_rate": 5.307606563017772e-06,
+      "loss": 0.0975,
+      "step": 2780
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 15.050796478633394,
+      "learning_rate": 5.304721836762717e-06,
+      "loss": 0.0774,
+      "step": 2781
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 6.720263367169756,
+      "learning_rate": 5.301837008694433e-06,
+      "loss": 0.0818,
+      "step": 2782
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.712277096253679,
+      "learning_rate": 5.298952079776794e-06,
+      "loss": 0.0989,
+      "step": 2783
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 7.4999136139393965,
+      "learning_rate": 5.296067050973709e-06,
+      "loss": 0.0798,
+      "step": 2784
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 16.593799394745258,
+      "learning_rate": 5.29318192324912e-06,
+      "loss": 0.0981,
+      "step": 2785
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.159019478221214,
+      "learning_rate": 5.290296697566999e-06,
+      "loss": 0.0942,
+      "step": 2786
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 12.250548068310936,
+      "learning_rate": 5.287411374891356e-06,
+      "loss": 0.0811,
+      "step": 2787
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 11.420608184090936,
+      "learning_rate": 5.284525956186231e-06,
+      "loss": 0.098,
+      "step": 2788
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 6.178523711245514,
+      "learning_rate": 5.281640442415695e-06,
+      "loss": 0.1224,
+      "step": 2789
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 14.432868868069074,
+      "learning_rate": 5.278754834543852e-06,
+      "loss": 0.0961,
+      "step": 2790
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 19.832143314823682,
+      "learning_rate": 5.275869133534838e-06,
+      "loss": 0.0934,
+      "step": 2791
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 6.173140621912715,
+      "learning_rate": 5.272983340352818e-06,
+      "loss": 0.0894,
+      "step": 2792
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.664241283983232,
+      "learning_rate": 5.270097455961991e-06,
+      "loss": 0.0939,
+      "step": 2793
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 6.199786160737331,
+      "learning_rate": 5.267211481326584e-06,
+      "loss": 0.0717,
+      "step": 2794
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 9.359083770279224,
+      "learning_rate": 5.264325417410854e-06,
+      "loss": 0.0862,
+      "step": 2795
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.999303639282669,
+      "learning_rate": 5.261439265179089e-06,
+      "loss": 0.0884,
+      "step": 2796
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 9.531777669594215,
+      "learning_rate": 5.258553025595605e-06,
+      "loss": 0.1085,
+      "step": 2797
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 10.564778248106444,
+      "learning_rate": 5.255666699624749e-06,
+      "loss": 0.0837,
+      "step": 2798
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.144429532283462,
+      "learning_rate": 5.252780288230899e-06,
+      "loss": 0.0817,
+      "step": 2799
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.479863646274058,
+      "learning_rate": 5.249893792378454e-06,
+      "loss": 0.1122,
+      "step": 2800
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.219194155167031,
+      "learning_rate": 5.24700721303185e-06,
+      "loss": 0.1039,
+      "step": 2801
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 16.285123515055172,
+      "learning_rate": 5.244120551155544e-06,
+      "loss": 0.1119,
+      "step": 2802
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 13.901589114079679,
+      "learning_rate": 5.241233807714024e-06,
+      "loss": 0.054,
+      "step": 2803
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.199611427046534,
+      "learning_rate": 5.238346983671805e-06,
+      "loss": 0.0583,
+      "step": 2804
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 7.9097903544474075,
+      "learning_rate": 5.235460079993429e-06,
+      "loss": 0.0524,
+      "step": 2805
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 9.10846903963757,
+      "learning_rate": 5.232573097643462e-06,
+      "loss": 0.0448,
+      "step": 2806
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 8.33526310227755,
+      "learning_rate": 5.229686037586502e-06,
+      "loss": 0.0567,
+      "step": 2807
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 14.233119308284538,
+      "learning_rate": 5.226798900787167e-06,
+      "loss": 0.0535,
+      "step": 2808
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 7.672664735913163,
+      "learning_rate": 5.223911688210104e-06,
+      "loss": 0.0562,
+      "step": 2809
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 11.614815160250359,
+      "learning_rate": 5.221024400819983e-06,
+      "loss": 0.0532,
+      "step": 2810
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 14.154573219522113,
+      "learning_rate": 5.218137039581504e-06,
+      "loss": 0.0424,
+      "step": 2811
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 6.576671964529106,
+      "learning_rate": 5.215249605459382e-06,
+      "loss": 0.0618,
+      "step": 2812
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 11.287896142648293,
+      "learning_rate": 5.212362099418369e-06,
+      "loss": 0.0522,
+      "step": 2813
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 6.582372740411856,
+      "learning_rate": 5.2094745224232306e-06,
+      "loss": 0.0511,
+      "step": 2814
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 6.17959564383128,
+      "learning_rate": 5.206586875438759e-06,
+      "loss": 0.0565,
+      "step": 2815
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 12.734580828635346,
+      "learning_rate": 5.203699159429773e-06,
+      "loss": 0.049,
+      "step": 2816
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 13.886150462876547,
+      "learning_rate": 5.200811375361112e-06,
+      "loss": 0.0618,
+      "step": 2817
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 9.75151912924947,
+      "learning_rate": 5.197923524197639e-06,
+      "loss": 0.0517,
+      "step": 2818
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 8.18582132062964,
+      "learning_rate": 5.195035606904237e-06,
+      "loss": 0.0461,
+      "step": 2819
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.949615940026982,
+      "learning_rate": 5.1921476244458135e-06,
+      "loss": 0.07,
+      "step": 2820
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.522744474789711,
+      "learning_rate": 5.189259577787297e-06,
+      "loss": 0.0469,
+      "step": 2821
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 13.237191723970465,
+      "learning_rate": 5.186371467893638e-06,
+      "loss": 0.0585,
+      "step": 2822
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 22.63811767687695,
+      "learning_rate": 5.1834832957298075e-06,
+      "loss": 0.0612,
+      "step": 2823
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 10.856725695302611,
+      "learning_rate": 5.180595062260797e-06,
+      "loss": 0.0328,
+      "step": 2824
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.140432943530161,
+      "learning_rate": 5.177706768451619e-06,
+      "loss": 0.042,
+      "step": 2825
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 14.324322382042538,
+      "learning_rate": 5.174818415267308e-06,
+      "loss": 0.0389,
+      "step": 2826
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 12.103738058031697,
+      "learning_rate": 5.1719300036729135e-06,
+      "loss": 0.0532,
+      "step": 2827
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 23.584660325094507,
+      "learning_rate": 5.169041534633511e-06,
+      "loss": 0.0512,
+      "step": 2828
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 10.668112848453543,
+      "learning_rate": 5.166153009114188e-06,
+      "loss": 0.0463,
+      "step": 2829
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.695483605532189,
+      "learning_rate": 5.163264428080057e-06,
+      "loss": 0.0368,
+      "step": 2830
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 12.550637521816691,
+      "learning_rate": 5.160375792496246e-06,
+      "loss": 0.0579,
+      "step": 2831
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 7.404223260758192,
+      "learning_rate": 5.157487103327901e-06,
+      "loss": 0.0544,
+      "step": 2832
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 14.804123391013425,
+      "learning_rate": 5.1545983615401885e-06,
+      "loss": 0.0443,
+      "step": 2833
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.9860623662205845,
+      "learning_rate": 5.151709568098289e-06,
+      "loss": 0.0437,
+      "step": 2834
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.4808084788201925,
+      "learning_rate": 5.1488207239674036e-06,
+      "loss": 0.048,
+      "step": 2835
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 9.244067790459159,
+      "learning_rate": 5.145931830112748e-06,
+      "loss": 0.0397,
+      "step": 2836
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 6.7362343327123035,
+      "learning_rate": 5.1430428874995554e-06,
+      "loss": 0.0555,
+      "step": 2837
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 9.004920379741428,
+      "learning_rate": 5.140153897093076e-06,
+      "loss": 0.0459,
+      "step": 2838
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 8.516109389469001,
+      "learning_rate": 5.1372648598585725e-06,
+      "loss": 0.0656,
+      "step": 2839
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 8.640361450205146,
+      "learning_rate": 5.134375776761329e-06,
+      "loss": 0.0576,
+      "step": 2840
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.993357106722759,
+      "learning_rate": 5.131486648766642e-06,
+      "loss": 0.0432,
+      "step": 2841
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 14.19048628173077,
+      "learning_rate": 5.1285974768398205e-06,
+      "loss": 0.0488,
+      "step": 2842
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 3.711900865433462,
+      "learning_rate": 5.125708261946192e-06,
+      "loss": 0.0494,
+      "step": 2843
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 7.5442087251704875,
+      "learning_rate": 5.122819005051096e-06,
+      "loss": 0.0534,
+      "step": 2844
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 20.77714075942429,
+      "learning_rate": 5.119929707119889e-06,
+      "loss": 0.057,
+      "step": 2845
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 24.696838420818487,
+      "learning_rate": 5.117040369117937e-06,
+      "loss": 0.071,
+      "step": 2846
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 12.90688500272758,
+      "learning_rate": 5.114150992010621e-06,
+      "loss": 0.0461,
+      "step": 2847
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 21.76698605488115,
+      "learning_rate": 5.1112615767633385e-06,
+      "loss": 0.0688,
+      "step": 2848
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 12.042476678664588,
+      "learning_rate": 5.108372124341494e-06,
+      "loss": 0.0603,
+      "step": 2849
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 8.711444498022269,
+      "learning_rate": 5.105482635710509e-06,
+      "loss": 0.045,
+      "step": 2850
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.965522711679408,
+      "learning_rate": 5.102593111835815e-06,
+      "loss": 0.0609,
+      "step": 2851
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.78715066740317,
+      "learning_rate": 5.099703553682854e-06,
+      "loss": 0.0474,
+      "step": 2852
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 14.612001845478728,
+      "learning_rate": 5.096813962217086e-06,
+      "loss": 0.0386,
+      "step": 2853
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 8.213767104804448,
+      "learning_rate": 5.093924338403971e-06,
+      "loss": 0.0436,
+      "step": 2854
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 14.21885718408981,
+      "learning_rate": 5.091034683208988e-06,
+      "loss": 0.0649,
+      "step": 2855
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 9.200406261001536,
+      "learning_rate": 5.088144997597627e-06,
+      "loss": 0.0344,
+      "step": 2856
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.7563067974727264,
+      "learning_rate": 5.085255282535383e-06,
+      "loss": 0.0523,
+      "step": 2857
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 17.092771383434464,
+      "learning_rate": 5.082365538987765e-06,
+      "loss": 0.0494,
+      "step": 2858
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 6.860949499099525,
+      "learning_rate": 5.079475767920289e-06,
+      "loss": 0.0499,
+      "step": 2859
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 22.180665634795407,
+      "learning_rate": 5.076585970298481e-06,
+      "loss": 0.0686,
+      "step": 2860
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 14.512557797350894,
+      "learning_rate": 5.073696147087878e-06,
+      "loss": 0.0486,
+      "step": 2861
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 6.190960927702063,
+      "learning_rate": 5.070806299254023e-06,
+      "loss": 0.0379,
+      "step": 2862
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 7.028780537576393,
+      "learning_rate": 5.067916427762466e-06,
+      "loss": 0.0535,
+      "step": 2863
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 22.0853084053586,
+      "learning_rate": 5.0650265335787685e-06,
+      "loss": 0.0513,
+      "step": 2864
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 18.82402778771914,
+      "learning_rate": 5.062136617668497e-06,
+      "loss": 0.0328,
+      "step": 2865
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 9.35572729496596,
+      "learning_rate": 5.059246680997228e-06,
+      "loss": 0.0499,
+      "step": 2866
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 15.494949337550196,
+      "learning_rate": 5.05635672453054e-06,
+      "loss": 0.0523,
+      "step": 2867
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 17.512400175601442,
+      "learning_rate": 5.053466749234023e-06,
+      "loss": 0.0384,
+      "step": 2868
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 40.02758074357572,
+      "learning_rate": 5.050576756073272e-06,
+      "loss": 0.0751,
+      "step": 2869
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 18.427268569788204,
+      "learning_rate": 5.047686746013888e-06,
+      "loss": 0.0413,
+      "step": 2870
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.541127274636648,
+      "learning_rate": 5.044796720021474e-06,
+      "loss": 0.0522,
+      "step": 2871
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 7.965720801969884,
+      "learning_rate": 5.041906679061643e-06,
+      "loss": 0.0359,
+      "step": 2872
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 19.01943256008662,
+      "learning_rate": 5.039016624100013e-06,
+      "loss": 0.0659,
+      "step": 2873
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 27.05389706378414,
+      "learning_rate": 5.036126556102202e-06,
+      "loss": 0.063,
+      "step": 2874
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.472693277113262,
+      "learning_rate": 5.033236476033838e-06,
+      "loss": 0.042,
+      "step": 2875
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.360766469879016,
+      "learning_rate": 5.0303463848605495e-06,
+      "loss": 0.0457,
+      "step": 2876
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 16.162091546785724,
+      "learning_rate": 5.027456283547969e-06,
+      "loss": 0.0628,
+      "step": 2877
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 21.94294781666631,
+      "learning_rate": 5.0245661730617344e-06,
+      "loss": 0.0573,
+      "step": 2878
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 12.700207444520526,
+      "learning_rate": 5.0216760543674855e-06,
+      "loss": 0.0488,
+      "step": 2879
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 10.814485131817554,
+      "learning_rate": 5.0187859284308635e-06,
+      "loss": 0.059,
+      "step": 2880
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 19.09796261660822,
+      "learning_rate": 5.015895796217514e-06,
+      "loss": 0.056,
+      "step": 2881
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 21.58846874899979,
+      "learning_rate": 5.013005658693083e-06,
+      "loss": 0.0521,
+      "step": 2882
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 16.698159335405045,
+      "learning_rate": 5.01011551682322e-06,
+      "loss": 0.0566,
+      "step": 2883
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 14.016096004860056,
+      "learning_rate": 5.007225371573573e-06,
+      "loss": 0.0602,
+      "step": 2884
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.06089635411483,
+      "learning_rate": 5.004335223909797e-06,
+      "loss": 0.0458,
+      "step": 2885
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 13.265950504045813,
+      "learning_rate": 5.0014450747975416e-06,
+      "loss": 0.0705,
+      "step": 2886
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 19.71288313562799,
+      "learning_rate": 4.998554925202459e-06,
+      "loss": 0.0456,
+      "step": 2887
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 18.997859716152803,
+      "learning_rate": 4.995664776090204e-06,
+      "loss": 0.0703,
+      "step": 2888
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 12.003644856154164,
+      "learning_rate": 4.9927746284264275e-06,
+      "loss": 0.0547,
+      "step": 2889
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.8612356586077965,
+      "learning_rate": 4.9898844831767826e-06,
+      "loss": 0.0421,
+      "step": 2890
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 3.857541041219327,
+      "learning_rate": 4.98699434130692e-06,
+      "loss": 0.0409,
+      "step": 2891
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 27.220331399521893,
+      "learning_rate": 4.984104203782488e-06,
+      "loss": 0.0518,
+      "step": 2892
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 17.76518467751359,
+      "learning_rate": 4.981214071569139e-06,
+      "loss": 0.0568,
+      "step": 2893
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 6.706506867503759,
+      "learning_rate": 4.978323945632515e-06,
+      "loss": 0.0346,
+      "step": 2894
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 6.00943424052001,
+      "learning_rate": 4.975433826938267e-06,
+      "loss": 0.049,
+      "step": 2895
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 6.204979629155921,
+      "learning_rate": 4.972543716452031e-06,
+      "loss": 0.0735,
+      "step": 2896
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 10.632788181873034,
+      "learning_rate": 4.969653615139452e-06,
+      "loss": 0.0454,
+      "step": 2897
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.839103370356773,
+      "learning_rate": 4.966763523966163e-06,
+      "loss": 0.0974,
+      "step": 2898
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 10.849055825875318,
+      "learning_rate": 4.963873443897799e-06,
+      "loss": 0.0521,
+      "step": 2899
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 7.708000075587648,
+      "learning_rate": 4.96098337589999e-06,
+      "loss": 0.0355,
+      "step": 2900
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.723507585122161,
+      "learning_rate": 4.958093320938358e-06,
+      "loss": 0.0539,
+      "step": 2901
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 8.158944443681586,
+      "learning_rate": 4.955203279978529e-06,
+      "loss": 0.0671,
+      "step": 2902
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 9.759906642189957,
+      "learning_rate": 4.952313253986114e-06,
+      "loss": 0.0531,
+      "step": 2903
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 6.11064821441867,
+      "learning_rate": 4.9494232439267296e-06,
+      "loss": 0.0496,
+      "step": 2904
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 15.84805318936541,
+      "learning_rate": 4.946533250765977e-06,
+      "loss": 0.0455,
+      "step": 2905
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 12.764198896161808,
+      "learning_rate": 4.943643275469461e-06,
+      "loss": 0.038,
+      "step": 2906
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 8.974850741601227,
+      "learning_rate": 4.940753319002773e-06,
+      "loss": 0.053,
+      "step": 2907
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 13.75551037130389,
+      "learning_rate": 4.937863382331504e-06,
+      "loss": 0.0544,
+      "step": 2908
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 16.32267799964506,
+      "learning_rate": 4.934973466421234e-06,
+      "loss": 0.062,
+      "step": 2909
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 10.103817381967463,
+      "learning_rate": 4.932083572237535e-06,
+      "loss": 0.0406,
+      "step": 2910
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 6.213672979461658,
+      "learning_rate": 4.92919370074598e-06,
+      "loss": 0.0564,
+      "step": 2911
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 9.834401409500478,
+      "learning_rate": 4.926303852912123e-06,
+      "loss": 0.0533,
+      "step": 2912
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 6.141094774831048,
+      "learning_rate": 4.9234140297015204e-06,
+      "loss": 0.0361,
+      "step": 2913
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 17.040575294136367,
+      "learning_rate": 4.920524232079712e-06,
+      "loss": 0.06,
+      "step": 2914
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 18.436793853047202,
+      "learning_rate": 4.917634461012238e-06,
+      "loss": 0.0632,
+      "step": 2915
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 11.958023047209824,
+      "learning_rate": 4.914744717464617e-06,
+      "loss": 0.0414,
+      "step": 2916
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 6.374684872099406,
+      "learning_rate": 4.911855002402375e-06,
+      "loss": 0.0558,
+      "step": 2917
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 8.56698629233553,
+      "learning_rate": 4.908965316791014e-06,
+      "loss": 0.0278,
+      "step": 2918
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 8.288813639711439,
+      "learning_rate": 4.906075661596031e-06,
+      "loss": 0.0467,
+      "step": 2919
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 25.748623981042726,
+      "learning_rate": 4.903186037782917e-06,
+      "loss": 0.0627,
+      "step": 2920
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.1547283981098655,
+      "learning_rate": 4.900296446317146e-06,
+      "loss": 0.0426,
+      "step": 2921
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 11.060180181270248,
+      "learning_rate": 4.897406888164187e-06,
+      "loss": 0.0411,
+      "step": 2922
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 9.894982430864497,
+      "learning_rate": 4.8945173642894915e-06,
+      "loss": 0.0428,
+      "step": 2923
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 10.318786752951437,
+      "learning_rate": 4.8916278756585074e-06,
+      "loss": 0.038,
+      "step": 2924
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 23.226668767699405,
+      "learning_rate": 4.888738423236664e-06,
+      "loss": 0.0451,
+      "step": 2925
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 7.663805162430875,
+      "learning_rate": 4.88584900798938e-06,
+      "loss": 0.0311,
+      "step": 2926
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 11.873525027460229,
+      "learning_rate": 4.882959630882066e-06,
+      "loss": 0.0649,
+      "step": 2927
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 21.51580574874802,
+      "learning_rate": 4.8800702928801124e-06,
+      "loss": 0.0541,
+      "step": 2928
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 9.563175048514058,
+      "learning_rate": 4.8771809949489056e-06,
+      "loss": 0.0491,
+      "step": 2929
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 9.177795863776534,
+      "learning_rate": 4.874291738053809e-06,
+      "loss": 0.058,
+      "step": 2930
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 14.671314431344912,
+      "learning_rate": 4.871402523160181e-06,
+      "loss": 0.0529,
+      "step": 2931
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 8.21054929459398,
+      "learning_rate": 4.868513351233359e-06,
+      "loss": 0.0407,
+      "step": 2932
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 14.943488995962824,
+      "learning_rate": 4.865624223238672e-06,
+      "loss": 0.0537,
+      "step": 2933
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 9.285607421042759,
+      "learning_rate": 4.862735140141428e-06,
+      "loss": 0.0526,
+      "step": 2934
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 12.545169261685402,
+      "learning_rate": 4.859846102906927e-06,
+      "loss": 0.0347,
+      "step": 2935
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 6.418869884194012,
+      "learning_rate": 4.856957112500446e-06,
+      "loss": 0.0605,
+      "step": 2936
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 10.91319808557055,
+      "learning_rate": 4.854068169887254e-06,
+      "loss": 0.052,
+      "step": 2937
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 7.162112299173172,
+      "learning_rate": 4.851179276032598e-06,
+      "loss": 0.0621,
+      "step": 2938
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 17.64808170091903,
+      "learning_rate": 4.848290431901712e-06,
+      "loss": 0.0616,
+      "step": 2939
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.721241146237817,
+      "learning_rate": 4.845401638459813e-06,
+      "loss": 0.0644,
+      "step": 2940
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.215058023030627,
+      "learning_rate": 4.8425128966721e-06,
+      "loss": 0.0445,
+      "step": 2941
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 8.470561168753651,
+      "learning_rate": 4.8396242075037555e-06,
+      "loss": 0.0458,
+      "step": 2942
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 12.193270887378144,
+      "learning_rate": 4.836735571919946e-06,
+      "loss": 0.04,
+      "step": 2943
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 6.51982708722715,
+      "learning_rate": 4.833846990885813e-06,
+      "loss": 0.0451,
+      "step": 2944
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 14.275220224337577,
+      "learning_rate": 4.830958465366492e-06,
+      "loss": 0.06,
+      "step": 2945
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 14.485618936801512,
+      "learning_rate": 4.828069996327088e-06,
+      "loss": 0.0452,
+      "step": 2946
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 27.79698221697965,
+      "learning_rate": 4.825181584732695e-06,
+      "loss": 0.0803,
+      "step": 2947
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 7.760801102621836,
+      "learning_rate": 4.822293231548382e-06,
+      "loss": 0.051,
+      "step": 2948
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 18.186535510480933,
+      "learning_rate": 4.819404937739205e-06,
+      "loss": 0.0798,
+      "step": 2949
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 13.657062142670078,
+      "learning_rate": 4.816516704270194e-06,
+      "loss": 0.0419,
+      "step": 2950
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 17.763665719452103,
+      "learning_rate": 4.813628532106363e-06,
+      "loss": 0.041,
+      "step": 2951
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 17.213024245970015,
+      "learning_rate": 4.810740422212705e-06,
+      "loss": 0.0514,
+      "step": 2952
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.415365069317704,
+      "learning_rate": 4.807852375554188e-06,
+      "loss": 0.0439,
+      "step": 2953
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.156693282692543,
+      "learning_rate": 4.804964393095765e-06,
+      "loss": 0.0477,
+      "step": 2954
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 11.394625354113346,
+      "learning_rate": 4.802076475802362e-06,
+      "loss": 0.0494,
+      "step": 2955
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.3842660382835024,
+      "learning_rate": 4.799188624638889e-06,
+      "loss": 0.0276,
+      "step": 2956
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.446181745984596,
+      "learning_rate": 4.796300840570227e-06,
+      "loss": 0.0522,
+      "step": 2957
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 15.29577470926494,
+      "learning_rate": 4.793413124561243e-06,
+      "loss": 0.0464,
+      "step": 2958
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 8.414125335615022,
+      "learning_rate": 4.790525477576773e-06,
+      "loss": 0.0498,
+      "step": 2959
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.031180542596354,
+      "learning_rate": 4.7876379005816325e-06,
+      "loss": 0.032,
+      "step": 2960
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.6969083162862804,
+      "learning_rate": 4.784750394540619e-06,
+      "loss": 0.0463,
+      "step": 2961
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 13.339709877305827,
+      "learning_rate": 4.781862960418498e-06,
+      "loss": 0.0376,
+      "step": 2962
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 18.500857965632857,
+      "learning_rate": 4.778975599180019e-06,
+      "loss": 0.0535,
+      "step": 2963
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 5.964132136696635,
+      "learning_rate": 4.776088311789897e-06,
+      "loss": 0.0497,
+      "step": 2964
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.411962079455377,
+      "learning_rate": 4.773201099212835e-06,
+      "loss": 0.0402,
+      "step": 2965
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 8.713256952124864,
+      "learning_rate": 4.770313962413499e-06,
+      "loss": 0.0499,
+      "step": 2966
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 14.98799544031958,
+      "learning_rate": 4.767426902356539e-06,
+      "loss": 0.0379,
+      "step": 2967
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 6.230504284482281,
+      "learning_rate": 4.7645399200065745e-06,
+      "loss": 0.0564,
+      "step": 2968
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 7.904911066689615,
+      "learning_rate": 4.761653016328197e-06,
+      "loss": 0.0477,
+      "step": 2969
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 15.071860947840747,
+      "learning_rate": 4.758766192285979e-06,
+      "loss": 0.0497,
+      "step": 2970
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 21.895501390170683,
+      "learning_rate": 4.755879448844458e-06,
+      "loss": 0.0709,
+      "step": 2971
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 13.52662158776065,
+      "learning_rate": 4.752992786968153e-06,
+      "loss": 0.0518,
+      "step": 2972
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 18.273989464770818,
+      "learning_rate": 4.750106207621546e-06,
+      "loss": 0.0556,
+      "step": 2973
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 12.57628275065163,
+      "learning_rate": 4.747219711769103e-06,
+      "loss": 0.0445,
+      "step": 2974
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 8.238153462757012,
+      "learning_rate": 4.74433330037525e-06,
+      "loss": 0.0358,
+      "step": 2975
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 5.815107502108913,
+      "learning_rate": 4.741446974404396e-06,
+      "loss": 0.0553,
+      "step": 2976
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 15.551938145304423,
+      "learning_rate": 4.738560734820914e-06,
+      "loss": 0.0397,
+      "step": 2977
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 11.898776772774422,
+      "learning_rate": 4.735674582589147e-06,
+      "loss": 0.0439,
+      "step": 2978
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 10.619558463221429,
+      "learning_rate": 4.732788518673418e-06,
+      "loss": 0.0492,
+      "step": 2979
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 9.889398543545939,
+      "learning_rate": 4.729902544038009e-06,
+      "loss": 0.0441,
+      "step": 2980
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.05100282019345,
+      "learning_rate": 4.7270166596471825e-06,
+      "loss": 0.0447,
+      "step": 2981
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 17.41976116286379,
+      "learning_rate": 4.724130866465163e-06,
+      "loss": 0.0475,
+      "step": 2982
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 11.266177641235146,
+      "learning_rate": 4.721245165456149e-06,
+      "loss": 0.0494,
+      "step": 2983
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 11.20597565365233,
+      "learning_rate": 4.7183595575843055e-06,
+      "loss": 0.069,
+      "step": 2984
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.259086726470608,
+      "learning_rate": 4.715474043813771e-06,
+      "loss": 0.0356,
+      "step": 2985
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.080850839428504,
+      "learning_rate": 4.712588625108645e-06,
+      "loss": 0.0481,
+      "step": 2986
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 7.956793520224343,
+      "learning_rate": 4.709703302433003e-06,
+      "loss": 0.0567,
+      "step": 2987
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 21.843191213401248,
+      "learning_rate": 4.706818076750883e-06,
+      "loss": 0.0645,
+      "step": 2988
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 18.694211673739613,
+      "learning_rate": 4.703932949026291e-06,
+      "loss": 0.0492,
+      "step": 2989
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 15.151416311688783,
+      "learning_rate": 4.701047920223207e-06,
+      "loss": 0.0446,
+      "step": 2990
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 10.890580913948755,
+      "learning_rate": 4.6981629913055674e-06,
+      "loss": 0.0569,
+      "step": 2991
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 7.892486306024457,
+      "learning_rate": 4.695278163237284e-06,
+      "loss": 0.0312,
+      "step": 2992
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 12.299530495762703,
+      "learning_rate": 4.692393436982229e-06,
+      "loss": 0.0308,
+      "step": 2993
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 9.153777049929731,
+      "learning_rate": 4.689508813504246e-06,
+      "loss": 0.0351,
+      "step": 2994
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.108518173770672,
+      "learning_rate": 4.686624293767138e-06,
+      "loss": 0.0349,
+      "step": 2995
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 17.853184369338187,
+      "learning_rate": 4.683739878734678e-06,
+      "loss": 0.0641,
+      "step": 2996
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 6.466717945483939,
+      "learning_rate": 4.6808555693706045e-06,
+      "loss": 0.0453,
+      "step": 2997
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.829172386888312,
+      "learning_rate": 4.677971366638616e-06,
+      "loss": 0.0391,
+      "step": 2998
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 6.808172030984788,
+      "learning_rate": 4.67508727150238e-06,
+      "loss": 0.0492,
+      "step": 2999
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 6.156095884400888,
+      "learning_rate": 4.672203284925525e-06,
+      "loss": 0.0414,
+      "step": 3000
+    },
+    {
+      "epoch": 2.14,
+      "eval_avg_AUC": 0.8227856069027939,
+      "eval_avg_Accuracy": 0.7180039787798409,
+      "eval_avg_Accuracy-right": 0.9098734837615756,
+      "eval_avg_Accuracy-wrong": 0.3834432567659768,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.6838678458638056,
+      "eval_last_AUC": 0.8253504247874757,
+      "eval_last_Accuracy": 0.7611903183023873,
+      "eval_last_Accuracy-right": 0.873288117907917,
+      "eval_last_Accuracy-wrong": 0.5657266317944053,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.6932099187920158,
+      "eval_max_AUC": 0.7682526969266754,
+      "eval_max_Accuracy": 0.6413710212201591,
+      "eval_max_Accuracy-right": 0.9846745793661145,
+      "eval_max_Accuracy-wrong": 0.042756424835114853,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.6326595141662732,
+      "eval_min_AUC": 0.8253761205386874,
+      "eval_min_Accuracy": 0.7610245358090185,
+      "eval_min_Accuracy-right": 0.808921351245598,
+      "eval_min_Accuracy-wrong": 0.6775073914032295,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.6877850984496956,
+      "eval_prod_AUC": 0.8274092551394246,
+      "eval_prod_Accuracy": 0.7385195623342176,
+      "eval_prod_Accuracy-right": 0.690556932307291,
+      "eval_prod_Accuracy-wrong": 0.8221514669092563,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.6858525438813741,
+      "eval_runtime": 246.8659,
+      "eval_samples_per_second": 97.737,
+      "eval_steps_per_second": 3.054,
+      "eval_sum_AUC": 0.6815349093354525,
+      "eval_sum_Accuracy": 0.6375994694960212,
+      "eval_sum_Accuracy-right": 0.9964784139820008,
+      "eval_sum_Accuracy-wrong": 0.011826245167159426,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6539070349976671,
+      "step": 3000
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 7.733951582678205,
+      "learning_rate": 4.669319407871647e-06,
+      "loss": 0.0526,
+      "step": 3001
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 13.960495751854996,
+      "learning_rate": 4.666435641304301e-06,
+      "loss": 0.046,
+      "step": 3002
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 8.77619391019296,
+      "learning_rate": 4.663551986187006e-06,
+      "loss": 0.0649,
+      "step": 3003
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 6.884626530449051,
+      "learning_rate": 4.660668443483248e-06,
+      "loss": 0.0432,
+      "step": 3004
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.3549337785724465,
+      "learning_rate": 4.657785014156468e-06,
+      "loss": 0.0498,
+      "step": 3005
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 10.946093760776968,
+      "learning_rate": 4.654901699170077e-06,
+      "loss": 0.0597,
+      "step": 3006
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 15.763447770059578,
+      "learning_rate": 4.652018499487442e-06,
+      "loss": 0.0491,
+      "step": 3007
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.809896767903205,
+      "learning_rate": 4.649135416071896e-06,
+      "loss": 0.061,
+      "step": 3008
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 15.298958843270297,
+      "learning_rate": 4.646252449886727e-06,
+      "loss": 0.0544,
+      "step": 3009
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 10.962570611172566,
+      "learning_rate": 4.6433696018951915e-06,
+      "loss": 0.0578,
+      "step": 3010
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 8.441821768445967,
+      "learning_rate": 4.640486873060501e-06,
+      "loss": 0.0522,
+      "step": 3011
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 12.827070504932639,
+      "learning_rate": 4.6376042643458254e-06,
+      "loss": 0.0376,
+      "step": 3012
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 9.437577892174438,
+      "learning_rate": 4.634721776714305e-06,
+      "loss": 0.0455,
+      "step": 3013
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 9.188265760228681,
+      "learning_rate": 4.631839411129025e-06,
+      "loss": 0.061,
+      "step": 3014
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 9.140831955733349,
+      "learning_rate": 4.628957168553044e-06,
+      "loss": 0.0463,
+      "step": 3015
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 6.621373831056885,
+      "learning_rate": 4.6260750499493665e-06,
+      "loss": 0.0577,
+      "step": 3016
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 16.423887046208293,
+      "learning_rate": 4.623193056280968e-06,
+      "loss": 0.0422,
+      "step": 3017
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 8.94491278041153,
+      "learning_rate": 4.6203111885107735e-06,
+      "loss": 0.0443,
+      "step": 3018
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 10.518796763313492,
+      "learning_rate": 4.617429447601665e-06,
+      "loss": 0.0434,
+      "step": 3019
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.614949678699768,
+      "learning_rate": 4.614547834516492e-06,
+      "loss": 0.0467,
+      "step": 3020
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 6.229969137081133,
+      "learning_rate": 4.6116663502180495e-06,
+      "loss": 0.0798,
+      "step": 3021
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.066868988185442,
+      "learning_rate": 4.6087849956691e-06,
+      "loss": 0.0394,
+      "step": 3022
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 11.331501605212363,
+      "learning_rate": 4.605903771832353e-06,
+      "loss": 0.0513,
+      "step": 3023
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.839781229557823,
+      "learning_rate": 4.603022679670482e-06,
+      "loss": 0.041,
+      "step": 3024
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 14.332881398684489,
+      "learning_rate": 4.6001417201461114e-06,
+      "loss": 0.0539,
+      "step": 3025
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 16.291207017760183,
+      "learning_rate": 4.597260894221826e-06,
+      "loss": 0.0546,
+      "step": 3026
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 11.921642342100183,
+      "learning_rate": 4.594380202860162e-06,
+      "loss": 0.0542,
+      "step": 3027
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 13.932638931859428,
+      "learning_rate": 4.5914996470236094e-06,
+      "loss": 0.0531,
+      "step": 3028
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 6.7911004552180465,
+      "learning_rate": 4.588619227674619e-06,
+      "loss": 0.031,
+      "step": 3029
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 19.56600048222998,
+      "learning_rate": 4.58573894577559e-06,
+      "loss": 0.0653,
+      "step": 3030
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.7997903160804594,
+      "learning_rate": 4.5828588022888815e-06,
+      "loss": 0.0432,
+      "step": 3031
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.36456708146258,
+      "learning_rate": 4.5799787981767975e-06,
+      "loss": 0.0533,
+      "step": 3032
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 12.061053723608204,
+      "learning_rate": 4.577098934401607e-06,
+      "loss": 0.0577,
+      "step": 3033
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.047403317404933,
+      "learning_rate": 4.57421921192552e-06,
+      "loss": 0.043,
+      "step": 3034
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 6.587631210630952,
+      "learning_rate": 4.5713396317107115e-06,
+      "loss": 0.0443,
+      "step": 3035
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.787472009592416,
+      "learning_rate": 4.568460194719299e-06,
+      "loss": 0.0341,
+      "step": 3036
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 6.7437175989326965,
+      "learning_rate": 4.565580901913356e-06,
+      "loss": 0.0533,
+      "step": 3037
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.747331536456431,
+      "learning_rate": 4.562701754254909e-06,
+      "loss": 0.0473,
+      "step": 3038
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.768433537381566,
+      "learning_rate": 4.559822752705933e-06,
+      "loss": 0.0473,
+      "step": 3039
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.3083313751743146,
+      "learning_rate": 4.556943898228358e-06,
+      "loss": 0.0335,
+      "step": 3040
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 14.995712674161572,
+      "learning_rate": 4.55406519178406e-06,
+      "loss": 0.0591,
+      "step": 3041
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 12.571671708096863,
+      "learning_rate": 4.551186634334873e-06,
+      "loss": 0.0516,
+      "step": 3042
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.8361711880156655,
+      "learning_rate": 4.54830822684257e-06,
+      "loss": 0.0463,
+      "step": 3043
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 6.242235542738953,
+      "learning_rate": 4.545429970268888e-06,
+      "loss": 0.0573,
+      "step": 3044
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.9572838572416,
+      "learning_rate": 4.542551865575499e-06,
+      "loss": 0.0381,
+      "step": 3045
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 7.465321058449582,
+      "learning_rate": 4.539673913724037e-06,
+      "loss": 0.0504,
+      "step": 3046
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 13.582177654022352,
+      "learning_rate": 4.5367961156760745e-06,
+      "loss": 0.0642,
+      "step": 3047
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 13.106849831794028,
+      "learning_rate": 4.533918472393141e-06,
+      "loss": 0.0544,
+      "step": 3048
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 8.258854894812476,
+      "learning_rate": 4.531040984836708e-06,
+      "loss": 0.0398,
+      "step": 3049
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 6.1961201449818875,
+      "learning_rate": 4.5281636539682e-06,
+      "loss": 0.0563,
+      "step": 3050
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 14.114915646585992,
+      "learning_rate": 4.5252864807489836e-06,
+      "loss": 0.0525,
+      "step": 3051
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.184899979478378,
+      "learning_rate": 4.522409466140379e-06,
+      "loss": 0.0417,
+      "step": 3052
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 15.445103578110677,
+      "learning_rate": 4.5195326111036475e-06,
+      "loss": 0.0515,
+      "step": 3053
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.519785034152788,
+      "learning_rate": 4.5166559166000035e-06,
+      "loss": 0.0558,
+      "step": 3054
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 11.233134348940904,
+      "learning_rate": 4.513779383590599e-06,
+      "loss": 0.0532,
+      "step": 3055
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 6.353767355142566,
+      "learning_rate": 4.510903013036542e-06,
+      "loss": 0.0411,
+      "step": 3056
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.602659776333459,
+      "learning_rate": 4.508026805898878e-06,
+      "loss": 0.0474,
+      "step": 3057
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.55271578684028,
+      "learning_rate": 4.505150763138604e-06,
+      "loss": 0.0438,
+      "step": 3058
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.4020523865517065,
+      "learning_rate": 4.502274885716656e-06,
+      "loss": 0.0441,
+      "step": 3059
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.974542469049749,
+      "learning_rate": 4.499399174593923e-06,
+      "loss": 0.038,
+      "step": 3060
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 13.183234894928061,
+      "learning_rate": 4.496523630731229e-06,
+      "loss": 0.0536,
+      "step": 3061
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 7.012776400774284,
+      "learning_rate": 4.493648255089347e-06,
+      "loss": 0.053,
+      "step": 3062
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.113694599229363,
+      "learning_rate": 4.490773048628997e-06,
+      "loss": 0.0441,
+      "step": 3063
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 6.999872520790414,
+      "learning_rate": 4.487898012310834e-06,
+      "loss": 0.0551,
+      "step": 3064
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 8.988865058697366,
+      "learning_rate": 4.485023147095466e-06,
+      "loss": 0.0759,
+      "step": 3065
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 8.608399626100027,
+      "learning_rate": 4.482148453943434e-06,
+      "loss": 0.0555,
+      "step": 3066
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.657772658940335,
+      "learning_rate": 4.479273933815232e-06,
+      "loss": 0.0548,
+      "step": 3067
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.243732810688884,
+      "learning_rate": 4.476399587671285e-06,
+      "loss": 0.0759,
+      "step": 3068
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 17.83601854064182,
+      "learning_rate": 4.47352541647197e-06,
+      "loss": 0.0443,
+      "step": 3069
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 8.347841250675605,
+      "learning_rate": 4.470651421177599e-06,
+      "loss": 0.0452,
+      "step": 3070
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 7.0436872734018765,
+      "learning_rate": 4.467777602748425e-06,
+      "loss": 0.0431,
+      "step": 3071
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 24.860758320056135,
+      "learning_rate": 4.4649039621446495e-06,
+      "loss": 0.0797,
+      "step": 3072
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 11.719607525022566,
+      "learning_rate": 4.462030500326403e-06,
+      "loss": 0.0443,
+      "step": 3073
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 11.724827346428865,
+      "learning_rate": 4.459157218253769e-06,
+      "loss": 0.0717,
+      "step": 3074
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 13.934654419830443,
+      "learning_rate": 4.456284116886758e-06,
+      "loss": 0.0496,
+      "step": 3075
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 7.660664913105194,
+      "learning_rate": 4.453411197185334e-06,
+      "loss": 0.0442,
+      "step": 3076
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 24.60532296154118,
+      "learning_rate": 4.450538460109384e-06,
+      "loss": 0.0543,
+      "step": 3077
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 9.440853316535325,
+      "learning_rate": 4.447665906618751e-06,
+      "loss": 0.037,
+      "step": 3078
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.5094571457438564,
+      "learning_rate": 4.444793537673204e-06,
+      "loss": 0.0406,
+      "step": 3079
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 14.051552491510163,
+      "learning_rate": 4.441921354232455e-06,
+      "loss": 0.0598,
+      "step": 3080
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.601130148270938,
+      "learning_rate": 4.439049357256156e-06,
+      "loss": 0.0391,
+      "step": 3081
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 16.768683231807135,
+      "learning_rate": 4.436177547703891e-06,
+      "loss": 0.0461,
+      "step": 3082
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 14.287404647048588,
+      "learning_rate": 4.433305926535189e-06,
+      "loss": 0.0694,
+      "step": 3083
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 10.160614532165187,
+      "learning_rate": 4.430434494709509e-06,
+      "loss": 0.0536,
+      "step": 3084
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 15.080982562292157,
+      "learning_rate": 4.427563253186253e-06,
+      "loss": 0.0343,
+      "step": 3085
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.675753477001895,
+      "learning_rate": 4.424692202924754e-06,
+      "loss": 0.0356,
+      "step": 3086
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.716909777073017,
+      "learning_rate": 4.421821344884281e-06,
+      "loss": 0.0341,
+      "step": 3087
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.431358994048849,
+      "learning_rate": 4.418950680024046e-06,
+      "loss": 0.0457,
+      "step": 3088
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 6.596427500698859,
+      "learning_rate": 4.416080209303187e-06,
+      "loss": 0.0395,
+      "step": 3089
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 7.187035201109175,
+      "learning_rate": 4.413209933680786e-06,
+      "loss": 0.0556,
+      "step": 3090
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 15.759502614264681,
+      "learning_rate": 4.410339854115849e-06,
+      "loss": 0.0658,
+      "step": 3091
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 8.965449381966314,
+      "learning_rate": 4.407469971567331e-06,
+      "loss": 0.0542,
+      "step": 3092
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 13.279415208911773,
+      "learning_rate": 4.4046002869941055e-06,
+      "loss": 0.0518,
+      "step": 3093
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.9208827638453485,
+      "learning_rate": 4.401730801354994e-06,
+      "loss": 0.0453,
+      "step": 3094
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.575198553252729,
+      "learning_rate": 4.39886151560874e-06,
+      "loss": 0.0413,
+      "step": 3095
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.725306026650709,
+      "learning_rate": 4.395992430714028e-06,
+      "loss": 0.0352,
+      "step": 3096
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 10.572986077833214,
+      "learning_rate": 4.393123547629472e-06,
+      "loss": 0.0573,
+      "step": 3097
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.788737593661607,
+      "learning_rate": 4.390254867313619e-06,
+      "loss": 0.049,
+      "step": 3098
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 14.988650115590175,
+      "learning_rate": 4.387386390724947e-06,
+      "loss": 0.041,
+      "step": 3099
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 15.407848641378886,
+      "learning_rate": 4.38451811882187e-06,
+      "loss": 0.0631,
+      "step": 3100
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 10.332737742350625,
+      "learning_rate": 4.3816500525627284e-06,
+      "loss": 0.0542,
+      "step": 3101
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.3046432737592766,
+      "learning_rate": 4.3787821929057985e-06,
+      "loss": 0.0524,
+      "step": 3102
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 21.62698520171683,
+      "learning_rate": 4.3759145408092855e-06,
+      "loss": 0.0541,
+      "step": 3103
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 6.318176228409699,
+      "learning_rate": 4.373047097231324e-06,
+      "loss": 0.047,
+      "step": 3104
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.161442887353655,
+      "learning_rate": 4.370179863129979e-06,
+      "loss": 0.0438,
+      "step": 3105
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 7.9377251511338205,
+      "learning_rate": 4.367312839463251e-06,
+      "loss": 0.0496,
+      "step": 3106
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.3131144070684595,
+      "learning_rate": 4.3644460271890614e-06,
+      "loss": 0.0415,
+      "step": 3107
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 8.363837704096479,
+      "learning_rate": 4.361579427265268e-06,
+      "loss": 0.0644,
+      "step": 3108
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 8.313467422358402,
+      "learning_rate": 4.358713040649654e-06,
+      "loss": 0.0499,
+      "step": 3109
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.1434368387419873,
+      "learning_rate": 4.3558468682999336e-06,
+      "loss": 0.0358,
+      "step": 3110
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.111253029616996,
+      "learning_rate": 4.352980911173747e-06,
+      "loss": 0.0492,
+      "step": 3111
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 8.357180537855855,
+      "learning_rate": 4.350115170228664e-06,
+      "loss": 0.0505,
+      "step": 3112
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 7.2716322758839755,
+      "learning_rate": 4.3472496464221845e-06,
+      "loss": 0.0474,
+      "step": 3113
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 7.2404915807490315,
+      "learning_rate": 4.344384340711728e-06,
+      "loss": 0.0434,
+      "step": 3114
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.148940982244169,
+      "learning_rate": 4.341519254054651e-06,
+      "loss": 0.0416,
+      "step": 3115
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.229546559770345,
+      "learning_rate": 4.338654387408229e-06,
+      "loss": 0.0396,
+      "step": 3116
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 20.48430154530679,
+      "learning_rate": 4.335789741729671e-06,
+      "loss": 0.0515,
+      "step": 3117
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 10.359553267941436,
+      "learning_rate": 4.332925317976104e-06,
+      "loss": 0.0451,
+      "step": 3118
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 12.04521250763658,
+      "learning_rate": 4.330061117104589e-06,
+      "loss": 0.0395,
+      "step": 3119
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 9.16643051448259,
+      "learning_rate": 4.327197140072108e-06,
+      "loss": 0.0539,
+      "step": 3120
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 11.188804325737475,
+      "learning_rate": 4.324333387835565e-06,
+      "loss": 0.048,
+      "step": 3121
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 10.924204030628253,
+      "learning_rate": 4.321469861351799e-06,
+      "loss": 0.0456,
+      "step": 3122
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 12.396543885172143,
+      "learning_rate": 4.318606561577562e-06,
+      "loss": 0.0711,
+      "step": 3123
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 7.358079264086069,
+      "learning_rate": 4.31574348946954e-06,
+      "loss": 0.0432,
+      "step": 3124
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 6.63910667690175,
+      "learning_rate": 4.312880645984334e-06,
+      "loss": 0.0473,
+      "step": 3125
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 6.926846866544162,
+      "learning_rate": 4.310018032078479e-06,
+      "loss": 0.0427,
+      "step": 3126
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.348927273635156,
+      "learning_rate": 4.307155648708421e-06,
+      "loss": 0.0398,
+      "step": 3127
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 8.293095802027796,
+      "learning_rate": 4.304293496830542e-06,
+      "loss": 0.0552,
+      "step": 3128
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.164416626873579,
+      "learning_rate": 4.301431577401136e-06,
+      "loss": 0.0558,
+      "step": 3129
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 10.31913313745745,
+      "learning_rate": 4.298569891376423e-06,
+      "loss": 0.0557,
+      "step": 3130
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 11.861633104275379,
+      "learning_rate": 4.2957084397125496e-06,
+      "loss": 0.0438,
+      "step": 3131
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 11.33706132619583,
+      "learning_rate": 4.292847223365574e-06,
+      "loss": 0.0434,
+      "step": 3132
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 15.917688181318276,
+      "learning_rate": 4.289986243291488e-06,
+      "loss": 0.0623,
+      "step": 3133
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.856782780631129,
+      "learning_rate": 4.287125500446193e-06,
+      "loss": 0.0336,
+      "step": 3134
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 16.31108504711644,
+      "learning_rate": 4.284264995785521e-06,
+      "loss": 0.0607,
+      "step": 3135
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 9.762647669216687,
+      "learning_rate": 4.2814047302652155e-06,
+      "loss": 0.0361,
+      "step": 3136
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 8.749071414210514,
+      "learning_rate": 4.278544704840948e-06,
+      "loss": 0.0448,
+      "step": 3137
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 6.27371687383982,
+      "learning_rate": 4.275684920468306e-06,
+      "loss": 0.0473,
+      "step": 3138
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.534335338717934,
+      "learning_rate": 4.272825378102791e-06,
+      "loss": 0.0475,
+      "step": 3139
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 6.9101311343506655,
+      "learning_rate": 4.269966078699836e-06,
+      "loss": 0.0396,
+      "step": 3140
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.184600894702316,
+      "learning_rate": 4.267107023214782e-06,
+      "loss": 0.0396,
+      "step": 3141
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 10.089000430219308,
+      "learning_rate": 4.264248212602896e-06,
+      "loss": 0.0527,
+      "step": 3142
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 6.60955809699945,
+      "learning_rate": 4.261389647819355e-06,
+      "loss": 0.0446,
+      "step": 3143
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 13.226538812036294,
+      "learning_rate": 4.258531329819264e-06,
+      "loss": 0.0481,
+      "step": 3144
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 10.865699000414827,
+      "learning_rate": 4.255673259557636e-06,
+      "loss": 0.0531,
+      "step": 3145
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 14.364449970953885,
+      "learning_rate": 4.252815437989408e-06,
+      "loss": 0.055,
+      "step": 3146
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 9.573885380724388,
+      "learning_rate": 4.24995786606943e-06,
+      "loss": 0.0589,
+      "step": 3147
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 20.58280975888398,
+      "learning_rate": 4.24710054475247e-06,
+      "loss": 0.0551,
+      "step": 3148
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.894893628781143,
+      "learning_rate": 4.244243474993214e-06,
+      "loss": 0.0312,
+      "step": 3149
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 10.294083577388148,
+      "learning_rate": 4.241386657746257e-06,
+      "loss": 0.0512,
+      "step": 3150
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 9.884297976636462,
+      "learning_rate": 4.2385300939661215e-06,
+      "loss": 0.0698,
+      "step": 3151
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 20.580492702452048,
+      "learning_rate": 4.2356737846072326e-06,
+      "loss": 0.0423,
+      "step": 3152
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 10.3250929805605,
+      "learning_rate": 4.232817730623941e-06,
+      "loss": 0.0622,
+      "step": 3153
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 7.020014929183152,
+      "learning_rate": 4.229961932970505e-06,
+      "loss": 0.029,
+      "step": 3154
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 7.624072768428424,
+      "learning_rate": 4.2271063926010995e-06,
+      "loss": 0.0357,
+      "step": 3155
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 10.226850797566772,
+      "learning_rate": 4.224251110469814e-06,
+      "loss": 0.0622,
+      "step": 3156
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 6.615838261687149,
+      "learning_rate": 4.221396087530652e-06,
+      "loss": 0.056,
+      "step": 3157
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 8.170230954284976,
+      "learning_rate": 4.218541324737529e-06,
+      "loss": 0.0552,
+      "step": 3158
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.928036811380873,
+      "learning_rate": 4.2156868230442756e-06,
+      "loss": 0.0472,
+      "step": 3159
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 15.829254203511692,
+      "learning_rate": 4.212832583404632e-06,
+      "loss": 0.0587,
+      "step": 3160
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 11.618419552979391,
+      "learning_rate": 4.2099786067722535e-06,
+      "loss": 0.0403,
+      "step": 3161
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.1222508462469003,
+      "learning_rate": 4.207124894100707e-06,
+      "loss": 0.0364,
+      "step": 3162
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.318748307419997,
+      "learning_rate": 4.2042714463434715e-06,
+      "loss": 0.0391,
+      "step": 3163
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 9.36228620240982,
+      "learning_rate": 4.201418264453935e-06,
+      "loss": 0.0319,
+      "step": 3164
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.3778091133940125,
+      "learning_rate": 4.198565349385402e-06,
+      "loss": 0.0536,
+      "step": 3165
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 17.14676052388145,
+      "learning_rate": 4.195712702091079e-06,
+      "loss": 0.0535,
+      "step": 3166
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.024792221872647,
+      "learning_rate": 4.192860323524094e-06,
+      "loss": 0.0481,
+      "step": 3167
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 6.294621583070762,
+      "learning_rate": 4.190008214637476e-06,
+      "loss": 0.0591,
+      "step": 3168
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 7.517307671125681,
+      "learning_rate": 4.187156376384171e-06,
+      "loss": 0.0526,
+      "step": 3169
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 8.592352297300314,
+      "learning_rate": 4.184304809717027e-06,
+      "loss": 0.0632,
+      "step": 3170
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 14.526658581579898,
+      "learning_rate": 4.18145351558881e-06,
+      "loss": 0.0461,
+      "step": 3171
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 16.837649318976588,
+      "learning_rate": 4.178602494952187e-06,
+      "loss": 0.0806,
+      "step": 3172
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 7.440201035737965,
+      "learning_rate": 4.175751748759737e-06,
+      "loss": 0.0474,
+      "step": 3173
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.692217557688077,
+      "learning_rate": 4.1729012779639495e-06,
+      "loss": 0.0359,
+      "step": 3174
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.1727632650733915,
+      "learning_rate": 4.170051083517217e-06,
+      "loss": 0.0648,
+      "step": 3175
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 8.946075258402107,
+      "learning_rate": 4.167201166371846e-06,
+      "loss": 0.0524,
+      "step": 3176
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 7.667399977486221,
+      "learning_rate": 4.164351527480042e-06,
+      "loss": 0.049,
+      "step": 3177
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.274714349541263,
+      "learning_rate": 4.161502167793928e-06,
+      "loss": 0.0706,
+      "step": 3178
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 10.214076104747047,
+      "learning_rate": 4.1586530882655226e-06,
+      "loss": 0.0504,
+      "step": 3179
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 7.550207984999468,
+      "learning_rate": 4.155804289846762e-06,
+      "loss": 0.0385,
+      "step": 3180
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 6.718996738912058,
+      "learning_rate": 4.152955773489479e-06,
+      "loss": 0.0577,
+      "step": 3181
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.097197065235624,
+      "learning_rate": 4.150107540145413e-06,
+      "loss": 0.0457,
+      "step": 3182
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 8.064906232599887,
+      "learning_rate": 4.147259590766219e-06,
+      "loss": 0.0574,
+      "step": 3183
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 6.518517725711126,
+      "learning_rate": 4.144411926303442e-06,
+      "loss": 0.0383,
+      "step": 3184
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.616507539227108,
+      "learning_rate": 4.141564547708546e-06,
+      "loss": 0.0521,
+      "step": 3185
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 6.556983372205476,
+      "learning_rate": 4.138717455932888e-06,
+      "loss": 0.0537,
+      "step": 3186
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.742026540119468,
+      "learning_rate": 4.13587065192774e-06,
+      "loss": 0.0479,
+      "step": 3187
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 16.34355762704361,
+      "learning_rate": 4.133024136644269e-06,
+      "loss": 0.0531,
+      "step": 3188
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 7.81542554393562,
+      "learning_rate": 4.130177911033546e-06,
+      "loss": 0.04,
+      "step": 3189
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 13.76182597376324,
+      "learning_rate": 4.127331976046553e-06,
+      "loss": 0.0431,
+      "step": 3190
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 6.879252527427723,
+      "learning_rate": 4.124486332634165e-06,
+      "loss": 0.067,
+      "step": 3191
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 9.10827224827748,
+      "learning_rate": 4.121640981747169e-06,
+      "loss": 0.0434,
+      "step": 3192
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 6.435745967654931,
+      "learning_rate": 4.118795924336245e-06,
+      "loss": 0.0596,
+      "step": 3193
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 6.612178222538567,
+      "learning_rate": 4.115951161351985e-06,
+      "loss": 0.0446,
+      "step": 3194
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 14.164355450798242,
+      "learning_rate": 4.113106693744871e-06,
+      "loss": 0.0479,
+      "step": 3195
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 7.157081086345673,
+      "learning_rate": 4.110262522465298e-06,
+      "loss": 0.0531,
+      "step": 3196
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 7.031758826153102,
+      "learning_rate": 4.107418648463553e-06,
+      "loss": 0.0307,
+      "step": 3197
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 13.865702651660351,
+      "learning_rate": 4.104575072689827e-06,
+      "loss": 0.0684,
+      "step": 3198
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.3223935942971883,
+      "learning_rate": 4.101731796094215e-06,
+      "loss": 0.0368,
+      "step": 3199
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 9.926947145875099,
+      "learning_rate": 4.098888819626704e-06,
+      "loss": 0.0475,
+      "step": 3200
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 13.764679326338504,
+      "learning_rate": 4.096046144237189e-06,
+      "loss": 0.0667,
+      "step": 3201
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.877567877322836,
+      "learning_rate": 4.093203770875458e-06,
+      "loss": 0.0318,
+      "step": 3202
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.0890455248333235,
+      "learning_rate": 4.090361700491203e-06,
+      "loss": 0.0391,
+      "step": 3203
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 11.307502997946623,
+      "learning_rate": 4.087519934034011e-06,
+      "loss": 0.0579,
+      "step": 3204
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 12.286021346963802,
+      "learning_rate": 4.084678472453371e-06,
+      "loss": 0.0425,
+      "step": 3205
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 20.05845646784748,
+      "learning_rate": 4.081837316698665e-06,
+      "loss": 0.0714,
+      "step": 3206
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 7.513715346915439,
+      "learning_rate": 4.078996467719179e-06,
+      "loss": 0.0566,
+      "step": 3207
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.871357352532569,
+      "learning_rate": 4.076155926464091e-06,
+      "loss": 0.0322,
+      "step": 3208
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 8.692473548212876,
+      "learning_rate": 4.07331569388248e-06,
+      "loss": 0.0452,
+      "step": 3209
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.348891593350171,
+      "learning_rate": 4.07047577092332e-06,
+      "loss": 0.0622,
+      "step": 3210
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 10.667994143010105,
+      "learning_rate": 4.067636158535483e-06,
+      "loss": 0.0429,
+      "step": 3211
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 14.33390498257431,
+      "learning_rate": 4.064796857667734e-06,
+      "loss": 0.0602,
+      "step": 3212
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 13.853874395384349,
+      "learning_rate": 4.0619578692687405e-06,
+      "loss": 0.0485,
+      "step": 3213
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 10.967333214518202,
+      "learning_rate": 4.059119194287056e-06,
+      "loss": 0.0618,
+      "step": 3214
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 19.491206872437058,
+      "learning_rate": 4.056280833671139e-06,
+      "loss": 0.0689,
+      "step": 3215
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 15.545263997390558,
+      "learning_rate": 4.053442788369334e-06,
+      "loss": 0.0463,
+      "step": 3216
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 8.404901220400674,
+      "learning_rate": 4.05060505932989e-06,
+      "loss": 0.049,
+      "step": 3217
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 10.955981738026576,
+      "learning_rate": 4.04776764750094e-06,
+      "loss": 0.0467,
+      "step": 3218
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 9.242743867660305,
+      "learning_rate": 4.04493055383052e-06,
+      "loss": 0.0492,
+      "step": 3219
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 18.269012387060222,
+      "learning_rate": 4.042093779266553e-06,
+      "loss": 0.0473,
+      "step": 3220
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.530310469645218,
+      "learning_rate": 4.0392573247568614e-06,
+      "loss": 0.0731,
+      "step": 3221
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 10.337877336593875,
+      "learning_rate": 4.036421191249155e-06,
+      "loss": 0.0583,
+      "step": 3222
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 11.665285153590094,
+      "learning_rate": 4.033585379691036e-06,
+      "loss": 0.101,
+      "step": 3223
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 10.20335646310635,
+      "learning_rate": 4.030749891030008e-06,
+      "loss": 0.0428,
+      "step": 3224
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 14.145496210000346,
+      "learning_rate": 4.0279147262134534e-06,
+      "loss": 0.0629,
+      "step": 3225
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 10.358245511220993,
+      "learning_rate": 4.025079886188661e-06,
+      "loss": 0.0611,
+      "step": 3226
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 11.004038035963342,
+      "learning_rate": 4.022245371902796e-06,
+      "loss": 0.0431,
+      "step": 3227
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 13.706420262135827,
+      "learning_rate": 4.01941118430293e-06,
+      "loss": 0.0503,
+      "step": 3228
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.96510652515189,
+      "learning_rate": 4.0165773243360105e-06,
+      "loss": 0.0313,
+      "step": 3229
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.516895579432337,
+      "learning_rate": 4.0137437929488885e-06,
+      "loss": 0.0425,
+      "step": 3230
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 6.163717649417404,
+      "learning_rate": 4.010910591088296e-06,
+      "loss": 0.0454,
+      "step": 3231
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 10.726656037501574,
+      "learning_rate": 4.008077719700859e-06,
+      "loss": 0.0592,
+      "step": 3232
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 13.5104992520412,
+      "learning_rate": 4.005245179733095e-06,
+      "loss": 0.045,
+      "step": 3233
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 7.879892748265249,
+      "learning_rate": 4.002412972131403e-06,
+      "loss": 0.0378,
+      "step": 3234
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 7.298147921669261,
+      "learning_rate": 3.999581097842082e-06,
+      "loss": 0.0468,
+      "step": 3235
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 8.052193186678828,
+      "learning_rate": 3.99674955781131e-06,
+      "loss": 0.0628,
+      "step": 3236
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.805780871133186,
+      "learning_rate": 3.99391835298516e-06,
+      "loss": 0.0336,
+      "step": 3237
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 11.265202041373872,
+      "learning_rate": 3.991087484309586e-06,
+      "loss": 0.0347,
+      "step": 3238
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 9.214043476468115,
+      "learning_rate": 3.988256952730439e-06,
+      "loss": 0.0771,
+      "step": 3239
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.8998163572166895,
+      "learning_rate": 3.985426759193449e-06,
+      "loss": 0.0495,
+      "step": 3240
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 10.144981153013562,
+      "learning_rate": 3.982596904644236e-06,
+      "loss": 0.0382,
+      "step": 3241
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.985568117991297,
+      "learning_rate": 3.979767390028309e-06,
+      "loss": 0.0441,
+      "step": 3242
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 17.97011714211926,
+      "learning_rate": 3.976938216291059e-06,
+      "loss": 0.0776,
+      "step": 3243
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.31899818255434,
+      "learning_rate": 3.974109384377768e-06,
+      "loss": 0.0443,
+      "step": 3244
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 10.545724201687834,
+      "learning_rate": 3.971280895233599e-06,
+      "loss": 0.0442,
+      "step": 3245
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.238792598733231,
+      "learning_rate": 3.968452749803605e-06,
+      "loss": 0.0577,
+      "step": 3246
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.7875985593995583,
+      "learning_rate": 3.965624949032723e-06,
+      "loss": 0.0409,
+      "step": 3247
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.594499938098831,
+      "learning_rate": 3.962797493865767e-06,
+      "loss": 0.0485,
+      "step": 3248
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.310912106765515,
+      "learning_rate": 3.959970385247451e-06,
+      "loss": 0.0536,
+      "step": 3249
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 8.359734666912926,
+      "learning_rate": 3.957143624122359e-06,
+      "loss": 0.0649,
+      "step": 3250
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.073298004365436,
+      "learning_rate": 3.954317211434966e-06,
+      "loss": 0.0496,
+      "step": 3251
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 6.362507284188202,
+      "learning_rate": 3.951491148129628e-06,
+      "loss": 0.0522,
+      "step": 3252
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.991116571447913,
+      "learning_rate": 3.948665435150589e-06,
+      "loss": 0.0389,
+      "step": 3253
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 10.386760657877408,
+      "learning_rate": 3.945840073441967e-06,
+      "loss": 0.0495,
+      "step": 3254
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.3677506790255665,
+      "learning_rate": 3.943015063947773e-06,
+      "loss": 0.0417,
+      "step": 3255
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 7.163430214164329,
+      "learning_rate": 3.940190407611891e-06,
+      "loss": 0.0495,
+      "step": 3256
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 17.871089845054883,
+      "learning_rate": 3.937366105378093e-06,
+      "loss": 0.0321,
+      "step": 3257
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 7.221258871476373,
+      "learning_rate": 3.93454215819003e-06,
+      "loss": 0.0586,
+      "step": 3258
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 8.203863757302269,
+      "learning_rate": 3.931718566991236e-06,
+      "loss": 0.0588,
+      "step": 3259
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 6.2318167707341505,
+      "learning_rate": 3.9288953327251265e-06,
+      "loss": 0.0539,
+      "step": 3260
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 8.561571820835148,
+      "learning_rate": 3.9260724563349935e-06,
+      "loss": 0.045,
+      "step": 3261
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 10.489240447227571,
+      "learning_rate": 3.923249938764016e-06,
+      "loss": 0.0543,
+      "step": 3262
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.63338501785195,
+      "learning_rate": 3.920427780955247e-06,
+      "loss": 0.0695,
+      "step": 3263
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 7.206481099887481,
+      "learning_rate": 3.917605983851622e-06,
+      "loss": 0.0625,
+      "step": 3264
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 12.970256281406726,
+      "learning_rate": 3.914784548395959e-06,
+      "loss": 0.048,
+      "step": 3265
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 16.605000076839822,
+      "learning_rate": 3.911963475530948e-06,
+      "loss": 0.0655,
+      "step": 3266
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 22.570657256087852,
+      "learning_rate": 3.909142766199163e-06,
+      "loss": 0.0626,
+      "step": 3267
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 7.757888466221637,
+      "learning_rate": 3.906322421343055e-06,
+      "loss": 0.0389,
+      "step": 3268
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 18.353788836285396,
+      "learning_rate": 3.903502441904956e-06,
+      "loss": 0.0423,
+      "step": 3269
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 18.335337367703037,
+      "learning_rate": 3.900682828827072e-06,
+      "loss": 0.041,
+      "step": 3270
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 20.560967998792574,
+      "learning_rate": 3.897863583051488e-06,
+      "loss": 0.0438,
+      "step": 3271
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 9.189481542895964,
+      "learning_rate": 3.895044705520167e-06,
+      "loss": 0.0408,
+      "step": 3272
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 6.240925968107035,
+      "learning_rate": 3.892226197174947e-06,
+      "loss": 0.0535,
+      "step": 3273
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 11.330018925248265,
+      "learning_rate": 3.889408058957547e-06,
+      "loss": 0.0564,
+      "step": 3274
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 22.462180664383062,
+      "learning_rate": 3.886590291809554e-06,
+      "loss": 0.0628,
+      "step": 3275
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 25.572962458851126,
+      "learning_rate": 3.883772896672443e-06,
+      "loss": 0.0505,
+      "step": 3276
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 16.781204089206188,
+      "learning_rate": 3.8809558744875534e-06,
+      "loss": 0.0385,
+      "step": 3277
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.374198239697441,
+      "learning_rate": 3.878139226196107e-06,
+      "loss": 0.0457,
+      "step": 3278
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 14.5149616122367,
+      "learning_rate": 3.875322952739196e-06,
+      "loss": 0.0605,
+      "step": 3279
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 13.155372455067319,
+      "learning_rate": 3.872507055057793e-06,
+      "loss": 0.0413,
+      "step": 3280
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 14.517540894737511,
+      "learning_rate": 3.8696915340927395e-06,
+      "loss": 0.0463,
+      "step": 3281
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 12.853273973170033,
+      "learning_rate": 3.866876390784752e-06,
+      "loss": 0.0489,
+      "step": 3282
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 16.08116245026929,
+      "learning_rate": 3.8640616260744266e-06,
+      "loss": 0.0584,
+      "step": 3283
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 13.158844571849507,
+      "learning_rate": 3.861247240902223e-06,
+      "loss": 0.0468,
+      "step": 3284
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 11.605770161229955,
+      "learning_rate": 3.858433236208485e-06,
+      "loss": 0.0745,
+      "step": 3285
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 14.696854573310642,
+      "learning_rate": 3.85561961293342e-06,
+      "loss": 0.0485,
+      "step": 3286
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 8.808771550272338,
+      "learning_rate": 3.852806372017115e-06,
+      "loss": 0.0525,
+      "step": 3287
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 12.28812481591032,
+      "learning_rate": 3.849993514399521e-06,
+      "loss": 0.0457,
+      "step": 3288
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 7.973779667671355,
+      "learning_rate": 3.847181041020472e-06,
+      "loss": 0.0442,
+      "step": 3289
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 19.096436336706248,
+      "learning_rate": 3.844368952819666e-06,
+      "loss": 0.0495,
+      "step": 3290
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.254182022740973,
+      "learning_rate": 3.84155725073667e-06,
+      "loss": 0.0621,
+      "step": 3291
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 8.2181153756375,
+      "learning_rate": 3.838745935710931e-06,
+      "loss": 0.0423,
+      "step": 3292
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 6.376471586863244,
+      "learning_rate": 3.835935008681757e-06,
+      "loss": 0.0704,
+      "step": 3293
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 12.63158037914216,
+      "learning_rate": 3.833124470588336e-06,
+      "loss": 0.0662,
+      "step": 3294
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 13.3387733700678,
+      "learning_rate": 3.830314322369717e-06,
+      "loss": 0.0459,
+      "step": 3295
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.133497815442831,
+      "learning_rate": 3.827504564964825e-06,
+      "loss": 0.038,
+      "step": 3296
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 8.710809474582618,
+      "learning_rate": 3.82469519931245e-06,
+      "loss": 0.0739,
+      "step": 3297
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 7.0462004661285835,
+      "learning_rate": 3.8218862263512565e-06,
+      "loss": 0.0383,
+      "step": 3298
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.6385692219029595,
+      "learning_rate": 3.819077647019772e-06,
+      "loss": 0.0709,
+      "step": 3299
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 24.502025525734307,
+      "learning_rate": 3.816269462256394e-06,
+      "loss": 0.0513,
+      "step": 3300
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 13.864706482022296,
+      "learning_rate": 3.813461672999394e-06,
+      "loss": 0.0684,
+      "step": 3301
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 6.156507954757023,
+      "learning_rate": 3.8106542801869007e-06,
+      "loss": 0.0494,
+      "step": 3302
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 6.807759895501928,
+      "learning_rate": 3.8078472847569215e-06,
+      "loss": 0.0603,
+      "step": 3303
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 7.8161472508869725,
+      "learning_rate": 3.805040687647321e-06,
+      "loss": 0.0661,
+      "step": 3304
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.692379712993815,
+      "learning_rate": 3.8022344897958402e-06,
+      "loss": 0.0519,
+      "step": 3305
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 13.215818470193987,
+      "learning_rate": 3.799428692140077e-06,
+      "loss": 0.0547,
+      "step": 3306
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 13.05567981160067,
+      "learning_rate": 3.7966232956175053e-06,
+      "loss": 0.0408,
+      "step": 3307
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 11.836967320563483,
+      "learning_rate": 3.793818301165457e-06,
+      "loss": 0.0607,
+      "step": 3308
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 24.41802793152467,
+      "learning_rate": 3.7910137097211345e-06,
+      "loss": 0.0675,
+      "step": 3309
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 13.008738936268362,
+      "learning_rate": 3.788209522221604e-06,
+      "loss": 0.0508,
+      "step": 3310
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 8.453167608671531,
+      "learning_rate": 3.7854057396037934e-06,
+      "loss": 0.0465,
+      "step": 3311
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 7.942453598993944,
+      "learning_rate": 3.7826023628045037e-06,
+      "loss": 0.0429,
+      "step": 3312
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 7.2021837650240546,
+      "learning_rate": 3.779799392760391e-06,
+      "loss": 0.0595,
+      "step": 3313
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 8.110336108153568,
+      "learning_rate": 3.7769968304079833e-06,
+      "loss": 0.0392,
+      "step": 3314
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.770513910427804,
+      "learning_rate": 3.7741946766836657e-06,
+      "loss": 0.0422,
+      "step": 3315
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.485390555701271,
+      "learning_rate": 3.771392932523691e-06,
+      "loss": 0.0442,
+      "step": 3316
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.776761654570514,
+      "learning_rate": 3.768591598864174e-06,
+      "loss": 0.0474,
+      "step": 3317
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 8.255921845791628,
+      "learning_rate": 3.765790676641092e-06,
+      "loss": 0.0593,
+      "step": 3318
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 7.84516772890703,
+      "learning_rate": 3.762990166790286e-06,
+      "loss": 0.0421,
+      "step": 3319
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 21.7601097912396,
+      "learning_rate": 3.760190070247458e-06,
+      "loss": 0.0898,
+      "step": 3320
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 8.031470132988193,
+      "learning_rate": 3.7573903879481714e-06,
+      "loss": 0.0466,
+      "step": 3321
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 11.109101536744378,
+      "learning_rate": 3.754591120827854e-06,
+      "loss": 0.0467,
+      "step": 3322
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 17.1446165094642,
+      "learning_rate": 3.7517922698217914e-06,
+      "loss": 0.0461,
+      "step": 3323
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 14.65645178329707,
+      "learning_rate": 3.7489938358651334e-06,
+      "loss": 0.0511,
+      "step": 3324
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 6.9909618648344605,
+      "learning_rate": 3.746195819892885e-06,
+      "loss": 0.0581,
+      "step": 3325
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 18.094397161249283,
+      "learning_rate": 3.7433982228399205e-06,
+      "loss": 0.0445,
+      "step": 3326
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 12.802062963646232,
+      "learning_rate": 3.7406010456409648e-06,
+      "loss": 0.05,
+      "step": 3327
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 16.396058272502454,
+      "learning_rate": 3.73780428923061e-06,
+      "loss": 0.0441,
+      "step": 3328
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 10.919646375990595,
+      "learning_rate": 3.7350079545433014e-06,
+      "loss": 0.0501,
+      "step": 3329
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.922454337057674,
+      "learning_rate": 3.7322120425133497e-06,
+      "loss": 0.0319,
+      "step": 3330
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 11.453873310916373,
+      "learning_rate": 3.729416554074917e-06,
+      "loss": 0.0383,
+      "step": 3331
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 13.08051920997564,
+      "learning_rate": 3.726621490162033e-06,
+      "loss": 0.0369,
+      "step": 3332
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 10.658265780826953,
+      "learning_rate": 3.7238268517085773e-06,
+      "loss": 0.0862,
+      "step": 3333
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 15.404614991290057,
+      "learning_rate": 3.7210326396482893e-06,
+      "loss": 0.0499,
+      "step": 3334
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 14.75043279711972,
+      "learning_rate": 3.718238854914771e-06,
+      "loss": 0.0703,
+      "step": 3335
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 15.503977295964852,
+      "learning_rate": 3.7154454984414733e-06,
+      "loss": 0.0573,
+      "step": 3336
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.946957800905822,
+      "learning_rate": 3.7126525711617135e-06,
+      "loss": 0.0673,
+      "step": 3337
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.259494768027386,
+      "learning_rate": 3.7098600740086555e-06,
+      "loss": 0.0547,
+      "step": 3338
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 11.546595002331136,
+      "learning_rate": 3.707068007915329e-06,
+      "loss": 0.056,
+      "step": 3339
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 17.967131882842132,
+      "learning_rate": 3.704276373814611e-06,
+      "loss": 0.0409,
+      "step": 3340
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 15.393071791424758,
+      "learning_rate": 3.7014851726392427e-06,
+      "loss": 0.0546,
+      "step": 3341
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 7.001550477108224,
+      "learning_rate": 3.6986944053218143e-06,
+      "loss": 0.0545,
+      "step": 3342
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.9519019922056784,
+      "learning_rate": 3.69590407279477e-06,
+      "loss": 0.043,
+      "step": 3343
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 8.960410735070326,
+      "learning_rate": 3.6931141759904175e-06,
+      "loss": 0.0607,
+      "step": 3344
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 14.847248534040741,
+      "learning_rate": 3.6903247158409077e-06,
+      "loss": 0.0394,
+      "step": 3345
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 17.388659127488623,
+      "learning_rate": 3.687535693278256e-06,
+      "loss": 0.0561,
+      "step": 3346
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 6.378086436756875,
+      "learning_rate": 3.6847471092343225e-06,
+      "loss": 0.049,
+      "step": 3347
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.734911486197315,
+      "learning_rate": 3.681958964640828e-06,
+      "loss": 0.052,
+      "step": 3348
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.652535209632719,
+      "learning_rate": 3.679171260429343e-06,
+      "loss": 0.0573,
+      "step": 3349
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 18.331070603639798,
+      "learning_rate": 3.676383997531288e-06,
+      "loss": 0.0636,
+      "step": 3350
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 10.55381413312245,
+      "learning_rate": 3.673597176877944e-06,
+      "loss": 0.0585,
+      "step": 3351
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 8.280879713430096,
+      "learning_rate": 3.670810799400435e-06,
+      "loss": 0.0447,
+      "step": 3352
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 6.519631412173078,
+      "learning_rate": 3.668024866029747e-06,
+      "loss": 0.0404,
+      "step": 3353
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 11.576104508421272,
+      "learning_rate": 3.665239377696706e-06,
+      "loss": 0.0468,
+      "step": 3354
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.760444512858353,
+      "learning_rate": 3.6624543353320006e-06,
+      "loss": 0.036,
+      "step": 3355
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.6760439313381905,
+      "learning_rate": 3.659669739866162e-06,
+      "loss": 0.0276,
+      "step": 3356
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.543274966456261,
+      "learning_rate": 3.6568855922295776e-06,
+      "loss": 0.0482,
+      "step": 3357
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.3086205701499,
+      "learning_rate": 3.654101893352482e-06,
+      "loss": 0.0451,
+      "step": 3358
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 13.06056766375821,
+      "learning_rate": 3.651318644164958e-06,
+      "loss": 0.0627,
+      "step": 3359
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 8.710508596248417,
+      "learning_rate": 3.6485358455969454e-06,
+      "loss": 0.043,
+      "step": 3360
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.694002302786408,
+      "learning_rate": 3.645753498578225e-06,
+      "loss": 0.0331,
+      "step": 3361
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 8.42561338264928,
+      "learning_rate": 3.6429716040384346e-06,
+      "loss": 0.034,
+      "step": 3362
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 6.1770297556861475,
+      "learning_rate": 3.6401901629070524e-06,
+      "loss": 0.0625,
+      "step": 3363
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.919884188773968,
+      "learning_rate": 3.6374091761134147e-06,
+      "loss": 0.0616,
+      "step": 3364
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.43484904633554,
+      "learning_rate": 3.6346286445866953e-06,
+      "loss": 0.0339,
+      "step": 3365
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.2904973516040323,
+      "learning_rate": 3.6318485692559263e-06,
+      "loss": 0.0532,
+      "step": 3366
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.091491571818965,
+      "learning_rate": 3.62906895104998e-06,
+      "loss": 0.0418,
+      "step": 3367
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 7.015427483416067,
+      "learning_rate": 3.6262897908975787e-06,
+      "loss": 0.0275,
+      "step": 3368
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.8745109759948946,
+      "learning_rate": 3.6235110897272917e-06,
+      "loss": 0.0345,
+      "step": 3369
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.67694078014422,
+      "learning_rate": 3.620732848467535e-06,
+      "loss": 0.041,
+      "step": 3370
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 7.963957573360893,
+      "learning_rate": 3.6179550680465703e-06,
+      "loss": 0.042,
+      "step": 3371
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 13.579331192890951,
+      "learning_rate": 3.615177749392506e-06,
+      "loss": 0.0472,
+      "step": 3372
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.9713676441301975,
+      "learning_rate": 3.6124008934332956e-06,
+      "loss": 0.0406,
+      "step": 3373
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.852127016193872,
+      "learning_rate": 3.609624501096739e-06,
+      "loss": 0.0449,
+      "step": 3374
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.550426276023454,
+      "learning_rate": 3.606848573310479e-06,
+      "loss": 0.0425,
+      "step": 3375
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 12.497366697285011,
+      "learning_rate": 3.6040731110020065e-06,
+      "loss": 0.0394,
+      "step": 3376
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 18.168602267642527,
+      "learning_rate": 3.6012981150986524e-06,
+      "loss": 0.0843,
+      "step": 3377
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 12.307783885049966,
+      "learning_rate": 3.598523586527599e-06,
+      "loss": 0.0384,
+      "step": 3378
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.59462797750005,
+      "learning_rate": 3.595749526215862e-06,
+      "loss": 0.0432,
+      "step": 3379
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 14.017885187186423,
+      "learning_rate": 3.5929759350903117e-06,
+      "loss": 0.046,
+      "step": 3380
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 24.792119713261677,
+      "learning_rate": 3.5902028140776524e-06,
+      "loss": 0.0671,
+      "step": 3381
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 12.244242993831454,
+      "learning_rate": 3.5874301641044386e-06,
+      "loss": 0.0345,
+      "step": 3382
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 7.018502392020034,
+      "learning_rate": 3.5846579860970632e-06,
+      "loss": 0.0583,
+      "step": 3383
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 16.228689702919265,
+      "learning_rate": 3.58188628098176e-06,
+      "loss": 0.0566,
+      "step": 3384
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 7.9160236332123395,
+      "learning_rate": 3.579115049684612e-06,
+      "loss": 0.0453,
+      "step": 3385
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 6.865601457118527,
+      "learning_rate": 3.576344293131533e-06,
+      "loss": 0.0482,
+      "step": 3386
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 14.727035098444215,
+      "learning_rate": 3.5735740122482896e-06,
+      "loss": 0.0439,
+      "step": 3387
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.831666188573751,
+      "learning_rate": 3.570804207960481e-06,
+      "loss": 0.0428,
+      "step": 3388
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 6.970567913151844,
+      "learning_rate": 3.5680348811935527e-06,
+      "loss": 0.0429,
+      "step": 3389
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.700224034319951,
+      "learning_rate": 3.565266032872785e-06,
+      "loss": 0.05,
+      "step": 3390
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.060450793558676,
+      "learning_rate": 3.5624976639233056e-06,
+      "loss": 0.0587,
+      "step": 3391
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 9.739022910580005,
+      "learning_rate": 3.559729775270076e-06,
+      "loss": 0.0383,
+      "step": 3392
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.282482394344644,
+      "learning_rate": 3.5569623678378972e-06,
+      "loss": 0.05,
+      "step": 3393
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 7.146211172306711,
+      "learning_rate": 3.554195442551416e-06,
+      "loss": 0.0346,
+      "step": 3394
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 7.117770381713717,
+      "learning_rate": 3.551429000335108e-06,
+      "loss": 0.0453,
+      "step": 3395
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.501430969155091,
+      "learning_rate": 3.5486630421132983e-06,
+      "loss": 0.0406,
+      "step": 3396
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 9.978686458274396,
+      "learning_rate": 3.5458975688101403e-06,
+      "loss": 0.0475,
+      "step": 3397
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 9.504820462744869,
+      "learning_rate": 3.5431325813496352e-06,
+      "loss": 0.055,
+      "step": 3398
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 13.297641172040194,
+      "learning_rate": 3.540368080655612e-06,
+      "loss": 0.064,
+      "step": 3399
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.145894409950023,
+      "learning_rate": 3.5376040676517443e-06,
+      "loss": 0.0508,
+      "step": 3400
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.6754745477101975,
+      "learning_rate": 3.5348405432615407e-06,
+      "loss": 0.0369,
+      "step": 3401
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 9.183504955229806,
+      "learning_rate": 3.5320775084083425e-06,
+      "loss": 0.0362,
+      "step": 3402
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 10.016899646390549,
+      "learning_rate": 3.529314964015336e-06,
+      "loss": 0.0524,
+      "step": 3403
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 7.225269060493574,
+      "learning_rate": 3.526552911005533e-06,
+      "loss": 0.0651,
+      "step": 3404
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 9.402695057707561,
+      "learning_rate": 3.523791350301793e-06,
+      "loss": 0.0508,
+      "step": 3405
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.528168693207084,
+      "learning_rate": 3.5210302828267984e-06,
+      "loss": 0.0591,
+      "step": 3406
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.978722816391329,
+      "learning_rate": 3.5182697095030795e-06,
+      "loss": 0.0323,
+      "step": 3407
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 11.785501724674626,
+      "learning_rate": 3.5155096312529913e-06,
+      "loss": 0.0605,
+      "step": 3408
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 8.863372588103983,
+      "learning_rate": 3.5127500489987252e-06,
+      "loss": 0.0441,
+      "step": 3409
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.336343653134836,
+      "learning_rate": 3.5099909636623148e-06,
+      "loss": 0.0493,
+      "step": 3410
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 9.86742725134051,
+      "learning_rate": 3.5072323761656163e-06,
+      "loss": 0.0493,
+      "step": 3411
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 7.055618254025233,
+      "learning_rate": 3.5044742874303297e-06,
+      "loss": 0.0541,
+      "step": 3412
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 7.776496254142825,
+      "learning_rate": 3.501716698377979e-06,
+      "loss": 0.0308,
+      "step": 3413
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 7.122528698892253,
+      "learning_rate": 3.4989596099299306e-06,
+      "loss": 0.0427,
+      "step": 3414
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 16.55113954313631,
+      "learning_rate": 3.496203023007374e-06,
+      "loss": 0.075,
+      "step": 3415
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 9.731003512719141,
+      "learning_rate": 3.4934469385313418e-06,
+      "loss": 0.0554,
+      "step": 3416
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 35.12656172388353,
+      "learning_rate": 3.490691357422689e-06,
+      "loss": 0.0719,
+      "step": 3417
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 7.754354703696009,
+      "learning_rate": 3.487936280602108e-06,
+      "loss": 0.0235,
+      "step": 3418
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.521163253822065,
+      "learning_rate": 3.4851817089901203e-06,
+      "loss": 0.033,
+      "step": 3419
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 26.224926062833664,
+      "learning_rate": 3.4824276435070804e-06,
+      "loss": 0.0563,
+      "step": 3420
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 15.61271153737091,
+      "learning_rate": 3.4796740850731716e-06,
+      "loss": 0.0356,
+      "step": 3421
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 9.031891603705114,
+      "learning_rate": 3.47692103460841e-06,
+      "loss": 0.0486,
+      "step": 3422
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 6.734795663073578,
+      "learning_rate": 3.474168493032641e-06,
+      "loss": 0.0525,
+      "step": 3423
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 6.755868031059958,
+      "learning_rate": 3.4714164612655387e-06,
+      "loss": 0.0458,
+      "step": 3424
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 7.708505272692726,
+      "learning_rate": 3.468664940226609e-06,
+      "loss": 0.0513,
+      "step": 3425
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.506075218775547,
+      "learning_rate": 3.4659139308351885e-06,
+      "loss": 0.0344,
+      "step": 3426
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.992509225070968,
+      "learning_rate": 3.4631634340104357e-06,
+      "loss": 0.044,
+      "step": 3427
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 7.003251664486951,
+      "learning_rate": 3.460413450671346e-06,
+      "loss": 0.0434,
+      "step": 3428
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 14.475401829587407,
+      "learning_rate": 3.457663981736739e-06,
+      "loss": 0.0854,
+      "step": 3429
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.483836265471178,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.0443,
+      "step": 3430
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 11.13154384738619,
+      "learning_rate": 3.4521665907553957e-06,
+      "loss": 0.0337,
+      "step": 3431
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.800442883625316,
+      "learning_rate": 3.4494186705454402e-06,
+      "loss": 0.0618,
+      "step": 3432
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 16.842487232611695,
+      "learning_rate": 3.446671268413528e-06,
+      "loss": 0.056,
+      "step": 3433
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 6.307882602308188,
+      "learning_rate": 3.443924385277617e-06,
+      "loss": 0.0401,
+      "step": 3434
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 12.347319578908353,
+      "learning_rate": 3.4411780220554937e-06,
+      "loss": 0.049,
+      "step": 3435
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 6.777317017761099,
+      "learning_rate": 3.4384321796647645e-06,
+      "loss": 0.044,
+      "step": 3436
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 8.578839854101997,
+      "learning_rate": 3.4356868590228727e-06,
+      "loss": 0.08,
+      "step": 3437
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 11.34847212257333,
+      "learning_rate": 3.4329420610470745e-06,
+      "loss": 0.0591,
+      "step": 3438
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 16.488323097081217,
+      "learning_rate": 3.4301977866544634e-06,
+      "loss": 0.0469,
+      "step": 3439
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.5345240893456635,
+      "learning_rate": 3.427454036761948e-06,
+      "loss": 0.037,
+      "step": 3440
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 10.70351554080197,
+      "learning_rate": 3.4247108122862703e-06,
+      "loss": 0.049,
+      "step": 3441
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 6.477152114166718,
+      "learning_rate": 3.4219681141439907e-06,
+      "loss": 0.0401,
+      "step": 3442
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 18.807079075717237,
+      "learning_rate": 3.4192259432514934e-06,
+      "loss": 0.0406,
+      "step": 3443
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 15.723153609547095,
+      "learning_rate": 3.4164843005249928e-06,
+      "loss": 0.0524,
+      "step": 3444
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 12.39846598919014,
+      "learning_rate": 3.413743186880519e-06,
+      "loss": 0.0407,
+      "step": 3445
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 7.3971140149986665,
+      "learning_rate": 3.4110026032339317e-06,
+      "loss": 0.0679,
+      "step": 3446
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 20.316958331448777,
+      "learning_rate": 3.408262550500908e-06,
+      "loss": 0.0699,
+      "step": 3447
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 19.17858909141001,
+      "learning_rate": 3.4055230295969556e-06,
+      "loss": 0.0484,
+      "step": 3448
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 10.402764837758005,
+      "learning_rate": 3.4027840414373924e-06,
+      "loss": 0.0569,
+      "step": 3449
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 7.135992378182201,
+      "learning_rate": 3.4000455869373716e-06,
+      "loss": 0.0426,
+      "step": 3450
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 14.292267710034482,
+      "learning_rate": 3.397307667011859e-06,
+      "loss": 0.0685,
+      "step": 3451
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 22.66042408586546,
+      "learning_rate": 3.394570282575642e-06,
+      "loss": 0.0729,
+      "step": 3452
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 7.338566737526984,
+      "learning_rate": 3.3918334345433367e-06,
+      "loss": 0.0495,
+      "step": 3453
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 8.259220087387108,
+      "learning_rate": 3.3890971238293703e-06,
+      "loss": 0.0544,
+      "step": 3454
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 7.349858298914431,
+      "learning_rate": 3.386361351347999e-06,
+      "loss": 0.0507,
+      "step": 3455
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.307434476542483,
+      "learning_rate": 3.3836261180132914e-06,
+      "loss": 0.0546,
+      "step": 3456
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.554073968741056,
+      "learning_rate": 3.3808914247391437e-06,
+      "loss": 0.0492,
+      "step": 3457
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 10.475716145946574,
+      "learning_rate": 3.3781572724392642e-06,
+      "loss": 0.0467,
+      "step": 3458
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 11.333875703512863,
+      "learning_rate": 3.3754236620271876e-06,
+      "loss": 0.0606,
+      "step": 3459
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 7.737852380637597,
+      "learning_rate": 3.3726905944162615e-06,
+      "loss": 0.0582,
+      "step": 3460
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 8.626067005065318,
+      "learning_rate": 3.3699580705196527e-06,
+      "loss": 0.0547,
+      "step": 3461
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 9.788902013697735,
+      "learning_rate": 3.367226091250353e-06,
+      "loss": 0.0514,
+      "step": 3462
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 7.494197930351465,
+      "learning_rate": 3.3644946575211634e-06,
+      "loss": 0.0322,
+      "step": 3463
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 6.72717147418577,
+      "learning_rate": 3.36176377024471e-06,
+      "loss": 0.0377,
+      "step": 3464
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 9.106408875639595,
+      "learning_rate": 3.3590334303334293e-06,
+      "loss": 0.0645,
+      "step": 3465
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.468657818143453,
+      "learning_rate": 3.356303638699583e-06,
+      "loss": 0.0349,
+      "step": 3466
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 10.178729399836424,
+      "learning_rate": 3.35357439625524e-06,
+      "loss": 0.0491,
+      "step": 3467
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 7.847892185842703,
+      "learning_rate": 3.3508457039122965e-06,
+      "loss": 0.0514,
+      "step": 3468
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 12.475199644318261,
+      "learning_rate": 3.348117562582457e-06,
+      "loss": 0.0379,
+      "step": 3469
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.425641585558723,
+      "learning_rate": 3.345389973177241e-06,
+      "loss": 0.0402,
+      "step": 3470
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.957786046660219,
+      "learning_rate": 3.342662936607992e-06,
+      "loss": 0.0685,
+      "step": 3471
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 10.682715401706243,
+      "learning_rate": 3.3399364537858594e-06,
+      "loss": 0.0541,
+      "step": 3472
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.124834331996581,
+      "learning_rate": 3.3372105256218153e-06,
+      "loss": 0.0482,
+      "step": 3473
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 6.155139065247913,
+      "learning_rate": 3.334485153026639e-06,
+      "loss": 0.0362,
+      "step": 3474
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 11.387503864889187,
+      "learning_rate": 3.3317603369109332e-06,
+      "loss": 0.0608,
+      "step": 3475
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 7.514201902512651,
+      "learning_rate": 3.3290360781851055e-06,
+      "loss": 0.0428,
+      "step": 3476
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 10.254928567111829,
+      "learning_rate": 3.326312377759383e-06,
+      "loss": 0.0479,
+      "step": 3477
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 8.581791189780155,
+      "learning_rate": 3.3235892365438038e-06,
+      "loss": 0.0291,
+      "step": 3478
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 9.933864421671199,
+      "learning_rate": 3.3208666554482216e-06,
+      "loss": 0.0492,
+      "step": 3479
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.1340261357847,
+      "learning_rate": 3.3181446353822997e-06,
+      "loss": 0.0361,
+      "step": 3480
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 10.211747713964122,
+      "learning_rate": 3.315423177255516e-06,
+      "loss": 0.0725,
+      "step": 3481
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 7.574056301175818,
+      "learning_rate": 3.312702281977161e-06,
+      "loss": 0.0604,
+      "step": 3482
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 20.440276470356725,
+      "learning_rate": 3.3099819504563356e-06,
+      "loss": 0.0413,
+      "step": 3483
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 11.787768676878581,
+      "learning_rate": 3.3072621836019535e-06,
+      "loss": 0.0383,
+      "step": 3484
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.320753627246697,
+      "learning_rate": 3.3045429823227405e-06,
+      "loss": 0.0366,
+      "step": 3485
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.0007273838143593,
+      "learning_rate": 3.3018243475272282e-06,
+      "loss": 0.0377,
+      "step": 3486
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 15.551358111361543,
+      "learning_rate": 3.2991062801237683e-06,
+      "loss": 0.0653,
+      "step": 3487
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 14.995269345180393,
+      "learning_rate": 3.296388781020513e-06,
+      "loss": 0.065,
+      "step": 3488
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 16.071603280041327,
+      "learning_rate": 3.293671851125434e-06,
+      "loss": 0.0832,
+      "step": 3489
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 11.916063649824876,
+      "learning_rate": 3.2909554913463034e-06,
+      "loss": 0.0473,
+      "step": 3490
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 21.15674028802956,
+      "learning_rate": 3.2882397025907114e-06,
+      "loss": 0.0777,
+      "step": 3491
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 16.156481807543038,
+      "learning_rate": 3.2855244857660497e-06,
+      "loss": 0.0443,
+      "step": 3492
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 16.88354258053397,
+      "learning_rate": 3.2828098417795267e-06,
+      "loss": 0.0409,
+      "step": 3493
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 22.697696452401754,
+      "learning_rate": 3.2800957715381537e-06,
+      "loss": 0.0652,
+      "step": 3494
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.743314683992511,
+      "learning_rate": 3.2773822759487497e-06,
+      "loss": 0.0425,
+      "step": 3495
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 14.222623429544777,
+      "learning_rate": 3.2746693559179483e-06,
+      "loss": 0.0468,
+      "step": 3496
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 36.49482472083385,
+      "learning_rate": 3.2719570123521816e-06,
+      "loss": 0.0753,
+      "step": 3497
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 18.21815675653801,
+      "learning_rate": 3.2692452461576997e-06,
+      "loss": 0.0572,
+      "step": 3498
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 12.83475452863996,
+      "learning_rate": 3.266534058240548e-06,
+      "loss": 0.0497,
+      "step": 3499
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.823480145703581,
+      "learning_rate": 3.2638234495065903e-06,
+      "loss": 0.0537,
+      "step": 3500
+    },
+    {
+      "epoch": 2.5,
+      "eval_avg_AUC": 0.832346751207196,
+      "eval_avg_Accuracy": 0.739597148541114,
+      "eval_avg_Accuracy-right": 0.8735489761314725,
+      "eval_avg_Accuracy-wrong": 0.5060268364794178,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.7006953165400592,
+      "eval_last_AUC": 0.8475061209281654,
+      "eval_last_Accuracy": 0.7807940981432361,
+      "eval_last_Accuracy-right": 0.8329855223685927,
+      "eval_last_Accuracy-wrong": 0.6897884921537412,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.7097495321532842,
+      "eval_max_AUC": 0.7701603116760396,
+      "eval_max_Accuracy": 0.6501574933687002,
+      "eval_max_Accuracy-right": 0.9781531237772271,
+      "eval_max_Accuracy-wrong": 0.07823516033659313,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.6369597827879677,
+      "eval_min_AUC": 0.8455461246669156,
+      "eval_min_Accuracy": 0.7731681034482759,
+      "eval_min_Accuracy-right": 0.7823138124429373,
+      "eval_min_Accuracy-wrong": 0.7572208323857176,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.7126465395097116,
+      "eval_prod_AUC": 0.8377160248849372,
+      "eval_prod_Accuracy": 0.7180868700265252,
+      "eval_prod_Accuracy-right": 0.6126255380200861,
+      "eval_prod_Accuracy-wrong": 0.9019786217875825,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.7055742832242811,
+      "eval_runtime": 247.3763,
+      "eval_samples_per_second": 97.536,
+      "eval_steps_per_second": 3.048,
+      "eval_sum_AUC": 0.712185339026256,
+      "eval_sum_Accuracy": 0.640873673740053,
+      "eval_sum_Accuracy-right": 0.9953697665318899,
+      "eval_sum_Accuracy-wrong": 0.02274277916761428,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6744582768230599,
+      "step": 3500
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 14.845003338865876,
+      "learning_rate": 3.261113420861487e-06,
+      "loss": 0.0569,
+      "step": 3501
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 20.476101635024914,
+      "learning_rate": 3.258403973210713e-06,
+      "loss": 0.0443,
+      "step": 3502
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 16.582413967496823,
+      "learning_rate": 3.2556951074595435e-06,
+      "loss": 0.0405,
+      "step": 3503
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 7.163367417609349,
+      "learning_rate": 3.2529868245130577e-06,
+      "loss": 0.0509,
+      "step": 3504
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 14.256405396539867,
+      "learning_rate": 3.250279125276148e-06,
+      "loss": 0.0443,
+      "step": 3505
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 9.355654318932615,
+      "learning_rate": 3.2475720106535036e-06,
+      "loss": 0.0351,
+      "step": 3506
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 15.4290622991359,
+      "learning_rate": 3.244865481549625e-06,
+      "loss": 0.0385,
+      "step": 3507
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.540807483931574,
+      "learning_rate": 3.24215953886881e-06,
+      "loss": 0.0711,
+      "step": 3508
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.2900577929188204,
+      "learning_rate": 3.2394541835151692e-06,
+      "loss": 0.0322,
+      "step": 3509
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.839214385110157,
+      "learning_rate": 3.2367494163926095e-06,
+      "loss": 0.0355,
+      "step": 3510
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 9.368493630197404,
+      "learning_rate": 3.234045238404841e-06,
+      "loss": 0.0588,
+      "step": 3511
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 13.586715948198096,
+      "learning_rate": 3.2313416504553852e-06,
+      "loss": 0.0579,
+      "step": 3512
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 7.231859040657891,
+      "learning_rate": 3.2286386534475568e-06,
+      "loss": 0.0381,
+      "step": 3513
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.101218107837911,
+      "learning_rate": 3.2259362482844803e-06,
+      "loss": 0.0369,
+      "step": 3514
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.651265778383636,
+      "learning_rate": 3.2232344358690765e-06,
+      "loss": 0.0352,
+      "step": 3515
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 12.961817876373003,
+      "learning_rate": 3.220533217104075e-06,
+      "loss": 0.0605,
+      "step": 3516
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 8.658443920435355,
+      "learning_rate": 3.217832592891999e-06,
+      "loss": 0.0306,
+      "step": 3517
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 7.589369821386444,
+      "learning_rate": 3.2151325641351817e-06,
+      "loss": 0.0788,
+      "step": 3518
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 7.192128941843156,
+      "learning_rate": 3.2124331317357506e-06,
+      "loss": 0.0422,
+      "step": 3519
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.335264955696329,
+      "learning_rate": 3.2097342965956334e-06,
+      "loss": 0.064,
+      "step": 3520
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 2.7148073824964096,
+      "learning_rate": 3.2070360596165667e-06,
+      "loss": 0.0297,
+      "step": 3521
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 14.262742389282423,
+      "learning_rate": 3.204338421700076e-06,
+      "loss": 0.0448,
+      "step": 3522
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 17.047759589845352,
+      "learning_rate": 3.201641383747498e-06,
+      "loss": 0.0615,
+      "step": 3523
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.1887465158901795,
+      "learning_rate": 3.1989449466599574e-06,
+      "loss": 0.0709,
+      "step": 3524
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 10.43183697384031,
+      "learning_rate": 3.1962491113383896e-06,
+      "loss": 0.0739,
+      "step": 3525
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 7.760472498332863,
+      "learning_rate": 3.1935538786835183e-06,
+      "loss": 0.0476,
+      "step": 3526
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 20.92702203320919,
+      "learning_rate": 3.1908592495958747e-06,
+      "loss": 0.053,
+      "step": 3527
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 11.407485827565694,
+      "learning_rate": 3.1881652249757823e-06,
+      "loss": 0.0372,
+      "step": 3528
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.8307280957120735,
+      "learning_rate": 3.185471805723365e-06,
+      "loss": 0.045,
+      "step": 3529
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 8.019327740566352,
+      "learning_rate": 3.1827789927385444e-06,
+      "loss": 0.0433,
+      "step": 3530
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.920055320148531,
+      "learning_rate": 3.18008678692104e-06,
+      "loss": 0.0533,
+      "step": 3531
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.226924007230088,
+      "learning_rate": 3.1773951891703668e-06,
+      "loss": 0.052,
+      "step": 3532
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 13.516443184870806,
+      "learning_rate": 3.1747042003858386e-06,
+      "loss": 0.0547,
+      "step": 3533
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 8.21547792246494,
+      "learning_rate": 3.1720138214665643e-06,
+      "loss": 0.0491,
+      "step": 3534
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 6.142068501852478,
+      "learning_rate": 3.1693240533114496e-06,
+      "loss": 0.0641,
+      "step": 3535
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.150763465920459,
+      "learning_rate": 3.1666348968191955e-06,
+      "loss": 0.0401,
+      "step": 3536
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 11.729987470097232,
+      "learning_rate": 3.1639463528883007e-06,
+      "loss": 0.0582,
+      "step": 3537
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 10.409400310191932,
+      "learning_rate": 3.161258422417055e-06,
+      "loss": 0.0383,
+      "step": 3538
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 6.526722578357467,
+      "learning_rate": 3.1585711063035496e-06,
+      "loss": 0.0496,
+      "step": 3539
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.236653388596862,
+      "learning_rate": 3.155884405445663e-06,
+      "loss": 0.0424,
+      "step": 3540
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.5839787734094095,
+      "learning_rate": 3.153198320741074e-06,
+      "loss": 0.0374,
+      "step": 3541
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 13.274151144936681,
+      "learning_rate": 3.150512853087253e-06,
+      "loss": 0.0564,
+      "step": 3542
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 9.510037031594369,
+      "learning_rate": 3.1478280033814657e-06,
+      "loss": 0.0466,
+      "step": 3543
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 6.862413054864041,
+      "learning_rate": 3.14514377252077e-06,
+      "loss": 0.0373,
+      "step": 3544
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.4795830545549435,
+      "learning_rate": 3.142460161402014e-06,
+      "loss": 0.0541,
+      "step": 3545
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 14.338633368147894,
+      "learning_rate": 3.139777170921847e-06,
+      "loss": 0.0584,
+      "step": 3546
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 8.975528387578118,
+      "learning_rate": 3.137094801976701e-06,
+      "loss": 0.0461,
+      "step": 3547
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 7.4199338093489215,
+      "learning_rate": 3.1344130554628104e-06,
+      "loss": 0.0346,
+      "step": 3548
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 10.49710065460678,
+      "learning_rate": 3.131731932276193e-06,
+      "loss": 0.0508,
+      "step": 3549
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.33530821774696,
+      "learning_rate": 3.129051433312664e-06,
+      "loss": 0.0599,
+      "step": 3550
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.796697346739547,
+      "learning_rate": 3.1263715594678257e-06,
+      "loss": 0.0528,
+      "step": 3551
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.240653694412567,
+      "learning_rate": 3.1236923116370764e-06,
+      "loss": 0.0381,
+      "step": 3552
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.9372123775383145,
+      "learning_rate": 3.121013690715601e-06,
+      "loss": 0.0376,
+      "step": 3553
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 7.609415911504751,
+      "learning_rate": 3.118335697598376e-06,
+      "loss": 0.0358,
+      "step": 3554
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 8.390856165835398,
+      "learning_rate": 3.1156583331801703e-06,
+      "loss": 0.0514,
+      "step": 3555
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 8.449747018529868,
+      "learning_rate": 3.1129815983555387e-06,
+      "loss": 0.0518,
+      "step": 3556
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 17.287771804155526,
+      "learning_rate": 3.1103054940188316e-06,
+      "loss": 0.0499,
+      "step": 3557
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.918114158102355,
+      "learning_rate": 3.1076300210641814e-06,
+      "loss": 0.0447,
+      "step": 3558
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 12.087731186438175,
+      "learning_rate": 3.1049551803855173e-06,
+      "loss": 0.0766,
+      "step": 3559
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.23804644083563,
+      "learning_rate": 3.1022809728765486e-06,
+      "loss": 0.0511,
+      "step": 3560
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 6.3009649089413236,
+      "learning_rate": 3.0996073994307825e-06,
+      "loss": 0.041,
+      "step": 3561
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.219391874758236,
+      "learning_rate": 3.0969344609415076e-06,
+      "loss": 0.0388,
+      "step": 3562
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.0531747625683945,
+      "learning_rate": 3.0942621583017994e-06,
+      "loss": 0.0349,
+      "step": 3563
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 9.97710030264506,
+      "learning_rate": 3.0915904924045294e-06,
+      "loss": 0.0431,
+      "step": 3564
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 9.397820259535381,
+      "learning_rate": 3.088919464142346e-06,
+      "loss": 0.0457,
+      "step": 3565
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.754455760260489,
+      "learning_rate": 3.0862490744076928e-06,
+      "loss": 0.0633,
+      "step": 3566
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.929988907403646,
+      "learning_rate": 3.0835793240927937e-06,
+      "loss": 0.0445,
+      "step": 3567
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.6203167310866937,
+      "learning_rate": 3.0809102140896652e-06,
+      "loss": 0.0346,
+      "step": 3568
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 8.143345190966631,
+      "learning_rate": 3.078241745290103e-06,
+      "loss": 0.0656,
+      "step": 3569
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 9.331281186466539,
+      "learning_rate": 3.075573918585696e-06,
+      "loss": 0.0452,
+      "step": 3570
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 10.05873032023487,
+      "learning_rate": 3.0729067348678127e-06,
+      "loss": 0.0308,
+      "step": 3571
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.45197339309329,
+      "learning_rate": 3.0702401950276066e-06,
+      "loss": 0.0505,
+      "step": 3572
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 6.511150541662428,
+      "learning_rate": 3.067574299956022e-06,
+      "loss": 0.0391,
+      "step": 3573
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 6.19790668856283,
+      "learning_rate": 3.0649090505437804e-06,
+      "loss": 0.0618,
+      "step": 3574
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.092533552558913,
+      "learning_rate": 3.062244447681396e-06,
+      "loss": 0.0362,
+      "step": 3575
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.328037295873963,
+      "learning_rate": 3.0595804922591564e-06,
+      "loss": 0.0398,
+      "step": 3576
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 11.908875157062429,
+      "learning_rate": 3.0569171851671436e-06,
+      "loss": 0.0689,
+      "step": 3577
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 10.261183867138291,
+      "learning_rate": 3.054254527295215e-06,
+      "loss": 0.0539,
+      "step": 3578
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.6021336026425312,
+      "learning_rate": 3.0515925195330148e-06,
+      "loss": 0.0393,
+      "step": 3579
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.4869403375119425,
+      "learning_rate": 3.048931162769969e-06,
+      "loss": 0.0507,
+      "step": 3580
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 16.354612895096743,
+      "learning_rate": 3.0462704578952874e-06,
+      "loss": 0.0511,
+      "step": 3581
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 7.463260414069788,
+      "learning_rate": 3.0436104057979604e-06,
+      "loss": 0.0478,
+      "step": 3582
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 8.65693289332545,
+      "learning_rate": 3.0409510073667602e-06,
+      "loss": 0.0355,
+      "step": 3583
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.322328157842049,
+      "learning_rate": 3.038292263490242e-06,
+      "loss": 0.0651,
+      "step": 3584
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 13.68926290517709,
+      "learning_rate": 3.035634175056742e-06,
+      "loss": 0.0544,
+      "step": 3585
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 10.244078321532088,
+      "learning_rate": 3.0329767429543767e-06,
+      "loss": 0.0561,
+      "step": 3586
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.0632698090854795,
+      "learning_rate": 3.030319968071043e-06,
+      "loss": 0.0345,
+      "step": 3587
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.8950494704080074,
+      "learning_rate": 3.0276638512944177e-06,
+      "loss": 0.0347,
+      "step": 3588
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 7.721157968690705,
+      "learning_rate": 3.025008393511961e-06,
+      "loss": 0.0644,
+      "step": 3589
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 14.006638808045752,
+      "learning_rate": 3.022353595610909e-06,
+      "loss": 0.0735,
+      "step": 3590
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.9140223429542447,
+      "learning_rate": 3.01969945847828e-06,
+      "loss": 0.043,
+      "step": 3591
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 9.900182325928498,
+      "learning_rate": 3.017045983000871e-06,
+      "loss": 0.0489,
+      "step": 3592
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.361887420728005,
+      "learning_rate": 3.014393170065256e-06,
+      "loss": 0.0556,
+      "step": 3593
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 13.67604306112046,
+      "learning_rate": 3.0117410205577903e-06,
+      "loss": 0.0477,
+      "step": 3594
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 9.916712836665532,
+      "learning_rate": 3.0090895353646053e-06,
+      "loss": 0.0385,
+      "step": 3595
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.3840417799657025,
+      "learning_rate": 3.006438715371614e-06,
+      "loss": 0.0473,
+      "step": 3596
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 7.2866366962959095,
+      "learning_rate": 3.0037885614645e-06,
+      "loss": 0.0517,
+      "step": 3597
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 10.141648265814117,
+      "learning_rate": 3.001139074528735e-06,
+      "loss": 0.0262,
+      "step": 3598
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.09068531568647,
+      "learning_rate": 2.9984902554495556e-06,
+      "loss": 0.0488,
+      "step": 3599
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 9.031353090048116,
+      "learning_rate": 2.995842105111987e-06,
+      "loss": 0.0409,
+      "step": 3600
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 6.990355483528864,
+      "learning_rate": 2.99319462440082e-06,
+      "loss": 0.0499,
+      "step": 3601
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.2774820742083435,
+      "learning_rate": 2.990547814200633e-06,
+      "loss": 0.0457,
+      "step": 3602
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 9.343295864443792,
+      "learning_rate": 2.987901675395771e-06,
+      "loss": 0.0399,
+      "step": 3603
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 9.517038015733123,
+      "learning_rate": 2.985256208870357e-06,
+      "loss": 0.0621,
+      "step": 3604
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 7.376829239076995,
+      "learning_rate": 2.982611415508294e-06,
+      "loss": 0.0432,
+      "step": 3605
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.701404340515294,
+      "learning_rate": 2.9799672961932525e-06,
+      "loss": 0.0607,
+      "step": 3606
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.540746888215145,
+      "learning_rate": 2.9773238518086866e-06,
+      "loss": 0.0315,
+      "step": 3607
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.28757599597712,
+      "learning_rate": 2.974681083237816e-06,
+      "loss": 0.0513,
+      "step": 3608
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 12.344327625926315,
+      "learning_rate": 2.972038991363643e-06,
+      "loss": 0.0467,
+      "step": 3609
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 10.626885297699232,
+      "learning_rate": 2.9693975770689344e-06,
+      "loss": 0.0414,
+      "step": 3610
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.12103907244829,
+      "learning_rate": 2.9667568412362415e-06,
+      "loss": 0.0318,
+      "step": 3611
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 13.006583054701983,
+      "learning_rate": 2.9641167847478797e-06,
+      "loss": 0.0516,
+      "step": 3612
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 7.09256099551172,
+      "learning_rate": 2.96147740848594e-06,
+      "loss": 0.0533,
+      "step": 3613
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 6.204069778734054,
+      "learning_rate": 2.9588387133322903e-06,
+      "loss": 0.0599,
+      "step": 3614
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 21.147619366359972,
+      "learning_rate": 2.9562007001685644e-06,
+      "loss": 0.0424,
+      "step": 3615
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 9.26229905936863,
+      "learning_rate": 2.9535633698761755e-06,
+      "loss": 0.0731,
+      "step": 3616
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 12.74094699586504,
+      "learning_rate": 2.9509267233363005e-06,
+      "loss": 0.0692,
+      "step": 3617
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.182529756900972,
+      "learning_rate": 2.948290761429895e-06,
+      "loss": 0.0459,
+      "step": 3618
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 13.62462515024402,
+      "learning_rate": 2.9456554850376805e-06,
+      "loss": 0.0695,
+      "step": 3619
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.9507955856152925,
+      "learning_rate": 2.943020895040155e-06,
+      "loss": 0.0305,
+      "step": 3620
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 34.83572138414297,
+      "learning_rate": 2.940386992317582e-06,
+      "loss": 0.073,
+      "step": 3621
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.677257365244086,
+      "learning_rate": 2.937753777749996e-06,
+      "loss": 0.032,
+      "step": 3622
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 13.762342920713616,
+      "learning_rate": 2.9351212522172056e-06,
+      "loss": 0.0811,
+      "step": 3623
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 10.635154432789744,
+      "learning_rate": 2.9324894165987837e-06,
+      "loss": 0.0372,
+      "step": 3624
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 11.482810263694958,
+      "learning_rate": 2.9298582717740797e-06,
+      "loss": 0.0728,
+      "step": 3625
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 22.37942524618786,
+      "learning_rate": 2.9272278186222025e-06,
+      "loss": 0.0446,
+      "step": 3626
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 9.388487880388006,
+      "learning_rate": 2.9245980580220405e-06,
+      "loss": 0.0342,
+      "step": 3627
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 6.033256714344545,
+      "learning_rate": 2.921968990852242e-06,
+      "loss": 0.0581,
+      "step": 3628
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.768022756457545,
+      "learning_rate": 2.9193406179912297e-06,
+      "loss": 0.0361,
+      "step": 3629
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 8.174036176719323,
+      "learning_rate": 2.91671294031719e-06,
+      "loss": 0.037,
+      "step": 3630
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 10.705237470263434,
+      "learning_rate": 2.91408595870808e-06,
+      "loss": 0.0402,
+      "step": 3631
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 9.964496232913909,
+      "learning_rate": 2.9114596740416224e-06,
+      "loss": 0.0462,
+      "step": 3632
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 10.145165458346774,
+      "learning_rate": 2.908834087195308e-06,
+      "loss": 0.046,
+      "step": 3633
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.38954700874282,
+      "learning_rate": 2.9062091990463935e-06,
+      "loss": 0.0441,
+      "step": 3634
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 13.879887451783997,
+      "learning_rate": 2.903585010471904e-06,
+      "loss": 0.0697,
+      "step": 3635
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 14.995211769676875,
+      "learning_rate": 2.9009615223486297e-06,
+      "loss": 0.0709,
+      "step": 3636
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 15.90824468650359,
+      "learning_rate": 2.898338735553128e-06,
+      "loss": 0.0475,
+      "step": 3637
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 15.039762855178544,
+      "learning_rate": 2.895716650961714e-06,
+      "loss": 0.0751,
+      "step": 3638
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 8.403369327327413,
+      "learning_rate": 2.8930952694504843e-06,
+      "loss": 0.0774,
+      "step": 3639
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 2.68859343539843,
+      "learning_rate": 2.8904745918952833e-06,
+      "loss": 0.031,
+      "step": 3640
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 7.863619642319471,
+      "learning_rate": 2.887854619171735e-06,
+      "loss": 0.0398,
+      "step": 3641
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 7.469206438001187,
+      "learning_rate": 2.8852353521552135e-06,
+      "loss": 0.0381,
+      "step": 3642
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 6.029345864226802,
+      "learning_rate": 2.8826167917208727e-06,
+      "loss": 0.0405,
+      "step": 3643
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.487371662761291,
+      "learning_rate": 2.8799989387436137e-06,
+      "loss": 0.0357,
+      "step": 3644
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 10.641719996018361,
+      "learning_rate": 2.8773817940981186e-06,
+      "loss": 0.0533,
+      "step": 3645
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 6.635229300228402,
+      "learning_rate": 2.8747653586588183e-06,
+      "loss": 0.0632,
+      "step": 3646
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 9.041375472119013,
+      "learning_rate": 2.872149633299913e-06,
+      "loss": 0.0475,
+      "step": 3647
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 7.537386679553337,
+      "learning_rate": 2.8695346188953666e-06,
+      "loss": 0.0461,
+      "step": 3648
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.14722877172696,
+      "learning_rate": 2.866920316318904e-06,
+      "loss": 0.048,
+      "step": 3649
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 14.424463819672413,
+      "learning_rate": 2.8643067264440116e-06,
+      "loss": 0.0525,
+      "step": 3650
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 13.853378771747693,
+      "learning_rate": 2.8616938501439384e-06,
+      "loss": 0.0692,
+      "step": 3651
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 18.270438337227063,
+      "learning_rate": 2.8590816882916948e-06,
+      "loss": 0.0568,
+      "step": 3652
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 18.947419081569944,
+      "learning_rate": 2.856470241760054e-06,
+      "loss": 0.0561,
+      "step": 3653
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.429111310023221,
+      "learning_rate": 2.8538595114215472e-06,
+      "loss": 0.0626,
+      "step": 3654
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.475739674756901,
+      "learning_rate": 2.8512494981484706e-06,
+      "loss": 0.0581,
+      "step": 3655
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 7.771125239389673,
+      "learning_rate": 2.848640202812872e-06,
+      "loss": 0.0417,
+      "step": 3656
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 23.062847695927893,
+      "learning_rate": 2.846031626286574e-06,
+      "loss": 0.0508,
+      "step": 3657
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 11.972621678603849,
+      "learning_rate": 2.8434237694411414e-06,
+      "loss": 0.0488,
+      "step": 3658
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 14.32588978216122,
+      "learning_rate": 2.840816633147917e-06,
+      "loss": 0.0611,
+      "step": 3659
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.884934135428629,
+      "learning_rate": 2.8382102182779846e-06,
+      "loss": 0.0441,
+      "step": 3660
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 9.351592698422403,
+      "learning_rate": 2.8356045257022037e-06,
+      "loss": 0.0354,
+      "step": 3661
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 6.957720102055877,
+      "learning_rate": 2.832999556291177e-06,
+      "loss": 0.0375,
+      "step": 3662
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 10.048379667158468,
+      "learning_rate": 2.8303953109152815e-06,
+      "loss": 0.0575,
+      "step": 3663
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 16.31872426637752,
+      "learning_rate": 2.827791790444638e-06,
+      "loss": 0.045,
+      "step": 3664
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.464392634847121,
+      "learning_rate": 2.8251889957491317e-06,
+      "loss": 0.0614,
+      "step": 3665
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 7.507048236348415,
+      "learning_rate": 2.822586927698407e-06,
+      "loss": 0.0433,
+      "step": 3666
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 9.366777000694293,
+      "learning_rate": 2.819985587161861e-06,
+      "loss": 0.0492,
+      "step": 3667
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 18.342694251690276,
+      "learning_rate": 2.8173849750086513e-06,
+      "loss": 0.0646,
+      "step": 3668
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 13.408885923969533,
+      "learning_rate": 2.8147850921076903e-06,
+      "loss": 0.072,
+      "step": 3669
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.859363792224824,
+      "learning_rate": 2.8121859393276475e-06,
+      "loss": 0.0466,
+      "step": 3670
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 6.544712210188832,
+      "learning_rate": 2.809587517536947e-06,
+      "loss": 0.0583,
+      "step": 3671
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 10.49401514565673,
+      "learning_rate": 2.806989827603771e-06,
+      "loss": 0.053,
+      "step": 3672
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.011757972273914,
+      "learning_rate": 2.8043928703960565e-06,
+      "loss": 0.0323,
+      "step": 3673
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 2.960978152701493,
+      "learning_rate": 2.8017966467814933e-06,
+      "loss": 0.0303,
+      "step": 3674
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 14.114822282619247,
+      "learning_rate": 2.7992011576275295e-06,
+      "loss": 0.0411,
+      "step": 3675
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 7.15626451156831,
+      "learning_rate": 2.7966064038013657e-06,
+      "loss": 0.0564,
+      "step": 3676
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 9.494513074846147,
+      "learning_rate": 2.7940123861699577e-06,
+      "loss": 0.0399,
+      "step": 3677
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.069632762089934,
+      "learning_rate": 2.7914191056000147e-06,
+      "loss": 0.0523,
+      "step": 3678
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 10.285203626675607,
+      "learning_rate": 2.788826562958e-06,
+      "loss": 0.0495,
+      "step": 3679
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 12.855691316396912,
+      "learning_rate": 2.7862347591101326e-06,
+      "loss": 0.051,
+      "step": 3680
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 8.43463699635779,
+      "learning_rate": 2.7836436949223755e-06,
+      "loss": 0.0451,
+      "step": 3681
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 6.876214234656952,
+      "learning_rate": 2.78105337126046e-06,
+      "loss": 0.0444,
+      "step": 3682
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 16.846348195822873,
+      "learning_rate": 2.7784637889898534e-06,
+      "loss": 0.0648,
+      "step": 3683
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.6815124763738485,
+      "learning_rate": 2.7758749489757914e-06,
+      "loss": 0.0646,
+      "step": 3684
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 10.381951130054908,
+      "learning_rate": 2.7732868520832455e-06,
+      "loss": 0.0453,
+      "step": 3685
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 11.46923666061598,
+      "learning_rate": 2.770699499176954e-06,
+      "loss": 0.051,
+      "step": 3686
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.726611333652352,
+      "learning_rate": 2.768112891121394e-06,
+      "loss": 0.0314,
+      "step": 3687
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 15.030148911582819,
+      "learning_rate": 2.7655270287808045e-06,
+      "loss": 0.0557,
+      "step": 3688
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 15.463593669759398,
+      "learning_rate": 2.762941913019166e-06,
+      "loss": 0.0581,
+      "step": 3689
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 6.6863472657093315,
+      "learning_rate": 2.760357544700215e-06,
+      "loss": 0.0718,
+      "step": 3690
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 14.381312588146585,
+      "learning_rate": 2.757773924687437e-06,
+      "loss": 0.0497,
+      "step": 3691
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 8.724171215718298,
+      "learning_rate": 2.755191053844068e-06,
+      "loss": 0.0458,
+      "step": 3692
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 8.172858308911552,
+      "learning_rate": 2.7526089330330925e-06,
+      "loss": 0.054,
+      "step": 3693
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 8.290284625175802,
+      "learning_rate": 2.7500275631172455e-06,
+      "loss": 0.0347,
+      "step": 3694
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.764092240236844,
+      "learning_rate": 2.74744694495901e-06,
+      "loss": 0.0269,
+      "step": 3695
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 8.50315536403876,
+      "learning_rate": 2.74486707942062e-06,
+      "loss": 0.0414,
+      "step": 3696
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 17.67269423715195,
+      "learning_rate": 2.7422879673640552e-06,
+      "loss": 0.0577,
+      "step": 3697
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 13.286685334149144,
+      "learning_rate": 2.7397096096510467e-06,
+      "loss": 0.0527,
+      "step": 3698
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.190362466075019,
+      "learning_rate": 2.7371320071430674e-06,
+      "loss": 0.0354,
+      "step": 3699
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 7.222587422386342,
+      "learning_rate": 2.7345551607013475e-06,
+      "loss": 0.04,
+      "step": 3700
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.454442361570336,
+      "learning_rate": 2.7319790711868545e-06,
+      "loss": 0.0438,
+      "step": 3701
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.525361617331926,
+      "learning_rate": 2.7294037394603135e-06,
+      "loss": 0.0638,
+      "step": 3702
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 6.9455554028374324,
+      "learning_rate": 2.7268291663821825e-06,
+      "loss": 0.0435,
+      "step": 3703
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.868096567440874,
+      "learning_rate": 2.7242553528126842e-06,
+      "loss": 0.0508,
+      "step": 3704
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 6.558699133066451,
+      "learning_rate": 2.72168229961177e-06,
+      "loss": 0.0406,
+      "step": 3705
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.402340494429636,
+      "learning_rate": 2.7191100076391473e-06,
+      "loss": 0.0454,
+      "step": 3706
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 12.420379796085895,
+      "learning_rate": 2.716538477754266e-06,
+      "loss": 0.0583,
+      "step": 3707
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 6.1149424905859675,
+      "learning_rate": 2.713967710816323e-06,
+      "loss": 0.0345,
+      "step": 3708
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 6.348271002030884,
+      "learning_rate": 2.7113977076842597e-06,
+      "loss": 0.0576,
+      "step": 3709
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 7.584921780363273,
+      "learning_rate": 2.7088284692167604e-06,
+      "loss": 0.0529,
+      "step": 3710
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 9.51983997797159,
+      "learning_rate": 2.7062599962722563e-06,
+      "loss": 0.0416,
+      "step": 3711
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 7.319440057301574,
+      "learning_rate": 2.703692289708922e-06,
+      "loss": 0.0438,
+      "step": 3712
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 26.01010056382804,
+      "learning_rate": 2.701125350384676e-06,
+      "loss": 0.0746,
+      "step": 3713
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 11.765662745979094,
+      "learning_rate": 2.69855917915718e-06,
+      "loss": 0.0467,
+      "step": 3714
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 23.418590205484847,
+      "learning_rate": 2.695993776883839e-06,
+      "loss": 0.0595,
+      "step": 3715
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 14.703328795567954,
+      "learning_rate": 2.693429144421803e-06,
+      "loss": 0.0302,
+      "step": 3716
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 23.578741481643345,
+      "learning_rate": 2.6908652826279623e-06,
+      "loss": 0.0658,
+      "step": 3717
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.857687742436122,
+      "learning_rate": 2.688302192358952e-06,
+      "loss": 0.0638,
+      "step": 3718
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 8.405158533466244,
+      "learning_rate": 2.6857398744711472e-06,
+      "loss": 0.0517,
+      "step": 3719
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 20.846618406862838,
+      "learning_rate": 2.683178329820666e-06,
+      "loss": 0.074,
+      "step": 3720
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 16.091758689157754,
+      "learning_rate": 2.680617559263368e-06,
+      "loss": 0.0647,
+      "step": 3721
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.701177069467726,
+      "learning_rate": 2.6780575636548544e-06,
+      "loss": 0.0526,
+      "step": 3722
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 13.59914895192763,
+      "learning_rate": 2.67549834385047e-06,
+      "loss": 0.0457,
+      "step": 3723
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 6.660108186626482,
+      "learning_rate": 2.67293990070529e-06,
+      "loss": 0.0396,
+      "step": 3724
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.818886071754114,
+      "learning_rate": 2.6703822350741483e-06,
+      "loss": 0.0388,
+      "step": 3725
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 7.17966247766626,
+      "learning_rate": 2.6678253478116e-06,
+      "loss": 0.0617,
+      "step": 3726
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 15.063926908466604,
+      "learning_rate": 2.665269239771953e-06,
+      "loss": 0.0401,
+      "step": 3727
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 17.589769650176894,
+      "learning_rate": 2.662713911809248e-06,
+      "loss": 0.0502,
+      "step": 3728
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 7.171926503654985,
+      "learning_rate": 2.6601593647772696e-06,
+      "loss": 0.0394,
+      "step": 3729
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 11.235108267848334,
+      "learning_rate": 2.657605599529538e-06,
+      "loss": 0.055,
+      "step": 3730
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 6.427120667435036,
+      "learning_rate": 2.6550526169193148e-06,
+      "loss": 0.0506,
+      "step": 3731
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.6786946918606946,
+      "learning_rate": 2.6525004177995984e-06,
+      "loss": 0.0367,
+      "step": 3732
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 11.498439922011594,
+      "learning_rate": 2.6499490030231255e-06,
+      "loss": 0.0499,
+      "step": 3733
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.251312011644968,
+      "learning_rate": 2.6473983734423725e-06,
+      "loss": 0.0386,
+      "step": 3734
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 8.575386085669063,
+      "learning_rate": 2.644848529909552e-06,
+      "loss": 0.0491,
+      "step": 3735
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 8.930496092355702,
+      "learning_rate": 2.6422994732766124e-06,
+      "loss": 0.0331,
+      "step": 3736
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 6.071195064464973,
+      "learning_rate": 2.6397512043952422e-06,
+      "loss": 0.0522,
+      "step": 3737
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 7.623010868928134,
+      "learning_rate": 2.637203724116865e-06,
+      "loss": 0.0363,
+      "step": 3738
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 9.317615708928468,
+      "learning_rate": 2.634657033292644e-06,
+      "loss": 0.0439,
+      "step": 3739
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.679344162012358,
+      "learning_rate": 2.6321111327734693e-06,
+      "loss": 0.041,
+      "step": 3740
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 10.628255304414862,
+      "learning_rate": 2.6295660234099816e-06,
+      "loss": 0.0474,
+      "step": 3741
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.7547535821491653,
+      "learning_rate": 2.6270217060525416e-06,
+      "loss": 0.0542,
+      "step": 3742
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 9.99181954191989,
+      "learning_rate": 2.624478181551261e-06,
+      "loss": 0.0369,
+      "step": 3743
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 7.006118145287793,
+      "learning_rate": 2.62193545075597e-06,
+      "loss": 0.0498,
+      "step": 3744
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.068279415859639,
+      "learning_rate": 2.6193935145162507e-06,
+      "loss": 0.0294,
+      "step": 3745
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 8.203047800359975,
+      "learning_rate": 2.6168523736814035e-06,
+      "loss": 0.0416,
+      "step": 3746
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 13.181075968942887,
+      "learning_rate": 2.6143120291004785e-06,
+      "loss": 0.0476,
+      "step": 3747
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.395303454942057,
+      "learning_rate": 2.611772481622246e-06,
+      "loss": 0.0324,
+      "step": 3748
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 8.82113177422998,
+      "learning_rate": 2.609233732095218e-06,
+      "loss": 0.0478,
+      "step": 3749
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 14.848385000891767,
+      "learning_rate": 2.6066957813676375e-06,
+      "loss": 0.0469,
+      "step": 3750
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.738432033718035,
+      "learning_rate": 2.604158630287482e-06,
+      "loss": 0.0406,
+      "step": 3751
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 6.340406715674136,
+      "learning_rate": 2.60162227970246e-06,
+      "loss": 0.0437,
+      "step": 3752
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 7.33847764367545,
+      "learning_rate": 2.5990867304600136e-06,
+      "loss": 0.0372,
+      "step": 3753
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.166395335908612,
+      "learning_rate": 2.5965519834073172e-06,
+      "loss": 0.0388,
+      "step": 3754
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 7.976267116665653,
+      "learning_rate": 2.5940180393912767e-06,
+      "loss": 0.0428,
+      "step": 3755
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 13.601712677500139,
+      "learning_rate": 2.5914848992585293e-06,
+      "loss": 0.0574,
+      "step": 3756
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 8.172580723390993,
+      "learning_rate": 2.588952563855448e-06,
+      "loss": 0.0438,
+      "step": 3757
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 7.078598667043399,
+      "learning_rate": 2.5864210340281247e-06,
+      "loss": 0.041,
+      "step": 3758
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 10.010293320385502,
+      "learning_rate": 2.5838903106224004e-06,
+      "loss": 0.0445,
+      "step": 3759
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 8.32877855610212,
+      "learning_rate": 2.5813603944838283e-06,
+      "loss": 0.0509,
+      "step": 3760
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 8.156187892342613,
+      "learning_rate": 2.578831286457708e-06,
+      "loss": 0.0385,
+      "step": 3761
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 8.634329209857926,
+      "learning_rate": 2.5763029873890542e-06,
+      "loss": 0.0607,
+      "step": 3762
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 10.967700021524228,
+      "learning_rate": 2.573775498122626e-06,
+      "loss": 0.0626,
+      "step": 3763
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 7.039557541057048,
+      "learning_rate": 2.5712488195028972e-06,
+      "loss": 0.0526,
+      "step": 3764
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.977594670286856,
+      "learning_rate": 2.5687229523740852e-06,
+      "loss": 0.0497,
+      "step": 3765
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 14.378683618047381,
+      "learning_rate": 2.566197897580124e-06,
+      "loss": 0.0786,
+      "step": 3766
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 11.70648699510383,
+      "learning_rate": 2.5636736559646824e-06,
+      "loss": 0.0514,
+      "step": 3767
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 9.415184037952974,
+      "learning_rate": 2.5611502283711576e-06,
+      "loss": 0.0461,
+      "step": 3768
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.61280952127481,
+      "learning_rate": 2.5586276156426726e-06,
+      "loss": 0.0415,
+      "step": 3769
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 10.422217664943227,
+      "learning_rate": 2.55610581862208e-06,
+      "loss": 0.0533,
+      "step": 3770
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 9.032558758185202,
+      "learning_rate": 2.553584838151959e-06,
+      "loss": 0.0513,
+      "step": 3771
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 9.894389545613095,
+      "learning_rate": 2.5510646750746154e-06,
+      "loss": 0.0457,
+      "step": 3772
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 9.655381132642049,
+      "learning_rate": 2.548545330232083e-06,
+      "loss": 0.0419,
+      "step": 3773
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.7174887879327745,
+      "learning_rate": 2.5460268044661215e-06,
+      "loss": 0.0477,
+      "step": 3774
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 16.20332294670294,
+      "learning_rate": 2.5435090986182176e-06,
+      "loss": 0.0663,
+      "step": 3775
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 7.352189227072673,
+      "learning_rate": 2.5409922135295827e-06,
+      "loss": 0.0466,
+      "step": 3776
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.547528471853324,
+      "learning_rate": 2.538476150041156e-06,
+      "loss": 0.032,
+      "step": 3777
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 9.148875990223138,
+      "learning_rate": 2.5359609089936006e-06,
+      "loss": 0.0935,
+      "step": 3778
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 7.890951821809732,
+      "learning_rate": 2.533446491227305e-06,
+      "loss": 0.0345,
+      "step": 3779
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 11.425432370389863,
+      "learning_rate": 2.5309328975823834e-06,
+      "loss": 0.051,
+      "step": 3780
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 7.139640801191179,
+      "learning_rate": 2.5284201288986744e-06,
+      "loss": 0.0556,
+      "step": 3781
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 12.38512234939895,
+      "learning_rate": 2.5259081860157418e-06,
+      "loss": 0.0362,
+      "step": 3782
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 6.283048395599437,
+      "learning_rate": 2.5233970697728673e-06,
+      "loss": 0.0679,
+      "step": 3783
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 7.575758374522927,
+      "learning_rate": 2.520886781009068e-06,
+      "loss": 0.0566,
+      "step": 3784
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.885508330549097,
+      "learning_rate": 2.5183773205630726e-06,
+      "loss": 0.0316,
+      "step": 3785
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 15.694594828315246,
+      "learning_rate": 2.515868689273344e-06,
+      "loss": 0.0845,
+      "step": 3786
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 9.405452769181828,
+      "learning_rate": 2.513360887978056e-06,
+      "loss": 0.0417,
+      "step": 3787
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.6408201726978895,
+      "learning_rate": 2.510853917515119e-06,
+      "loss": 0.0554,
+      "step": 3788
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.7474944897018405,
+      "learning_rate": 2.50834777872215e-06,
+      "loss": 0.036,
+      "step": 3789
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 6.32046437072879,
+      "learning_rate": 2.505842472436506e-06,
+      "loss": 0.0375,
+      "step": 3790
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.876168319824173,
+      "learning_rate": 2.5033379994952493e-06,
+      "loss": 0.0405,
+      "step": 3791
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.4487084790696407,
+      "learning_rate": 2.5008343607351733e-06,
+      "loss": 0.0408,
+      "step": 3792
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 21.368691291832302,
+      "learning_rate": 2.4983315569927895e-06,
+      "loss": 0.0717,
+      "step": 3793
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.394868150765753,
+      "learning_rate": 2.495829589104333e-06,
+      "loss": 0.0406,
+      "step": 3794
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 7.701013264562421,
+      "learning_rate": 2.493328457905755e-06,
+      "loss": 0.0374,
+      "step": 3795
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 9.241059362038095,
+      "learning_rate": 2.490828164232732e-06,
+      "loss": 0.0476,
+      "step": 3796
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 10.070938365629022,
+      "learning_rate": 2.4883287089206582e-06,
+      "loss": 0.0363,
+      "step": 3797
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 9.776212143814869,
+      "learning_rate": 2.48583009280465e-06,
+      "loss": 0.0464,
+      "step": 3798
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.2565285582268926,
+      "learning_rate": 2.483332316719535e-06,
+      "loss": 0.0363,
+      "step": 3799
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.0244463261051457,
+      "learning_rate": 2.4808353814998747e-06,
+      "loss": 0.0432,
+      "step": 3800
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 11.23437776694193,
+      "learning_rate": 2.4783392879799345e-06,
+      "loss": 0.0754,
+      "step": 3801
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 7.504406874948566,
+      "learning_rate": 2.4758440369937125e-06,
+      "loss": 0.0468,
+      "step": 3802
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 11.839846713260169,
+      "learning_rate": 2.4733496293749116e-06,
+      "loss": 0.0382,
+      "step": 3803
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 15.580980042235241,
+      "learning_rate": 2.4708560659569665e-06,
+      "loss": 0.0377,
+      "step": 3804
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.915771833170678,
+      "learning_rate": 2.4683633475730158e-06,
+      "loss": 0.0426,
+      "step": 3805
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.7801116481824035,
+      "learning_rate": 2.465871475055931e-06,
+      "loss": 0.0416,
+      "step": 3806
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 13.257844295682492,
+      "learning_rate": 2.4633804492382866e-06,
+      "loss": 0.0469,
+      "step": 3807
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 2.321691675555728,
+      "learning_rate": 2.460890270952383e-06,
+      "loss": 0.0398,
+      "step": 3808
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.925469089728795,
+      "learning_rate": 2.4584009410302357e-06,
+      "loss": 0.0483,
+      "step": 3809
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.729798068275378,
+      "learning_rate": 2.4559124603035744e-06,
+      "loss": 0.0408,
+      "step": 3810
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 8.12961700972875,
+      "learning_rate": 2.4534248296038488e-06,
+      "loss": 0.0488,
+      "step": 3811
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.6114805207114986,
+      "learning_rate": 2.4509380497622208e-06,
+      "loss": 0.031,
+      "step": 3812
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 10.032594319995873,
+      "learning_rate": 2.448452121609571e-06,
+      "loss": 0.0432,
+      "step": 3813
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 6.816652452453811,
+      "learning_rate": 2.445967045976493e-06,
+      "loss": 0.0637,
+      "step": 3814
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 8.219953176187644,
+      "learning_rate": 2.443482823693298e-06,
+      "loss": 0.0431,
+      "step": 3815
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 6.300399021489175,
+      "learning_rate": 2.4409994555900125e-06,
+      "loss": 0.0393,
+      "step": 3816
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 10.553580034207647,
+      "learning_rate": 2.4385169424963696e-06,
+      "loss": 0.0486,
+      "step": 3817
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.250056961602972,
+      "learning_rate": 2.4360352852418305e-06,
+      "loss": 0.038,
+      "step": 3818
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 2.5109869107834233,
+      "learning_rate": 2.4335544846555564e-06,
+      "loss": 0.0341,
+      "step": 3819
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.141809200082095,
+      "learning_rate": 2.431074541566436e-06,
+      "loss": 0.0346,
+      "step": 3820
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 8.93333143773801,
+      "learning_rate": 2.4285954568030566e-06,
+      "loss": 0.0596,
+      "step": 3821
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 2.8270654820116854,
+      "learning_rate": 2.426117231193735e-06,
+      "loss": 0.0378,
+      "step": 3822
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 6.114877716065037,
+      "learning_rate": 2.4236398655664834e-06,
+      "loss": 0.0361,
+      "step": 3823
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 11.797891191634468,
+      "learning_rate": 2.4211633607490442e-06,
+      "loss": 0.0575,
+      "step": 3824
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 10.73316573754624,
+      "learning_rate": 2.4186877175688576e-06,
+      "loss": 0.0678,
+      "step": 3825
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 5.071528261076872,
+      "learning_rate": 2.4162129368530848e-06,
+      "loss": 0.053,
+      "step": 3826
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 5.82366753919729,
+      "learning_rate": 2.413739019428595e-06,
+      "loss": 0.0371,
+      "step": 3827
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.408441529976167,
+      "learning_rate": 2.41126596612197e-06,
+      "loss": 0.0352,
+      "step": 3828
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 2.9553768058089136,
+      "learning_rate": 2.408793777759504e-06,
+      "loss": 0.0439,
+      "step": 3829
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 7.364755485236182,
+      "learning_rate": 2.4063224551672e-06,
+      "loss": 0.0334,
+      "step": 3830
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 7.684527850853966,
+      "learning_rate": 2.4038519991707725e-06,
+      "loss": 0.0483,
+      "step": 3831
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 9.183703665017836,
+      "learning_rate": 2.4013824105956483e-06,
+      "loss": 0.054,
+      "step": 3832
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.171026788107983,
+      "learning_rate": 2.3989136902669614e-06,
+      "loss": 0.0352,
+      "step": 3833
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 19.60009405514915,
+      "learning_rate": 2.396445839009558e-06,
+      "loss": 0.0358,
+      "step": 3834
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.711556978156083,
+      "learning_rate": 2.3939788576479926e-06,
+      "loss": 0.0421,
+      "step": 3835
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 22.38928410529173,
+      "learning_rate": 2.39151274700653e-06,
+      "loss": 0.0849,
+      "step": 3836
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 10.767567534815475,
+      "learning_rate": 2.389047507909143e-06,
+      "loss": 0.0384,
+      "step": 3837
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 12.836118570888207,
+      "learning_rate": 2.3865831411795137e-06,
+      "loss": 0.0568,
+      "step": 3838
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 15.689055796114143,
+      "learning_rate": 2.3841196476410337e-06,
+      "loss": 0.0585,
+      "step": 3839
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 7.494309674312284,
+      "learning_rate": 2.3816570281168016e-06,
+      "loss": 0.0571,
+      "step": 3840
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 16.871108256802874,
+      "learning_rate": 2.379195283429626e-06,
+      "loss": 0.0378,
+      "step": 3841
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 9.679241876905133,
+      "learning_rate": 2.3767344144020164e-06,
+      "loss": 0.0595,
+      "step": 3842
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 2.73030324100972,
+      "learning_rate": 2.374274421856202e-06,
+      "loss": 0.0332,
+      "step": 3843
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.984926898926288,
+      "learning_rate": 2.371815306614104e-06,
+      "loss": 0.0441,
+      "step": 3844
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 6.959555864531863,
+      "learning_rate": 2.3693570694973673e-06,
+      "loss": 0.0458,
+      "step": 3845
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 19.08144651468584,
+      "learning_rate": 2.366899711327326e-06,
+      "loss": 0.0403,
+      "step": 3846
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 7.685927254780913,
+      "learning_rate": 2.3644432329250374e-06,
+      "loss": 0.0418,
+      "step": 3847
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.733683829988181,
+      "learning_rate": 2.3619876351112486e-06,
+      "loss": 0.0475,
+      "step": 3848
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 7.2819284808845905,
+      "learning_rate": 2.3595329187064282e-06,
+      "loss": 0.041,
+      "step": 3849
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 8.790602910243651,
+      "learning_rate": 2.3570790845307367e-06,
+      "loss": 0.0455,
+      "step": 3850
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.762808139392179,
+      "learning_rate": 2.3546261334040475e-06,
+      "loss": 0.0621,
+      "step": 3851
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 18.647824678033906,
+      "learning_rate": 2.352174066145938e-06,
+      "loss": 0.0724,
+      "step": 3852
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 7.320240282215148,
+      "learning_rate": 2.3497228835756887e-06,
+      "loss": 0.0392,
+      "step": 3853
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 12.389189567383427,
+      "learning_rate": 2.3472725865122854e-06,
+      "loss": 0.0609,
+      "step": 3854
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 10.228654260988792,
+      "learning_rate": 2.344823175774418e-06,
+      "loss": 0.0463,
+      "step": 3855
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 6.773649330102428,
+      "learning_rate": 2.3423746521804796e-06,
+      "loss": 0.0524,
+      "step": 3856
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.435268080793936,
+      "learning_rate": 2.339927016548568e-06,
+      "loss": 0.0579,
+      "step": 3857
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 7.41772531916488,
+      "learning_rate": 2.3374802696964842e-06,
+      "loss": 0.0454,
+      "step": 3858
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 10.022754338495929,
+      "learning_rate": 2.3350344124417336e-06,
+      "loss": 0.035,
+      "step": 3859
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 10.610656647476898,
+      "learning_rate": 2.3325894456015154e-06,
+      "loss": 0.0523,
+      "step": 3860
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 18.39360281841152,
+      "learning_rate": 2.3301453699927477e-06,
+      "loss": 0.0579,
+      "step": 3861
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 6.048783311715747,
+      "learning_rate": 2.3277021864320332e-06,
+      "loss": 0.0506,
+      "step": 3862
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 7.751672566932973,
+      "learning_rate": 2.325259895735693e-06,
+      "loss": 0.0463,
+      "step": 3863
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 18.50386797364024,
+      "learning_rate": 2.322818498719734e-06,
+      "loss": 0.044,
+      "step": 3864
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 12.855762559723866,
+      "learning_rate": 2.3203779961998795e-06,
+      "loss": 0.0297,
+      "step": 3865
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 13.336159487903306,
+      "learning_rate": 2.317938388991541e-06,
+      "loss": 0.0365,
+      "step": 3866
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 16.833109489814944,
+      "learning_rate": 2.3154996779098405e-06,
+      "loss": 0.058,
+      "step": 3867
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 3.4464387255304665,
+      "learning_rate": 2.313061863769594e-06,
+      "loss": 0.0407,
+      "step": 3868
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 5.236811429649089,
+      "learning_rate": 2.310624947385322e-06,
+      "loss": 0.0554,
+      "step": 3869
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 17.42376851125036,
+      "learning_rate": 2.3081889295712434e-06,
+      "loss": 0.0373,
+      "step": 3870
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 28.16530722294004,
+      "learning_rate": 2.3057538111412765e-06,
+      "loss": 0.0786,
+      "step": 3871
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 3.5925463205431445,
+      "learning_rate": 2.3033195929090404e-06,
+      "loss": 0.0395,
+      "step": 3872
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 9.615138697340234,
+      "learning_rate": 2.300886275687852e-06,
+      "loss": 0.0433,
+      "step": 3873
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 18.44213000447938,
+      "learning_rate": 2.298453860290728e-06,
+      "loss": 0.0474,
+      "step": 3874
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 14.67959710316391,
+      "learning_rate": 2.296022347530384e-06,
+      "loss": 0.0471,
+      "step": 3875
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 10.905366471260221,
+      "learning_rate": 2.293591738219233e-06,
+      "loss": 0.0513,
+      "step": 3876
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 11.816111298594338,
+      "learning_rate": 2.2911620331693867e-06,
+      "loss": 0.0495,
+      "step": 3877
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 10.254614678819744,
+      "learning_rate": 2.2887332331926555e-06,
+      "loss": 0.06,
+      "step": 3878
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 16.736637604289378,
+      "learning_rate": 2.2863053391005462e-06,
+      "loss": 0.0427,
+      "step": 3879
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 18.759961357816216,
+      "learning_rate": 2.2838783517042628e-06,
+      "loss": 0.0543,
+      "step": 3880
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 17.80455205698793,
+      "learning_rate": 2.281452271814708e-06,
+      "loss": 0.063,
+      "step": 3881
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 19.60942208840727,
+      "learning_rate": 2.2790271002424794e-06,
+      "loss": 0.0605,
+      "step": 3882
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 12.453488405455872,
+      "learning_rate": 2.276602837797872e-06,
+      "loss": 0.0385,
+      "step": 3883
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 6.709907947321404,
+      "learning_rate": 2.274179485290879e-06,
+      "loss": 0.0788,
+      "step": 3884
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 20.038024659232907,
+      "learning_rate": 2.271757043531184e-06,
+      "loss": 0.056,
+      "step": 3885
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 9.050003866061806,
+      "learning_rate": 2.2693355133281706e-06,
+      "loss": 0.0639,
+      "step": 3886
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 18.58808666026702,
+      "learning_rate": 2.266914895490918e-06,
+      "loss": 0.0564,
+      "step": 3887
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 27.52059173829719,
+      "learning_rate": 2.2644951908282e-06,
+      "loss": 0.0483,
+      "step": 3888
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 10.592134606963889,
+      "learning_rate": 2.262076400148484e-06,
+      "loss": 0.0627,
+      "step": 3889
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 8.496491370108698,
+      "learning_rate": 2.2596585242599333e-06,
+      "loss": 0.0794,
+      "step": 3890
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 17.493489913800556,
+      "learning_rate": 2.257241563970405e-06,
+      "loss": 0.0555,
+      "step": 3891
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 13.568365004774797,
+      "learning_rate": 2.254825520087451e-06,
+      "loss": 0.064,
+      "step": 3892
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.508400191683394,
+      "learning_rate": 2.2524103934183154e-06,
+      "loss": 0.0353,
+      "step": 3893
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 12.32194315514444,
+      "learning_rate": 2.249996184769938e-06,
+      "loss": 0.0521,
+      "step": 3894
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 18.4710969837349,
+      "learning_rate": 2.2475828949489504e-06,
+      "loss": 0.0694,
+      "step": 3895
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 10.766594698490506,
+      "learning_rate": 2.2451705247616774e-06,
+      "loss": 0.0493,
+      "step": 3896
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 9.107591867740075,
+      "learning_rate": 2.2427590750141364e-06,
+      "loss": 0.0403,
+      "step": 3897
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 10.375009405030395,
+      "learning_rate": 2.240348546512039e-06,
+      "loss": 0.0589,
+      "step": 3898
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 16.177260041162715,
+      "learning_rate": 2.237938940060786e-06,
+      "loss": 0.0407,
+      "step": 3899
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 10.180497058246122,
+      "learning_rate": 2.235530256465474e-06,
+      "loss": 0.0477,
+      "step": 3900
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.248521888185715,
+      "learning_rate": 2.233122496530884e-06,
+      "loss": 0.0443,
+      "step": 3901
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 3.581699527389943,
+      "learning_rate": 2.2307156610615e-06,
+      "loss": 0.0369,
+      "step": 3902
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 9.5196018314941,
+      "learning_rate": 2.2283097508614837e-06,
+      "loss": 0.0368,
+      "step": 3903
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 3.2933709812401295,
+      "learning_rate": 2.225904766734702e-06,
+      "loss": 0.0442,
+      "step": 3904
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 12.630550369303542,
+      "learning_rate": 2.2235007094846963e-06,
+      "loss": 0.0475,
+      "step": 3905
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 10.408411930140902,
+      "learning_rate": 2.2210975799147143e-06,
+      "loss": 0.0349,
+      "step": 3906
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.291974830052436,
+      "learning_rate": 2.21869537882768e-06,
+      "loss": 0.0429,
+      "step": 3907
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 7.91993089844047,
+      "learning_rate": 2.21629410702622e-06,
+      "loss": 0.0311,
+      "step": 3908
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 7.232659660731117,
+      "learning_rate": 2.2138937653126393e-06,
+      "loss": 0.0378,
+      "step": 3909
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 6.846272738804037,
+      "learning_rate": 2.2114943544889366e-06,
+      "loss": 0.0432,
+      "step": 3910
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 6.8102817293655065,
+      "learning_rate": 2.2090958753568013e-06,
+      "loss": 0.0545,
+      "step": 3911
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 15.66677503356628,
+      "learning_rate": 2.206698328717609e-06,
+      "loss": 0.0679,
+      "step": 3912
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 8.851017334282632,
+      "learning_rate": 2.2043017153724253e-06,
+      "loss": 0.0357,
+      "step": 3913
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.348470099639396,
+      "learning_rate": 2.2019060361220036e-06,
+      "loss": 0.0399,
+      "step": 3914
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 18.351757414727047,
+      "learning_rate": 2.199511291766783e-06,
+      "loss": 0.0485,
+      "step": 3915
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 13.228376315595304,
+      "learning_rate": 2.1971174831068944e-06,
+      "loss": 0.0478,
+      "step": 3916
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.989234913705305,
+      "learning_rate": 2.1947246109421514e-06,
+      "loss": 0.0687,
+      "step": 3917
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 7.578293430391057,
+      "learning_rate": 2.192332676072061e-06,
+      "loss": 0.0698,
+      "step": 3918
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 19.50726319750263,
+      "learning_rate": 2.189941679295807e-06,
+      "loss": 0.0401,
+      "step": 3919
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 7.056050545567341,
+      "learning_rate": 2.1875516214122723e-06,
+      "loss": 0.0625,
+      "step": 3920
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 6.43155113737614,
+      "learning_rate": 2.185162503220013e-06,
+      "loss": 0.0531,
+      "step": 3921
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 7.53621423016585,
+      "learning_rate": 2.182774325517285e-06,
+      "loss": 0.0531,
+      "step": 3922
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 5.175804512791806,
+      "learning_rate": 2.180387089102016e-06,
+      "loss": 0.0406,
+      "step": 3923
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 5.263521350393166,
+      "learning_rate": 2.1780007947718336e-06,
+      "loss": 0.0337,
+      "step": 3924
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 6.01139176635299,
+      "learning_rate": 2.175615443324035e-06,
+      "loss": 0.0613,
+      "step": 3925
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 11.818360574772607,
+      "learning_rate": 2.173231035555618e-06,
+      "loss": 0.0646,
+      "step": 3926
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 3.039048127733819,
+      "learning_rate": 2.170847572263252e-06,
+      "loss": 0.0482,
+      "step": 3927
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 6.788911571364,
+      "learning_rate": 2.1684650542432985e-06,
+      "loss": 0.0548,
+      "step": 3928
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.333329054284375,
+      "learning_rate": 2.166083482291801e-06,
+      "loss": 0.0405,
+      "step": 3929
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 7.014460540954454,
+      "learning_rate": 2.1637028572044867e-06,
+      "loss": 0.0438,
+      "step": 3930
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 5.847965777925317,
+      "learning_rate": 2.1613231797767668e-06,
+      "loss": 0.066,
+      "step": 3931
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 8.636014633570216,
+      "learning_rate": 2.158944450803736e-06,
+      "loss": 0.0629,
+      "step": 3932
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 29.011178688159877,
+      "learning_rate": 2.1565666710801714e-06,
+      "loss": 0.0831,
+      "step": 3933
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.645775512919015,
+      "learning_rate": 2.1541898414005343e-06,
+      "loss": 0.0553,
+      "step": 3934
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.5723835589756465,
+      "learning_rate": 2.1518139625589663e-06,
+      "loss": 0.0328,
+      "step": 3935
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 9.046555105138385,
+      "learning_rate": 2.1494390353492935e-06,
+      "loss": 0.0498,
+      "step": 3936
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 5.551419544385709,
+      "learning_rate": 2.1470650605650235e-06,
+      "loss": 0.0357,
+      "step": 3937
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 5.766470083781369,
+      "learning_rate": 2.144692038999345e-06,
+      "loss": 0.0416,
+      "step": 3938
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 11.936429313135362,
+      "learning_rate": 2.142319971445129e-06,
+      "loss": 0.0496,
+      "step": 3939
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.825676884200282,
+      "learning_rate": 2.139948858694926e-06,
+      "loss": 0.0468,
+      "step": 3940
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.708123288700695,
+      "learning_rate": 2.137578701540971e-06,
+      "loss": 0.0472,
+      "step": 3941
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 3.6296186806200073,
+      "learning_rate": 2.1352095007751754e-06,
+      "loss": 0.045,
+      "step": 3942
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 5.596968703398852,
+      "learning_rate": 2.132841257189137e-06,
+      "loss": 0.0561,
+      "step": 3943
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 3.7089513216445256,
+      "learning_rate": 2.1304739715741235e-06,
+      "loss": 0.0627,
+      "step": 3944
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 3.45323572738163,
+      "learning_rate": 2.128107644721096e-06,
+      "loss": 0.0333,
+      "step": 3945
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 5.788593095757397,
+      "learning_rate": 2.1257422774206816e-06,
+      "loss": 0.045,
+      "step": 3946
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 9.321850934523075,
+      "learning_rate": 2.1233778704632002e-06,
+      "loss": 0.0383,
+      "step": 3947
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 10.186691315669988,
+      "learning_rate": 2.1210144246386378e-06,
+      "loss": 0.0892,
+      "step": 3948
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 8.396952907923813,
+      "learning_rate": 2.1186519407366725e-06,
+      "loss": 0.0443,
+      "step": 3949
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 2.555619165023534,
+      "learning_rate": 2.1162904195466455e-06,
+      "loss": 0.0361,
+      "step": 3950
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 8.059103832940904,
+      "learning_rate": 2.113929861857594e-06,
+      "loss": 0.0584,
+      "step": 3951
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 5.033265961432398,
+      "learning_rate": 2.1115702684582177e-06,
+      "loss": 0.0424,
+      "step": 3952
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.060885661594244,
+      "learning_rate": 2.1092116401369033e-06,
+      "loss": 0.0406,
+      "step": 3953
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 2.7540242910146486,
+      "learning_rate": 2.1068539776817115e-06,
+      "loss": 0.0282,
+      "step": 3954
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 4.013085063098618,
+      "learning_rate": 2.1044972818803816e-06,
+      "loss": 0.0482,
+      "step": 3955
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 11.249338957476022,
+      "learning_rate": 2.1021415535203294e-06,
+      "loss": 0.0404,
+      "step": 3956
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 11.67130080350898,
+      "learning_rate": 2.0997867933886467e-06,
+      "loss": 0.0558,
+      "step": 3957
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 11.149614862400087,
+      "learning_rate": 2.0974330022721044e-06,
+      "loss": 0.0512,
+      "step": 3958
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 3.687014650829844,
+      "learning_rate": 2.0950801809571466e-06,
+      "loss": 0.0612,
+      "step": 3959
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.395001654489367,
+      "learning_rate": 2.0927283302298944e-06,
+      "loss": 0.0432,
+      "step": 3960
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.284072609548135,
+      "learning_rate": 2.0903774508761477e-06,
+      "loss": 0.0447,
+      "step": 3961
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 3.343071674546592,
+      "learning_rate": 2.0880275436813726e-06,
+      "loss": 0.0218,
+      "step": 3962
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 6.360795545549215,
+      "learning_rate": 2.0856786094307247e-06,
+      "loss": 0.053,
+      "step": 3963
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 6.794862698130726,
+      "learning_rate": 2.0833306489090186e-06,
+      "loss": 0.0465,
+      "step": 3964
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.79082889481297,
+      "learning_rate": 2.08098366290076e-06,
+      "loss": 0.0308,
+      "step": 3965
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 10.3999370599744,
+      "learning_rate": 2.078637652190112e-06,
+      "loss": 0.0356,
+      "step": 3966
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.198526561829731,
+      "learning_rate": 2.0762926175609287e-06,
+      "loss": 0.045,
+      "step": 3967
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 7.195220899911224,
+      "learning_rate": 2.0739485597967237e-06,
+      "loss": 0.064,
+      "step": 3968
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.666052330346906,
+      "learning_rate": 2.0716054796806916e-06,
+      "loss": 0.0467,
+      "step": 3969
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.662416911777877,
+      "learning_rate": 2.0692633779956998e-06,
+      "loss": 0.0317,
+      "step": 3970
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 8.017085740979171,
+      "learning_rate": 2.0669222555242884e-06,
+      "loss": 0.0516,
+      "step": 3971
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 7.265789210058505,
+      "learning_rate": 2.064582113048669e-06,
+      "loss": 0.0457,
+      "step": 3972
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 8.795311756133641,
+      "learning_rate": 2.0622429513507275e-06,
+      "loss": 0.0427,
+      "step": 3973
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 6.9365340050039865,
+      "learning_rate": 2.05990477121202e-06,
+      "loss": 0.063,
+      "step": 3974
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 3.914484245239362,
+      "learning_rate": 2.0575675734137773e-06,
+      "loss": 0.0345,
+      "step": 3975
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 5.314654954698931,
+      "learning_rate": 2.0552313587369003e-06,
+      "loss": 0.0421,
+      "step": 3976
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 3.3533779142183926,
+      "learning_rate": 2.052896127961963e-06,
+      "loss": 0.0424,
+      "step": 3977
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 19.476559355360912,
+      "learning_rate": 2.050561881869205e-06,
+      "loss": 0.0509,
+      "step": 3978
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 6.938628294930286,
+      "learning_rate": 2.048228621238547e-06,
+      "loss": 0.0466,
+      "step": 3979
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 10.94306529980738,
+      "learning_rate": 2.0458963468495692e-06,
+      "loss": 0.0442,
+      "step": 3980
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.791706216599774,
+      "learning_rate": 2.0435650594815338e-06,
+      "loss": 0.0298,
+      "step": 3981
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 20.151160656020064,
+      "learning_rate": 2.0412347599133607e-06,
+      "loss": 0.0598,
+      "step": 3982
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 12.255216096956687,
+      "learning_rate": 2.0389054489236534e-06,
+      "loss": 0.0381,
+      "step": 3983
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.855045793700611,
+      "learning_rate": 2.03657712729067e-06,
+      "loss": 0.0604,
+      "step": 3984
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 8.182350191418786,
+      "learning_rate": 2.034249795792355e-06,
+      "loss": 0.0321,
+      "step": 3985
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 7.707765374797664,
+      "learning_rate": 2.031923455206306e-06,
+      "loss": 0.0521,
+      "step": 3986
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 6.104221297033886,
+      "learning_rate": 2.0295981063098e-06,
+      "loss": 0.0394,
+      "step": 3987
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 11.93087451550472,
+      "learning_rate": 2.027273749879777e-06,
+      "loss": 0.0406,
+      "step": 3988
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 5.2406447108361265,
+      "learning_rate": 2.02495038669285e-06,
+      "loss": 0.0346,
+      "step": 3989
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 13.604224730402509,
+      "learning_rate": 2.0226280175252966e-06,
+      "loss": 0.0383,
+      "step": 3990
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 2.555103971529308,
+      "learning_rate": 2.020306643153063e-06,
+      "loss": 0.0337,
+      "step": 3991
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 6.369185477762121,
+      "learning_rate": 2.0179862643517657e-06,
+      "loss": 0.0542,
+      "step": 3992
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 3.4873107821184797,
+      "learning_rate": 2.015666881896684e-06,
+      "loss": 0.0335,
+      "step": 3993
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 6.806890531404817,
+      "learning_rate": 2.0133484965627683e-06,
+      "loss": 0.0486,
+      "step": 3994
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 11.163092400988585,
+      "learning_rate": 2.0110311091246333e-06,
+      "loss": 0.0443,
+      "step": 3995
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 8.775320024586614,
+      "learning_rate": 2.0087147203565614e-06,
+      "loss": 0.0558,
+      "step": 3996
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.638711681258063,
+      "learning_rate": 2.0063993310325013e-06,
+      "loss": 0.0591,
+      "step": 3997
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 11.201379197569311,
+      "learning_rate": 2.0040849419260682e-06,
+      "loss": 0.0336,
+      "step": 3998
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 8.662046945406965,
+      "learning_rate": 2.0017715538105416e-06,
+      "loss": 0.05,
+      "step": 3999
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 22.627187773041392,
+      "learning_rate": 1.9994591674588677e-06,
+      "loss": 0.0756,
+      "step": 4000
+    },
+    {
+      "epoch": 2.86,
+      "eval_avg_AUC": 0.8313536271084199,
+      "eval_avg_Accuracy": 0.7335875331564987,
+      "eval_avg_Accuracy-right": 0.8922655536715794,
+      "eval_avg_Accuracy-wrong": 0.4569024334773709,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.7065878301129395,
+      "eval_last_AUC": 0.8360746856019163,
+      "eval_last_Accuracy": 0.7791777188328912,
+      "eval_last_Accuracy-right": 0.8478544411112561,
+      "eval_last_Accuracy-wrong": 0.6594268819649761,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.7097758926367477,
+      "eval_max_AUC": 0.779076192285081,
+      "eval_max_Accuracy": 0.6495358090185677,
+      "eval_max_Accuracy-right": 0.9808921351245597,
+      "eval_max_Accuracy-wrong": 0.07175346827382306,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.6517312365875962,
+      "eval_min_AUC": 0.8398533477848513,
+      "eval_min_Accuracy": 0.7703498010610079,
+      "eval_min_Accuracy-right": 0.7896178427024912,
+      "eval_min_Accuracy-wrong": 0.7367523311348647,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.7123435959578684,
+      "eval_prod_AUC": 0.834872802991024,
+      "eval_prod_Accuracy": 0.7446949602122016,
+      "eval_prod_Accuracy-right": 0.6746445806704057,
+      "eval_prod_Accuracy-wrong": 0.8668410279736184,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.707421683014454,
+      "eval_runtime": 248.5729,
+      "eval_samples_per_second": 97.066,
+      "eval_steps_per_second": 3.033,
+      "eval_sum_AUC": 0.7054043711817337,
+      "eval_sum_Accuracy": 0.6395059681697612,
+      "eval_sum_Accuracy-right": 0.9969349158732229,
+      "eval_sum_Accuracy-wrong": 0.016261087104844214,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.682774019703003,
+      "step": 4000
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 11.448506614643406,
+      "learning_rate": 1.9971477836436575e-06,
+      "loss": 0.0907,
+      "step": 4001
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 17.106042357096115,
+      "learning_rate": 1.99483740313719e-06,
+      "loss": 0.046,
+      "step": 4002
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 15.306913920279596,
+      "learning_rate": 1.9925280267114e-06,
+      "loss": 0.0353,
+      "step": 4003
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.772152424590818,
+      "learning_rate": 1.9902196551379006e-06,
+      "loss": 0.0455,
+      "step": 4004
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 15.68768042844194,
+      "learning_rate": 1.987912289187954e-06,
+      "loss": 0.0629,
+      "step": 4005
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 8.371248909494858,
+      "learning_rate": 1.9856059296325027e-06,
+      "loss": 0.0479,
+      "step": 4006
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 6.182577891454594,
+      "learning_rate": 1.9833005772421354e-06,
+      "loss": 0.0281,
+      "step": 4007
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 10.517434691180428,
+      "learning_rate": 1.980996232787121e-06,
+      "loss": 0.044,
+      "step": 4008
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 12.669315091660465,
+      "learning_rate": 1.978692897037377e-06,
+      "loss": 0.0345,
+      "step": 4009
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 5.040911189426195,
+      "learning_rate": 1.9763905707624975e-06,
+      "loss": 0.0225,
+      "step": 4010
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 6.772416865182909,
+      "learning_rate": 1.974089254731727e-06,
+      "loss": 0.0538,
+      "step": 4011
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 6.111698799497366,
+      "learning_rate": 1.97178894971398e-06,
+      "loss": 0.0297,
+      "step": 4012
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 5.376778629697909,
+      "learning_rate": 1.9694896564778317e-06,
+      "loss": 0.0508,
+      "step": 4013
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 12.684429026836513,
+      "learning_rate": 1.9671913757915173e-06,
+      "loss": 0.0394,
+      "step": 4014
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 10.673329195030192,
+      "learning_rate": 1.964894108422936e-06,
+      "loss": 0.0573,
+      "step": 4015
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 9.962999370202702,
+      "learning_rate": 1.962597855139648e-06,
+      "loss": 0.0633,
+      "step": 4016
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.930744399839919,
+      "learning_rate": 1.960302616708873e-06,
+      "loss": 0.0561,
+      "step": 4017
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 2.99617559332668,
+      "learning_rate": 1.9580083938974937e-06,
+      "loss": 0.0344,
+      "step": 4018
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 6.613066828727661,
+      "learning_rate": 1.9557151874720526e-06,
+      "loss": 0.0646,
+      "step": 4019
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 14.130923755778449,
+      "learning_rate": 1.953422998198754e-06,
+      "loss": 0.0476,
+      "step": 4020
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.093142217656671,
+      "learning_rate": 1.9511318268434554e-06,
+      "loss": 0.0422,
+      "step": 4021
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.700589187453384,
+      "learning_rate": 1.9488416741716877e-06,
+      "loss": 0.0416,
+      "step": 4022
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 11.290562714679753,
+      "learning_rate": 1.946552540948625e-06,
+      "loss": 0.0374,
+      "step": 4023
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 11.076994659685992,
+      "learning_rate": 1.944264427939118e-06,
+      "loss": 0.064,
+      "step": 4024
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 5.073760434659358,
+      "learning_rate": 1.941977335907659e-06,
+      "loss": 0.034,
+      "step": 4025
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 12.555237380180236,
+      "learning_rate": 1.939691265618417e-06,
+      "loss": 0.0356,
+      "step": 4026
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 13.260024757677835,
+      "learning_rate": 1.9374062178352036e-06,
+      "loss": 0.043,
+      "step": 4027
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 8.049256823016284,
+      "learning_rate": 1.935122193321499e-06,
+      "loss": 0.0456,
+      "step": 4028
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.427445039334935,
+      "learning_rate": 1.932839192840436e-06,
+      "loss": 0.0403,
+      "step": 4029
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 11.153086971868019,
+      "learning_rate": 1.930557217154809e-06,
+      "loss": 0.0361,
+      "step": 4030
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 6.049221219619234,
+      "learning_rate": 1.9282762670270693e-06,
+      "loss": 0.0492,
+      "step": 4031
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 7.07032061230794,
+      "learning_rate": 1.925996343219323e-06,
+      "loss": 0.0557,
+      "step": 4032
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 11.430527196674731,
+      "learning_rate": 1.923717446493336e-06,
+      "loss": 0.054,
+      "step": 4033
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 16.819939441212135,
+      "learning_rate": 1.9214395776105297e-06,
+      "loss": 0.0526,
+      "step": 4034
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 12.959453782895224,
+      "learning_rate": 1.919162737331983e-06,
+      "loss": 0.047,
+      "step": 4035
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 4.702665094500624,
+      "learning_rate": 1.9168869264184296e-06,
+      "loss": 0.0423,
+      "step": 4036
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 8.969495139042449,
+      "learning_rate": 1.9146121456302613e-06,
+      "loss": 0.038,
+      "step": 4037
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 5.982997771608221,
+      "learning_rate": 1.9123383957275237e-06,
+      "loss": 0.0415,
+      "step": 4038
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 21.08738392970205,
+      "learning_rate": 1.91006567746992e-06,
+      "loss": 0.0711,
+      "step": 4039
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 6.035083987282505,
+      "learning_rate": 1.907793991616806e-06,
+      "loss": 0.0494,
+      "step": 4040
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 8.001816643118843,
+      "learning_rate": 1.9055233389271955e-06,
+      "loss": 0.0515,
+      "step": 4041
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.808783157897872,
+      "learning_rate": 1.9032537201597556e-06,
+      "loss": 0.0374,
+      "step": 4042
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 7.463408759803654,
+      "learning_rate": 1.9009851360728077e-06,
+      "loss": 0.0522,
+      "step": 4043
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 7.169003885468974,
+      "learning_rate": 1.898717587424328e-06,
+      "loss": 0.0566,
+      "step": 4044
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 6.208200904132249,
+      "learning_rate": 1.8964510749719484e-06,
+      "loss": 0.0567,
+      "step": 4045
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 16.35211171549844,
+      "learning_rate": 1.8941855994729497e-06,
+      "loss": 0.0486,
+      "step": 4046
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.281197291589123,
+      "learning_rate": 1.8919211616842703e-06,
+      "loss": 0.0331,
+      "step": 4047
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.747918718572736,
+      "learning_rate": 1.8896577623625017e-06,
+      "loss": 0.037,
+      "step": 4048
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 12.046638045605734,
+      "learning_rate": 1.887395402263888e-06,
+      "loss": 0.0693,
+      "step": 4049
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 27.115553942502387,
+      "learning_rate": 1.8851340821443248e-06,
+      "loss": 0.0374,
+      "step": 4050
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 18.92193070578465,
+      "learning_rate": 1.882873802759362e-06,
+      "loss": 0.0794,
+      "step": 4051
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.295273630927773,
+      "learning_rate": 1.8806145648642005e-06,
+      "loss": 0.04,
+      "step": 4052
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 3.412350246216868,
+      "learning_rate": 1.8783563692136936e-06,
+      "loss": 0.0385,
+      "step": 4053
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 17.67225398516106,
+      "learning_rate": 1.8760992165623465e-06,
+      "loss": 0.0724,
+      "step": 4054
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 8.36568199802075,
+      "learning_rate": 1.873843107664316e-06,
+      "loss": 0.0547,
+      "step": 4055
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 9.290828968139262,
+      "learning_rate": 1.87158804327341e-06,
+      "loss": 0.0535,
+      "step": 4056
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 2.802306481599526,
+      "learning_rate": 1.8693340241430874e-06,
+      "loss": 0.0308,
+      "step": 4057
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 10.43756004493309,
+      "learning_rate": 1.867081051026458e-06,
+      "loss": 0.057,
+      "step": 4058
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 6.681088881904337,
+      "learning_rate": 1.8648291246762818e-06,
+      "loss": 0.0321,
+      "step": 4059
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 8.252176413766616,
+      "learning_rate": 1.8625782458449693e-06,
+      "loss": 0.0511,
+      "step": 4060
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 8.406152709881708,
+      "learning_rate": 1.860328415284583e-06,
+      "loss": 0.0521,
+      "step": 4061
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 2.896995014403766,
+      "learning_rate": 1.8580796337468276e-06,
+      "loss": 0.0422,
+      "step": 4062
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 3.8791001120200477,
+      "learning_rate": 1.8558319019830695e-06,
+      "loss": 0.0373,
+      "step": 4063
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 5.53865330305396,
+      "learning_rate": 1.853585220744311e-06,
+      "loss": 0.0411,
+      "step": 4064
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 12.29757774570903,
+      "learning_rate": 1.851339590781217e-06,
+      "loss": 0.0444,
+      "step": 4065
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 17.39424270516312,
+      "learning_rate": 1.8490950128440877e-06,
+      "loss": 0.0834,
+      "step": 4066
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 14.625296601227284,
+      "learning_rate": 1.8468514876828847e-06,
+      "loss": 0.0585,
+      "step": 4067
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 18.641308415923643,
+      "learning_rate": 1.844609016047204e-06,
+      "loss": 0.0481,
+      "step": 4068
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 9.87551268358642,
+      "learning_rate": 1.8423675986863054e-06,
+      "loss": 0.0361,
+      "step": 4069
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 3.7267052490862635,
+      "learning_rate": 1.8401272363490818e-06,
+      "loss": 0.0443,
+      "step": 4070
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 10.562034474063116,
+      "learning_rate": 1.8378879297840818e-06,
+      "loss": 0.0477,
+      "step": 4071
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 5.802811004992916,
+      "learning_rate": 1.8356496797395002e-06,
+      "loss": 0.0434,
+      "step": 4072
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 6.436346892902528,
+      "learning_rate": 1.8334124869631765e-06,
+      "loss": 0.0345,
+      "step": 4073
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 5.010512163189531,
+      "learning_rate": 1.8311763522025994e-06,
+      "loss": 0.0364,
+      "step": 4074
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 7.935289328219277,
+      "learning_rate": 1.828941276204903e-06,
+      "loss": 0.0379,
+      "step": 4075
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 5.923070187953769,
+      "learning_rate": 1.8267072597168673e-06,
+      "loss": 0.0304,
+      "step": 4076
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 3.8603791640597245,
+      "learning_rate": 1.8244743034849193e-06,
+      "loss": 0.0446,
+      "step": 4077
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 3.915093269488006,
+      "learning_rate": 1.8222424082551303e-06,
+      "loss": 0.0691,
+      "step": 4078
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 5.595291860210755,
+      "learning_rate": 1.820011574773221e-06,
+      "loss": 0.0426,
+      "step": 4079
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 6.802755351874992,
+      "learning_rate": 1.8177818037845485e-06,
+      "loss": 0.0476,
+      "step": 4080
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 3.6813348946660613,
+      "learning_rate": 1.8155530960341273e-06,
+      "loss": 0.0359,
+      "step": 4081
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 18.71702922304564,
+      "learning_rate": 1.8133254522666033e-06,
+      "loss": 0.0569,
+      "step": 4082
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 7.6921438088763985,
+      "learning_rate": 1.8110988732262808e-06,
+      "loss": 0.0419,
+      "step": 4083
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 10.044378122355903,
+      "learning_rate": 1.8088733596570945e-06,
+      "loss": 0.0382,
+      "step": 4084
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 6.114486237395475,
+      "learning_rate": 1.806648912302636e-06,
+      "loss": 0.061,
+      "step": 4085
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 3.3399112721858772,
+      "learning_rate": 1.8044255319061287e-06,
+      "loss": 0.0387,
+      "step": 4086
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 12.412698383313344,
+      "learning_rate": 1.8022032192104517e-06,
+      "loss": 0.0405,
+      "step": 4087
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 30.82021867460914,
+      "learning_rate": 1.7999819749581154e-06,
+      "loss": 0.0749,
+      "step": 4088
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 11.480240630642303,
+      "learning_rate": 1.797761799891281e-06,
+      "loss": 0.0581,
+      "step": 4089
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 7.582474162510632,
+      "learning_rate": 1.7955426947517507e-06,
+      "loss": 0.059,
+      "step": 4090
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 10.63345521267209,
+      "learning_rate": 1.793324660280968e-06,
+      "loss": 0.0526,
+      "step": 4091
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 8.103239978569194,
+      "learning_rate": 1.7911076972200193e-06,
+      "loss": 0.0411,
+      "step": 4092
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 11.925157615703743,
+      "learning_rate": 1.7888918063096334e-06,
+      "loss": 0.0408,
+      "step": 4093
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 14.374780754417847,
+      "learning_rate": 1.7866769882901814e-06,
+      "loss": 0.0412,
+      "step": 4094
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 6.109178998467284,
+      "learning_rate": 1.784463243901674e-06,
+      "loss": 0.0528,
+      "step": 4095
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 9.609524368121408,
+      "learning_rate": 1.7822505738837648e-06,
+      "loss": 0.0651,
+      "step": 4096
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 5.040866179019186,
+      "learning_rate": 1.7800389789757483e-06,
+      "loss": 0.0445,
+      "step": 4097
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 5.714985544615724,
+      "learning_rate": 1.7778284599165597e-06,
+      "loss": 0.0487,
+      "step": 4098
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 7.619378628354265,
+      "learning_rate": 1.7756190174447734e-06,
+      "loss": 0.0436,
+      "step": 4099
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 3.2474864738073133,
+      "learning_rate": 1.7734106522986061e-06,
+      "loss": 0.0462,
+      "step": 4100
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 4.3973092959361875,
+      "learning_rate": 1.7712033652159133e-06,
+      "loss": 0.0531,
+      "step": 4101
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 7.4959152699641765,
+      "learning_rate": 1.7689971569341907e-06,
+      "loss": 0.0576,
+      "step": 4102
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 3.749459966128013,
+      "learning_rate": 1.7667920281905738e-06,
+      "loss": 0.0277,
+      "step": 4103
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 7.533700501363081,
+      "learning_rate": 1.764587979721838e-06,
+      "loss": 0.0647,
+      "step": 4104
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 10.29526487200459,
+      "learning_rate": 1.7623850122643926e-06,
+      "loss": 0.0401,
+      "step": 4105
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 12.913824847457555,
+      "learning_rate": 1.7601831265542968e-06,
+      "loss": 0.0405,
+      "step": 4106
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 6.04341070181785,
+      "learning_rate": 1.7579823233272337e-06,
+      "loss": 0.0359,
+      "step": 4107
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 19.592526482210516,
+      "learning_rate": 1.7557826033185404e-06,
+      "loss": 0.054,
+      "step": 4108
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 5.540253325530484,
+      "learning_rate": 1.7535839672631772e-06,
+      "loss": 0.0508,
+      "step": 4109
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 7.452509099451598,
+      "learning_rate": 1.7513864158957556e-06,
+      "loss": 0.0364,
+      "step": 4110
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 20.761584265848015,
+      "learning_rate": 1.7491899499505122e-06,
+      "loss": 0.0438,
+      "step": 4111
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 13.58680016006121,
+      "learning_rate": 1.746994570161334e-06,
+      "loss": 0.0384,
+      "step": 4112
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.603506950318464,
+      "learning_rate": 1.7448002772617324e-06,
+      "loss": 0.0438,
+      "step": 4113
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 3.8154557106353284,
+      "learning_rate": 1.7426070719848632e-06,
+      "loss": 0.0257,
+      "step": 4114
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 5.747171185189265,
+      "learning_rate": 1.7404149550635173e-06,
+      "loss": 0.0524,
+      "step": 4115
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 7.9740227323975335,
+      "learning_rate": 1.7382239272301221e-06,
+      "loss": 0.077,
+      "step": 4116
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 10.989166783225153,
+      "learning_rate": 1.7360339892167404e-06,
+      "loss": 0.0374,
+      "step": 4117
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 6.7143857340659165,
+      "learning_rate": 1.7338451417550712e-06,
+      "loss": 0.0756,
+      "step": 4118
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 12.19837267726741,
+      "learning_rate": 1.7316573855764485e-06,
+      "loss": 0.092,
+      "step": 4119
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 7.153980131842924,
+      "learning_rate": 1.7294707214118434e-06,
+      "loss": 0.0359,
+      "step": 4120
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 7.588739529742861,
+      "learning_rate": 1.7272851499918603e-06,
+      "loss": 0.0444,
+      "step": 4121
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 4.8112730950323765,
+      "learning_rate": 1.725100672046741e-06,
+      "loss": 0.0451,
+      "step": 4122
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 7.260957042480349,
+      "learning_rate": 1.7229172883063556e-06,
+      "loss": 0.0417,
+      "step": 4123
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 7.1770527506341235,
+      "learning_rate": 1.7207349995002192e-06,
+      "loss": 0.0321,
+      "step": 4124
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 8.493330479432425,
+      "learning_rate": 1.7185538063574692e-06,
+      "loss": 0.0701,
+      "step": 4125
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 18.78832049114435,
+      "learning_rate": 1.7163737096068883e-06,
+      "loss": 0.0322,
+      "step": 4126
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 3.9921302247739954,
+      "learning_rate": 1.7141947099768818e-06,
+      "loss": 0.0453,
+      "step": 4127
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 10.454142818755402,
+      "learning_rate": 1.7120168081955001e-06,
+      "loss": 0.0321,
+      "step": 4128
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 7.925392060998192,
+      "learning_rate": 1.7098400049904163e-06,
+      "loss": 0.0514,
+      "step": 4129
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 6.926450108002052,
+      "learning_rate": 1.707664301088941e-06,
+      "loss": 0.0338,
+      "step": 4130
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 3.1024288350216556,
+      "learning_rate": 1.705489697218019e-06,
+      "loss": 0.0466,
+      "step": 4131
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 9.17160382357441,
+      "learning_rate": 1.7033161941042248e-06,
+      "loss": 0.0503,
+      "step": 4132
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 9.016353852398503,
+      "learning_rate": 1.7011437924737666e-06,
+      "loss": 0.0489,
+      "step": 4133
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 9.854573725493205,
+      "learning_rate": 1.6989724930524843e-06,
+      "loss": 0.0579,
+      "step": 4134
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 6.671723566772397,
+      "learning_rate": 1.6968022965658492e-06,
+      "loss": 0.0335,
+      "step": 4135
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 2.720192606101922,
+      "learning_rate": 1.694633203738964e-06,
+      "loss": 0.0402,
+      "step": 4136
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 5.995599861484541,
+      "learning_rate": 1.6924652152965632e-06,
+      "loss": 0.0549,
+      "step": 4137
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 12.266299701613004,
+      "learning_rate": 1.690298331963014e-06,
+      "loss": 0.0572,
+      "step": 4138
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 3.4813446495338654,
+      "learning_rate": 1.6881325544623067e-06,
+      "loss": 0.0452,
+      "step": 4139
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 5.891263821479523,
+      "learning_rate": 1.6859678835180749e-06,
+      "loss": 0.0515,
+      "step": 4140
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 3.6553475031782474,
+      "learning_rate": 1.6838043198535693e-06,
+      "loss": 0.0418,
+      "step": 4141
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 5.408411863091127,
+      "learning_rate": 1.681641864191682e-06,
+      "loss": 0.0557,
+      "step": 4142
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 5.8114627598252895,
+      "learning_rate": 1.6794805172549244e-06,
+      "loss": 0.0398,
+      "step": 4143
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 7.53548148027711,
+      "learning_rate": 1.6773202797654486e-06,
+      "loss": 0.061,
+      "step": 4144
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 5.913855035387853,
+      "learning_rate": 1.6751611524450235e-06,
+      "loss": 0.0335,
+      "step": 4145
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 2.024414954794077,
+      "learning_rate": 1.6730031360150605e-06,
+      "loss": 0.0341,
+      "step": 4146
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 9.593789286345464,
+      "learning_rate": 1.670846231196588e-06,
+      "loss": 0.0404,
+      "step": 4147
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.308719832906545,
+      "learning_rate": 1.6686904387102692e-06,
+      "loss": 0.0405,
+      "step": 4148
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.491244505461696,
+      "learning_rate": 1.6665357592763948e-06,
+      "loss": 0.0345,
+      "step": 4149
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 14.320074989143352,
+      "learning_rate": 1.6643821936148834e-06,
+      "loss": 0.0335,
+      "step": 4150
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.632814125620873,
+      "learning_rate": 1.6622297424452817e-06,
+      "loss": 0.0333,
+      "step": 4151
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 8.547410451451574,
+      "learning_rate": 1.6600784064867625e-06,
+      "loss": 0.049,
+      "step": 4152
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 11.406023385749469,
+      "learning_rate": 1.6579281864581275e-06,
+      "loss": 0.0543,
+      "step": 4153
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 6.388068135398717,
+      "learning_rate": 1.6557790830778058e-06,
+      "loss": 0.0488,
+      "step": 4154
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.4616777204213145,
+      "learning_rate": 1.6536310970638525e-06,
+      "loss": 0.0606,
+      "step": 4155
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 20.563153827350366,
+      "learning_rate": 1.6514842291339494e-06,
+      "loss": 0.0524,
+      "step": 4156
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 10.870505302773477,
+      "learning_rate": 1.6493384800054052e-06,
+      "loss": 0.0561,
+      "step": 4157
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 9.278625304721448,
+      "learning_rate": 1.6471938503951546e-06,
+      "loss": 0.0536,
+      "step": 4158
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.924475025051709,
+      "learning_rate": 1.6450503410197582e-06,
+      "loss": 0.0427,
+      "step": 4159
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 14.546159414440977,
+      "learning_rate": 1.6429079525954023e-06,
+      "loss": 0.0796,
+      "step": 4160
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 3.274576485929536,
+      "learning_rate": 1.6407666858378985e-06,
+      "loss": 0.0368,
+      "step": 4161
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.750934336503444,
+      "learning_rate": 1.6386265414626834e-06,
+      "loss": 0.0559,
+      "step": 4162
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 3.540640136850486,
+      "learning_rate": 1.636487520184822e-06,
+      "loss": 0.0367,
+      "step": 4163
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 4.692046580807771,
+      "learning_rate": 1.6343496227189948e-06,
+      "loss": 0.0359,
+      "step": 4164
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 10.53276192290032,
+      "learning_rate": 1.632212849779521e-06,
+      "loss": 0.0316,
+      "step": 4165
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 5.302753116947668,
+      "learning_rate": 1.630077202080328e-06,
+      "loss": 0.0267,
+      "step": 4166
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 5.259703313914846,
+      "learning_rate": 1.6279426803349828e-06,
+      "loss": 0.0547,
+      "step": 4167
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.082893040513792,
+      "learning_rate": 1.6258092852566625e-06,
+      "loss": 0.0217,
+      "step": 4168
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 8.23019416830836,
+      "learning_rate": 1.6236770175581807e-06,
+      "loss": 0.0509,
+      "step": 4169
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 12.438101804745394,
+      "learning_rate": 1.62154587795196e-06,
+      "loss": 0.055,
+      "step": 4170
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 5.427420456057769,
+      "learning_rate": 1.6194158671500616e-06,
+      "loss": 0.0518,
+      "step": 4171
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 10.415775308376247,
+      "learning_rate": 1.6172869858641554e-06,
+      "loss": 0.0432,
+      "step": 4172
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 8.480843745125458,
+      "learning_rate": 1.6151592348055433e-06,
+      "loss": 0.0427,
+      "step": 4173
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 4.617772973040294,
+      "learning_rate": 1.6130326146851455e-06,
+      "loss": 0.0438,
+      "step": 4174
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 5.919133113007816,
+      "learning_rate": 1.6109071262135056e-06,
+      "loss": 0.0423,
+      "step": 4175
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 3.689039215465472,
+      "learning_rate": 1.608782770100789e-06,
+      "loss": 0.0372,
+      "step": 4176
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 14.067237284232494,
+      "learning_rate": 1.6066595470567825e-06,
+      "loss": 0.0435,
+      "step": 4177
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 5.854433151920558,
+      "learning_rate": 1.6045374577908944e-06,
+      "loss": 0.0504,
+      "step": 4178
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 11.6646480285659,
+      "learning_rate": 1.6024165030121542e-06,
+      "loss": 0.0554,
+      "step": 4179
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 6.011438841677108,
+      "learning_rate": 1.6002966834292116e-06,
+      "loss": 0.0504,
+      "step": 4180
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 6.227447353492944,
+      "learning_rate": 1.5981779997503405e-06,
+      "loss": 0.0476,
+      "step": 4181
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 10.349021587684007,
+      "learning_rate": 1.5960604526834266e-06,
+      "loss": 0.077,
+      "step": 4182
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 6.211039507384463,
+      "learning_rate": 1.5939440429359888e-06,
+      "loss": 0.0554,
+      "step": 4183
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 2.6513931652813705,
+      "learning_rate": 1.591828771215152e-06,
+      "loss": 0.0348,
+      "step": 4184
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 9.179016643100248,
+      "learning_rate": 1.5897146382276752e-06,
+      "loss": 0.0577,
+      "step": 4185
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 7.051485770699399,
+      "learning_rate": 1.587601644679922e-06,
+      "loss": 0.0298,
+      "step": 4186
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 7.952697092704008,
+      "learning_rate": 1.58548979127789e-06,
+      "loss": 0.0339,
+      "step": 4187
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 3.6748750294465,
+      "learning_rate": 1.5833790787271819e-06,
+      "loss": 0.0346,
+      "step": 4188
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 5.611862793582695,
+      "learning_rate": 1.5812695077330325e-06,
+      "loss": 0.056,
+      "step": 4189
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 11.09492276918743,
+      "learning_rate": 1.5791610790002838e-06,
+      "loss": 0.0451,
+      "step": 4190
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 8.006779281484823,
+      "learning_rate": 1.577053793233403e-06,
+      "loss": 0.049,
+      "step": 4191
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 10.760492669788933,
+      "learning_rate": 1.5749476511364726e-06,
+      "loss": 0.0446,
+      "step": 4192
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 10.776948111144355,
+      "learning_rate": 1.5728426534131946e-06,
+      "loss": 0.0305,
+      "step": 4193
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 4.0188505044934155,
+      "learning_rate": 1.5707388007668877e-06,
+      "loss": 0.0481,
+      "step": 4194
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 8.979036207018861,
+      "learning_rate": 1.568636093900488e-06,
+      "loss": 0.0398,
+      "step": 4195
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 6.689484772564502,
+      "learning_rate": 1.5665345335165488e-06,
+      "loss": 0.0412,
+      "step": 4196
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.5182528489425104,
+      "learning_rate": 1.5644341203172415e-06,
+      "loss": 0.0522,
+      "step": 4197
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 7.494580739110864,
+      "learning_rate": 1.5623348550043516e-06,
+      "loss": 0.0384,
+      "step": 4198
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 6.568879027287965,
+      "learning_rate": 1.5602367382792839e-06,
+      "loss": 0.0331,
+      "step": 4199
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 14.033675146688019,
+      "learning_rate": 1.5581397708430578e-06,
+      "loss": 0.0726,
+      "step": 4200
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 4.394766382636879,
+      "learning_rate": 1.556043953396309e-06,
+      "loss": 0.0311,
+      "step": 4201
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 11.599669876581697,
+      "learning_rate": 1.5539492866392891e-06,
+      "loss": 0.0349,
+      "step": 4202
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 5.095777532033208,
+      "learning_rate": 1.551855771271865e-06,
+      "loss": 0.0338,
+      "step": 4203
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 8.662438814706617,
+      "learning_rate": 1.5497634079935198e-06,
+      "loss": 0.0439,
+      "step": 4204
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.5906055347775252,
+      "learning_rate": 1.5476721975033498e-06,
+      "loss": 0.0178,
+      "step": 4205
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 3.918653409467006,
+      "learning_rate": 1.5455821405000703e-06,
+      "loss": 0.0265,
+      "step": 4206
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 5.99063662385307,
+      "learning_rate": 1.5434932376820039e-06,
+      "loss": 0.034,
+      "step": 4207
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 5.838954154742759,
+      "learning_rate": 1.5414054897470942e-06,
+      "loss": 0.0314,
+      "step": 4208
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.5387521453912725,
+      "learning_rate": 1.5393188973928957e-06,
+      "loss": 0.0201,
+      "step": 4209
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 3.3458673133915355,
+      "learning_rate": 1.5372334613165784e-06,
+      "loss": 0.0305,
+      "step": 4210
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 5.3408721579010185,
+      "learning_rate": 1.5351491822149255e-06,
+      "loss": 0.0208,
+      "step": 4211
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 5.0928277444132934,
+      "learning_rate": 1.533066060784333e-06,
+      "loss": 0.0292,
+      "step": 4212
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 4.046163091652899,
+      "learning_rate": 1.5309840977208096e-06,
+      "loss": 0.0229,
+      "step": 4213
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 2.3567429229007706,
+      "learning_rate": 1.5289032937199793e-06,
+      "loss": 0.0251,
+      "step": 4214
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 4.662301159431529,
+      "learning_rate": 1.5268236494770772e-06,
+      "loss": 0.0248,
+      "step": 4215
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 5.949991688960428,
+      "learning_rate": 1.5247451656869499e-06,
+      "loss": 0.0298,
+      "step": 4216
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 4.549139241488894,
+      "learning_rate": 1.5226678430440588e-06,
+      "loss": 0.0263,
+      "step": 4217
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 3.180555537523223,
+      "learning_rate": 1.5205916822424755e-06,
+      "loss": 0.0239,
+      "step": 4218
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 3.443474683242361,
+      "learning_rate": 1.5185166839758836e-06,
+      "loss": 0.0295,
+      "step": 4219
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 1.5504955901675597,
+      "learning_rate": 1.5164428489375789e-06,
+      "loss": 0.0229,
+      "step": 4220
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 2.295258883006284,
+      "learning_rate": 1.5143701778204683e-06,
+      "loss": 0.0263,
+      "step": 4221
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 3.551031768131113,
+      "learning_rate": 1.5122986713170712e-06,
+      "loss": 0.0274,
+      "step": 4222
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 10.191936949992497,
+      "learning_rate": 1.510228330119512e-06,
+      "loss": 0.0294,
+      "step": 4223
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 6.993474554478619,
+      "learning_rate": 1.5081591549195357e-06,
+      "loss": 0.0232,
+      "step": 4224
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 9.562045171559191,
+      "learning_rate": 1.5060911464084864e-06,
+      "loss": 0.0357,
+      "step": 4225
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 3.0924209278860673,
+      "learning_rate": 1.5040243052773312e-06,
+      "loss": 0.0326,
+      "step": 4226
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 4.420905003070142,
+      "learning_rate": 1.5019586322166323e-06,
+      "loss": 0.0173,
+      "step": 4227
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 4.811580679582127,
+      "learning_rate": 1.4998941279165773e-06,
+      "loss": 0.0158,
+      "step": 4228
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 9.851045271456961,
+      "learning_rate": 1.4978307930669483e-06,
+      "loss": 0.024,
+      "step": 4229
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 9.584372906933657,
+      "learning_rate": 1.4957686283571498e-06,
+      "loss": 0.0256,
+      "step": 4230
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 6.770213359448928,
+      "learning_rate": 1.4937076344761858e-06,
+      "loss": 0.0271,
+      "step": 4231
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 3.7727215155076355,
+      "learning_rate": 1.4916478121126732e-06,
+      "loss": 0.0224,
+      "step": 4232
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 1.893841240458152,
+      "learning_rate": 1.4895891619548374e-06,
+      "loss": 0.0232,
+      "step": 4233
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 5.161001637713427,
+      "learning_rate": 1.4875316846905113e-06,
+      "loss": 0.023,
+      "step": 4234
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 11.061852811785181,
+      "learning_rate": 1.4854753810071364e-06,
+      "loss": 0.028,
+      "step": 4235
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 8.109821052482838,
+      "learning_rate": 1.4834202515917628e-06,
+      "loss": 0.0312,
+      "step": 4236
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 11.014754015899051,
+      "learning_rate": 1.4813662971310465e-06,
+      "loss": 0.032,
+      "step": 4237
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 2.2753130698707484,
+      "learning_rate": 1.4793135183112523e-06,
+      "loss": 0.027,
+      "step": 4238
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 4.690230867022709,
+      "learning_rate": 1.477261915818251e-06,
+      "loss": 0.0206,
+      "step": 4239
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 5.281295741354212,
+      "learning_rate": 1.4752114903375243e-06,
+      "loss": 0.0131,
+      "step": 4240
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 4.3490688386339516,
+      "learning_rate": 1.473162242554151e-06,
+      "loss": 0.0299,
+      "step": 4241
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 8.960682672351583,
+      "learning_rate": 1.47111417315283e-06,
+      "loss": 0.0282,
+      "step": 4242
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 8.66598782769347,
+      "learning_rate": 1.4690672828178532e-06,
+      "loss": 0.0362,
+      "step": 4243
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 5.361898466333456,
+      "learning_rate": 1.467021572233131e-06,
+      "loss": 0.0323,
+      "step": 4244
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 3.731815327650255,
+      "learning_rate": 1.4649770420821663e-06,
+      "loss": 0.0297,
+      "step": 4245
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 5.073322490267396,
+      "learning_rate": 1.4629336930480813e-06,
+      "loss": 0.0276,
+      "step": 4246
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 2.1603771617345915,
+      "learning_rate": 1.4608915258135914e-06,
+      "loss": 0.0354,
+      "step": 4247
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 6.080643211063934,
+      "learning_rate": 1.4588505410610283e-06,
+      "loss": 0.0316,
+      "step": 4248
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 4.017000843022467,
+      "learning_rate": 1.4568107394723175e-06,
+      "loss": 0.0245,
+      "step": 4249
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 2.6670606172944034,
+      "learning_rate": 1.4547721217289972e-06,
+      "loss": 0.0292,
+      "step": 4250
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 8.81886940865451,
+      "learning_rate": 1.4527346885122073e-06,
+      "loss": 0.0335,
+      "step": 4251
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 6.820384061010361,
+      "learning_rate": 1.450698440502692e-06,
+      "loss": 0.022,
+      "step": 4252
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 3.2055510470449873,
+      "learning_rate": 1.4486633783807997e-06,
+      "loss": 0.0227,
+      "step": 4253
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 6.073201859461071,
+      "learning_rate": 1.4466295028264822e-06,
+      "loss": 0.0244,
+      "step": 4254
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 2.993722905445915,
+      "learning_rate": 1.4445968145192951e-06,
+      "loss": 0.023,
+      "step": 4255
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 2.6344293218691632,
+      "learning_rate": 1.4425653141383977e-06,
+      "loss": 0.0262,
+      "step": 4256
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 3.3473254018552536,
+      "learning_rate": 1.4405350023625514e-06,
+      "loss": 0.0179,
+      "step": 4257
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 2.0005642649862208,
+      "learning_rate": 1.4385058798701223e-06,
+      "loss": 0.0231,
+      "step": 4258
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 2.6183840757874544,
+      "learning_rate": 1.4364779473390767e-06,
+      "loss": 0.0266,
+      "step": 4259
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 3.7496408204859124,
+      "learning_rate": 1.4344512054469855e-06,
+      "loss": 0.0322,
+      "step": 4260
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 5.046695349286583,
+      "learning_rate": 1.4324256548710202e-06,
+      "loss": 0.0212,
+      "step": 4261
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 4.538256274604029,
+      "learning_rate": 1.430401296287955e-06,
+      "loss": 0.0237,
+      "step": 4262
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 3.4393732090188815,
+      "learning_rate": 1.4283781303741662e-06,
+      "loss": 0.0169,
+      "step": 4263
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 4.128065352323907,
+      "learning_rate": 1.4263561578056307e-06,
+      "loss": 0.0239,
+      "step": 4264
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 3.7030098567807346,
+      "learning_rate": 1.4243353792579285e-06,
+      "loss": 0.028,
+      "step": 4265
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 2.497292358728816,
+      "learning_rate": 1.4223157954062344e-06,
+      "loss": 0.0199,
+      "step": 4266
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 6.153106787419073,
+      "learning_rate": 1.4202974069253362e-06,
+      "loss": 0.0301,
+      "step": 4267
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 1.5508727812000602,
+      "learning_rate": 1.418280214489608e-06,
+      "loss": 0.0194,
+      "step": 4268
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 3.9753775875375252,
+      "learning_rate": 1.416264218773038e-06,
+      "loss": 0.0207,
+      "step": 4269
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 4.719703036216151,
+      "learning_rate": 1.4142494204492007e-06,
+      "loss": 0.0307,
+      "step": 4270
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 1.8739011534547396,
+      "learning_rate": 1.412235820191285e-06,
+      "loss": 0.014,
+      "step": 4271
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 1.6215340748185867,
+      "learning_rate": 1.4102234186720653e-06,
+      "loss": 0.0198,
+      "step": 4272
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 1.1853722180193103,
+      "learning_rate": 1.4082122165639285e-06,
+      "loss": 0.0182,
+      "step": 4273
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 1.4639779224320038,
+      "learning_rate": 1.4062022145388503e-06,
+      "loss": 0.0155,
+      "step": 4274
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 3.0442440975692535,
+      "learning_rate": 1.4041934132684116e-06,
+      "loss": 0.034,
+      "step": 4275
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 1.485809174565446,
+      "learning_rate": 1.4021858134237892e-06,
+      "loss": 0.0212,
+      "step": 4276
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 2.050713547548853,
+      "learning_rate": 1.4001794156757598e-06,
+      "loss": 0.0276,
+      "step": 4277
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 3.065013650984198,
+      "learning_rate": 1.398174220694699e-06,
+      "loss": 0.0283,
+      "step": 4278
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 3.4201497917182877,
+      "learning_rate": 1.3961702291505791e-06,
+      "loss": 0.0275,
+      "step": 4279
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 2.7541672840549887,
+      "learning_rate": 1.3941674417129714e-06,
+      "loss": 0.023,
+      "step": 4280
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 2.246885890731216,
+      "learning_rate": 1.3921658590510434e-06,
+      "loss": 0.0336,
+      "step": 4281
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 3.022641932858734,
+      "learning_rate": 1.3901654818335618e-06,
+      "loss": 0.0173,
+      "step": 4282
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 6.8132174445899665,
+      "learning_rate": 1.3881663107288918e-06,
+      "loss": 0.0199,
+      "step": 4283
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 1.867430595204692,
+      "learning_rate": 1.386168346404988e-06,
+      "loss": 0.0254,
+      "step": 4284
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 3.233486849230603,
+      "learning_rate": 1.3841715895294138e-06,
+      "loss": 0.0185,
+      "step": 4285
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 4.830562376779117,
+      "learning_rate": 1.3821760407693175e-06,
+      "loss": 0.0284,
+      "step": 4286
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 1.5354005181744397,
+      "learning_rate": 1.3801817007914543e-06,
+      "loss": 0.0244,
+      "step": 4287
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 3.8781541269057374,
+      "learning_rate": 1.3781885702621644e-06,
+      "loss": 0.0317,
+      "step": 4288
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 3.305553379082568,
+      "learning_rate": 1.3761966498473956e-06,
+      "loss": 0.017,
+      "step": 4289
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 5.335432721094165,
+      "learning_rate": 1.3742059402126818e-06,
+      "loss": 0.0297,
+      "step": 4290
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 3.030662176498976,
+      "learning_rate": 1.3722164420231565e-06,
+      "loss": 0.0234,
+      "step": 4291
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 2.2382330008234748,
+      "learning_rate": 1.370228155943548e-06,
+      "loss": 0.0209,
+      "step": 4292
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 2.1132694656038105,
+      "learning_rate": 1.3682410826381816e-06,
+      "loss": 0.0181,
+      "step": 4293
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 2.882412146996378,
+      "learning_rate": 1.366255222770973e-06,
+      "loss": 0.0223,
+      "step": 4294
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 4.0879767931219435,
+      "learning_rate": 1.364270577005436e-06,
+      "loss": 0.0446,
+      "step": 4295
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 3.898872989029239,
+      "learning_rate": 1.3622871460046778e-06,
+      "loss": 0.0242,
+      "step": 4296
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 2.6033162498479463,
+      "learning_rate": 1.3603049304313992e-06,
+      "loss": 0.0349,
+      "step": 4297
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 3.5165832344769536,
+      "learning_rate": 1.3583239309478953e-06,
+      "loss": 0.0233,
+      "step": 4298
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 6.573066298874758,
+      "learning_rate": 1.3563441482160562e-06,
+      "loss": 0.0214,
+      "step": 4299
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 2.235003481206468,
+      "learning_rate": 1.35436558289736e-06,
+      "loss": 0.0264,
+      "step": 4300
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 5.948089488451964,
+      "learning_rate": 1.3523882356528883e-06,
+      "loss": 0.0163,
+      "step": 4301
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 7.51398793823194,
+      "learning_rate": 1.350412107143303e-06,
+      "loss": 0.031,
+      "step": 4302
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 2.0676717040990757,
+      "learning_rate": 1.3484371980288712e-06,
+      "loss": 0.0195,
+      "step": 4303
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 2.0443453458688623,
+      "learning_rate": 1.3464635089694416e-06,
+      "loss": 0.0204,
+      "step": 4304
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 2.4356846280920754,
+      "learning_rate": 1.344491040624466e-06,
+      "loss": 0.0255,
+      "step": 4305
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 4.725705710515471,
+      "learning_rate": 1.3425197936529766e-06,
+      "loss": 0.0226,
+      "step": 4306
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 4.276820785445097,
+      "learning_rate": 1.3405497687136098e-06,
+      "loss": 0.0295,
+      "step": 4307
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 1.4908882433698298,
+      "learning_rate": 1.3385809664645827e-06,
+      "loss": 0.0192,
+      "step": 4308
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 2.5380484947293613,
+      "learning_rate": 1.336613387563711e-06,
+      "loss": 0.0266,
+      "step": 4309
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 2.2945673276968894,
+      "learning_rate": 1.3346470326683986e-06,
+      "loss": 0.0289,
+      "step": 4310
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 2.4304115620560154,
+      "learning_rate": 1.3326819024356413e-06,
+      "loss": 0.0208,
+      "step": 4311
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 2.8002395851902047,
+      "learning_rate": 1.3307179975220264e-06,
+      "loss": 0.0295,
+      "step": 4312
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 2.275129136772214,
+      "learning_rate": 1.3287553185837298e-06,
+      "loss": 0.0194,
+      "step": 4313
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 7.423216970876992,
+      "learning_rate": 1.3267938662765206e-06,
+      "loss": 0.0204,
+      "step": 4314
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 2.6068477417666607,
+      "learning_rate": 1.324833641255755e-06,
+      "loss": 0.0209,
+      "step": 4315
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 2.474988243701278,
+      "learning_rate": 1.3228746441763813e-06,
+      "loss": 0.0143,
+      "step": 4316
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 2.441680844619571,
+      "learning_rate": 1.3209168756929363e-06,
+      "loss": 0.0305,
+      "step": 4317
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 7.651310891463135,
+      "learning_rate": 1.3189603364595483e-06,
+      "loss": 0.0259,
+      "step": 4318
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 1.8489347124819762,
+      "learning_rate": 1.3170050271299316e-06,
+      "loss": 0.0206,
+      "step": 4319
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 2.050853437133708,
+      "learning_rate": 1.315050948357392e-06,
+      "loss": 0.0201,
+      "step": 4320
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 3.8678126664888266,
+      "learning_rate": 1.3130981007948247e-06,
+      "loss": 0.0286,
+      "step": 4321
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 3.5527708798067263,
+      "learning_rate": 1.3111464850947103e-06,
+      "loss": 0.0284,
+      "step": 4322
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 6.799812210909939,
+      "learning_rate": 1.3091961019091216e-06,
+      "loss": 0.0285,
+      "step": 4323
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 3.718954096429571,
+      "learning_rate": 1.3072469518897184e-06,
+      "loss": 0.0201,
+      "step": 4324
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 7.454710537462165,
+      "learning_rate": 1.3052990356877444e-06,
+      "loss": 0.0204,
+      "step": 4325
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 5.810987162547036,
+      "learning_rate": 1.3033523539540394e-06,
+      "loss": 0.0218,
+      "step": 4326
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 2.6929964518246527,
+      "learning_rate": 1.3014069073390206e-06,
+      "loss": 0.0237,
+      "step": 4327
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 1.6508887017292961,
+      "learning_rate": 1.2994626964927042e-06,
+      "loss": 0.0233,
+      "step": 4328
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 1.694127839814189,
+      "learning_rate": 1.2975197220646807e-06,
+      "loss": 0.0146,
+      "step": 4329
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 8.55682892697071,
+      "learning_rate": 1.29557798470414e-06,
+      "loss": 0.0198,
+      "step": 4330
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 4.464695362177069,
+      "learning_rate": 1.293637485059847e-06,
+      "loss": 0.0292,
+      "step": 4331
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 1.3157782950293537,
+      "learning_rate": 1.291698223780164e-06,
+      "loss": 0.0184,
+      "step": 4332
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 2.6685124060299343,
+      "learning_rate": 1.2897602015130306e-06,
+      "loss": 0.0253,
+      "step": 4333
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 3.4823603776992815,
+      "learning_rate": 1.287823418905977e-06,
+      "loss": 0.0304,
+      "step": 4334
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 5.911506432130707,
+      "learning_rate": 1.2858878766061178e-06,
+      "loss": 0.0257,
+      "step": 4335
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 2.086775123511386,
+      "learning_rate": 1.2839535752601551e-06,
+      "loss": 0.0324,
+      "step": 4336
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 2.029837014664569,
+      "learning_rate": 1.2820205155143738e-06,
+      "loss": 0.0204,
+      "step": 4337
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 7.437040162360332,
+      "learning_rate": 1.2800886980146453e-06,
+      "loss": 0.0263,
+      "step": 4338
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 12.488964061218802,
+      "learning_rate": 1.2781581234064256e-06,
+      "loss": 0.033,
+      "step": 4339
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 5.81702543197951,
+      "learning_rate": 1.276228792334756e-06,
+      "loss": 0.0384,
+      "step": 4340
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 2.7625144524398144,
+      "learning_rate": 1.274300705444262e-06,
+      "loss": 0.0206,
+      "step": 4341
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 1.9870680625000081,
+      "learning_rate": 1.2723738633791538e-06,
+      "loss": 0.0257,
+      "step": 4342
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 1.9775795029648342,
+      "learning_rate": 1.2704482667832218e-06,
+      "loss": 0.0261,
+      "step": 4343
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 7.743239827097478,
+      "learning_rate": 1.2685239162998485e-06,
+      "loss": 0.026,
+      "step": 4344
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 3.6982912516993784,
+      "learning_rate": 1.2666008125719904e-06,
+      "loss": 0.0281,
+      "step": 4345
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 5.896769242939257,
+      "learning_rate": 1.2646789562421975e-06,
+      "loss": 0.0206,
+      "step": 4346
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 5.284872574990886,
+      "learning_rate": 1.2627583479525913e-06,
+      "loss": 0.0313,
+      "step": 4347
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 6.165105344259164,
+      "learning_rate": 1.2608389883448896e-06,
+      "loss": 0.0283,
+      "step": 4348
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 3.7245284863056223,
+      "learning_rate": 1.2589208780603795e-06,
+      "loss": 0.025,
+      "step": 4349
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 2.567851032035511,
+      "learning_rate": 1.2570040177399435e-06,
+      "loss": 0.0274,
+      "step": 4350
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 4.0847654157063555,
+      "learning_rate": 1.255088408024036e-06,
+      "loss": 0.0256,
+      "step": 4351
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 4.834804601853921,
+      "learning_rate": 1.2531740495526989e-06,
+      "loss": 0.0204,
+      "step": 4352
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 3.9081054310744254,
+      "learning_rate": 1.2512609429655553e-06,
+      "loss": 0.0241,
+      "step": 4353
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 3.942714359609263,
+      "learning_rate": 1.249349088901809e-06,
+      "loss": 0.0239,
+      "step": 4354
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 6.164893142313988,
+      "learning_rate": 1.247438488000247e-06,
+      "loss": 0.0227,
+      "step": 4355
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 6.8002871593043235,
+      "learning_rate": 1.245529140899236e-06,
+      "loss": 0.0285,
+      "step": 4356
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 14.128621503244876,
+      "learning_rate": 1.2436210482367245e-06,
+      "loss": 0.0278,
+      "step": 4357
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 4.892012988671743,
+      "learning_rate": 1.2417142106502418e-06,
+      "loss": 0.0212,
+      "step": 4358
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 6.258033264830363,
+      "learning_rate": 1.2398086287768969e-06,
+      "loss": 0.0276,
+      "step": 4359
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 2.533324048190474,
+      "learning_rate": 1.237904303253381e-06,
+      "loss": 0.0209,
+      "step": 4360
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 7.187459746139216,
+      "learning_rate": 1.236001234715965e-06,
+      "loss": 0.021,
+      "step": 4361
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 3.648792787441624,
+      "learning_rate": 1.2340994238004987e-06,
+      "loss": 0.0202,
+      "step": 4362
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 8.691909830938794,
+      "learning_rate": 1.2321988711424132e-06,
+      "loss": 0.0224,
+      "step": 4363
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 6.995988388162933,
+      "learning_rate": 1.2302995773767174e-06,
+      "loss": 0.0349,
+      "step": 4364
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 11.008840944917822,
+      "learning_rate": 1.2284015431380015e-06,
+      "loss": 0.0285,
+      "step": 4365
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 3.373639756983513,
+      "learning_rate": 1.2265047690604354e-06,
+      "loss": 0.0184,
+      "step": 4366
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 3.803315273323329,
+      "learning_rate": 1.2246092557777633e-06,
+      "loss": 0.0263,
+      "step": 4367
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 3.4078700572104674,
+      "learning_rate": 1.2227150039233132e-06,
+      "loss": 0.032,
+      "step": 4368
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 3.581845625792135,
+      "learning_rate": 1.2208220141299893e-06,
+      "loss": 0.0294,
+      "step": 4369
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 10.077585671022081,
+      "learning_rate": 1.2189302870302755e-06,
+      "loss": 0.0233,
+      "step": 4370
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 6.127715448474595,
+      "learning_rate": 1.2170398232562324e-06,
+      "loss": 0.0158,
+      "step": 4371
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 8.996189920092641,
+      "learning_rate": 1.2151506234395e-06,
+      "loss": 0.0471,
+      "step": 4372
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 8.082755597103926,
+      "learning_rate": 1.2132626882112935e-06,
+      "loss": 0.0246,
+      "step": 4373
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 4.081561675170032,
+      "learning_rate": 1.211376018202408e-06,
+      "loss": 0.0271,
+      "step": 4374
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 7.991360997398173,
+      "learning_rate": 1.2094906140432155e-06,
+      "loss": 0.0232,
+      "step": 4375
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 5.043322011112299,
+      "learning_rate": 1.2076064763636641e-06,
+      "loss": 0.0167,
+      "step": 4376
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 3.5060059412563933,
+      "learning_rate": 1.205723605793279e-06,
+      "loss": 0.0177,
+      "step": 4377
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 3.3044991967413786,
+      "learning_rate": 1.2038420029611625e-06,
+      "loss": 0.0185,
+      "step": 4378
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 2.5519361887925407,
+      "learning_rate": 1.2019616684959934e-06,
+      "loss": 0.0158,
+      "step": 4379
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 1.5711224232775167,
+      "learning_rate": 1.2000826030260254e-06,
+      "loss": 0.0196,
+      "step": 4380
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 3.845792898604567,
+      "learning_rate": 1.1982048071790903e-06,
+      "loss": 0.0327,
+      "step": 4381
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 1.807799902555709,
+      "learning_rate": 1.1963282815825938e-06,
+      "loss": 0.0217,
+      "step": 4382
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 3.5441110376934706,
+      "learning_rate": 1.194453026863519e-06,
+      "loss": 0.0205,
+      "step": 4383
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 2.2644135155675387,
+      "learning_rate": 1.1925790436484219e-06,
+      "loss": 0.0182,
+      "step": 4384
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 7.720254102788194,
+      "learning_rate": 1.1907063325634376e-06,
+      "loss": 0.0249,
+      "step": 4385
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 2.967984156703408,
+      "learning_rate": 1.1888348942342697e-06,
+      "loss": 0.0239,
+      "step": 4386
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 5.143287739884021,
+      "learning_rate": 1.1869647292862051e-06,
+      "loss": 0.0336,
+      "step": 4387
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 6.7684831711996045,
+      "learning_rate": 1.1850958383440957e-06,
+      "loss": 0.0294,
+      "step": 4388
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 8.714645017881613,
+      "learning_rate": 1.183228222032378e-06,
+      "loss": 0.0316,
+      "step": 4389
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 5.477500456705119,
+      "learning_rate": 1.181361880975052e-06,
+      "loss": 0.0294,
+      "step": 4390
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 3.500435025541258,
+      "learning_rate": 1.1794968157957026e-06,
+      "loss": 0.0218,
+      "step": 4391
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 4.026778952227533,
+      "learning_rate": 1.1776330271174786e-06,
+      "loss": 0.0235,
+      "step": 4392
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 1.910801405729195,
+      "learning_rate": 1.1757705155631072e-06,
+      "loss": 0.0224,
+      "step": 4393
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 4.7870288481139776,
+      "learning_rate": 1.1739092817548887e-06,
+      "loss": 0.0263,
+      "step": 4394
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 3.7988605929737616,
+      "learning_rate": 1.172049326314696e-06,
+      "loss": 0.0221,
+      "step": 4395
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 4.220102301388364,
+      "learning_rate": 1.1701906498639741e-06,
+      "loss": 0.0193,
+      "step": 4396
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 1.7825031992726328,
+      "learning_rate": 1.1683332530237423e-06,
+      "loss": 0.0278,
+      "step": 4397
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 2.9360054998312166,
+      "learning_rate": 1.1664771364145905e-06,
+      "loss": 0.0293,
+      "step": 4398
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 4.303354411436024,
+      "learning_rate": 1.1646223006566827e-06,
+      "loss": 0.0363,
+      "step": 4399
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 2.194395308626148,
+      "learning_rate": 1.162768746369753e-06,
+      "loss": 0.0237,
+      "step": 4400
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 2.116716075883209,
+      "learning_rate": 1.1609164741731105e-06,
+      "loss": 0.0273,
+      "step": 4401
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 3.210931319165583,
+      "learning_rate": 1.1590654846856291e-06,
+      "loss": 0.0263,
+      "step": 4402
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 2.119091545167854,
+      "learning_rate": 1.1572157785257643e-06,
+      "loss": 0.0143,
+      "step": 4403
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 4.324074704437676,
+      "learning_rate": 1.1553673563115325e-06,
+      "loss": 0.0336,
+      "step": 4404
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 4.170772991278056,
+      "learning_rate": 1.153520218660531e-06,
+      "loss": 0.038,
+      "step": 4405
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 4.213138084530538,
+      "learning_rate": 1.1516743661899172e-06,
+      "loss": 0.0237,
+      "step": 4406
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 11.656757681338371,
+      "learning_rate": 1.1498297995164305e-06,
+      "loss": 0.0331,
+      "step": 4407
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 3.4983502997263547,
+      "learning_rate": 1.1479865192563683e-06,
+      "loss": 0.0318,
+      "step": 4408
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 2.8807252835615254,
+      "learning_rate": 1.146144526025612e-06,
+      "loss": 0.0242,
+      "step": 4409
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 2.6306843568728038,
+      "learning_rate": 1.1443038204396007e-06,
+      "loss": 0.0258,
+      "step": 4410
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 4.192920419278408,
+      "learning_rate": 1.1424644031133502e-06,
+      "loss": 0.0187,
+      "step": 4411
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 7.177668219309729,
+      "learning_rate": 1.1406262746614433e-06,
+      "loss": 0.026,
+      "step": 4412
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 4.873333282003289,
+      "learning_rate": 1.1387894356980334e-06,
+      "loss": 0.0289,
+      "step": 4413
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 2.9519728668908227,
+      "learning_rate": 1.1369538868368424e-06,
+      "loss": 0.0196,
+      "step": 4414
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 3.8848350672996834,
+      "learning_rate": 1.1351196286911615e-06,
+      "loss": 0.0194,
+      "step": 4415
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 4.635840605259712,
+      "learning_rate": 1.1332866618738498e-06,
+      "loss": 0.0232,
+      "step": 4416
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 1.5581337612715147,
+      "learning_rate": 1.1314549869973363e-06,
+      "loss": 0.0232,
+      "step": 4417
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 1.7583630718679997,
+      "learning_rate": 1.1296246046736176e-06,
+      "loss": 0.0185,
+      "step": 4418
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 6.283893780931881,
+      "learning_rate": 1.1277955155142578e-06,
+      "loss": 0.0232,
+      "step": 4419
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 7.862048061035702,
+      "learning_rate": 1.1259677201303905e-06,
+      "loss": 0.0239,
+      "step": 4420
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 1.6230472483193787,
+      "learning_rate": 1.1241412191327155e-06,
+      "loss": 0.0157,
+      "step": 4421
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 2.755073615396238,
+      "learning_rate": 1.1223160131315008e-06,
+      "loss": 0.0298,
+      "step": 4422
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 2.73806500409641,
+      "learning_rate": 1.1204921027365818e-06,
+      "loss": 0.03,
+      "step": 4423
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 3.1203954227660895,
+      "learning_rate": 1.1186694885573602e-06,
+      "loss": 0.0197,
+      "step": 4424
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 3.5066645012673496,
+      "learning_rate": 1.1168481712028061e-06,
+      "loss": 0.0297,
+      "step": 4425
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 4.2675581356015915,
+      "learning_rate": 1.115028151281457e-06,
+      "loss": 0.022,
+      "step": 4426
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 3.4456080030453426,
+      "learning_rate": 1.1132094294014106e-06,
+      "loss": 0.0126,
+      "step": 4427
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 6.579809040631279,
+      "learning_rate": 1.1113920061703416e-06,
+      "loss": 0.0318,
+      "step": 4428
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 2.475266437209171,
+      "learning_rate": 1.1095758821954788e-06,
+      "loss": 0.0166,
+      "step": 4429
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 2.0190295382576133,
+      "learning_rate": 1.107761058083629e-06,
+      "loss": 0.0327,
+      "step": 4430
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 1.9898004038566273,
+      "learning_rate": 1.1059475344411535e-06,
+      "loss": 0.0185,
+      "step": 4431
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 3.6520762370965887,
+      "learning_rate": 1.104135311873989e-06,
+      "loss": 0.012,
+      "step": 4432
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 1.790944349770655,
+      "learning_rate": 1.1023243909876275e-06,
+      "loss": 0.0175,
+      "step": 4433
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 1.4622862012678017,
+      "learning_rate": 1.1005147723871374e-06,
+      "loss": 0.0174,
+      "step": 4434
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 2.29471929364547,
+      "learning_rate": 1.0987064566771405e-06,
+      "loss": 0.021,
+      "step": 4435
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 6.60079633509637,
+      "learning_rate": 1.0968994444618313e-06,
+      "loss": 0.0331,
+      "step": 4436
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 1.9722105559545458,
+      "learning_rate": 1.0950937363449659e-06,
+      "loss": 0.0176,
+      "step": 4437
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 9.226928463154826,
+      "learning_rate": 1.0932893329298643e-06,
+      "loss": 0.0339,
+      "step": 4438
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 2.7596617157692784,
+      "learning_rate": 1.0914862348194121e-06,
+      "loss": 0.0234,
+      "step": 4439
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 5.353123299645156,
+      "learning_rate": 1.0896844426160575e-06,
+      "loss": 0.0313,
+      "step": 4440
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 7.939115234782665,
+      "learning_rate": 1.0878839569218124e-06,
+      "loss": 0.0187,
+      "step": 4441
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 3.5070933721087356,
+      "learning_rate": 1.0860847783382534e-06,
+      "loss": 0.0319,
+      "step": 4442
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 7.566534679772082,
+      "learning_rate": 1.0842869074665186e-06,
+      "loss": 0.0226,
+      "step": 4443
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 2.5892901109879563,
+      "learning_rate": 1.0824903449073115e-06,
+      "loss": 0.0241,
+      "step": 4444
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 4.0522225755260655,
+      "learning_rate": 1.0806950912608937e-06,
+      "loss": 0.0407,
+      "step": 4445
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 4.042715326868905,
+      "learning_rate": 1.0789011471270983e-06,
+      "loss": 0.0237,
+      "step": 4446
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 5.5516387804921985,
+      "learning_rate": 1.0771085131053087e-06,
+      "loss": 0.0225,
+      "step": 4447
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 3.6164402815909362,
+      "learning_rate": 1.0753171897944835e-06,
+      "loss": 0.0215,
+      "step": 4448
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 6.6222474603434405,
+      "learning_rate": 1.0735271777931322e-06,
+      "loss": 0.022,
+      "step": 4449
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 2.592791914105326,
+      "learning_rate": 1.0717384776993356e-06,
+      "loss": 0.0322,
+      "step": 4450
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 5.227418309409531,
+      "learning_rate": 1.069951090110728e-06,
+      "loss": 0.0263,
+      "step": 4451
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 2.77396491744881,
+      "learning_rate": 1.06816501562451e-06,
+      "loss": 0.0195,
+      "step": 4452
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 1.8546536944682173,
+      "learning_rate": 1.0663802548374424e-06,
+      "loss": 0.0214,
+      "step": 4453
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 3.6321220341074585,
+      "learning_rate": 1.064596808345847e-06,
+      "loss": 0.0237,
+      "step": 4454
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 9.088425273383129,
+      "learning_rate": 1.0628146767456066e-06,
+      "loss": 0.031,
+      "step": 4455
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 2.244296281743473,
+      "learning_rate": 1.061033860632164e-06,
+      "loss": 0.0244,
+      "step": 4456
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 1.8747595056637634,
+      "learning_rate": 1.0592543606005235e-06,
+      "loss": 0.0254,
+      "step": 4457
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 3.2313945845341676,
+      "learning_rate": 1.0574761772452486e-06,
+      "loss": 0.0342,
+      "step": 4458
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 2.0815746855400103,
+      "learning_rate": 1.0556993111604635e-06,
+      "loss": 0.0226,
+      "step": 4459
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 8.359714562424623,
+      "learning_rate": 1.0539237629398536e-06,
+      "loss": 0.0335,
+      "step": 4460
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 3.709979629335406,
+      "learning_rate": 1.052149533176659e-06,
+      "loss": 0.0193,
+      "step": 4461
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 2.239188178419034,
+      "learning_rate": 1.050376622463688e-06,
+      "loss": 0.0217,
+      "step": 4462
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 4.238067345733504,
+      "learning_rate": 1.0486050313932972e-06,
+      "loss": 0.0246,
+      "step": 4463
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 2.8338637835412612,
+      "learning_rate": 1.0468347605574137e-06,
+      "loss": 0.0266,
+      "step": 4464
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 4.869316442569333,
+      "learning_rate": 1.0450658105475126e-06,
+      "loss": 0.0232,
+      "step": 4465
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 3.5665750122251625,
+      "learning_rate": 1.0432981819546384e-06,
+      "loss": 0.0309,
+      "step": 4466
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 8.936241009064924,
+      "learning_rate": 1.0415318753693837e-06,
+      "loss": 0.0236,
+      "step": 4467
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 3.101588523930688,
+      "learning_rate": 1.0397668913819086e-06,
+      "loss": 0.0239,
+      "step": 4468
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 4.295993293606268,
+      "learning_rate": 1.0380032305819243e-06,
+      "loss": 0.0223,
+      "step": 4469
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 2.3409226213317025,
+      "learning_rate": 1.0362408935587026e-06,
+      "loss": 0.0309,
+      "step": 4470
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 2.8493429852061487,
+      "learning_rate": 1.0344798809010748e-06,
+      "loss": 0.0246,
+      "step": 4471
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 4.991729680693538,
+      "learning_rate": 1.0327201931974262e-06,
+      "loss": 0.0165,
+      "step": 4472
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 1.8954580681443631,
+      "learning_rate": 1.0309618310357023e-06,
+      "loss": 0.0196,
+      "step": 4473
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 10.27608130286604,
+      "learning_rate": 1.0292047950034046e-06,
+      "loss": 0.0312,
+      "step": 4474
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 2.153159817327546,
+      "learning_rate": 1.0274490856875908e-06,
+      "loss": 0.0109,
+      "step": 4475
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 1.5007061694216104,
+      "learning_rate": 1.0256947036748766e-06,
+      "loss": 0.0186,
+      "step": 4476
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 1.7574699954655115,
+      "learning_rate": 1.0239416495514331e-06,
+      "loss": 0.0267,
+      "step": 4477
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 3.1371640873189954,
+      "learning_rate": 1.0221899239029887e-06,
+      "loss": 0.0185,
+      "step": 4478
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 5.019779692114384,
+      "learning_rate": 1.0204395273148277e-06,
+      "loss": 0.0157,
+      "step": 4479
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 7.540699269714138,
+      "learning_rate": 1.0186904603717894e-06,
+      "loss": 0.027,
+      "step": 4480
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 3.9237182973169595,
+      "learning_rate": 1.0169427236582702e-06,
+      "loss": 0.0377,
+      "step": 4481
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 4.278480950923269,
+      "learning_rate": 1.0151963177582208e-06,
+      "loss": 0.0245,
+      "step": 4482
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 2.312231473624651,
+      "learning_rate": 1.0134512432551492e-06,
+      "loss": 0.0183,
+      "step": 4483
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 2.0986535225249416,
+      "learning_rate": 1.0117075007321152e-06,
+      "loss": 0.0143,
+      "step": 4484
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 5.975648412284588,
+      "learning_rate": 1.009965090771739e-06,
+      "loss": 0.0197,
+      "step": 4485
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 4.374236320725866,
+      "learning_rate": 1.0082240139561866e-06,
+      "loss": 0.0254,
+      "step": 4486
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 6.532426395412988,
+      "learning_rate": 1.0064842708671908e-06,
+      "loss": 0.0208,
+      "step": 4487
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 3.6018002154036113,
+      "learning_rate": 1.0047458620860251e-06,
+      "loss": 0.0195,
+      "step": 4488
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 2.4807601063417293,
+      "learning_rate": 1.0030087881935308e-06,
+      "loss": 0.0158,
+      "step": 4489
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 3.4131438838142536,
+      "learning_rate": 1.0012730497700912e-06,
+      "loss": 0.0293,
+      "step": 4490
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 7.077550750630359,
+      "learning_rate": 9.995386473956531e-07,
+      "loss": 0.0251,
+      "step": 4491
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 2.258377721601201,
+      "learning_rate": 9.978055816497084e-07,
+      "loss": 0.0227,
+      "step": 4492
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 2.229914063381325,
+      "learning_rate": 9.960738531113118e-07,
+      "loss": 0.0209,
+      "step": 4493
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 2.1564691341307487,
+      "learning_rate": 9.94343462359061e-07,
+      "loss": 0.014,
+      "step": 4494
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 5.382295516557594,
+      "learning_rate": 9.926144099711138e-07,
+      "loss": 0.0277,
+      "step": 4495
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 2.516122869741995,
+      "learning_rate": 9.90886696525179e-07,
+      "loss": 0.022,
+      "step": 4496
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 7.685796145048834,
+      "learning_rate": 9.89160322598517e-07,
+      "loss": 0.0294,
+      "step": 4497
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 2.327309331754003,
+      "learning_rate": 9.874352887679416e-07,
+      "loss": 0.0273,
+      "step": 4498
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 2.8309636158087965,
+      "learning_rate": 9.857115956098196e-07,
+      "loss": 0.0287,
+      "step": 4499
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 3.2811544904994556,
+      "learning_rate": 9.839892437000675e-07,
+      "loss": 0.0135,
+      "step": 4500
+    },
+    {
+      "epoch": 3.21,
+      "eval_avg_AUC": 0.8345286134092101,
+      "eval_avg_Accuracy": 0.7458968832891246,
+      "eval_avg_Accuracy-right": 0.888352680318247,
+      "eval_avg_Accuracy-wrong": 0.49749829429156245,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.7117713077832741,
+      "eval_last_AUC": 0.8350567927270806,
+      "eval_last_Accuracy": 0.7807526525198939,
+      "eval_last_Accuracy-right": 0.8413982000782575,
+      "eval_last_Accuracy-wrong": 0.6750056856947919,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.7121076559312289,
+      "eval_max_AUC": 0.7854278482735377,
+      "eval_max_Accuracy": 0.6475878647214854,
+      "eval_max_Accuracy-right": 0.9867614451545585,
+      "eval_max_Accuracy-wrong": 0.056174664544007276,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.6558197091469862,
+      "eval_min_AUC": 0.843139889983326,
+      "eval_min_Accuracy": 0.7734996684350133,
+      "eval_min_Accuracy-right": 0.7857049693491587,
+      "eval_min_Accuracy-wrong": 0.7522174209688424,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.715089706106881,
+      "eval_prod_AUC": 0.8407684428232383,
+      "eval_prod_Accuracy": 0.7391826923076923,
+      "eval_prod_Accuracy-right": 0.6629059606104083,
+      "eval_prod_Accuracy-wrong": 0.8721855810780077,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.7128298178825854,
+      "eval_runtime": 247.0161,
+      "eval_samples_per_second": 97.678,
+      "eval_steps_per_second": 3.052,
+      "eval_sum_AUC": 0.7186416616330681,
+      "eval_sum_Accuracy": 0.6397546419098143,
+      "eval_sum_Accuracy-right": 0.9967392722055562,
+      "eval_sum_Accuracy-wrong": 0.017284512167386856,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6893739044164864,
+      "step": 4500
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 1.4722866410217152,
+      "learning_rate": 9.822682336141558e-07,
+      "loss": 0.0178,
+      "step": 4501
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 2.181579380218437,
+      "learning_rate": 9.805485659271064e-07,
+      "loss": 0.0168,
+      "step": 4502
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 7.752924249109985,
+      "learning_rate": 9.788302412134931e-07,
+      "loss": 0.0352,
+      "step": 4503
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 1.9281551566524828,
+      "learning_rate": 9.77113260047436e-07,
+      "loss": 0.0192,
+      "step": 4504
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 1.9477930911053865,
+      "learning_rate": 9.753976230026158e-07,
+      "loss": 0.0234,
+      "step": 4505
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 6.050182109498725,
+      "learning_rate": 9.736833306522537e-07,
+      "loss": 0.0257,
+      "step": 4506
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 5.560668219829083,
+      "learning_rate": 9.719703835691314e-07,
+      "loss": 0.0216,
+      "step": 4507
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 10.427582712165258,
+      "learning_rate": 9.702587823255715e-07,
+      "loss": 0.0413,
+      "step": 4508
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 2.6181714476401536,
+      "learning_rate": 9.685485274934576e-07,
+      "loss": 0.0179,
+      "step": 4509
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 9.156873614600924,
+      "learning_rate": 9.66839619644211e-07,
+      "loss": 0.0251,
+      "step": 4510
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 3.4673271118101643,
+      "learning_rate": 9.651320593488162e-07,
+      "loss": 0.0191,
+      "step": 4511
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 1.6137388839865046,
+      "learning_rate": 9.634258471777958e-07,
+      "loss": 0.0184,
+      "step": 4512
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 3.6836288992334048,
+      "learning_rate": 9.617209837012287e-07,
+      "loss": 0.0264,
+      "step": 4513
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 5.5935012176996635,
+      "learning_rate": 9.600174694887421e-07,
+      "loss": 0.0247,
+      "step": 4514
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 7.298134909789238,
+      "learning_rate": 9.583153051095107e-07,
+      "loss": 0.0247,
+      "step": 4515
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 5.308498768012288,
+      "learning_rate": 9.5661449113226e-07,
+      "loss": 0.0287,
+      "step": 4516
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 1.8179009486270912,
+      "learning_rate": 9.549150281252633e-07,
+      "loss": 0.0206,
+      "step": 4517
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 3.3811622578866727,
+      "learning_rate": 9.532169166563426e-07,
+      "loss": 0.0284,
+      "step": 4518
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 3.8222062654617406,
+      "learning_rate": 9.515201572928689e-07,
+      "loss": 0.028,
+      "step": 4519
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 3.0854257738751163,
+      "learning_rate": 9.49824750601761e-07,
+      "loss": 0.0316,
+      "step": 4520
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 2.5397879239379098,
+      "learning_rate": 9.481306971494858e-07,
+      "loss": 0.0265,
+      "step": 4521
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 3.2717124935629953,
+      "learning_rate": 9.464379975020576e-07,
+      "loss": 0.0278,
+      "step": 4522
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 9.377789209888471,
+      "learning_rate": 9.447466522250393e-07,
+      "loss": 0.0318,
+      "step": 4523
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 6.441122861364104,
+      "learning_rate": 9.430566618835407e-07,
+      "loss": 0.0314,
+      "step": 4524
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 2.3160747535915265,
+      "learning_rate": 9.413680270422187e-07,
+      "loss": 0.0289,
+      "step": 4525
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 1.6669925746710255,
+      "learning_rate": 9.396807482652775e-07,
+      "loss": 0.0241,
+      "step": 4526
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 2.8847819825771777,
+      "learning_rate": 9.3799482611647e-07,
+      "loss": 0.0266,
+      "step": 4527
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 4.401247157463592,
+      "learning_rate": 9.363102611590918e-07,
+      "loss": 0.0206,
+      "step": 4528
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 4.810249161783182,
+      "learning_rate": 9.346270539559882e-07,
+      "loss": 0.0213,
+      "step": 4529
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 7.9036297509881095,
+      "learning_rate": 9.329452050695497e-07,
+      "loss": 0.0221,
+      "step": 4530
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 3.570781343969614,
+      "learning_rate": 9.312647150617144e-07,
+      "loss": 0.0207,
+      "step": 4531
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 1.6397653229972227,
+      "learning_rate": 9.295855844939639e-07,
+      "loss": 0.012,
+      "step": 4532
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 2.258414374329704,
+      "learning_rate": 9.279078139273279e-07,
+      "loss": 0.0181,
+      "step": 4533
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 8.950313191107599,
+      "learning_rate": 9.262314039223802e-07,
+      "loss": 0.0312,
+      "step": 4534
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 2.1815871351729617,
+      "learning_rate": 9.245563550392406e-07,
+      "loss": 0.0223,
+      "step": 4535
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 4.859815723424763,
+      "learning_rate": 9.22882667837574e-07,
+      "loss": 0.0266,
+      "step": 4536
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 2.355286650208844,
+      "learning_rate": 9.212103428765912e-07,
+      "loss": 0.0203,
+      "step": 4537
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 2.0779801100488684,
+      "learning_rate": 9.19539380715046e-07,
+      "loss": 0.0234,
+      "step": 4538
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 9.338632225920227,
+      "learning_rate": 9.178697819112381e-07,
+      "loss": 0.0253,
+      "step": 4539
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 2.2899098228257464,
+      "learning_rate": 9.162015470230123e-07,
+      "loss": 0.0175,
+      "step": 4540
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 1.5376658110359513,
+      "learning_rate": 9.145346766077562e-07,
+      "loss": 0.025,
+      "step": 4541
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 1.9588815283252403,
+      "learning_rate": 9.128691712224025e-07,
+      "loss": 0.0208,
+      "step": 4542
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 1.793777776979931,
+      "learning_rate": 9.112050314234272e-07,
+      "loss": 0.0204,
+      "step": 4543
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 3.048832843843268,
+      "learning_rate": 9.0954225776685e-07,
+      "loss": 0.0146,
+      "step": 4544
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 2.3438063416343393,
+      "learning_rate": 9.078808508082354e-07,
+      "loss": 0.0206,
+      "step": 4545
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 1.9916762447835414,
+      "learning_rate": 9.06220811102691e-07,
+      "loss": 0.0257,
+      "step": 4546
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 5.85438556591717,
+      "learning_rate": 9.045621392048637e-07,
+      "loss": 0.0235,
+      "step": 4547
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 2.46668861130742,
+      "learning_rate": 9.029048356689507e-07,
+      "loss": 0.0255,
+      "step": 4548
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 4.512065904072692,
+      "learning_rate": 9.012489010486835e-07,
+      "loss": 0.0242,
+      "step": 4549
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 3.86233637448217,
+      "learning_rate": 8.995943358973463e-07,
+      "loss": 0.0244,
+      "step": 4550
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 3.2961520762406575,
+      "learning_rate": 8.979411407677535e-07,
+      "loss": 0.0241,
+      "step": 4551
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 4.216390230976483,
+      "learning_rate": 8.962893162122749e-07,
+      "loss": 0.0201,
+      "step": 4552
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 8.348781851060899,
+      "learning_rate": 8.946388627828106e-07,
+      "loss": 0.0311,
+      "step": 4553
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 6.084513848361836,
+      "learning_rate": 8.929897810308102e-07,
+      "loss": 0.025,
+      "step": 4554
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 8.2559434065546,
+      "learning_rate": 8.913420715072619e-07,
+      "loss": 0.0396,
+      "step": 4555
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 4.631280709242924,
+      "learning_rate": 8.896957347626966e-07,
+      "loss": 0.0321,
+      "step": 4556
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 3.139609791677315,
+      "learning_rate": 8.880507713471853e-07,
+      "loss": 0.0157,
+      "step": 4557
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 3.308998251970711,
+      "learning_rate": 8.864071818103415e-07,
+      "loss": 0.0179,
+      "step": 4558
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 2.4597231253581624,
+      "learning_rate": 8.847649667013187e-07,
+      "loss": 0.0248,
+      "step": 4559
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 4.1940520085889625,
+      "learning_rate": 8.831241265688112e-07,
+      "loss": 0.0284,
+      "step": 4560
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 9.58472105522457,
+      "learning_rate": 8.814846619610545e-07,
+      "loss": 0.0199,
+      "step": 4561
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 4.313842958685549,
+      "learning_rate": 8.79846573425826e-07,
+      "loss": 0.0267,
+      "step": 4562
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 3.509297605663404,
+      "learning_rate": 8.782098615104373e-07,
+      "loss": 0.0182,
+      "step": 4563
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 1.6077932936017227,
+      "learning_rate": 8.765745267617487e-07,
+      "loss": 0.0247,
+      "step": 4564
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 2.9038110707158453,
+      "learning_rate": 8.749405697261515e-07,
+      "loss": 0.0228,
+      "step": 4565
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 3.162924718811875,
+      "learning_rate": 8.733079909495868e-07,
+      "loss": 0.0229,
+      "step": 4566
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 2.0279848737464015,
+      "learning_rate": 8.716767909775231e-07,
+      "loss": 0.028,
+      "step": 4567
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 6.216485879376879,
+      "learning_rate": 8.700469703549802e-07,
+      "loss": 0.0333,
+      "step": 4568
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 7.520962408278871,
+      "learning_rate": 8.684185296265074e-07,
+      "loss": 0.0197,
+      "step": 4569
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 8.629919683440207,
+      "learning_rate": 8.667914693362006e-07,
+      "loss": 0.0187,
+      "step": 4570
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 7.682867297333057,
+      "learning_rate": 8.651657900276878e-07,
+      "loss": 0.0235,
+      "step": 4571
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 6.700794151870362,
+      "learning_rate": 8.635414922441398e-07,
+      "loss": 0.0226,
+      "step": 4572
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 1.3609985270641234,
+      "learning_rate": 8.61918576528265e-07,
+      "loss": 0.0191,
+      "step": 4573
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 8.333651139592504,
+      "learning_rate": 8.60297043422309e-07,
+      "loss": 0.0236,
+      "step": 4574
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 5.2970319558403975,
+      "learning_rate": 8.586768934680572e-07,
+      "loss": 0.0227,
+      "step": 4575
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 8.021689608056526,
+      "learning_rate": 8.570581272068307e-07,
+      "loss": 0.0183,
+      "step": 4576
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 2.30192703714936,
+      "learning_rate": 8.554407451794905e-07,
+      "loss": 0.0168,
+      "step": 4577
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 10.68080512987844,
+      "learning_rate": 8.538247479264327e-07,
+      "loss": 0.0336,
+      "step": 4578
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 4.080574857143387,
+      "learning_rate": 8.522101359875934e-07,
+      "loss": 0.0234,
+      "step": 4579
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 6.2548733609393015,
+      "learning_rate": 8.505969099024436e-07,
+      "loss": 0.0318,
+      "step": 4580
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 2.452170699337819,
+      "learning_rate": 8.489850702099922e-07,
+      "loss": 0.0318,
+      "step": 4581
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 5.3176103182624805,
+      "learning_rate": 8.473746174487846e-07,
+      "loss": 0.0326,
+      "step": 4582
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 4.033919981368345,
+      "learning_rate": 8.457655521569036e-07,
+      "loss": 0.0287,
+      "step": 4583
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 8.624185466236282,
+      "learning_rate": 8.441578748719676e-07,
+      "loss": 0.0274,
+      "step": 4584
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 2.78523165440918,
+      "learning_rate": 8.425515861311312e-07,
+      "loss": 0.0197,
+      "step": 4585
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 6.283027986139463,
+      "learning_rate": 8.409466864710858e-07,
+      "loss": 0.0248,
+      "step": 4586
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 5.122071684243946,
+      "learning_rate": 8.393431764280591e-07,
+      "loss": 0.0175,
+      "step": 4587
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 7.677661344943328,
+      "learning_rate": 8.377410565378097e-07,
+      "loss": 0.026,
+      "step": 4588
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 6.845604358260919,
+      "learning_rate": 8.361403273356411e-07,
+      "loss": 0.0262,
+      "step": 4589
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 1.9332179724262128,
+      "learning_rate": 8.345409893563816e-07,
+      "loss": 0.0208,
+      "step": 4590
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 4.828115067765296,
+      "learning_rate": 8.329430431344043e-07,
+      "loss": 0.0283,
+      "step": 4591
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 4.041522711176144,
+      "learning_rate": 8.313464892036083e-07,
+      "loss": 0.0182,
+      "step": 4592
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 3.5931461402812284,
+      "learning_rate": 8.297513280974362e-07,
+      "loss": 0.0205,
+      "step": 4593
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 7.322773741524985,
+      "learning_rate": 8.281575603488573e-07,
+      "loss": 0.0245,
+      "step": 4594
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 2.2037482548752654,
+      "learning_rate": 8.265651864903823e-07,
+      "loss": 0.0228,
+      "step": 4595
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 1.8543264539363935,
+      "learning_rate": 8.249742070540506e-07,
+      "loss": 0.0205,
+      "step": 4596
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 4.841219706984167,
+      "learning_rate": 8.233846225714386e-07,
+      "loss": 0.0315,
+      "step": 4597
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 1.9559510186530535,
+      "learning_rate": 8.217964335736556e-07,
+      "loss": 0.0196,
+      "step": 4598
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 2.369932136393158,
+      "learning_rate": 8.202096405913462e-07,
+      "loss": 0.0211,
+      "step": 4599
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 3.149653056045046,
+      "learning_rate": 8.186242441546866e-07,
+      "loss": 0.0208,
+      "step": 4600
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 5.1208147167576294,
+      "learning_rate": 8.170402447933873e-07,
+      "loss": 0.0377,
+      "step": 4601
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 2.50638995160429,
+      "learning_rate": 8.154576430366922e-07,
+      "loss": 0.0223,
+      "step": 4602
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 5.384939207639812,
+      "learning_rate": 8.13876439413378e-07,
+      "loss": 0.028,
+      "step": 4603
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 1.762765510713304,
+      "learning_rate": 8.122966344517536e-07,
+      "loss": 0.0239,
+      "step": 4604
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 3.0740858313842443,
+      "learning_rate": 8.107182286796633e-07,
+      "loss": 0.0244,
+      "step": 4605
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 3.9397427075031013,
+      "learning_rate": 8.091412226244771e-07,
+      "loss": 0.0172,
+      "step": 4606
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 3.5915667776547204,
+      "learning_rate": 8.07565616813108e-07,
+      "loss": 0.0344,
+      "step": 4607
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 3.9272905067394714,
+      "learning_rate": 8.059914117719897e-07,
+      "loss": 0.0253,
+      "step": 4608
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 3.2510066913863462,
+      "learning_rate": 8.044186080270983e-07,
+      "loss": 0.0248,
+      "step": 4609
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 1.2617444867153567,
+      "learning_rate": 8.028472061039322e-07,
+      "loss": 0.0192,
+      "step": 4610
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 1.939700065819408,
+      "learning_rate": 8.012772065275304e-07,
+      "loss": 0.021,
+      "step": 4611
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 2.1320831911446376,
+      "learning_rate": 7.997086098224555e-07,
+      "loss": 0.0199,
+      "step": 4612
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 1.884317978445364,
+      "learning_rate": 7.981414165128065e-07,
+      "loss": 0.0151,
+      "step": 4613
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 4.5621264328140905,
+      "learning_rate": 7.965756271222108e-07,
+      "loss": 0.0208,
+      "step": 4614
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 1.440610499433646,
+      "learning_rate": 7.950112421738282e-07,
+      "loss": 0.0178,
+      "step": 4615
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 2.258538996875376,
+      "learning_rate": 7.934482621903494e-07,
+      "loss": 0.0262,
+      "step": 4616
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 3.0165939984057832,
+      "learning_rate": 7.91886687693994e-07,
+      "loss": 0.0165,
+      "step": 4617
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 5.695857517771355,
+      "learning_rate": 7.903265192065141e-07,
+      "loss": 0.0197,
+      "step": 4618
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 1.7055585895666883,
+      "learning_rate": 7.887677572491903e-07,
+      "loss": 0.0187,
+      "step": 4619
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 7.5887897339828685,
+      "learning_rate": 7.872104023428339e-07,
+      "loss": 0.0238,
+      "step": 4620
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 3.374951107007746,
+      "learning_rate": 7.856544550077883e-07,
+      "loss": 0.0257,
+      "step": 4621
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 1.855484086420173,
+      "learning_rate": 7.840999157639195e-07,
+      "loss": 0.019,
+      "step": 4622
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 2.792593588176287,
+      "learning_rate": 7.825467851306335e-07,
+      "loss": 0.0207,
+      "step": 4623
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 9.742513544372374,
+      "learning_rate": 7.809950636268554e-07,
+      "loss": 0.0224,
+      "step": 4624
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 5.9915718445374795,
+      "learning_rate": 7.794447517710485e-07,
+      "loss": 0.0223,
+      "step": 4625
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 2.4603505001707378,
+      "learning_rate": 7.778958500811961e-07,
+      "loss": 0.0317,
+      "step": 4626
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 2.349151356305383,
+      "learning_rate": 7.7634835907482e-07,
+      "loss": 0.0232,
+      "step": 4627
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 5.788557646206046,
+      "learning_rate": 7.748022792689613e-07,
+      "loss": 0.0233,
+      "step": 4628
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 4.051836245160571,
+      "learning_rate": 7.732576111801982e-07,
+      "loss": 0.0125,
+      "step": 4629
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 2.8753818346406246,
+      "learning_rate": 7.717143553246298e-07,
+      "loss": 0.0306,
+      "step": 4630
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 2.577591283510448,
+      "learning_rate": 7.701725122178871e-07,
+      "loss": 0.0186,
+      "step": 4631
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 11.542060297374746,
+      "learning_rate": 7.686320823751298e-07,
+      "loss": 0.0478,
+      "step": 4632
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 2.371685944489164,
+      "learning_rate": 7.670930663110426e-07,
+      "loss": 0.0251,
+      "step": 4633
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 1.7468794803617154,
+      "learning_rate": 7.655554645398405e-07,
+      "loss": 0.0228,
+      "step": 4634
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 1.245714311313327,
+      "learning_rate": 7.640192775752647e-07,
+      "loss": 0.0189,
+      "step": 4635
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 4.9245303461145715,
+      "learning_rate": 7.624845059305836e-07,
+      "loss": 0.028,
+      "step": 4636
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 3.7927382042335367,
+      "learning_rate": 7.609511501185929e-07,
+      "loss": 0.0305,
+      "step": 4637
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 1.8226386480619419,
+      "learning_rate": 7.594192106516151e-07,
+      "loss": 0.024,
+      "step": 4638
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 4.11296383175397,
+      "learning_rate": 7.578886880414999e-07,
+      "loss": 0.0342,
+      "step": 4639
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 2.9392679931067014,
+      "learning_rate": 7.563595827996235e-07,
+      "loss": 0.0247,
+      "step": 4640
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 3.0949822502136284,
+      "learning_rate": 7.548318954368883e-07,
+      "loss": 0.0252,
+      "step": 4641
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 2.384999555074518,
+      "learning_rate": 7.533056264637228e-07,
+      "loss": 0.0207,
+      "step": 4642
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 1.6244482154616275,
+      "learning_rate": 7.51780776390082e-07,
+      "loss": 0.0199,
+      "step": 4643
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 5.849177252436853,
+      "learning_rate": 7.50257345725447e-07,
+      "loss": 0.0227,
+      "step": 4644
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 7.411666454715126,
+      "learning_rate": 7.487353349788234e-07,
+      "loss": 0.0223,
+      "step": 4645
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 2.944448161435966,
+      "learning_rate": 7.472147446587452e-07,
+      "loss": 0.0213,
+      "step": 4646
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 4.2919752995505585,
+      "learning_rate": 7.456955752732659e-07,
+      "loss": 0.0292,
+      "step": 4647
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 1.7148082801364484,
+      "learning_rate": 7.441778273299738e-07,
+      "loss": 0.0146,
+      "step": 4648
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 5.5113167274528125,
+      "learning_rate": 7.426615013359706e-07,
+      "loss": 0.0242,
+      "step": 4649
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 3.4930471463189185,
+      "learning_rate": 7.411465977978949e-07,
+      "loss": 0.0214,
+      "step": 4650
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 2.7583759156057948,
+      "learning_rate": 7.396331172218996e-07,
+      "loss": 0.0263,
+      "step": 4651
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 3.0835376005766184,
+      "learning_rate": 7.381210601136702e-07,
+      "loss": 0.0226,
+      "step": 4652
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 7.939488044101162,
+      "learning_rate": 7.366104269784086e-07,
+      "loss": 0.0222,
+      "step": 4653
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 7.71559173914037,
+      "learning_rate": 7.351012183208511e-07,
+      "loss": 0.0246,
+      "step": 4654
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 3.3505208772325212,
+      "learning_rate": 7.335934346452484e-07,
+      "loss": 0.022,
+      "step": 4655
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 2.522427789926864,
+      "learning_rate": 7.320870764553795e-07,
+      "loss": 0.0258,
+      "step": 4656
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 4.625965296359435,
+      "learning_rate": 7.305821442545474e-07,
+      "loss": 0.0318,
+      "step": 4657
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 1.666715669670613,
+      "learning_rate": 7.290786385455778e-07,
+      "loss": 0.0267,
+      "step": 4658
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 7.423398252798393,
+      "learning_rate": 7.275765598308199e-07,
+      "loss": 0.0457,
+      "step": 4659
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 4.94580570567993,
+      "learning_rate": 7.26075908612146e-07,
+      "loss": 0.0223,
+      "step": 4660
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 2.6200308676820714,
+      "learning_rate": 7.245766853909519e-07,
+      "loss": 0.0302,
+      "step": 4661
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 2.565484956644103,
+      "learning_rate": 7.230788906681558e-07,
+      "loss": 0.0139,
+      "step": 4662
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 7.519074271081959,
+      "learning_rate": 7.215825249441982e-07,
+      "loss": 0.0438,
+      "step": 4663
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 2.9349056048339865,
+      "learning_rate": 7.200875887190445e-07,
+      "loss": 0.0319,
+      "step": 4664
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 1.7802836631328636,
+      "learning_rate": 7.185940824921772e-07,
+      "loss": 0.0256,
+      "step": 4665
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 2.3165391573905887,
+      "learning_rate": 7.171020067626089e-07,
+      "loss": 0.0241,
+      "step": 4666
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 10.891239008200467,
+      "learning_rate": 7.156113620288646e-07,
+      "loss": 0.0393,
+      "step": 4667
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 8.083097463271788,
+      "learning_rate": 7.141221487890027e-07,
+      "loss": 0.0271,
+      "step": 4668
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 4.11813397569581,
+      "learning_rate": 7.126343675405905e-07,
+      "loss": 0.0164,
+      "step": 4669
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 5.230913755558933,
+      "learning_rate": 7.111480187807296e-07,
+      "loss": 0.0227,
+      "step": 4670
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 2.4065813670459244,
+      "learning_rate": 7.096631030060308e-07,
+      "loss": 0.0265,
+      "step": 4671
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 2.9175150140266624,
+      "learning_rate": 7.081796207126373e-07,
+      "loss": 0.0167,
+      "step": 4672
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 5.43837835587297,
+      "learning_rate": 7.06697572396205e-07,
+      "loss": 0.0166,
+      "step": 4673
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 5.613308008082632,
+      "learning_rate": 7.052169585519142e-07,
+      "loss": 0.0162,
+      "step": 4674
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 4.7970690996692165,
+      "learning_rate": 7.037377796744666e-07,
+      "loss": 0.0214,
+      "step": 4675
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 6.004092278249741,
+      "learning_rate": 7.022600362580817e-07,
+      "loss": 0.0234,
+      "step": 4676
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 4.179910056219882,
+      "learning_rate": 7.007837287965024e-07,
+      "loss": 0.014,
+      "step": 4677
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 5.786550535023316,
+      "learning_rate": 6.993088577829904e-07,
+      "loss": 0.0287,
+      "step": 4678
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 8.49339045448722,
+      "learning_rate": 6.978354237103264e-07,
+      "loss": 0.0291,
+      "step": 4679
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 1.9362629401618683,
+      "learning_rate": 6.963634270708137e-07,
+      "loss": 0.0211,
+      "step": 4680
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 3.2084607601706954,
+      "learning_rate": 6.948928683562722e-07,
+      "loss": 0.0145,
+      "step": 4681
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 6.388831706059872,
+      "learning_rate": 6.934237480580435e-07,
+      "loss": 0.0261,
+      "step": 4682
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 8.92008557471272,
+      "learning_rate": 6.919560666669889e-07,
+      "loss": 0.025,
+      "step": 4683
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 3.5596242184271523,
+      "learning_rate": 6.904898246734864e-07,
+      "loss": 0.024,
+      "step": 4684
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 5.813474706944849,
+      "learning_rate": 6.890250225674361e-07,
+      "loss": 0.0238,
+      "step": 4685
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 6.456049406535201,
+      "learning_rate": 6.875616608382562e-07,
+      "loss": 0.0235,
+      "step": 4686
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 4.27019288004158,
+      "learning_rate": 6.860997399748792e-07,
+      "loss": 0.0241,
+      "step": 4687
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 1.4468000545374635,
+      "learning_rate": 6.846392604657653e-07,
+      "loss": 0.0173,
+      "step": 4688
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 2.0063895434320953,
+      "learning_rate": 6.831802227988843e-07,
+      "loss": 0.0128,
+      "step": 4689
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 4.125314951911715,
+      "learning_rate": 6.817226274617283e-07,
+      "loss": 0.0184,
+      "step": 4690
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 4.390828758784832,
+      "learning_rate": 6.802664749413079e-07,
+      "loss": 0.0222,
+      "step": 4691
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 2.617942356146691,
+      "learning_rate": 6.788117657241506e-07,
+      "loss": 0.017,
+      "step": 4692
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 5.244438256633183,
+      "learning_rate": 6.773585002963007e-07,
+      "loss": 0.0287,
+      "step": 4693
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 5.848459710599114,
+      "learning_rate": 6.759066791433228e-07,
+      "loss": 0.025,
+      "step": 4694
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 4.5353239904491645,
+      "learning_rate": 6.744563027502959e-07,
+      "loss": 0.0229,
+      "step": 4695
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 6.973704437407207,
+      "learning_rate": 6.730073716018187e-07,
+      "loss": 0.0313,
+      "step": 4696
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 3.2864116478612178,
+      "learning_rate": 6.715598861820055e-07,
+      "loss": 0.0179,
+      "step": 4697
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 2.120343882460884,
+      "learning_rate": 6.701138469744883e-07,
+      "loss": 0.0195,
+      "step": 4698
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 2.2826363640967333,
+      "learning_rate": 6.686692544624157e-07,
+      "loss": 0.0249,
+      "step": 4699
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 2.2453281994267424,
+      "learning_rate": 6.672261091284526e-07,
+      "loss": 0.0298,
+      "step": 4700
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 4.527463652856993,
+      "learning_rate": 6.657844114547812e-07,
+      "loss": 0.0332,
+      "step": 4701
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 12.362619407093279,
+      "learning_rate": 6.643441619230989e-07,
+      "loss": 0.0338,
+      "step": 4702
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 4.943448636464124,
+      "learning_rate": 6.629053610146202e-07,
+      "loss": 0.0238,
+      "step": 4703
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 3.9420677318165227,
+      "learning_rate": 6.61468009210075e-07,
+      "loss": 0.0211,
+      "step": 4704
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 2.069578664244523,
+      "learning_rate": 6.600321069897097e-07,
+      "loss": 0.0155,
+      "step": 4705
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 6.509957366481949,
+      "learning_rate": 6.585976548332856e-07,
+      "loss": 0.0441,
+      "step": 4706
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 3.550669987246946,
+      "learning_rate": 6.571646532200815e-07,
+      "loss": 0.0255,
+      "step": 4707
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 2.07784447599088,
+      "learning_rate": 6.557331026288855e-07,
+      "loss": 0.0137,
+      "step": 4708
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 2.33838698403416,
+      "learning_rate": 6.543030035380099e-07,
+      "loss": 0.017,
+      "step": 4709
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 4.926223440067987,
+      "learning_rate": 6.528743564252737e-07,
+      "loss": 0.023,
+      "step": 4710
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 3.0636490626676856,
+      "learning_rate": 6.514471617680184e-07,
+      "loss": 0.0229,
+      "step": 4711
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 2.4741516925135745,
+      "learning_rate": 6.500214200430921e-07,
+      "loss": 0.017,
+      "step": 4712
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 2.4767551496918587,
+      "learning_rate": 6.485971317268658e-07,
+      "loss": 0.0192,
+      "step": 4713
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 1.7457587448524006,
+      "learning_rate": 6.471742972952172e-07,
+      "loss": 0.0236,
+      "step": 4714
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 3.6635714256934637,
+      "learning_rate": 6.457529172235427e-07,
+      "loss": 0.0201,
+      "step": 4715
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 3.108350647844174,
+      "learning_rate": 6.44332991986753e-07,
+      "loss": 0.0288,
+      "step": 4716
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 1.5565324965470344,
+      "learning_rate": 6.429145220592703e-07,
+      "loss": 0.0173,
+      "step": 4717
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 5.2053739429829005,
+      "learning_rate": 6.414975079150321e-07,
+      "loss": 0.0235,
+      "step": 4718
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 6.228664595722526,
+      "learning_rate": 6.400819500274891e-07,
+      "loss": 0.0341,
+      "step": 4719
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 6.550911531984451,
+      "learning_rate": 6.386678488696057e-07,
+      "loss": 0.0226,
+      "step": 4720
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 1.3109577383123228,
+      "learning_rate": 6.372552049138591e-07,
+      "loss": 0.0179,
+      "step": 4721
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 4.114856675378657,
+      "learning_rate": 6.358440186322401e-07,
+      "loss": 0.0256,
+      "step": 4722
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 2.4993011122011786,
+      "learning_rate": 6.344342904962536e-07,
+      "loss": 0.0167,
+      "step": 4723
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 2.1082760871594544,
+      "learning_rate": 6.330260209769124e-07,
+      "loss": 0.0158,
+      "step": 4724
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 3.5954228865011024,
+      "learning_rate": 6.316192105447499e-07,
+      "loss": 0.0181,
+      "step": 4725
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 3.250561893307712,
+      "learning_rate": 6.302138596698032e-07,
+      "loss": 0.0179,
+      "step": 4726
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 5.034063607262308,
+      "learning_rate": 6.288099688216309e-07,
+      "loss": 0.0226,
+      "step": 4727
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 3.382398513985597,
+      "learning_rate": 6.27407538469294e-07,
+      "loss": 0.0223,
+      "step": 4728
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 3.5480387834752256,
+      "learning_rate": 6.260065690813754e-07,
+      "loss": 0.0244,
+      "step": 4729
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 3.894566614171884,
+      "learning_rate": 6.246070611259603e-07,
+      "loss": 0.038,
+      "step": 4730
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 4.908257487051644,
+      "learning_rate": 6.232090150706555e-07,
+      "loss": 0.0319,
+      "step": 4731
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 2.2356350192362955,
+      "learning_rate": 6.218124313825696e-07,
+      "loss": 0.0118,
+      "step": 4732
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 2.7833053267514285,
+      "learning_rate": 6.204173105283295e-07,
+      "loss": 0.0305,
+      "step": 4733
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 3.0358977610764937,
+      "learning_rate": 6.190236529740701e-07,
+      "loss": 0.0213,
+      "step": 4734
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 2.7799160284276017,
+      "learning_rate": 6.176314591854388e-07,
+      "loss": 0.024,
+      "step": 4735
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 2.411428363895045,
+      "learning_rate": 6.162407296275936e-07,
+      "loss": 0.024,
+      "step": 4736
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 4.987033320862642,
+      "learning_rate": 6.148514647652026e-07,
+      "loss": 0.0213,
+      "step": 4737
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 7.294162310830876,
+      "learning_rate": 6.134636650624448e-07,
+      "loss": 0.0306,
+      "step": 4738
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 1.9285035219081978,
+      "learning_rate": 6.120773309830108e-07,
+      "loss": 0.0151,
+      "step": 4739
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 2.9779680103281083,
+      "learning_rate": 6.106924629900996e-07,
+      "loss": 0.023,
+      "step": 4740
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 2.802445642894432,
+      "learning_rate": 6.09309061546422e-07,
+      "loss": 0.0204,
+      "step": 4741
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 3.5712560891174503,
+      "learning_rate": 6.079271271141979e-07,
+      "loss": 0.0258,
+      "step": 4742
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 2.7284199278884085,
+      "learning_rate": 6.065466601551578e-07,
+      "loss": 0.0249,
+      "step": 4743
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 2.1930907915124753,
+      "learning_rate": 6.051676611305401e-07,
+      "loss": 0.0327,
+      "step": 4744
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 3.0801008957709364,
+      "learning_rate": 6.037901305010951e-07,
+      "loss": 0.0156,
+      "step": 4745
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 2.2322852894555716,
+      "learning_rate": 6.024140687270813e-07,
+      "loss": 0.0134,
+      "step": 4746
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 3.7566488570881544,
+      "learning_rate": 6.010394762682659e-07,
+      "loss": 0.018,
+      "step": 4747
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 2.266574212516343,
+      "learning_rate": 5.996663535839275e-07,
+      "loss": 0.0201,
+      "step": 4748
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 2.644461947400835,
+      "learning_rate": 5.982947011328489e-07,
+      "loss": 0.0243,
+      "step": 4749
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 2.973314350873358,
+      "learning_rate": 5.969245193733275e-07,
+      "loss": 0.0241,
+      "step": 4750
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 4.151307716784977,
+      "learning_rate": 5.955558087631641e-07,
+      "loss": 0.021,
+      "step": 4751
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 4.048781625741955,
+      "learning_rate": 5.941885697596734e-07,
+      "loss": 0.0216,
+      "step": 4752
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 3.429703276468199,
+      "learning_rate": 5.928228028196714e-07,
+      "loss": 0.0236,
+      "step": 4753
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 2.739052446340753,
+      "learning_rate": 5.914585083994906e-07,
+      "loss": 0.017,
+      "step": 4754
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 3.04638508596632,
+      "learning_rate": 5.900956869549629e-07,
+      "loss": 0.0281,
+      "step": 4755
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 4.775164840894267,
+      "learning_rate": 5.887343389414363e-07,
+      "loss": 0.0363,
+      "step": 4756
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 1.9766077837059972,
+      "learning_rate": 5.873744648137592e-07,
+      "loss": 0.032,
+      "step": 4757
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 1.932804696724512,
+      "learning_rate": 5.860160650262925e-07,
+      "loss": 0.0171,
+      "step": 4758
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 2.4446932774769956,
+      "learning_rate": 5.846591400329021e-07,
+      "loss": 0.0123,
+      "step": 4759
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 2.8494754763154027,
+      "learning_rate": 5.833036902869626e-07,
+      "loss": 0.0229,
+      "step": 4760
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 3.070788144846567,
+      "learning_rate": 5.81949716241354e-07,
+      "loss": 0.0193,
+      "step": 4761
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 3.3717997476125032,
+      "learning_rate": 5.805972183484654e-07,
+      "loss": 0.0272,
+      "step": 4762
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 1.891084092238694,
+      "learning_rate": 5.792461970601903e-07,
+      "loss": 0.02,
+      "step": 4763
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 1.6405267772372447,
+      "learning_rate": 5.778966528279306e-07,
+      "loss": 0.016,
+      "step": 4764
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 3.494793807973005,
+      "learning_rate": 5.765485861025944e-07,
+      "loss": 0.0191,
+      "step": 4765
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 2.096825436091188,
+      "learning_rate": 5.752019973345963e-07,
+      "loss": 0.0208,
+      "step": 4766
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 1.678751370669293,
+      "learning_rate": 5.738568869738537e-07,
+      "loss": 0.0174,
+      "step": 4767
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 4.166473159072722,
+      "learning_rate": 5.725132554697971e-07,
+      "loss": 0.024,
+      "step": 4768
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 4.3633134523358335,
+      "learning_rate": 5.711711032713547e-07,
+      "loss": 0.0182,
+      "step": 4769
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 2.4438998290312313,
+      "learning_rate": 5.698304308269686e-07,
+      "loss": 0.0225,
+      "step": 4770
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 1.6826909960550536,
+      "learning_rate": 5.684912385845786e-07,
+      "loss": 0.0165,
+      "step": 4771
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 4.223897905047371,
+      "learning_rate": 5.671535269916373e-07,
+      "loss": 0.0189,
+      "step": 4772
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 4.451136477185219,
+      "learning_rate": 5.658172964950953e-07,
+      "loss": 0.0172,
+      "step": 4773
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 2.8782735665276964,
+      "learning_rate": 5.644825475414162e-07,
+      "loss": 0.0149,
+      "step": 4774
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 5.622361288092509,
+      "learning_rate": 5.631492805765609e-07,
+      "loss": 0.0232,
+      "step": 4775
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 2.0095858180774,
+      "learning_rate": 5.618174960459999e-07,
+      "loss": 0.0343,
+      "step": 4776
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 4.312258137394328,
+      "learning_rate": 5.604871943947071e-07,
+      "loss": 0.0214,
+      "step": 4777
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 12.714949832938647,
+      "learning_rate": 5.591583760671609e-07,
+      "loss": 0.0416,
+      "step": 4778
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 6.098583691916675,
+      "learning_rate": 5.578310415073451e-07,
+      "loss": 0.0201,
+      "step": 4779
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 3.0148079241593546,
+      "learning_rate": 5.565051911587455e-07,
+      "loss": 0.0257,
+      "step": 4780
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 2.1858188497490474,
+      "learning_rate": 5.551808254643543e-07,
+      "loss": 0.0237,
+      "step": 4781
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 7.095925230827684,
+      "learning_rate": 5.538579448666675e-07,
+      "loss": 0.0175,
+      "step": 4782
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 2.3749779288397534,
+      "learning_rate": 5.525365498076807e-07,
+      "loss": 0.0241,
+      "step": 4783
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 2.1169352892214444,
+      "learning_rate": 5.51216640728901e-07,
+      "loss": 0.0263,
+      "step": 4784
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 1.5113690124988577,
+      "learning_rate": 5.498982180713308e-07,
+      "loss": 0.0244,
+      "step": 4785
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 5.743204991318629,
+      "learning_rate": 5.485812822754826e-07,
+      "loss": 0.0365,
+      "step": 4786
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 5.539838806774285,
+      "learning_rate": 5.472658337813664e-07,
+      "loss": 0.0237,
+      "step": 4787
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 2.172371017801228,
+      "learning_rate": 5.459518730285007e-07,
+      "loss": 0.0176,
+      "step": 4788
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 7.723251710056527,
+      "learning_rate": 5.446394004559008e-07,
+      "loss": 0.0264,
+      "step": 4789
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 3.6273060070583125,
+      "learning_rate": 5.43328416502093e-07,
+      "loss": 0.0194,
+      "step": 4790
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 5.07385094981908,
+      "learning_rate": 5.420189216050969e-07,
+      "loss": 0.0348,
+      "step": 4791
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 2.3283643337319133,
+      "learning_rate": 5.407109162024409e-07,
+      "loss": 0.0178,
+      "step": 4792
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 1.9835055011656801,
+      "learning_rate": 5.394044007311544e-07,
+      "loss": 0.0307,
+      "step": 4793
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 8.575120390262455,
+      "learning_rate": 5.380993756277675e-07,
+      "loss": 0.0278,
+      "step": 4794
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 2.5516786025822182,
+      "learning_rate": 5.367958413283141e-07,
+      "loss": 0.0219,
+      "step": 4795
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 9.296882286540368,
+      "learning_rate": 5.354937982683283e-07,
+      "loss": 0.0288,
+      "step": 4796
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 3.5951696364880306,
+      "learning_rate": 5.341932468828481e-07,
+      "loss": 0.0207,
+      "step": 4797
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 3.7362598014081576,
+      "learning_rate": 5.328941876064114e-07,
+      "loss": 0.0259,
+      "step": 4798
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 2.4075304327331613,
+      "learning_rate": 5.315966208730578e-07,
+      "loss": 0.0219,
+      "step": 4799
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 2.245070060364752,
+      "learning_rate": 5.30300547116328e-07,
+      "loss": 0.024,
+      "step": 4800
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 4.6692741589089675,
+      "learning_rate": 5.290059667692655e-07,
+      "loss": 0.0308,
+      "step": 4801
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 3.000431411370373,
+      "learning_rate": 5.277128802644133e-07,
+      "loss": 0.0195,
+      "step": 4802
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 8.965965626231066,
+      "learning_rate": 5.264212880338154e-07,
+      "loss": 0.0414,
+      "step": 4803
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 5.876899527347979,
+      "learning_rate": 5.251311905090167e-07,
+      "loss": 0.0178,
+      "step": 4804
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 4.667587955450332,
+      "learning_rate": 5.238425881210624e-07,
+      "loss": 0.0205,
+      "step": 4805
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 20.266107099324124,
+      "learning_rate": 5.225554813004996e-07,
+      "loss": 0.0611,
+      "step": 4806
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 7.450685723335502,
+      "learning_rate": 5.21269870477375e-07,
+      "loss": 0.0396,
+      "step": 4807
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 2.8697958072570726,
+      "learning_rate": 5.199857560812316e-07,
+      "loss": 0.0218,
+      "step": 4808
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 5.772178641403427,
+      "learning_rate": 5.187031385411206e-07,
+      "loss": 0.0254,
+      "step": 4809
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 6.398504975526057,
+      "learning_rate": 5.174220182855844e-07,
+      "loss": 0.0217,
+      "step": 4810
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 7.595861533557351,
+      "learning_rate": 5.161423957426725e-07,
+      "loss": 0.0276,
+      "step": 4811
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 10.706839343586863,
+      "learning_rate": 5.148642713399272e-07,
+      "loss": 0.0207,
+      "step": 4812
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 9.248483629993162,
+      "learning_rate": 5.13587645504397e-07,
+      "loss": 0.0163,
+      "step": 4813
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 4.5853381448697315,
+      "learning_rate": 5.123125186626227e-07,
+      "loss": 0.032,
+      "step": 4814
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 2.6956221870281567,
+      "learning_rate": 5.110388912406517e-07,
+      "loss": 0.0199,
+      "step": 4815
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 6.193292360315304,
+      "learning_rate": 5.097667636640241e-07,
+      "loss": 0.0251,
+      "step": 4816
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 4.970186592154489,
+      "learning_rate": 5.084961363577817e-07,
+      "loss": 0.0285,
+      "step": 4817
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 2.2733355420940127,
+      "learning_rate": 5.072270097464649e-07,
+      "loss": 0.0255,
+      "step": 4818
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 1.282955595029532,
+      "learning_rate": 5.059593842541127e-07,
+      "loss": 0.0167,
+      "step": 4819
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 2.9173904214804534,
+      "learning_rate": 5.04693260304262e-07,
+      "loss": 0.0229,
+      "step": 4820
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 8.307122745953784,
+      "learning_rate": 5.034286383199488e-07,
+      "loss": 0.0229,
+      "step": 4821
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 10.234033109502867,
+      "learning_rate": 5.021655187237067e-07,
+      "loss": 0.0214,
+      "step": 4822
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 5.625721576167736,
+      "learning_rate": 5.009039019375672e-07,
+      "loss": 0.0244,
+      "step": 4823
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 6.9193873894305,
+      "learning_rate": 4.996437883830596e-07,
+      "loss": 0.0136,
+      "step": 4824
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 1.9958571703345365,
+      "learning_rate": 4.983851784812127e-07,
+      "loss": 0.0194,
+      "step": 4825
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 3.457883216533768,
+      "learning_rate": 4.97128072652549e-07,
+      "loss": 0.0174,
+      "step": 4826
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 3.264883698161827,
+      "learning_rate": 4.958724713170943e-07,
+      "loss": 0.0245,
+      "step": 4827
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 1.6502594031454794,
+      "learning_rate": 4.946183748943639e-07,
+      "loss": 0.0254,
+      "step": 4828
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 1.992537904277229,
+      "learning_rate": 4.933657838033795e-07,
+      "loss": 0.0189,
+      "step": 4829
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 2.8779533325492452,
+      "learning_rate": 4.921146984626507e-07,
+      "loss": 0.0241,
+      "step": 4830
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 3.3861004105405037,
+      "learning_rate": 4.908651192901926e-07,
+      "loss": 0.022,
+      "step": 4831
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 1.7445944302458545,
+      "learning_rate": 4.896170467035089e-07,
+      "loss": 0.0226,
+      "step": 4832
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 3.5655653643743537,
+      "learning_rate": 4.883704811196072e-07,
+      "loss": 0.0161,
+      "step": 4833
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 8.66061468381288,
+      "learning_rate": 4.871254229549855e-07,
+      "loss": 0.0523,
+      "step": 4834
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 2.657255822757317,
+      "learning_rate": 4.858818726256425e-07,
+      "loss": 0.028,
+      "step": 4835
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 5.472995401509639,
+      "learning_rate": 4.846398305470712e-07,
+      "loss": 0.0291,
+      "step": 4836
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 5.524011687090876,
+      "learning_rate": 4.833992971342604e-07,
+      "loss": 0.0233,
+      "step": 4837
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 8.501747778235401,
+      "learning_rate": 4.821602728016955e-07,
+      "loss": 0.0182,
+      "step": 4838
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 4.991085189378154,
+      "learning_rate": 4.809227579633585e-07,
+      "loss": 0.0291,
+      "step": 4839
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 5.373771634991806,
+      "learning_rate": 4.796867530327249e-07,
+      "loss": 0.0316,
+      "step": 4840
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 2.6554569867124536,
+      "learning_rate": 4.784522584227675e-07,
+      "loss": 0.0186,
+      "step": 4841
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 5.663613995175922,
+      "learning_rate": 4.772192745459536e-07,
+      "loss": 0.0249,
+      "step": 4842
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 2.1898976437413404,
+      "learning_rate": 4.7598780181424666e-07,
+      "loss": 0.0228,
+      "step": 4843
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 4.016259077763735,
+      "learning_rate": 4.7475784063910404e-07,
+      "loss": 0.0166,
+      "step": 4844
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 1.9221511010470675,
+      "learning_rate": 4.7352939143147927e-07,
+      "loss": 0.0231,
+      "step": 4845
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 3.563250640079929,
+      "learning_rate": 4.72302454601819e-07,
+      "loss": 0.0238,
+      "step": 4846
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 7.844911546765337,
+      "learning_rate": 4.7107703056006706e-07,
+      "loss": 0.0254,
+      "step": 4847
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 5.403934146700979,
+      "learning_rate": 4.6985311971565806e-07,
+      "loss": 0.0185,
+      "step": 4848
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 5.167903062966448,
+      "learning_rate": 4.6863072247752664e-07,
+      "loss": 0.016,
+      "step": 4849
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 2.1577262913144835,
+      "learning_rate": 4.67409839254096e-07,
+      "loss": 0.0212,
+      "step": 4850
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 1.3631800246611254,
+      "learning_rate": 4.66190470453286e-07,
+      "loss": 0.0197,
+      "step": 4851
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 3.344043147865704,
+      "learning_rate": 4.6497261648251134e-07,
+      "loss": 0.02,
+      "step": 4852
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 3.0874123639808926,
+      "learning_rate": 4.6375627774867925e-07,
+      "loss": 0.0212,
+      "step": 4853
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 7.328165344305696,
+      "learning_rate": 4.6254145465819134e-07,
+      "loss": 0.0213,
+      "step": 4854
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 3.224518234244968,
+      "learning_rate": 4.6132814761694234e-07,
+      "loss": 0.016,
+      "step": 4855
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 4.097008960876435,
+      "learning_rate": 4.6011635703032075e-07,
+      "loss": 0.0252,
+      "step": 4856
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 2.807322653482841,
+      "learning_rate": 4.589060833032083e-07,
+      "loss": 0.0268,
+      "step": 4857
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 2.3328664724594423,
+      "learning_rate": 4.5769732683997983e-07,
+      "loss": 0.0319,
+      "step": 4858
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 1.7226625611795265,
+      "learning_rate": 4.564900880445039e-07,
+      "loss": 0.021,
+      "step": 4859
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 4.033978626071074,
+      "learning_rate": 4.552843673201407e-07,
+      "loss": 0.0255,
+      "step": 4860
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 2.467384163882045,
+      "learning_rate": 4.540801650697446e-07,
+      "loss": 0.0245,
+      "step": 4861
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 2.9446504009525656,
+      "learning_rate": 4.528774816956616e-07,
+      "loss": 0.0178,
+      "step": 4862
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 2.235471595072864,
+      "learning_rate": 4.516763175997302e-07,
+      "loss": 0.0237,
+      "step": 4863
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 3.5832326050990857,
+      "learning_rate": 4.5047667318328215e-07,
+      "loss": 0.0277,
+      "step": 4864
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 7.581065600283904,
+      "learning_rate": 4.492785488471413e-07,
+      "loss": 0.0321,
+      "step": 4865
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 2.6477077947678427,
+      "learning_rate": 4.480819449916224e-07,
+      "loss": 0.0253,
+      "step": 4866
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 3.2013043262055714,
+      "learning_rate": 4.468868620165334e-07,
+      "loss": 0.0266,
+      "step": 4867
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 7.4880088058210115,
+      "learning_rate": 4.4569330032117496e-07,
+      "loss": 0.033,
+      "step": 4868
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 4.464721969976531,
+      "learning_rate": 4.445012603043347e-07,
+      "loss": 0.0327,
+      "step": 4869
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 3.1361654089894224,
+      "learning_rate": 4.4331074236430014e-07,
+      "loss": 0.0221,
+      "step": 4870
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 2.418337459504903,
+      "learning_rate": 4.421217468988409e-07,
+      "loss": 0.0188,
+      "step": 4871
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 2.973188602787579,
+      "learning_rate": 4.409342743052264e-07,
+      "loss": 0.0176,
+      "step": 4872
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 4.845221165052784,
+      "learning_rate": 4.3974832498020983e-07,
+      "loss": 0.0243,
+      "step": 4873
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 2.687507603812843,
+      "learning_rate": 4.385638993200425e-07,
+      "loss": 0.0227,
+      "step": 4874
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 5.577237802892938,
+      "learning_rate": 4.3738099772045963e-07,
+      "loss": 0.0234,
+      "step": 4875
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 1.9561421206253133,
+      "learning_rate": 4.3619962057669216e-07,
+      "loss": 0.025,
+      "step": 4876
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 4.175169583976657,
+      "learning_rate": 4.350197682834606e-07,
+      "loss": 0.0178,
+      "step": 4877
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 5.29437999575849,
+      "learning_rate": 4.338414412349745e-07,
+      "loss": 0.0204,
+      "step": 4878
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 2.0279366468727096,
+      "learning_rate": 4.3266463982493566e-07,
+      "loss": 0.0251,
+      "step": 4879
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 3.208336780233766,
+      "learning_rate": 4.314893644465351e-07,
+      "loss": 0.0241,
+      "step": 4880
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 3.0834492531396256,
+      "learning_rate": 4.303156154924537e-07,
+      "loss": 0.0139,
+      "step": 4881
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 3.2142530911093683,
+      "learning_rate": 4.291433933548633e-07,
+      "loss": 0.0255,
+      "step": 4882
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 2.640834983845967,
+      "learning_rate": 4.279726984254251e-07,
+      "loss": 0.0262,
+      "step": 4883
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 3.651611286643107,
+      "learning_rate": 4.268035310952906e-07,
+      "loss": 0.0197,
+      "step": 4884
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 2.2783581009506872,
+      "learning_rate": 4.256358917550979e-07,
+      "loss": 0.0192,
+      "step": 4885
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 5.428102406876919,
+      "learning_rate": 4.244697807949805e-07,
+      "loss": 0.0196,
+      "step": 4886
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 4.979308584900138,
+      "learning_rate": 4.2330519860455446e-07,
+      "loss": 0.0229,
+      "step": 4887
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 4.1409739409071635,
+      "learning_rate": 4.2214214557293133e-07,
+      "loss": 0.0306,
+      "step": 4888
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 4.79268503474881,
+      "learning_rate": 4.209806220887053e-07,
+      "loss": 0.03,
+      "step": 4889
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 2.352736577614645,
+      "learning_rate": 4.1982062853996695e-07,
+      "loss": 0.0215,
+      "step": 4890
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 3.0169446795812007,
+      "learning_rate": 4.1866216531428806e-07,
+      "loss": 0.0196,
+      "step": 4891
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 4.629541748577467,
+      "learning_rate": 4.1750523279873613e-07,
+      "loss": 0.0195,
+      "step": 4892
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 1.271554392634293,
+      "learning_rate": 4.1634983137986083e-07,
+      "loss": 0.0134,
+      "step": 4893
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 2.0119672947069476,
+      "learning_rate": 4.151959614437046e-07,
+      "loss": 0.0247,
+      "step": 4894
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 5.712184762810063,
+      "learning_rate": 4.1404362337579716e-07,
+      "loss": 0.029,
+      "step": 4895
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 5.945272020870616,
+      "learning_rate": 4.128928175611546e-07,
+      "loss": 0.029,
+      "step": 4896
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 4.408542849416339,
+      "learning_rate": 4.1174354438428434e-07,
+      "loss": 0.0188,
+      "step": 4897
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 7.5322367428462025,
+      "learning_rate": 4.105958042291791e-07,
+      "loss": 0.0363,
+      "step": 4898
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 1.6314161372537972,
+      "learning_rate": 4.0944959747931945e-07,
+      "loss": 0.0179,
+      "step": 4899
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 3.334468123285942,
+      "learning_rate": 4.0830492451767566e-07,
+      "loss": 0.0211,
+      "step": 4900
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 2.5645975778823953,
+      "learning_rate": 4.0716178572670405e-07,
+      "loss": 0.0224,
+      "step": 4901
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 4.513557316836511,
+      "learning_rate": 4.060201814883474e-07,
+      "loss": 0.0137,
+      "step": 4902
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 3.9386081706521936,
+      "learning_rate": 4.0488011218403844e-07,
+      "loss": 0.0279,
+      "step": 4903
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 2.9682175362012773,
+      "learning_rate": 4.0374157819469406e-07,
+      "loss": 0.0221,
+      "step": 4904
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 4.816340081088018,
+      "learning_rate": 4.0260457990072113e-07,
+      "loss": 0.0267,
+      "step": 4905
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 1.652181073212692,
+      "learning_rate": 4.014691176820107e-07,
+      "loss": 0.0142,
+      "step": 4906
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 2.860067524660608,
+      "learning_rate": 4.003351919179421e-07,
+      "loss": 0.0188,
+      "step": 4907
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 1.120466782402688,
+      "learning_rate": 3.9920280298738125e-07,
+      "loss": 0.0195,
+      "step": 4908
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 3.1345045016083892,
+      "learning_rate": 3.980719512686809e-07,
+      "loss": 0.0175,
+      "step": 4909
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 3.7019204611349212,
+      "learning_rate": 3.969426371396773e-07,
+      "loss": 0.0179,
+      "step": 4910
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 2.976700351913577,
+      "learning_rate": 3.9581486097769905e-07,
+      "loss": 0.0188,
+      "step": 4911
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 2.486198283586228,
+      "learning_rate": 3.946886231595526e-07,
+      "loss": 0.0215,
+      "step": 4912
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 2.020627642013814,
+      "learning_rate": 3.935639240615396e-07,
+      "loss": 0.0218,
+      "step": 4913
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 4.969087632628006,
+      "learning_rate": 3.924407640594391e-07,
+      "loss": 0.0237,
+      "step": 4914
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 2.887660959477525,
+      "learning_rate": 3.913191435285224e-07,
+      "loss": 0.0196,
+      "step": 4915
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 1.867078635117958,
+      "learning_rate": 3.9019906284354145e-07,
+      "loss": 0.0118,
+      "step": 4916
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 1.9465176519470353,
+      "learning_rate": 3.8908052237873863e-07,
+      "loss": 0.0207,
+      "step": 4917
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 9.8498836097476,
+      "learning_rate": 3.879635225078371e-07,
+      "loss": 0.0282,
+      "step": 4918
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 2.9397169943458836,
+      "learning_rate": 3.868480636040484e-07,
+      "loss": 0.0266,
+      "step": 4919
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 4.960618492091106,
+      "learning_rate": 3.857341460400665e-07,
+      "loss": 0.0177,
+      "step": 4920
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 4.712244580795838,
+      "learning_rate": 3.846217701880739e-07,
+      "loss": 0.0237,
+      "step": 4921
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 1.6971508399719888,
+      "learning_rate": 3.835109364197348e-07,
+      "loss": 0.018,
+      "step": 4922
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 1.7472712527944807,
+      "learning_rate": 3.8240164510620017e-07,
+      "loss": 0.0211,
+      "step": 4923
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 3.5480848152651516,
+      "learning_rate": 3.81293896618104e-07,
+      "loss": 0.0257,
+      "step": 4924
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 3.256572425168427,
+      "learning_rate": 3.8018769132556644e-07,
+      "loss": 0.0149,
+      "step": 4925
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 2.4597451706002413,
+      "learning_rate": 3.790830295981912e-07,
+      "loss": 0.0182,
+      "step": 4926
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 3.5212984506844682,
+      "learning_rate": 3.7797991180506643e-07,
+      "loss": 0.019,
+      "step": 4927
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 3.4875509288086013,
+      "learning_rate": 3.768783383147623e-07,
+      "loss": 0.0197,
+      "step": 4928
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 3.163210180212613,
+      "learning_rate": 3.757783094953382e-07,
+      "loss": 0.0292,
+      "step": 4929
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 2.5198870342212296,
+      "learning_rate": 3.746798257143314e-07,
+      "loss": 0.029,
+      "step": 4930
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 3.5847902090339847,
+      "learning_rate": 3.735828873387681e-07,
+      "loss": 0.0183,
+      "step": 4931
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 3.338194522323422,
+      "learning_rate": 3.724874947351531e-07,
+      "loss": 0.0196,
+      "step": 4932
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 1.9680640652690908,
+      "learning_rate": 3.7139364826948077e-07,
+      "loss": 0.0177,
+      "step": 4933
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 1.7834906752501898,
+      "learning_rate": 3.7030134830722207e-07,
+      "loss": 0.0166,
+      "step": 4934
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 2.341415129554604,
+      "learning_rate": 3.692105952133379e-07,
+      "loss": 0.0249,
+      "step": 4935
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 4.6407785758751166,
+      "learning_rate": 3.681213893522667e-07,
+      "loss": 0.0352,
+      "step": 4936
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 4.540938320445459,
+      "learning_rate": 3.670337310879335e-07,
+      "loss": 0.0237,
+      "step": 4937
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 6.387135784803516,
+      "learning_rate": 3.6594762078374536e-07,
+      "loss": 0.0277,
+      "step": 4938
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 3.665487214795183,
+      "learning_rate": 3.6486305880259085e-07,
+      "loss": 0.0257,
+      "step": 4939
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 4.212800882402541,
+      "learning_rate": 3.6378004550684355e-07,
+      "loss": 0.0213,
+      "step": 4940
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 3.2600196073500043,
+      "learning_rate": 3.626985812583572e-07,
+      "loss": 0.0193,
+      "step": 4941
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 2.0264755974293966,
+      "learning_rate": 3.6161866641847007e-07,
+      "loss": 0.0271,
+      "step": 4942
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 3.4707348585734024,
+      "learning_rate": 3.6054030134800243e-07,
+      "loss": 0.0162,
+      "step": 4943
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 2.0961417452701587,
+      "learning_rate": 3.594634864072527e-07,
+      "loss": 0.0232,
+      "step": 4944
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 4.063463726526555,
+      "learning_rate": 3.583882219560092e-07,
+      "loss": 0.032,
+      "step": 4945
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 2.098516311167729,
+      "learning_rate": 3.57314508353534e-07,
+      "loss": 0.0345,
+      "step": 4946
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 3.9001687600424275,
+      "learning_rate": 3.5624234595857787e-07,
+      "loss": 0.0237,
+      "step": 4947
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 1.3522189099210975,
+      "learning_rate": 3.551717351293676e-07,
+      "loss": 0.0137,
+      "step": 4948
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 5.448938058753399,
+      "learning_rate": 3.541026762236166e-07,
+      "loss": 0.0335,
+      "step": 4949
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 3.39809976716535,
+      "learning_rate": 3.5303516959851405e-07,
+      "loss": 0.0267,
+      "step": 4950
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 2.0973599945236923,
+      "learning_rate": 3.519692156107379e-07,
+      "loss": 0.02,
+      "step": 4951
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 4.928976500976884,
+      "learning_rate": 3.509048146164401e-07,
+      "loss": 0.0232,
+      "step": 4952
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 4.341400781737448,
+      "learning_rate": 3.4984196697125827e-07,
+      "loss": 0.021,
+      "step": 4953
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 1.5588549990820906,
+      "learning_rate": 3.4878067303030836e-07,
+      "loss": 0.0135,
+      "step": 4954
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 5.866237503465065,
+      "learning_rate": 3.4772093314818957e-07,
+      "loss": 0.0292,
+      "step": 4955
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 2.889730819504992,
+      "learning_rate": 3.4666274767897967e-07,
+      "loss": 0.0207,
+      "step": 4956
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 6.8785021783137665,
+      "learning_rate": 3.456061169762392e-07,
+      "loss": 0.0226,
+      "step": 4957
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 2.4293314184044568,
+      "learning_rate": 3.44551041393007e-07,
+      "loss": 0.0219,
+      "step": 4958
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 2.5630792149658586,
+      "learning_rate": 3.434975212818048e-07,
+      "loss": 0.0239,
+      "step": 4959
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 1.6349579915872139,
+      "learning_rate": 3.424455569946317e-07,
+      "loss": 0.0134,
+      "step": 4960
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 6.31085497252182,
+      "learning_rate": 3.4139514888296975e-07,
+      "loss": 0.0312,
+      "step": 4961
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 5.08757040363162,
+      "learning_rate": 3.403462972977789e-07,
+      "loss": 0.0244,
+      "step": 4962
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 4.263621591200577,
+      "learning_rate": 3.392990025895004e-07,
+      "loss": 0.0243,
+      "step": 4963
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 2.1468229060472144,
+      "learning_rate": 3.3825326510805556e-07,
+      "loss": 0.0219,
+      "step": 4964
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 3.37958648447092,
+      "learning_rate": 3.372090852028437e-07,
+      "loss": 0.0281,
+      "step": 4965
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 1.7362602943935712,
+      "learning_rate": 3.361664632227446e-07,
+      "loss": 0.0188,
+      "step": 4966
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 2.274637522937686,
+      "learning_rate": 3.3512539951611856e-07,
+      "loss": 0.0234,
+      "step": 4967
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 5.0068330161997805,
+      "learning_rate": 3.3408589443080395e-07,
+      "loss": 0.0206,
+      "step": 4968
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 2.6474551601173517,
+      "learning_rate": 3.3304794831411804e-07,
+      "loss": 0.0234,
+      "step": 4969
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 2.496187749477123,
+      "learning_rate": 3.3201156151285994e-07,
+      "loss": 0.0258,
+      "step": 4970
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 4.010290134183546,
+      "learning_rate": 3.309767343733028e-07,
+      "loss": 0.0176,
+      "step": 4971
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 1.5675425735480626,
+      "learning_rate": 3.299434672412044e-07,
+      "loss": 0.0194,
+      "step": 4972
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 2.539100465424783,
+      "learning_rate": 3.2891176046179583e-07,
+      "loss": 0.019,
+      "step": 4973
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 2.0917188211506823,
+      "learning_rate": 3.278816143797919e-07,
+      "loss": 0.0194,
+      "step": 4974
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 4.350767461798503,
+      "learning_rate": 3.2685302933938177e-07,
+      "loss": 0.0386,
+      "step": 4975
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 3.8928169385095948,
+      "learning_rate": 3.2582600568423715e-07,
+      "loss": 0.0244,
+      "step": 4976
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 5.787909162323301,
+      "learning_rate": 3.2480054375750305e-07,
+      "loss": 0.0193,
+      "step": 4977
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 2.4001819336533567,
+      "learning_rate": 3.237766439018064e-07,
+      "loss": 0.02,
+      "step": 4978
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 1.8709562877855663,
+      "learning_rate": 3.227543064592514e-07,
+      "loss": 0.0191,
+      "step": 4979
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 1.9065701916885103,
+      "learning_rate": 3.2173353177142044e-07,
+      "loss": 0.0203,
+      "step": 4980
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 2.5120744516072513,
+      "learning_rate": 3.207143201793722e-07,
+      "loss": 0.0187,
+      "step": 4981
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 1.8222263981036755,
+      "learning_rate": 3.1969667202364496e-07,
+      "loss": 0.0223,
+      "step": 4982
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 3.4179200252243804,
+      "learning_rate": 3.1868058764425337e-07,
+      "loss": 0.0249,
+      "step": 4983
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 5.7498072907102635,
+      "learning_rate": 3.1766606738069084e-07,
+      "loss": 0.0275,
+      "step": 4984
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 5.041392321955332,
+      "learning_rate": 3.166531115719268e-07,
+      "loss": 0.021,
+      "step": 4985
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 1.822728818625865,
+      "learning_rate": 3.1564172055640994e-07,
+      "loss": 0.0185,
+      "step": 4986
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 2.7917425319945153,
+      "learning_rate": 3.1463189467206166e-07,
+      "loss": 0.023,
+      "step": 4987
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 4.366628592986311,
+      "learning_rate": 3.1362363425628763e-07,
+      "loss": 0.0171,
+      "step": 4988
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 3.3852483744219595,
+      "learning_rate": 3.1261693964596275e-07,
+      "loss": 0.0244,
+      "step": 4989
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 3.7515129198680555,
+      "learning_rate": 3.116118111774452e-07,
+      "loss": 0.0148,
+      "step": 4990
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 1.9817429833762845,
+      "learning_rate": 3.106082491865647e-07,
+      "loss": 0.0157,
+      "step": 4991
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 2.414884784646765,
+      "learning_rate": 3.0960625400863253e-07,
+      "loss": 0.0179,
+      "step": 4992
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 5.250163439928367,
+      "learning_rate": 3.0860582597843137e-07,
+      "loss": 0.0252,
+      "step": 4993
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 4.698750314158885,
+      "learning_rate": 3.0760696543022496e-07,
+      "loss": 0.0187,
+      "step": 4994
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 12.932454054533897,
+      "learning_rate": 3.066096726977502e-07,
+      "loss": 0.0597,
+      "step": 4995
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 6.094441081177688,
+      "learning_rate": 3.056139481142206e-07,
+      "loss": 0.0201,
+      "step": 4996
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 7.143468273686904,
+      "learning_rate": 3.0461979201232674e-07,
+      "loss": 0.0393,
+      "step": 4997
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 1.6228527593437256,
+      "learning_rate": 3.0362720472423503e-07,
+      "loss": 0.018,
+      "step": 4998
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 8.117921391666709,
+      "learning_rate": 3.026361865815869e-07,
+      "loss": 0.036,
+      "step": 4999
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 4.951489289600751,
+      "learning_rate": 3.016467379154997e-07,
+      "loss": 0.0274,
+      "step": 5000
+    },
+    {
+      "epoch": 3.57,
+      "eval_avg_AUC": 0.8360170574106709,
+      "eval_avg_Accuracy": 0.743824602122016,
+      "eval_avg_Accuracy-right": 0.8943524194600235,
+      "eval_avg_Accuracy-wrong": 0.4813509210825563,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.7115880873199548,
+      "eval_last_AUC": 0.8253922573221322,
+      "eval_last_Accuracy": 0.7799651856763926,
+      "eval_last_Accuracy-right": 0.8460284335463676,
+      "eval_last_Accuracy-wrong": 0.6647714350693654,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.706694551574502,
+      "eval_max_AUC": 0.7861393054105478,
+      "eval_max_Accuracy": 0.647256299734748,
+      "eval_max_Accuracy-right": 0.988065736272336,
+      "eval_max_Accuracy-wrong": 0.05299067546054128,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.6590420344459055,
+      "eval_min_AUC": 0.8438178165751404,
+      "eval_min_Accuracy": 0.7740384615384616,
+      "eval_min_Accuracy-right": 0.7930089996087126,
+      "eval_min_Accuracy-wrong": 0.7409597452808733,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.7097302187281208,
+      "eval_prod_AUC": 0.8398013852657344,
+      "eval_prod_Accuracy": 0.7416279840848806,
+      "eval_prod_Accuracy-right": 0.6696882744228512,
+      "eval_prod_Accuracy-wrong": 0.8670684557652946,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.7074705074291826,
+      "eval_runtime": 246.8796,
+      "eval_samples_per_second": 97.732,
+      "eval_steps_per_second": 3.054,
+      "eval_sum_AUC": 0.7124423224936764,
+      "eval_sum_Accuracy": 0.6394645225464191,
+      "eval_sum_Accuracy-right": 0.9969349158732229,
+      "eval_sum_Accuracy-wrong": 0.016147373209006142,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6847078223460834,
+      "step": 5000
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 3.654880951851883,
+      "learning_rate": 3.0065885905656733e-07,
+      "loss": 0.0203,
+      "step": 5001
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 5.243962114831206,
+      "learning_rate": 2.99672550334858e-07,
+      "loss": 0.0173,
+      "step": 5002
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 1.4092832840135632,
+      "learning_rate": 2.986878120799158e-07,
+      "loss": 0.0208,
+      "step": 5003
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 1.554011978745569,
+      "learning_rate": 2.977046446207604e-07,
+      "loss": 0.0176,
+      "step": 5004
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 4.669169960672686,
+      "learning_rate": 2.967230482858863e-07,
+      "loss": 0.0252,
+      "step": 5005
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 6.6094996745112695,
+      "learning_rate": 2.957430234032627e-07,
+      "loss": 0.0215,
+      "step": 5006
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 11.110930559363911,
+      "learning_rate": 2.947645703003338e-07,
+      "loss": 0.0281,
+      "step": 5007
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 5.9334704383066175,
+      "learning_rate": 2.937876893040209e-07,
+      "loss": 0.0171,
+      "step": 5008
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 4.423469270196388,
+      "learning_rate": 2.9281238074071463e-07,
+      "loss": 0.0315,
+      "step": 5009
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 1.592156374591865,
+      "learning_rate": 2.9183864493628756e-07,
+      "loss": 0.0227,
+      "step": 5010
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 4.119354327917944,
+      "learning_rate": 2.908664822160806e-07,
+      "loss": 0.0254,
+      "step": 5011
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 4.884003650845199,
+      "learning_rate": 2.898958929049117e-07,
+      "loss": 0.0186,
+      "step": 5012
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 1.961721413748107,
+      "learning_rate": 2.889268773270731e-07,
+      "loss": 0.0222,
+      "step": 5013
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 2.7963462297389414,
+      "learning_rate": 2.879594358063303e-07,
+      "loss": 0.0233,
+      "step": 5014
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 2.468003197724322,
+      "learning_rate": 2.869935686659248e-07,
+      "loss": 0.0169,
+      "step": 5015
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 8.136562695225999,
+      "learning_rate": 2.8602927622856935e-07,
+      "loss": 0.0215,
+      "step": 5016
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 5.485663261521494,
+      "learning_rate": 2.8506655881645305e-07,
+      "loss": 0.0159,
+      "step": 5017
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 2.0505850995875723,
+      "learning_rate": 2.841054167512369e-07,
+      "loss": 0.0294,
+      "step": 5018
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 7.481500531403321,
+      "learning_rate": 2.8314585035405683e-07,
+      "loss": 0.0284,
+      "step": 5019
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 2.9777680960797746,
+      "learning_rate": 2.8218785994552136e-07,
+      "loss": 0.028,
+      "step": 5020
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 1.8732685907666995,
+      "learning_rate": 2.8123144584571326e-07,
+      "loss": 0.0196,
+      "step": 5021
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 5.086031806559671,
+      "learning_rate": 2.8027660837418813e-07,
+      "loss": 0.0253,
+      "step": 5022
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 6.7694477711519685,
+      "learning_rate": 2.793233478499752e-07,
+      "loss": 0.0291,
+      "step": 5023
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 3.4396884785592605,
+      "learning_rate": 2.7837166459157625e-07,
+      "loss": 0.0202,
+      "step": 5024
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 5.189513945664995,
+      "learning_rate": 2.77421558916966e-07,
+      "loss": 0.0231,
+      "step": 5025
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 2.550855581854903,
+      "learning_rate": 2.764730311435931e-07,
+      "loss": 0.014,
+      "step": 5026
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 4.1416552070916355,
+      "learning_rate": 2.755260815883781e-07,
+      "loss": 0.0367,
+      "step": 5027
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 2.011663446639588,
+      "learning_rate": 2.745807105677145e-07,
+      "loss": 0.0251,
+      "step": 5028
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 7.231248835410802,
+      "learning_rate": 2.736369183974685e-07,
+      "loss": 0.02,
+      "step": 5029
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 2.63783442077024,
+      "learning_rate": 2.726947053929768e-07,
+      "loss": 0.0271,
+      "step": 5030
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 3.454747329266388,
+      "learning_rate": 2.7175407186905367e-07,
+      "loss": 0.017,
+      "step": 5031
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 5.611368469459556,
+      "learning_rate": 2.708150181399788e-07,
+      "loss": 0.0202,
+      "step": 5032
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 2.2270437823391966,
+      "learning_rate": 2.698775445195101e-07,
+      "loss": 0.0245,
+      "step": 5033
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 2.6227149241256744,
+      "learning_rate": 2.689416513208726e-07,
+      "loss": 0.0145,
+      "step": 5034
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 3.930124224787002,
+      "learning_rate": 2.6800733885676833e-07,
+      "loss": 0.0425,
+      "step": 5035
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 3.6236506163310427,
+      "learning_rate": 2.6707460743936653e-07,
+      "loss": 0.0212,
+      "step": 5036
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 7.174749354653931,
+      "learning_rate": 2.6614345738031014e-07,
+      "loss": 0.0288,
+      "step": 5037
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 3.9949991370218254,
+      "learning_rate": 2.6521388899071467e-07,
+      "loss": 0.0237,
+      "step": 5038
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 1.4321127479377347,
+      "learning_rate": 2.642859025811656e-07,
+      "loss": 0.017,
+      "step": 5039
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 4.085364605840262,
+      "learning_rate": 2.633594984617199e-07,
+      "loss": 0.0229,
+      "step": 5040
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 4.510319037477354,
+      "learning_rate": 2.624346769419078e-07,
+      "loss": 0.0276,
+      "step": 5041
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 2.265080203783568,
+      "learning_rate": 2.6151143833072824e-07,
+      "loss": 0.0195,
+      "step": 5042
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 2.381010118389464,
+      "learning_rate": 2.605897829366527e-07,
+      "loss": 0.0164,
+      "step": 5043
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 2.1997765716885223,
+      "learning_rate": 2.596697110676233e-07,
+      "loss": 0.0253,
+      "step": 5044
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 2.80303572995716,
+      "learning_rate": 2.5875122303105403e-07,
+      "loss": 0.0303,
+      "step": 5045
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 2.042229652025857,
+      "learning_rate": 2.5783431913382673e-07,
+      "loss": 0.0197,
+      "step": 5046
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 3.4939362138330483,
+      "learning_rate": 2.5691899968229904e-07,
+      "loss": 0.0174,
+      "step": 5047
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 1.4919612882154958,
+      "learning_rate": 2.560052649822925e-07,
+      "loss": 0.0164,
+      "step": 5048
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 2.9015314648170483,
+      "learning_rate": 2.5509311533910674e-07,
+      "loss": 0.0202,
+      "step": 5049
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 2.6111830304001855,
+      "learning_rate": 2.5418255105750465e-07,
+      "loss": 0.0182,
+      "step": 5050
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 3.25436373070333,
+      "learning_rate": 2.532735724417251e-07,
+      "loss": 0.0296,
+      "step": 5051
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 2.0050448829571654,
+      "learning_rate": 2.52366179795473e-07,
+      "loss": 0.0205,
+      "step": 5052
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 5.903582016575689,
+      "learning_rate": 2.5146037342192673e-07,
+      "loss": 0.0338,
+      "step": 5053
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 3.84120253557257,
+      "learning_rate": 2.505561536237311e-07,
+      "loss": 0.0205,
+      "step": 5054
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 3.1135264434051866,
+      "learning_rate": 2.496535207030043e-07,
+      "loss": 0.0112,
+      "step": 5055
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 1.5988271065115593,
+      "learning_rate": 2.4875247496133234e-07,
+      "loss": 0.0239,
+      "step": 5056
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 2.5038282984009146,
+      "learning_rate": 2.4785301669977116e-07,
+      "loss": 0.0233,
+      "step": 5057
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 3.2800353162346583,
+      "learning_rate": 2.469551462188463e-07,
+      "loss": 0.0207,
+      "step": 5058
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 2.5254055880544684,
+      "learning_rate": 2.460588638185535e-07,
+      "loss": 0.0244,
+      "step": 5059
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 2.135851694029928,
+      "learning_rate": 2.45164169798357e-07,
+      "loss": 0.0271,
+      "step": 5060
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 5.307448036798728,
+      "learning_rate": 2.4427106445719053e-07,
+      "loss": 0.0141,
+      "step": 5061
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 3.9909825038257694,
+      "learning_rate": 2.4337954809345807e-07,
+      "loss": 0.0143,
+      "step": 5062
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 5.963517100815372,
+      "learning_rate": 2.4248962100503095e-07,
+      "loss": 0.0257,
+      "step": 5063
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 3.314407673144177,
+      "learning_rate": 2.416012834892506e-07,
+      "loss": 0.0271,
+      "step": 5064
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 3.1641223902931577,
+      "learning_rate": 2.4071453584292693e-07,
+      "loss": 0.0187,
+      "step": 5065
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 2.1701515875255235,
+      "learning_rate": 2.3982937836233954e-07,
+      "loss": 0.0261,
+      "step": 5066
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 3.6782775860952124,
+      "learning_rate": 2.389458113432347e-07,
+      "loss": 0.0218,
+      "step": 5067
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 3.2993387317839593,
+      "learning_rate": 2.380638350808301e-07,
+      "loss": 0.0245,
+      "step": 5068
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 1.5026356633345086,
+      "learning_rate": 2.371834498698089e-07,
+      "loss": 0.0189,
+      "step": 5069
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 3.729960127563879,
+      "learning_rate": 2.363046560043264e-07,
+      "loss": 0.0256,
+      "step": 5070
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 3.429117472534288,
+      "learning_rate": 2.3542745377800046e-07,
+      "loss": 0.0197,
+      "step": 5071
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 2.2548825921747713,
+      "learning_rate": 2.3455184348392446e-07,
+      "loss": 0.014,
+      "step": 5072
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 5.749312446063484,
+      "learning_rate": 2.3367782541465268e-07,
+      "loss": 0.0181,
+      "step": 5073
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 4.623364041936797,
+      "learning_rate": 2.3280539986221317e-07,
+      "loss": 0.0201,
+      "step": 5074
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 2.473930184609748,
+      "learning_rate": 2.3193456711809837e-07,
+      "loss": 0.025,
+      "step": 5075
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 3.597217833878548,
+      "learning_rate": 2.3106532747327104e-07,
+      "loss": 0.0147,
+      "step": 5076
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 2.8255040242530125,
+      "learning_rate": 2.3019768121815777e-07,
+      "loss": 0.0132,
+      "step": 5077
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 2.3758633766254955,
+      "learning_rate": 2.2933162864265836e-07,
+      "loss": 0.015,
+      "step": 5078
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 4.64748860266476,
+      "learning_rate": 2.2846717003613462e-07,
+      "loss": 0.0245,
+      "step": 5079
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 3.860781440775316,
+      "learning_rate": 2.2760430568741943e-07,
+      "loss": 0.021,
+      "step": 5080
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 2.042023533674793,
+      "learning_rate": 2.2674303588481162e-07,
+      "loss": 0.0194,
+      "step": 5081
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 1.7607599316113298,
+      "learning_rate": 2.258833609160771e-07,
+      "loss": 0.021,
+      "step": 5082
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 4.441214972475646,
+      "learning_rate": 2.2502528106845e-07,
+      "loss": 0.0197,
+      "step": 5083
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 4.545608016353152,
+      "learning_rate": 2.241687966286299e-07,
+      "loss": 0.0257,
+      "step": 5084
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 1.6754502037667738,
+      "learning_rate": 2.233139078827845e-07,
+      "loss": 0.0189,
+      "step": 5085
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 1.0124530351530259,
+      "learning_rate": 2.2246061511654816e-07,
+      "loss": 0.013,
+      "step": 5086
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 1.3560170545749979,
+      "learning_rate": 2.2160891861502165e-07,
+      "loss": 0.0203,
+      "step": 5087
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 3.614265986080177,
+      "learning_rate": 2.2075881866277348e-07,
+      "loss": 0.0145,
+      "step": 5088
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 1.8367535225634963,
+      "learning_rate": 2.199103155438359e-07,
+      "loss": 0.0218,
+      "step": 5089
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 2.112045433383258,
+      "learning_rate": 2.1906340954171212e-07,
+      "loss": 0.0324,
+      "step": 5090
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 2.89243344264716,
+      "learning_rate": 2.1821810093936636e-07,
+      "loss": 0.0231,
+      "step": 5091
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 2.5341079427529376,
+      "learning_rate": 2.1737439001923488e-07,
+      "loss": 0.0171,
+      "step": 5092
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 12.959994215228567,
+      "learning_rate": 2.1653227706321388e-07,
+      "loss": 0.055,
+      "step": 5093
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 2.8515096352994016,
+      "learning_rate": 2.156917623526722e-07,
+      "loss": 0.0296,
+      "step": 5094
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 3.541543190563925,
+      "learning_rate": 2.1485284616843904e-07,
+      "loss": 0.0184,
+      "step": 5095
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 3.7930353728065533,
+      "learning_rate": 2.140155287908141e-07,
+      "loss": 0.0248,
+      "step": 5096
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 3.481361108056802,
+      "learning_rate": 2.131798104995586e-07,
+      "loss": 0.0275,
+      "step": 5097
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 6.486064903589152,
+      "learning_rate": 2.123456915739025e-07,
+      "loss": 0.0203,
+      "step": 5098
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 3.4582584714064426,
+      "learning_rate": 2.115131722925401e-07,
+      "loss": 0.0294,
+      "step": 5099
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 1.3294381184143647,
+      "learning_rate": 2.1068225293363166e-07,
+      "loss": 0.023,
+      "step": 5100
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 2.312292253078772,
+      "learning_rate": 2.0985293377480342e-07,
+      "loss": 0.0206,
+      "step": 5101
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 3.2619168492415973,
+      "learning_rate": 2.0902521509314543e-07,
+      "loss": 0.0173,
+      "step": 5102
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 3.81893103016964,
+      "learning_rate": 2.0819909716521426e-07,
+      "loss": 0.0235,
+      "step": 5103
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 2.3150610495838744,
+      "learning_rate": 2.0737458026703182e-07,
+      "loss": 0.0168,
+      "step": 5104
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 4.629763159973222,
+      "learning_rate": 2.0655166467408283e-07,
+      "loss": 0.0433,
+      "step": 5105
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 5.583557149885652,
+      "learning_rate": 2.057303506613212e-07,
+      "loss": 0.0238,
+      "step": 5106
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 3.0910682729647805,
+      "learning_rate": 2.049106385031602e-07,
+      "loss": 0.0221,
+      "step": 5107
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 4.763394187633355,
+      "learning_rate": 2.0409252847348404e-07,
+      "loss": 0.022,
+      "step": 5108
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 1.864205678587097,
+      "learning_rate": 2.032760208456358e-07,
+      "loss": 0.0174,
+      "step": 5109
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 5.291056945540742,
+      "learning_rate": 2.0246111589242835e-07,
+      "loss": 0.0265,
+      "step": 5110
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 1.921185917208169,
+      "learning_rate": 2.0164781388613386e-07,
+      "loss": 0.0202,
+      "step": 5111
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 7.495862039740579,
+      "learning_rate": 2.0083611509849443e-07,
+      "loss": 0.0235,
+      "step": 5112
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 1.3689942096771623,
+      "learning_rate": 2.0002601980071145e-07,
+      "loss": 0.0245,
+      "step": 5113
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 4.366828293872557,
+      "learning_rate": 1.9921752826345397e-07,
+      "loss": 0.0237,
+      "step": 5114
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 1.9223888609093516,
+      "learning_rate": 1.9841064075685367e-07,
+      "loss": 0.0229,
+      "step": 5115
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 1.2881973687247583,
+      "learning_rate": 1.9760535755050715e-07,
+      "loss": 0.0125,
+      "step": 5116
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 2.6967817037967867,
+      "learning_rate": 1.9680167891347356e-07,
+      "loss": 0.016,
+      "step": 5117
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 3.439604203890696,
+      "learning_rate": 1.9599960511427761e-07,
+      "loss": 0.0323,
+      "step": 5118
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 1.5596126183501706,
+      "learning_rate": 1.9519913642090715e-07,
+      "loss": 0.0179,
+      "step": 5119
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 2.9106021839335043,
+      "learning_rate": 1.9440027310081323e-07,
+      "loss": 0.0309,
+      "step": 5120
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 4.111574300372222,
+      "learning_rate": 1.9360301542091065e-07,
+      "loss": 0.0214,
+      "step": 5121
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 3.430689782203096,
+      "learning_rate": 1.9280736364757912e-07,
+      "loss": 0.0167,
+      "step": 5122
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 2.7827580688892324,
+      "learning_rate": 1.9201331804665934e-07,
+      "loss": 0.0135,
+      "step": 5123
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 1.750804909510771,
+      "learning_rate": 1.9122087888345798e-07,
+      "loss": 0.0179,
+      "step": 5124
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 3.102904067177262,
+      "learning_rate": 1.9043004642274266e-07,
+      "loss": 0.0181,
+      "step": 5125
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 1.4913641627966314,
+      "learning_rate": 1.896408209287459e-07,
+      "loss": 0.0184,
+      "step": 5126
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 1.2369411725617043,
+      "learning_rate": 1.888532026651624e-07,
+      "loss": 0.0197,
+      "step": 5127
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 3.1080100506244572,
+      "learning_rate": 1.880671918951499e-07,
+      "loss": 0.016,
+      "step": 5128
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 3.60563971987711,
+      "learning_rate": 1.8728278888132944e-07,
+      "loss": 0.0275,
+      "step": 5129
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 2.013124995048135,
+      "learning_rate": 1.864999938857842e-07,
+      "loss": 0.0272,
+      "step": 5130
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 3.569223411896495,
+      "learning_rate": 1.8571880717006218e-07,
+      "loss": 0.013,
+      "step": 5131
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 2.717475672756776,
+      "learning_rate": 1.8493922899516902e-07,
+      "loss": 0.0257,
+      "step": 5132
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 3.2812818109675796,
+      "learning_rate": 1.8416125962157971e-07,
+      "loss": 0.0374,
+      "step": 5133
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 6.769023739742816,
+      "learning_rate": 1.8338489930922632e-07,
+      "loss": 0.0213,
+      "step": 5134
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 5.479076444685034,
+      "learning_rate": 1.8261014831750633e-07,
+      "loss": 0.0232,
+      "step": 5135
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 3.1296442552646995,
+      "learning_rate": 1.8183700690527717e-07,
+      "loss": 0.0199,
+      "step": 5136
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 2.5439097741169814,
+      "learning_rate": 1.810654753308616e-07,
+      "loss": 0.0301,
+      "step": 5137
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 1.373119819713574,
+      "learning_rate": 1.8029555385204067e-07,
+      "loss": 0.0166,
+      "step": 5138
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 2.135208763283895,
+      "learning_rate": 1.795272427260608e-07,
+      "loss": 0.0168,
+      "step": 5139
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 2.092430470504436,
+      "learning_rate": 1.7876054220962835e-07,
+      "loss": 0.0161,
+      "step": 5140
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 2.8211956696294695,
+      "learning_rate": 1.779954525589128e-07,
+      "loss": 0.0222,
+      "step": 5141
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 1.7382385750874663,
+      "learning_rate": 1.7723197402954419e-07,
+      "loss": 0.0143,
+      "step": 5142
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 3.548820464246968,
+      "learning_rate": 1.7647010687661558e-07,
+      "loss": 0.0152,
+      "step": 5143
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 1.9248454205191103,
+      "learning_rate": 1.757098513546801e-07,
+      "loss": 0.0228,
+      "step": 5144
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 5.791954592906748,
+      "learning_rate": 1.74951207717754e-07,
+      "loss": 0.0251,
+      "step": 5145
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 3.654095689991998,
+      "learning_rate": 1.7419417621931388e-07,
+      "loss": 0.021,
+      "step": 5146
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 3.678062581923196,
+      "learning_rate": 1.7343875711229864e-07,
+      "loss": 0.0274,
+      "step": 5147
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 2.4221492030173604,
+      "learning_rate": 1.7268495064910574e-07,
+      "loss": 0.0193,
+      "step": 5148
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 2.665876583116908,
+      "learning_rate": 1.719327570815993e-07,
+      "loss": 0.0284,
+      "step": 5149
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 4.023954591813659,
+      "learning_rate": 1.711821766610977e-07,
+      "loss": 0.0265,
+      "step": 5150
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 2.666367457998923,
+      "learning_rate": 1.704332096383865e-07,
+      "loss": 0.0207,
+      "step": 5151
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 4.801752871559416,
+      "learning_rate": 1.696858562637077e-07,
+      "loss": 0.014,
+      "step": 5152
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 2.7854192881731974,
+      "learning_rate": 1.689401167867677e-07,
+      "loss": 0.0282,
+      "step": 5153
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 2.964894722808635,
+      "learning_rate": 1.6819599145672993e-07,
+      "loss": 0.028,
+      "step": 5154
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 3.7778238501101375,
+      "learning_rate": 1.674534805222222e-07,
+      "loss": 0.0338,
+      "step": 5155
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 2.277752038612983,
+      "learning_rate": 1.667125842313305e-07,
+      "loss": 0.0182,
+      "step": 5156
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 1.2367806497125986,
+      "learning_rate": 1.6597330283160184e-07,
+      "loss": 0.0113,
+      "step": 5157
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 8.140836309513228,
+      "learning_rate": 1.6523563657004416e-07,
+      "loss": 0.0375,
+      "step": 5158
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 2.5710975609049305,
+      "learning_rate": 1.644995856931253e-07,
+      "loss": 0.02,
+      "step": 5159
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 2.969233743080048,
+      "learning_rate": 1.6376515044677354e-07,
+      "loss": 0.0219,
+      "step": 5160
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 2.177218001293877,
+      "learning_rate": 1.630323310763776e-07,
+      "loss": 0.0217,
+      "step": 5161
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 2.1841237881191446,
+      "learning_rate": 1.6230112782678608e-07,
+      "loss": 0.0266,
+      "step": 5162
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 2.4817841006206276,
+      "learning_rate": 1.6157154094230744e-07,
+      "loss": 0.0182,
+      "step": 5163
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 4.016931727249433,
+      "learning_rate": 1.6084357066670997e-07,
+      "loss": 0.0272,
+      "step": 5164
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 1.879190602814266,
+      "learning_rate": 1.601172172432225e-07,
+      "loss": 0.0209,
+      "step": 5165
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 2.778061083524743,
+      "learning_rate": 1.5939248091453252e-07,
+      "loss": 0.0218,
+      "step": 5166
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 5.732208780209183,
+      "learning_rate": 1.5866936192278915e-07,
+      "loss": 0.0325,
+      "step": 5167
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 1.4706232698161006,
+      "learning_rate": 1.5794786050959797e-07,
+      "loss": 0.0134,
+      "step": 5168
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 6.98594884392636,
+      "learning_rate": 1.5722797691602842e-07,
+      "loss": 0.0319,
+      "step": 5169
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 3.349426832853079,
+      "learning_rate": 1.5650971138260473e-07,
+      "loss": 0.0248,
+      "step": 5170
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 4.976579295378491,
+      "learning_rate": 1.5579306414931493e-07,
+      "loss": 0.0274,
+      "step": 5171
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 2.608275894549004,
+      "learning_rate": 1.5507803545560195e-07,
+      "loss": 0.0272,
+      "step": 5172
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 3.1503623838936474,
+      "learning_rate": 1.543646255403719e-07,
+      "loss": 0.0266,
+      "step": 5173
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 1.5556403344359921,
+      "learning_rate": 1.5365283464198743e-07,
+      "loss": 0.0189,
+      "step": 5174
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 8.45265240922781,
+      "learning_rate": 1.529426629982711e-07,
+      "loss": 0.0317,
+      "step": 5175
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 4.4790627118344535,
+      "learning_rate": 1.5223411084650476e-07,
+      "loss": 0.02,
+      "step": 5176
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 2.2062655844358336,
+      "learning_rate": 1.5152717842342845e-07,
+      "loss": 0.0246,
+      "step": 5177
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 2.6205269218486786,
+      "learning_rate": 1.5082186596524218e-07,
+      "loss": 0.0175,
+      "step": 5178
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 3.0051366479489072,
+      "learning_rate": 1.501181737076035e-07,
+      "loss": 0.0238,
+      "step": 5179
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 4.819673804209162,
+      "learning_rate": 1.4941610188562884e-07,
+      "loss": 0.0199,
+      "step": 5180
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 5.138577134268833,
+      "learning_rate": 1.4871565073389382e-07,
+      "loss": 0.0246,
+      "step": 5181
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 3.8138304855316023,
+      "learning_rate": 1.4801682048643183e-07,
+      "loss": 0.0242,
+      "step": 5182
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 1.4542929428683837,
+      "learning_rate": 1.4731961137673555e-07,
+      "loss": 0.016,
+      "step": 5183
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 4.104948196555313,
+      "learning_rate": 1.466240236377553e-07,
+      "loss": 0.0257,
+      "step": 5184
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 3.5735418575307967,
+      "learning_rate": 1.4593005750189958e-07,
+      "loss": 0.0232,
+      "step": 5185
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 1.8507033348331547,
+      "learning_rate": 1.4523771320103574e-07,
+      "loss": 0.0195,
+      "step": 5186
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 1.8413134451311026,
+      "learning_rate": 1.4454699096648873e-07,
+      "loss": 0.0345,
+      "step": 5187
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 3.3278075671092697,
+      "learning_rate": 1.4385789102904168e-07,
+      "loss": 0.0301,
+      "step": 5188
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 2.3248996332848746,
+      "learning_rate": 1.4317041361893546e-07,
+      "loss": 0.0295,
+      "step": 5189
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 3.4264176035418945,
+      "learning_rate": 1.4248455896587022e-07,
+      "loss": 0.0327,
+      "step": 5190
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 1.728019296230289,
+      "learning_rate": 1.418003272990004e-07,
+      "loss": 0.0189,
+      "step": 5191
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 1.489162413811022,
+      "learning_rate": 1.4111771884694315e-07,
+      "loss": 0.0203,
+      "step": 5192
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 1.9524657898968014,
+      "learning_rate": 1.4043673383776825e-07,
+      "loss": 0.0123,
+      "step": 5193
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 2.415123271352204,
+      "learning_rate": 1.3975737249900812e-07,
+      "loss": 0.0266,
+      "step": 5194
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 2.8692636393398847,
+      "learning_rate": 1.3907963505764731e-07,
+      "loss": 0.0243,
+      "step": 5195
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 2.8502413687478807,
+      "learning_rate": 1.384035217401325e-07,
+      "loss": 0.0139,
+      "step": 5196
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 2.862718769125452,
+      "learning_rate": 1.3772903277236404e-07,
+      "loss": 0.0235,
+      "step": 5197
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 2.1513904659754792,
+      "learning_rate": 1.370561683797028e-07,
+      "loss": 0.0271,
+      "step": 5198
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 6.519218790872213,
+      "learning_rate": 1.363849287869645e-07,
+      "loss": 0.0378,
+      "step": 5199
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 4.489049467119617,
+      "learning_rate": 1.3571531421842256e-07,
+      "loss": 0.0273,
+      "step": 5200
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 2.3577172209252852,
+      "learning_rate": 1.3504732489780849e-07,
+      "loss": 0.0219,
+      "step": 5201
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 5.7833943763627795,
+      "learning_rate": 1.3438096104830879e-07,
+      "loss": 0.0217,
+      "step": 5202
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 1.988371201831818,
+      "learning_rate": 1.3371622289256869e-07,
+      "loss": 0.0202,
+      "step": 5203
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 2.425503910065707,
+      "learning_rate": 1.3305311065269e-07,
+      "loss": 0.0198,
+      "step": 5204
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 3.090722053136764,
+      "learning_rate": 1.323916245502299e-07,
+      "loss": 0.0245,
+      "step": 5205
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 4.9627962283694345,
+      "learning_rate": 1.3173176480620442e-07,
+      "loss": 0.0244,
+      "step": 5206
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 9.206592423133095,
+      "learning_rate": 1.3107353164108273e-07,
+      "loss": 0.0316,
+      "step": 5207
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 2.7039312687504373,
+      "learning_rate": 1.3041692527479556e-07,
+      "loss": 0.022,
+      "step": 5208
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 1.3752954379070126,
+      "learning_rate": 1.2976194592672465e-07,
+      "loss": 0.0183,
+      "step": 5209
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 4.6190385954775905,
+      "learning_rate": 1.2910859381571327e-07,
+      "loss": 0.0233,
+      "step": 5210
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 3.210019534830187,
+      "learning_rate": 1.284568691600563e-07,
+      "loss": 0.0186,
+      "step": 5211
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 3.4280035373439843,
+      "learning_rate": 1.2780677217750949e-07,
+      "loss": 0.0217,
+      "step": 5212
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 2.9802093187192633,
+      "learning_rate": 1.271583030852791e-07,
+      "loss": 0.0247,
+      "step": 5213
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 1.385873022404817,
+      "learning_rate": 1.2651146210003406e-07,
+      "loss": 0.0176,
+      "step": 5214
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 5.529817438941661,
+      "learning_rate": 1.2586624943789372e-07,
+      "loss": 0.0199,
+      "step": 5215
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 2.3117896694300617,
+      "learning_rate": 1.2522266531443616e-07,
+      "loss": 0.0269,
+      "step": 5216
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 7.010813791691153,
+      "learning_rate": 1.245807099446955e-07,
+      "loss": 0.0401,
+      "step": 5217
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 6.309724081611123,
+      "learning_rate": 1.239403835431602e-07,
+      "loss": 0.0204,
+      "step": 5218
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 3.991131983188111,
+      "learning_rate": 1.2330168632377514e-07,
+      "loss": 0.021,
+      "step": 5219
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 3.3782189157497045,
+      "learning_rate": 1.2266461849994138e-07,
+      "loss": 0.0224,
+      "step": 5220
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 4.944132812084641,
+      "learning_rate": 1.2202918028451527e-07,
+      "loss": 0.0171,
+      "step": 5221
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 2.3020823968537423,
+      "learning_rate": 1.2139537188980753e-07,
+      "loss": 0.0197,
+      "step": 5222
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 2.252902650158216,
+      "learning_rate": 1.207631935275866e-07,
+      "loss": 0.0306,
+      "step": 5223
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 4.783919269935325,
+      "learning_rate": 1.2013264540907455e-07,
+      "loss": 0.0215,
+      "step": 5224
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 2.5909512837118926,
+      "learning_rate": 1.1950372774494846e-07,
+      "loss": 0.0176,
+      "step": 5225
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 4.136611119846169,
+      "learning_rate": 1.1887644074534244e-07,
+      "loss": 0.0186,
+      "step": 5226
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 2.72318296618443,
+      "learning_rate": 1.182507846198444e-07,
+      "loss": 0.0245,
+      "step": 5227
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 7.516987600595989,
+      "learning_rate": 1.1762675957749769e-07,
+      "loss": 0.0422,
+      "step": 5228
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 2.5276403915062757,
+      "learning_rate": 1.1700436582680108e-07,
+      "loss": 0.0191,
+      "step": 5229
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 4.752529791718225,
+      "learning_rate": 1.1638360357570654e-07,
+      "loss": 0.0293,
+      "step": 5230
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 2.400911193761536,
+      "learning_rate": 1.157644730316243e-07,
+      "loss": 0.0157,
+      "step": 5231
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 2.326937968268249,
+      "learning_rate": 1.1514697440141498e-07,
+      "loss": 0.0182,
+      "step": 5232
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 2.4877722821251047,
+      "learning_rate": 1.1453110789139855e-07,
+      "loss": 0.0208,
+      "step": 5233
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 3.9936599844023553,
+      "learning_rate": 1.1391687370734594e-07,
+      "loss": 0.018,
+      "step": 5234
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 3.6287267410137187,
+      "learning_rate": 1.1330427205448579e-07,
+      "loss": 0.0285,
+      "step": 5235
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 2.4006517456771665,
+      "learning_rate": 1.1269330313749715e-07,
+      "loss": 0.0214,
+      "step": 5236
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 4.142343883359804,
+      "learning_rate": 1.1208396716051895e-07,
+      "loss": 0.0247,
+      "step": 5237
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 2.711787066445167,
+      "learning_rate": 1.1147626432713943e-07,
+      "loss": 0.017,
+      "step": 5238
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 1.7231843572349768,
+      "learning_rate": 1.1087019484040562e-07,
+      "loss": 0.0136,
+      "step": 5239
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 6.191379324297485,
+      "learning_rate": 1.1026575890281443e-07,
+      "loss": 0.0314,
+      "step": 5240
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 2.822860417056047,
+      "learning_rate": 1.0966295671632043e-07,
+      "loss": 0.0251,
+      "step": 5241
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 2.203361342127882,
+      "learning_rate": 1.0906178848233029e-07,
+      "loss": 0.0172,
+      "step": 5242
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 2.8531402032960096,
+      "learning_rate": 1.0846225440170611e-07,
+      "loss": 0.0256,
+      "step": 5243
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 1.698947522501088,
+      "learning_rate": 1.0786435467476264e-07,
+      "loss": 0.0149,
+      "step": 5244
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 4.130192879839065,
+      "learning_rate": 1.072680895012701e-07,
+      "loss": 0.0208,
+      "step": 5245
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 4.923493442779139,
+      "learning_rate": 1.0667345908045135e-07,
+      "loss": 0.0206,
+      "step": 5246
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 1.6193563517025997,
+      "learning_rate": 1.0608046361098356e-07,
+      "loss": 0.021,
+      "step": 5247
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 5.093530894764852,
+      "learning_rate": 1.0548910329099771e-07,
+      "loss": 0.0214,
+      "step": 5248
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 3.1583303655468944,
+      "learning_rate": 1.048993783180785e-07,
+      "loss": 0.0204,
+      "step": 5249
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 4.86486432471581,
+      "learning_rate": 1.0431128888926222e-07,
+      "loss": 0.0346,
+      "step": 5250
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 3.9913107188739105,
+      "learning_rate": 1.0372483520104337e-07,
+      "loss": 0.0359,
+      "step": 5251
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 2.2381666578064294,
+      "learning_rate": 1.0314001744936409e-07,
+      "loss": 0.0213,
+      "step": 5252
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 4.551116852075309,
+      "learning_rate": 1.0255683582962583e-07,
+      "loss": 0.0211,
+      "step": 5253
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 1.057164791329309,
+      "learning_rate": 1.0197529053667721e-07,
+      "loss": 0.0178,
+      "step": 5254
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 3.1358364708701503,
+      "learning_rate": 1.013953817648261e-07,
+      "loss": 0.0205,
+      "step": 5255
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 4.521051074903339,
+      "learning_rate": 1.008171097078292e-07,
+      "loss": 0.0273,
+      "step": 5256
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 2.9904025566820267,
+      "learning_rate": 1.0024047455889918e-07,
+      "loss": 0.0304,
+      "step": 5257
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 3.778324379127207,
+      "learning_rate": 9.966547651069913e-08,
+      "loss": 0.0278,
+      "step": 5258
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 10.272189356491223,
+      "learning_rate": 9.909211575534705e-08,
+      "loss": 0.0461,
+      "step": 5259
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 2.2283715681606227,
+      "learning_rate": 9.852039248441414e-08,
+      "loss": 0.0171,
+      "step": 5260
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 4.722409299384118,
+      "learning_rate": 9.79503068889226e-08,
+      "loss": 0.0255,
+      "step": 5261
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 3.4508766889297706,
+      "learning_rate": 9.738185915935005e-08,
+      "loss": 0.0292,
+      "step": 5262
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 4.496268757517527,
+      "learning_rate": 9.681504948562403e-08,
+      "loss": 0.0217,
+      "step": 5263
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 5.404507854655809,
+      "learning_rate": 9.624987805712749e-08,
+      "loss": 0.0197,
+      "step": 5264
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 2.622445035917951,
+      "learning_rate": 9.568634506269381e-08,
+      "loss": 0.0259,
+      "step": 5265
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 4.884116982411068,
+      "learning_rate": 9.51244506906096e-08,
+      "loss": 0.0258,
+      "step": 5266
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 1.8627897365580712,
+      "learning_rate": 9.45641951286158e-08,
+      "loss": 0.0194,
+      "step": 5267
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 2.7000828431244783,
+      "learning_rate": 9.400557856390158e-08,
+      "loss": 0.0199,
+      "step": 5268
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 3.956492664965638,
+      "learning_rate": 9.344860118311427e-08,
+      "loss": 0.0234,
+      "step": 5269
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 1.425209166481016,
+      "learning_rate": 9.289326317234726e-08,
+      "loss": 0.0231,
+      "step": 5270
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 4.097267096486783,
+      "learning_rate": 9.23395647171521e-08,
+      "loss": 0.0235,
+      "step": 5271
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 2.028670905892367,
+      "learning_rate": 9.178750600252695e-08,
+      "loss": 0.0255,
+      "step": 5272
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 4.966736021039718,
+      "learning_rate": 9.123708721292756e-08,
+      "loss": 0.028,
+      "step": 5273
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 3.197737197596851,
+      "learning_rate": 9.06883085322574e-08,
+      "loss": 0.0216,
+      "step": 5274
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 4.080727773730197,
+      "learning_rate": 9.014117014387424e-08,
+      "loss": 0.0352,
+      "step": 5275
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 1.1296264280256079,
+      "learning_rate": 8.95956722305874e-08,
+      "loss": 0.0169,
+      "step": 5276
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 4.057116883266143,
+      "learning_rate": 8.905181497465664e-08,
+      "loss": 0.0221,
+      "step": 5277
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 3.9647187605506953,
+      "learning_rate": 8.850959855779662e-08,
+      "loss": 0.0233,
+      "step": 5278
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 2.1040774256861408,
+      "learning_rate": 8.796902316117018e-08,
+      "loss": 0.0213,
+      "step": 5279
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 4.65723542116145,
+      "learning_rate": 8.743008896539451e-08,
+      "loss": 0.0258,
+      "step": 5280
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 4.876295934301376,
+      "learning_rate": 8.68927961505378e-08,
+      "loss": 0.0257,
+      "step": 5281
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 2.8627960416897555,
+      "learning_rate": 8.635714489611868e-08,
+      "loss": 0.0314,
+      "step": 5282
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 2.273725154988319,
+      "learning_rate": 8.582313538110898e-08,
+      "loss": 0.0161,
+      "step": 5283
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 1.9282706264267362,
+      "learning_rate": 8.529076778393097e-08,
+      "loss": 0.0254,
+      "step": 5284
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 2.628970273375051,
+      "learning_rate": 8.476004228245848e-08,
+      "loss": 0.0231,
+      "step": 5285
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 3.0795335082550763,
+      "learning_rate": 8.42309590540169e-08,
+      "loss": 0.0233,
+      "step": 5286
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 2.4895174177627926,
+      "learning_rate": 8.370351827538259e-08,
+      "loss": 0.0258,
+      "step": 5287
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 7.5059982137655075,
+      "learning_rate": 8.317772012278347e-08,
+      "loss": 0.0257,
+      "step": 5288
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 2.287349453785497,
+      "learning_rate": 8.26535647718979e-08,
+      "loss": 0.0153,
+      "step": 5289
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 4.007946391371989,
+      "learning_rate": 8.213105239785691e-08,
+      "loss": 0.0142,
+      "step": 5290
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 3.325150468656642,
+      "learning_rate": 8.161018317524139e-08,
+      "loss": 0.0186,
+      "step": 5291
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 2.679684188837439,
+      "learning_rate": 8.109095727808269e-08,
+      "loss": 0.0303,
+      "step": 5292
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 3.923277021778115,
+      "learning_rate": 8.057337487986427e-08,
+      "loss": 0.0271,
+      "step": 5293
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 2.9481719065310883,
+      "learning_rate": 8.005743615352057e-08,
+      "loss": 0.022,
+      "step": 5294
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 11.41032621498422,
+      "learning_rate": 7.954314127143481e-08,
+      "loss": 0.0336,
+      "step": 5295
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 8.438022620378145,
+      "learning_rate": 7.903049040544453e-08,
+      "loss": 0.0272,
+      "step": 5296
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 1.5690249006305896,
+      "learning_rate": 7.851948372683382e-08,
+      "loss": 0.0239,
+      "step": 5297
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 1.8028194196876857,
+      "learning_rate": 7.801012140634167e-08,
+      "loss": 0.02,
+      "step": 5298
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 1.8809847137659024,
+      "learning_rate": 7.750240361415362e-08,
+      "loss": 0.0178,
+      "step": 5299
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 4.070256377359856,
+      "learning_rate": 7.69963305199084e-08,
+      "loss": 0.0211,
+      "step": 5300
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 4.871741592906539,
+      "learning_rate": 7.64919022926941e-08,
+      "loss": 0.0213,
+      "step": 5301
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 2.1231914716571154,
+      "learning_rate": 7.598911910105033e-08,
+      "loss": 0.0186,
+      "step": 5302
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 2.0372035432976916,
+      "learning_rate": 7.548798111296552e-08,
+      "loss": 0.0222,
+      "step": 5303
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 3.43083865516561,
+      "learning_rate": 7.498848849588015e-08,
+      "loss": 0.0245,
+      "step": 5304
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 4.2401607366865335,
+      "learning_rate": 7.449064141668238e-08,
+      "loss": 0.0262,
+      "step": 5305
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 2.638289900596031,
+      "learning_rate": 7.399444004171364e-08,
+      "loss": 0.0216,
+      "step": 5306
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 5.203760317859933,
+      "learning_rate": 7.349988453676349e-08,
+      "loss": 0.0264,
+      "step": 5307
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 1.1903840699355497,
+      "learning_rate": 7.300697506707254e-08,
+      "loss": 0.0174,
+      "step": 5308
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 2.386179668964028,
+      "learning_rate": 7.251571179732963e-08,
+      "loss": 0.0213,
+      "step": 5309
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 5.227249096322965,
+      "learning_rate": 7.202609489167734e-08,
+      "loss": 0.0252,
+      "step": 5310
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 1.9320796097209536,
+      "learning_rate": 7.153812451370312e-08,
+      "loss": 0.0148,
+      "step": 5311
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 3.182049411942898,
+      "learning_rate": 7.10518008264488e-08,
+      "loss": 0.0228,
+      "step": 5312
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 3.3538825883283443,
+      "learning_rate": 7.056712399240274e-08,
+      "loss": 0.0275,
+      "step": 5313
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 2.1138391051762473,
+      "learning_rate": 7.008409417350648e-08,
+      "loss": 0.0259,
+      "step": 5314
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 4.7881945597014255,
+      "learning_rate": 6.960271153114706e-08,
+      "loss": 0.0272,
+      "step": 5315
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 1.3152877442321909,
+      "learning_rate": 6.912297622616526e-08,
+      "loss": 0.0118,
+      "step": 5316
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 2.201245036810093,
+      "learning_rate": 6.864488841884786e-08,
+      "loss": 0.0182,
+      "step": 5317
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 1.7396078183985866,
+      "learning_rate": 6.816844826893431e-08,
+      "loss": 0.0211,
+      "step": 5318
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 6.429755754943171,
+      "learning_rate": 6.769365593561117e-08,
+      "loss": 0.0239,
+      "step": 5319
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 2.4797877328826865,
+      "learning_rate": 6.722051157751597e-08,
+      "loss": 0.0221,
+      "step": 5320
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 4.190112748365354,
+      "learning_rate": 6.674901535273448e-08,
+      "loss": 0.0164,
+      "step": 5321
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 3.4184760808635124,
+      "learning_rate": 6.627916741880291e-08,
+      "loss": 0.0369,
+      "step": 5322
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 2.141574310849085,
+      "learning_rate": 6.581096793270625e-08,
+      "loss": 0.0137,
+      "step": 5323
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 2.740988732465579,
+      "learning_rate": 6.534441705087768e-08,
+      "loss": 0.0214,
+      "step": 5324
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 3.057761817664433,
+      "learning_rate": 6.487951492920141e-08,
+      "loss": 0.0271,
+      "step": 5325
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 4.367102135231493,
+      "learning_rate": 6.441626172300986e-08,
+      "loss": 0.0225,
+      "step": 5326
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 5.584248486877098,
+      "learning_rate": 6.395465758708419e-08,
+      "loss": 0.018,
+      "step": 5327
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 5.53656132887622,
+      "learning_rate": 6.349470267565549e-08,
+      "loss": 0.0262,
+      "step": 5328
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 2.128469356840069,
+      "learning_rate": 6.303639714240196e-08,
+      "loss": 0.0127,
+      "step": 5329
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 4.588339566836977,
+      "learning_rate": 6.257974114045385e-08,
+      "loss": 0.0219,
+      "step": 5330
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 1.3886881130197628,
+      "learning_rate": 6.212473482238635e-08,
+      "loss": 0.0196,
+      "step": 5331
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 11.056473846052311,
+      "learning_rate": 6.167137834022785e-08,
+      "loss": 0.0398,
+      "step": 5332
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 1.7596324104793302,
+      "learning_rate": 6.121967184545107e-08,
+      "loss": 0.0129,
+      "step": 5333
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 1.6553378546687454,
+      "learning_rate": 6.076961548898086e-08,
+      "loss": 0.0172,
+      "step": 5334
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 3.0913951303287965,
+      "learning_rate": 6.032120942118858e-08,
+      "loss": 0.0193,
+      "step": 5335
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 3.909975477030834,
+      "learning_rate": 5.98744537918955e-08,
+      "loss": 0.0185,
+      "step": 5336
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 1.5964964171735954,
+      "learning_rate": 5.9429348750371097e-08,
+      "loss": 0.02,
+      "step": 5337
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 3.4138357917288413,
+      "learning_rate": 5.898589444533254e-08,
+      "loss": 0.0198,
+      "step": 5338
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 4.010010339656211,
+      "learning_rate": 5.85440910249474e-08,
+      "loss": 0.0468,
+      "step": 5339
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 2.884838335613086,
+      "learning_rate": 5.810393863682873e-08,
+      "loss": 0.0236,
+      "step": 5340
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 2.5632337910539778,
+      "learning_rate": 5.7665437428041096e-08,
+      "loss": 0.0254,
+      "step": 5341
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 6.461929942388153,
+      "learning_rate": 5.722858754509564e-08,
+      "loss": 0.0225,
+      "step": 5342
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 2.358122356416235,
+      "learning_rate": 5.679338913395116e-08,
+      "loss": 0.0181,
+      "step": 5343
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 7.677224673110656,
+      "learning_rate": 5.6359842340016904e-08,
+      "loss": 0.0262,
+      "step": 5344
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 2.168935169589492,
+      "learning_rate": 5.5927947308147545e-08,
+      "loss": 0.0177,
+      "step": 5345
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 3.466736981204486,
+      "learning_rate": 5.549770418264766e-08,
+      "loss": 0.0177,
+      "step": 5346
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 3.4138530734232755,
+      "learning_rate": 5.5069113107270034e-08,
+      "loss": 0.0244,
+      "step": 5347
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 2.065047272470145,
+      "learning_rate": 5.464217422521456e-08,
+      "loss": 0.0221,
+      "step": 5348
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 2.8946037809002765,
+      "learning_rate": 5.421688767912936e-08,
+      "loss": 0.0186,
+      "step": 5349
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 4.394964370403741,
+      "learning_rate": 5.3793253611110206e-08,
+      "loss": 0.0197,
+      "step": 5350
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 2.608996497277368,
+      "learning_rate": 5.3371272162702214e-08,
+      "loss": 0.0177,
+      "step": 5351
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 1.8025304112213554,
+      "learning_rate": 5.295094347489593e-08,
+      "loss": 0.0207,
+      "step": 5352
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 2.5613629695229787,
+      "learning_rate": 5.253226768813235e-08,
+      "loss": 0.0164,
+      "step": 5353
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 4.54003532307264,
+      "learning_rate": 5.211524494229736e-08,
+      "loss": 0.023,
+      "step": 5354
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 1.904588615161526,
+      "learning_rate": 5.169987537672727e-08,
+      "loss": 0.0179,
+      "step": 5355
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 1.9393310412596467,
+      "learning_rate": 5.128615913020385e-08,
+      "loss": 0.0222,
+      "step": 5356
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 3.359468134735287,
+      "learning_rate": 5.087409634095819e-08,
+      "loss": 0.0238,
+      "step": 5357
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 1.1449746630362878,
+      "learning_rate": 5.046368714666683e-08,
+      "loss": 0.0117,
+      "step": 5358
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 4.222518213830071,
+      "learning_rate": 5.0054931684457296e-08,
+      "loss": 0.0383,
+      "step": 5359
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 3.743647719670252,
+      "learning_rate": 4.964783009090035e-08,
+      "loss": 0.0293,
+      "step": 5360
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 3.833891477000365,
+      "learning_rate": 4.9242382502017185e-08,
+      "loss": 0.0229,
+      "step": 5361
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 4.168894379967887,
+      "learning_rate": 4.883858905327499e-08,
+      "loss": 0.019,
+      "step": 5362
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 1.913254240680341,
+      "learning_rate": 4.843644987958862e-08,
+      "loss": 0.0188,
+      "step": 5363
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 2.110753327008818,
+      "learning_rate": 4.8035965115320604e-08,
+      "loss": 0.0197,
+      "step": 5364
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 8.56246472776577,
+      "learning_rate": 4.763713489428001e-08,
+      "loss": 0.0337,
+      "step": 5365
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 3.119553272465091,
+      "learning_rate": 4.723995934972414e-08,
+      "loss": 0.0235,
+      "step": 5366
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 3.8853086683100098,
+      "learning_rate": 4.684443861435572e-08,
+      "loss": 0.0244,
+      "step": 5367
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 4.4695595709236855,
+      "learning_rate": 4.6450572820325727e-08,
+      "loss": 0.0252,
+      "step": 5368
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 2.7874075602983672,
+      "learning_rate": 4.605836209923331e-08,
+      "loss": 0.0192,
+      "step": 5369
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 2.203531606811125,
+      "learning_rate": 4.566780658212144e-08,
+      "loss": 0.0206,
+      "step": 5370
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 3.160499631682135,
+      "learning_rate": 4.5278906399483516e-08,
+      "loss": 0.0169,
+      "step": 5371
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 4.6089108849374805,
+      "learning_rate": 4.489166168125725e-08,
+      "loss": 0.0271,
+      "step": 5372
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 4.866432991585118,
+      "learning_rate": 4.4506072556829704e-08,
+      "loss": 0.0189,
+      "step": 5373
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 3.197986454347981,
+      "learning_rate": 4.4122139155031717e-08,
+      "loss": 0.0219,
+      "step": 5374
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 4.3140025616477145,
+      "learning_rate": 4.373986160414345e-08,
+      "loss": 0.04,
+      "step": 5375
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.9589388375969385,
+      "learning_rate": 4.335924003189107e-08,
+      "loss": 0.0173,
+      "step": 5376
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 2.9736420672090085,
+      "learning_rate": 4.298027456544674e-08,
+      "loss": 0.0215,
+      "step": 5377
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 3.0076067234022306,
+      "learning_rate": 4.260296533143027e-08,
+      "loss": 0.0311,
+      "step": 5378
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 2.9163352666589164,
+      "learning_rate": 4.22273124559075e-08,
+      "loss": 0.0179,
+      "step": 5379
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 5.350361549448945,
+      "learning_rate": 4.185331606439136e-08,
+      "loss": 0.0296,
+      "step": 5380
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 3.5143808177875666,
+      "learning_rate": 4.148097628184078e-08,
+      "loss": 0.0268,
+      "step": 5381
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 1.6488078338220415,
+      "learning_rate": 4.111029323266125e-08,
+      "loss": 0.0212,
+      "step": 5382
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 2.9180480292411337,
+      "learning_rate": 4.07412670407048e-08,
+      "loss": 0.0167,
+      "step": 5383
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 3.426743137102039,
+      "learning_rate": 4.037389782927059e-08,
+      "loss": 0.0261,
+      "step": 5384
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 2.626920868414007,
+      "learning_rate": 4.000818572110265e-08,
+      "loss": 0.0187,
+      "step": 5385
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 4.358340512942245,
+      "learning_rate": 3.964413083839269e-08,
+      "loss": 0.024,
+      "step": 5386
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 2.5771043456304734,
+      "learning_rate": 3.9281733302778404e-08,
+      "loss": 0.0135,
+      "step": 5387
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 3.7108362612772945,
+      "learning_rate": 3.892099323534293e-08,
+      "loss": 0.0251,
+      "step": 5388
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 3.2641134315951765,
+      "learning_rate": 3.856191075661708e-08,
+      "loss": 0.0207,
+      "step": 5389
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 4.7954663584761,
+      "learning_rate": 3.8204485986576e-08,
+      "loss": 0.0233,
+      "step": 5390
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 3.0039195291483587,
+      "learning_rate": 3.784871904464249e-08,
+      "loss": 0.0214,
+      "step": 5391
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 4.781319594897229,
+      "learning_rate": 3.7494610049684796e-08,
+      "loss": 0.0202,
+      "step": 5392
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 2.2181342509203716,
+      "learning_rate": 3.714215912001773e-08,
+      "loss": 0.0237,
+      "step": 5393
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 1.5307656638034723,
+      "learning_rate": 3.6791366373400974e-08,
+      "loss": 0.016,
+      "step": 5394
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 2.9980957129316477,
+      "learning_rate": 3.6442231927041324e-08,
+      "loss": 0.0209,
+      "step": 5395
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 2.614325470570068,
+      "learning_rate": 3.609475589759104e-08,
+      "loss": 0.0195,
+      "step": 5396
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 1.2342786417833946,
+      "learning_rate": 3.574893840114835e-08,
+      "loss": 0.0195,
+      "step": 5397
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 3.13513680493031,
+      "learning_rate": 3.5404779553257494e-08,
+      "loss": 0.0198,
+      "step": 5398
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 15.078842855499811,
+      "learning_rate": 3.506227946890761e-08,
+      "loss": 0.042,
+      "step": 5399
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 4.968249293917528,
+      "learning_rate": 3.4721438262534935e-08,
+      "loss": 0.0286,
+      "step": 5400
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 2.300715233738611,
+      "learning_rate": 3.438225604802115e-08,
+      "loss": 0.0247,
+      "step": 5401
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 3.2215687096264367,
+      "learning_rate": 3.404473293869226e-08,
+      "loss": 0.0225,
+      "step": 5402
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 1.3402505655817432,
+      "learning_rate": 3.370886904732196e-08,
+      "loss": 0.0137,
+      "step": 5403
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 1.0509369564653657,
+      "learning_rate": 3.33746644861288e-08,
+      "loss": 0.0152,
+      "step": 5404
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 1.5200283370650576,
+      "learning_rate": 3.30421193667757e-08,
+      "loss": 0.0172,
+      "step": 5405
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 6.978782111965761,
+      "learning_rate": 3.271123380037322e-08,
+      "loss": 0.0288,
+      "step": 5406
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 4.070184829965542,
+      "learning_rate": 3.2382007897475695e-08,
+      "loss": 0.0289,
+      "step": 5407
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 4.409008744759003,
+      "learning_rate": 3.2054441768083477e-08,
+      "loss": 0.0241,
+      "step": 5408
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 2.2078652879223797,
+      "learning_rate": 3.1728535521643454e-08,
+      "loss": 0.0198,
+      "step": 5409
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 2.187261175257129,
+      "learning_rate": 3.1404289267046305e-08,
+      "loss": 0.0185,
+      "step": 5410
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 1.8104547975037084,
+      "learning_rate": 3.1081703112628146e-08,
+      "loss": 0.0207,
+      "step": 5411
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 2.108463142618941,
+      "learning_rate": 3.0760777166172206e-08,
+      "loss": 0.0204,
+      "step": 5412
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 4.41782265356589,
+      "learning_rate": 3.0441511534904934e-08,
+      "loss": 0.0266,
+      "step": 5413
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 2.1563829583102483,
+      "learning_rate": 3.012390632549933e-08,
+      "loss": 0.0298,
+      "step": 5414
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 7.78806885480321,
+      "learning_rate": 2.9807961644073294e-08,
+      "loss": 0.026,
+      "step": 5415
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 3.120989169250971,
+      "learning_rate": 2.9493677596189595e-08,
+      "loss": 0.0237,
+      "step": 5416
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 4.358905498727036,
+      "learning_rate": 2.9181054286855916e-08,
+      "loss": 0.0214,
+      "step": 5417
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 4.6477516202899,
+      "learning_rate": 2.887009182052647e-08,
+      "loss": 0.0307,
+      "step": 5418
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 6.1082992817988115,
+      "learning_rate": 2.8560790301098705e-08,
+      "loss": 0.0244,
+      "step": 5419
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 1.275610383748102,
+      "learning_rate": 2.825314983191718e-08,
+      "loss": 0.0116,
+      "step": 5420
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 8.547679089993268,
+      "learning_rate": 2.7947170515768562e-08,
+      "loss": 0.0333,
+      "step": 5421
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 2.957300537052848,
+      "learning_rate": 2.7642852454887736e-08,
+      "loss": 0.0221,
+      "step": 5422
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 2.033116201364948,
+      "learning_rate": 2.7340195750952813e-08,
+      "loss": 0.0148,
+      "step": 5423
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 5.438523966512901,
+      "learning_rate": 2.703920050508624e-08,
+      "loss": 0.0304,
+      "step": 5424
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 5.690418670617803,
+      "learning_rate": 2.673986681785645e-08,
+      "loss": 0.0202,
+      "step": 5425
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 6.5518385685801555,
+      "learning_rate": 2.6442194789277342e-08,
+      "loss": 0.0322,
+      "step": 5426
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 1.2401661063067848,
+      "learning_rate": 2.6146184518804908e-08,
+      "loss": 0.014,
+      "step": 5427
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 3.209009421866879,
+      "learning_rate": 2.5851836105343363e-08,
+      "loss": 0.0168,
+      "step": 5428
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 2.2479236401385623,
+      "learning_rate": 2.555914964723849e-08,
+      "loss": 0.0192,
+      "step": 5429
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 3.664875088278122,
+      "learning_rate": 2.5268125242283724e-08,
+      "loss": 0.0293,
+      "step": 5430
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 4.403075650273632,
+      "learning_rate": 2.4978762987714067e-08,
+      "loss": 0.0368,
+      "step": 5431
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 6.9310695166002025,
+      "learning_rate": 2.469106298021273e-08,
+      "loss": 0.0316,
+      "step": 5432
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 2.423140496516659,
+      "learning_rate": 2.4405025315904495e-08,
+      "loss": 0.024,
+      "step": 5433
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 4.022916173686561,
+      "learning_rate": 2.412065009036013e-08,
+      "loss": 0.0212,
+      "step": 5434
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 3.837813048775894,
+      "learning_rate": 2.3837937398594747e-08,
+      "loss": 0.0298,
+      "step": 5435
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 2.1748941776797377,
+      "learning_rate": 2.3556887335067223e-08,
+      "loss": 0.0262,
+      "step": 5436
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 6.561398321137626,
+      "learning_rate": 2.3277499993682452e-08,
+      "loss": 0.036,
+      "step": 5437
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 4.357304631831137,
+      "learning_rate": 2.2999775467788532e-08,
+      "loss": 0.0223,
+      "step": 5438
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 3.486845904406258,
+      "learning_rate": 2.272371385017902e-08,
+      "loss": 0.0272,
+      "step": 5439
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 4.662953700822703,
+      "learning_rate": 2.244931523309013e-08,
+      "loss": 0.0278,
+      "step": 5440
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 2.054534031716196,
+      "learning_rate": 2.2176579708204636e-08,
+      "loss": 0.0244,
+      "step": 5441
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 3.9539255224153704,
+      "learning_rate": 2.190550736664798e-08,
+      "loss": 0.0359,
+      "step": 5442
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 5.221943937345425,
+      "learning_rate": 2.163609829898994e-08,
+      "loss": 0.0316,
+      "step": 5443
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 2.6919999185285426,
+      "learning_rate": 2.136835259524628e-08,
+      "loss": 0.0281,
+      "step": 5444
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 3.252062666573824,
+      "learning_rate": 2.1102270344874887e-08,
+      "loss": 0.0428,
+      "step": 5445
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 2.2687059601810353,
+      "learning_rate": 2.083785163677965e-08,
+      "loss": 0.0194,
+      "step": 5446
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 3.2991800662002784,
+      "learning_rate": 2.0575096559306564e-08,
+      "loss": 0.0274,
+      "step": 5447
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 3.6298689814703224,
+      "learning_rate": 2.0314005200248178e-08,
+      "loss": 0.0195,
+      "step": 5448
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 3.218939239822081,
+      "learning_rate": 2.0054577646839156e-08,
+      "loss": 0.0142,
+      "step": 5449
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 2.868578007944478,
+      "learning_rate": 1.979681398575961e-08,
+      "loss": 0.0172,
+      "step": 5450
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 4.0210996974880056,
+      "learning_rate": 1.954071430313287e-08,
+      "loss": 0.0219,
+      "step": 5451
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 2.3082378385282873,
+      "learning_rate": 1.9286278684526593e-08,
+      "loss": 0.0243,
+      "step": 5452
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 3.362288093740942,
+      "learning_rate": 1.9033507214952784e-08,
+      "loss": 0.017,
+      "step": 5453
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 4.605772008765052,
+      "learning_rate": 1.878239997886666e-08,
+      "loss": 0.023,
+      "step": 5454
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 5.6829958574551505,
+      "learning_rate": 1.853295706016778e-08,
+      "loss": 0.0274,
+      "step": 5455
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 2.357378781581144,
+      "learning_rate": 1.8285178542200022e-08,
+      "loss": 0.0222,
+      "step": 5456
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 2.7214489224447185,
+      "learning_rate": 1.8039064507750503e-08,
+      "loss": 0.0131,
+      "step": 5457
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 2.8725088439600706,
+      "learning_rate": 1.7794615039050665e-08,
+      "loss": 0.0144,
+      "step": 5458
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 4.16619545847278,
+      "learning_rate": 1.7551830217775734e-08,
+      "loss": 0.0164,
+      "step": 5459
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 2.0022748959715058,
+      "learning_rate": 1.7310710125044707e-08,
+      "loss": 0.0226,
+      "step": 5460
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 2.3611659620346592,
+      "learning_rate": 1.7071254841419805e-08,
+      "loss": 0.0202,
+      "step": 5461
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 3.864100176432943,
+      "learning_rate": 1.6833464446907588e-08,
+      "loss": 0.0208,
+      "step": 5462
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 2.0424463543342584,
+      "learning_rate": 1.6597339020958393e-08,
+      "loss": 0.0192,
+      "step": 5463
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 2.3858731170790857,
+      "learning_rate": 1.6362878642466328e-08,
+      "loss": 0.0205,
+      "step": 5464
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 2.110909016473477,
+      "learning_rate": 1.6130083389768735e-08,
+      "loss": 0.0182,
+      "step": 5465
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 3.7439060845341365,
+      "learning_rate": 1.5898953340646728e-08,
+      "loss": 0.029,
+      "step": 5466
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 4.647959641308654,
+      "learning_rate": 1.5669488572325197e-08,
+      "loss": 0.0206,
+      "step": 5467
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 2.5611865882124523,
+      "learning_rate": 1.5441689161472816e-08,
+      "loss": 0.0283,
+      "step": 5468
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 3.131129293544498,
+      "learning_rate": 1.521555518420148e-08,
+      "loss": 0.0255,
+      "step": 5469
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 2.906211714511769,
+      "learning_rate": 1.499108671606686e-08,
+      "loss": 0.0222,
+      "step": 5470
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 1.3921197046262057,
+      "learning_rate": 1.4768283832067853e-08,
+      "loss": 0.0206,
+      "step": 5471
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 3.859282967632713,
+      "learning_rate": 1.4547146606646578e-08,
+      "loss": 0.0236,
+      "step": 5472
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 5.407318739007233,
+      "learning_rate": 1.4327675113690598e-08,
+      "loss": 0.0219,
+      "step": 5473
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 4.811797317082379,
+      "learning_rate": 1.4109869426527368e-08,
+      "loss": 0.0205,
+      "step": 5474
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 2.252930648915975,
+      "learning_rate": 1.3893729617931451e-08,
+      "loss": 0.0241,
+      "step": 5475
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 1.4816278252113144,
+      "learning_rate": 1.3679255760118415e-08,
+      "loss": 0.0227,
+      "step": 5476
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 6.986746169818905,
+      "learning_rate": 1.3466447924748716e-08,
+      "loss": 0.0249,
+      "step": 5477
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 2.1997925480493876,
+      "learning_rate": 1.3255306182924365e-08,
+      "loss": 0.0179,
+      "step": 5478
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 4.030533592119653,
+      "learning_rate": 1.3045830605192266e-08,
+      "loss": 0.0458,
+      "step": 5479
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 3.088240734799316,
+      "learning_rate": 1.2838021261541988e-08,
+      "loss": 0.019,
+      "step": 5480
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 3.0526266172301084,
+      "learning_rate": 1.263187822140688e-08,
+      "loss": 0.0228,
+      "step": 5481
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 2.888270689172121,
+      "learning_rate": 1.2427401553662955e-08,
+      "loss": 0.0192,
+      "step": 5482
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 2.4369666827080487,
+      "learning_rate": 1.2224591326628898e-08,
+      "loss": 0.015,
+      "step": 5483
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 2.2435601031926296,
+      "learning_rate": 1.2023447608068283e-08,
+      "loss": 0.0208,
+      "step": 5484
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 5.144528764350548,
+      "learning_rate": 1.182397046518735e-08,
+      "loss": 0.0224,
+      "step": 5485
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 2.2211905772296,
+      "learning_rate": 1.1626159964633899e-08,
+      "loss": 0.0147,
+      "step": 5486
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 2.2902390229673197,
+      "learning_rate": 1.1430016172501169e-08,
+      "loss": 0.0149,
+      "step": 5487
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 4.665203417246369,
+      "learning_rate": 1.1235539154323405e-08,
+      "loss": 0.0213,
+      "step": 5488
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 6.05200035827132,
+      "learning_rate": 1.1042728975079741e-08,
+      "loss": 0.0254,
+      "step": 5489
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 6.912668006593784,
+      "learning_rate": 1.0851585699191425e-08,
+      "loss": 0.0291,
+      "step": 5490
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 2.8237061465451463,
+      "learning_rate": 1.0662109390522924e-08,
+      "loss": 0.0197,
+      "step": 5491
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 5.594205018270895,
+      "learning_rate": 1.047430011238193e-08,
+      "loss": 0.0262,
+      "step": 5492
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 2.5381484863031885,
+      "learning_rate": 1.028815792751936e-08,
+      "loss": 0.0226,
+      "step": 5493
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 3.127384381139182,
+      "learning_rate": 1.0103682898128241e-08,
+      "loss": 0.0184,
+      "step": 5494
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 3.2830582442340375,
+      "learning_rate": 9.920875085845383e-09,
+      "loss": 0.0213,
+      "step": 5495
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 2.7865544882843336,
+      "learning_rate": 9.739734551749703e-09,
+      "loss": 0.0185,
+      "step": 5496
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 6.719662127295909,
+      "learning_rate": 9.560261356364452e-09,
+      "loss": 0.0387,
+      "step": 5497
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 3.0048549069559707,
+      "learning_rate": 9.382455559654446e-09,
+      "loss": 0.026,
+      "step": 5498
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 2.031697797024053,
+      "learning_rate": 9.206317221027717e-09,
+      "loss": 0.023,
+      "step": 5499
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 3.903823119328278,
+      "learning_rate": 9.031846399336075e-09,
+      "loss": 0.0327,
+      "step": 5500
+    },
+    {
+      "epoch": 3.93,
+      "eval_avg_AUC": 0.8371702589944845,
+      "eval_avg_Accuracy": 0.7440318302387268,
+      "eval_avg_Accuracy-right": 0.8942872049041346,
+      "eval_avg_Accuracy-wrong": 0.48203320445758474,
+      "eval_avg_Num questions with both labels": 523,
+      "eval_avg_Question-wise AUC": 0.7103294303171083,
+      "eval_last_AUC": 0.8319023333640545,
+      "eval_last_Accuracy": 0.7803796419098143,
+      "eval_last_Accuracy-right": 0.8463545063258119,
+      "eval_last_Accuracy-wrong": 0.6653400045485558,
+      "eval_last_Num questions with both labels": 523,
+      "eval_last_Question-wise AUC": 0.7067685458768546,
+      "eval_max_AUC": 0.7867669380385187,
+      "eval_max_Accuracy": 0.6470490716180372,
+      "eval_max_Accuracy-right": 0.9887178818312248,
+      "eval_max_Accuracy-wrong": 0.05128496702297021,
+      "eval_max_Num questions with both labels": 523,
+      "eval_max_Question-wise AUC": 0.6553866744173231,
+      "eval_min_AUC": 0.8460472659735542,
+      "eval_min_Accuracy": 0.7755305039787799,
+      "eval_min_Accuracy-right": 0.7935307160558237,
+      "eval_min_Accuracy-wrong": 0.7441437343643393,
+      "eval_min_Num questions with both labels": 523,
+      "eval_min_Question-wise AUC": 0.711460753125292,
+      "eval_prod_AUC": 0.8437307587767714,
+      "eval_prod_Accuracy": 0.7427470159151194,
+      "eval_prod_Accuracy-right": 0.6701447763140733,
+      "eval_prod_Accuracy-wrong": 0.869342733682056,
+      "eval_prod_Num questions with both labels": 523,
+      "eval_prod_Question-wise AUC": 0.7105555644172705,
+      "eval_runtime": 247.3548,
+      "eval_samples_per_second": 97.544,
+      "eval_steps_per_second": 3.048,
+      "eval_sum_AUC": 0.7143932084431329,
+      "eval_sum_Accuracy": 0.6395474137931034,
+      "eval_sum_Accuracy-right": 0.996869701317334,
+      "eval_sum_Accuracy-wrong": 0.016488514896520354,
+      "eval_sum_Num questions with both labels": 523,
+      "eval_sum_Question-wise AUC": 0.6877217306105692,
+      "step": 5500
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 3.3371950491903637,
+      "learning_rate": 8.859043152872892e-09,
+      "loss": 0.0194,
+      "step": 5501
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 1.9530134838227862,
+      "learning_rate": 8.687907539375318e-09,
+      "loss": 0.0204,
+      "step": 5502
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 1.9341448030769988,
+      "learning_rate": 8.518439616022057e-09,
+      "loss": 0.0219,
+      "step": 5503
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 2.087075389938105,
+      "learning_rate": 8.350639439436703e-09,
+      "loss": 0.0242,
+      "step": 5504
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 3.430222763743477,
+      "learning_rate": 8.184507065683855e-09,
+      "loss": 0.0174,
+      "step": 5505
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 5.6908272831133395,
+      "learning_rate": 8.020042550271889e-09,
+      "loss": 0.02,
+      "step": 5506
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 2.351915576486446,
+      "learning_rate": 7.857245948150183e-09,
+      "loss": 0.0186,
+      "step": 5507
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 4.173408896062741,
+      "learning_rate": 7.696117313713559e-09,
+      "loss": 0.0271,
+      "step": 5508
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 1.7164931573536757,
+      "learning_rate": 7.536656700797284e-09,
+      "loss": 0.0169,
+      "step": 5509
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 2.99059556731782,
+      "learning_rate": 7.37886416268041e-09,
+      "loss": 0.0201,
+      "step": 5510
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 3.3403066040208462,
+      "learning_rate": 7.222739752084096e-09,
+      "loss": 0.0251,
+      "step": 5511
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 8.053052196324328,
+      "learning_rate": 7.068283521172725e-09,
+      "loss": 0.0359,
+      "step": 5512
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 4.416499688532894,
+      "learning_rate": 6.915495521552795e-09,
+      "loss": 0.018,
+      "step": 5513
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 4.387845883879492,
+      "learning_rate": 6.764375804274026e-09,
+      "loss": 0.0239,
+      "step": 5514
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 5.03141411349809,
+      "learning_rate": 6.61492441982714e-09,
+      "loss": 0.0377,
+      "step": 5515
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 3.757977750589518,
+      "learning_rate": 6.467141418147748e-09,
+      "loss": 0.0122,
+      "step": 5516
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 2.001295426759492,
+      "learning_rate": 6.321026848613021e-09,
+      "loss": 0.0153,
+      "step": 5517
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 3.9821414447648973,
+      "learning_rate": 6.176580760041684e-09,
+      "loss": 0.0194,
+      "step": 5518
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 1.7433347366769623,
+      "learning_rate": 6.033803200696242e-09,
+      "loss": 0.018,
+      "step": 5519
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 4.301461630452348,
+      "learning_rate": 5.892694218281869e-09,
+      "loss": 0.0371,
+      "step": 5520
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 4.386923072523771,
+      "learning_rate": 5.753253859944741e-09,
+      "loss": 0.0165,
+      "step": 5521
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 3.0793034972539677,
+      "learning_rate": 5.615482172275366e-09,
+      "loss": 0.0387,
+      "step": 5522
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 6.778878857726523,
+      "learning_rate": 5.479379201305257e-09,
+      "loss": 0.0304,
+      "step": 5523
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 4.863627634913964,
+      "learning_rate": 5.344944992509149e-09,
+      "loss": 0.0256,
+      "step": 5524
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 2.509596156570167,
+      "learning_rate": 5.212179590803335e-09,
+      "loss": 0.0306,
+      "step": 5525
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 2.796887352552083,
+      "learning_rate": 5.08108304054844e-09,
+      "loss": 0.0224,
+      "step": 5526
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 4.106611953848664,
+      "learning_rate": 4.9516553855455395e-09,
+      "loss": 0.0308,
+      "step": 5527
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 2.2897445226023216,
+      "learning_rate": 4.82389666903893e-09,
+      "loss": 0.0221,
+      "step": 5528
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 2.7080187516169496,
+      "learning_rate": 4.697806933715021e-09,
+      "loss": 0.0269,
+      "step": 5529
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 4.343144051787935,
+      "learning_rate": 4.573386221703446e-09,
+      "loss": 0.0267,
+      "step": 5530
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 2.7596315781432397,
+      "learning_rate": 4.450634574574286e-09,
+      "loss": 0.0331,
+      "step": 5531
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 4.324439345233048,
+      "learning_rate": 4.329552033341955e-09,
+      "loss": 0.0182,
+      "step": 5532
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 2.8977509062535747,
+      "learning_rate": 4.210138638462424e-09,
+      "loss": 0.0135,
+      "step": 5533
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 4.537391924016334,
+      "learning_rate": 4.0923944298337796e-09,
+      "loss": 0.0325,
+      "step": 5534
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 5.661417948871261,
+      "learning_rate": 3.976319446795662e-09,
+      "loss": 0.0379,
+      "step": 5535
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 2.8349607349929005,
+      "learning_rate": 3.8619137281326044e-09,
+      "loss": 0.0227,
+      "step": 5536
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 5.200658373149944,
+      "learning_rate": 3.749177312068475e-09,
+      "loss": 0.0279,
+      "step": 5537
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 1.8628376484778557,
+      "learning_rate": 3.63811023627092e-09,
+      "loss": 0.0193,
+      "step": 5538
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 4.724184543652983,
+      "learning_rate": 3.528712537849144e-09,
+      "loss": 0.0267,
+      "step": 5539
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 3.6749826001937764,
+      "learning_rate": 3.42098425335613e-09,
+      "loss": 0.0311,
+      "step": 5540
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 4.909671343908744,
+      "learning_rate": 3.3149254187841985e-09,
+      "loss": 0.0288,
+      "step": 5541
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 3.8677490201938785,
+      "learning_rate": 3.210536069571113e-09,
+      "loss": 0.024,
+      "step": 5542
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 3.608231101955727,
+      "learning_rate": 3.1078162405939747e-09,
+      "loss": 0.0213,
+      "step": 5543
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 3.272735238444045,
+      "learning_rate": 3.006765966174774e-09,
+      "loss": 0.0318,
+      "step": 5544
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 3.1790589293641482,
+      "learning_rate": 2.907385280075392e-09,
+      "loss": 0.0182,
+      "step": 5545
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 3.649186670073757,
+      "learning_rate": 2.80967421550038e-09,
+      "loss": 0.0286,
+      "step": 5546
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 3.138795296256432,
+      "learning_rate": 2.7136328050980654e-09,
+      "loss": 0.0284,
+      "step": 5547
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 3.9420052048363736,
+      "learning_rate": 2.6192610809566697e-09,
+      "loss": 0.0237,
+      "step": 5548
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 2.182296768945293,
+      "learning_rate": 2.5265590746076373e-09,
+      "loss": 0.0113,
+      "step": 5549
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 2.6773104562420236,
+      "learning_rate": 2.43552681702508e-09,
+      "loss": 0.0203,
+      "step": 5550
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 3.367481619945359,
+      "learning_rate": 2.346164338624113e-09,
+      "loss": 0.0228,
+      "step": 5551
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 5.067110242743501,
+      "learning_rate": 2.2584716692619636e-09,
+      "loss": 0.0237,
+      "step": 5552
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 2.0718286609725105,
+      "learning_rate": 2.172448838239083e-09,
+      "loss": 0.0221,
+      "step": 5553
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 2.662417149164622,
+      "learning_rate": 2.08809587429748e-09,
+      "loss": 0.0202,
+      "step": 5554
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 3.989820634216304,
+      "learning_rate": 2.0054128056201662e-09,
+      "loss": 0.0137,
+      "step": 5555
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 1.8775842103565406,
+      "learning_rate": 1.924399659833376e-09,
+      "loss": 0.014,
+      "step": 5556
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 7.413735859817457,
+      "learning_rate": 1.8450564640054569e-09,
+      "loss": 0.046,
+      "step": 5557
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 3.243264854700322,
+      "learning_rate": 1.7673832446463146e-09,
+      "loss": 0.0276,
+      "step": 5558
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 7.71185976345551,
+      "learning_rate": 1.6913800277085225e-09,
+      "loss": 0.031,
+      "step": 5559
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 6.885871743475731,
+      "learning_rate": 1.6170468385845462e-09,
+      "loss": 0.0285,
+      "step": 5560
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 3.6615117870884992,
+      "learning_rate": 1.5443837021122954e-09,
+      "loss": 0.0207,
+      "step": 5561
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 4.28734845963075,
+      "learning_rate": 1.473390642569017e-09,
+      "loss": 0.0173,
+      "step": 5562
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 2.702589837418001,
+      "learning_rate": 1.4040676836746259e-09,
+      "loss": 0.0251,
+      "step": 5563
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 5.090936212710113,
+      "learning_rate": 1.336414848591705e-09,
+      "loss": 0.0137,
+      "step": 5564
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 5.178772005848657,
+      "learning_rate": 1.2704321599243951e-09,
+      "loss": 0.0318,
+      "step": 5565
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 3.0897135807835494,
+      "learning_rate": 1.206119639718395e-09,
+      "loss": 0.0284,
+      "step": 5566
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 6.956302961997545,
+      "learning_rate": 1.1434773094615158e-09,
+      "loss": 0.0202,
+      "step": 5567
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 5.512949716160351,
+      "learning_rate": 1.0825051900842377e-09,
+      "loss": 0.0299,
+      "step": 5568
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 4.34290757998301,
+      "learning_rate": 1.0232033019580423e-09,
+      "loss": 0.0247,
+      "step": 5569
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 1.9398553518199555,
+      "learning_rate": 9.655716648970804e-10,
+      "loss": 0.0212,
+      "step": 5570
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 3.452078753978198,
+      "learning_rate": 9.096102981570598e-10,
+      "loss": 0.0181,
+      "step": 5571
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 1.1011444564361026,
+      "learning_rate": 8.553192204358018e-10,
+      "loss": 0.0116,
+      "step": 5572
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 2.2777409632755012,
+      "learning_rate": 8.026984498726853e-10,
+      "loss": 0.0277,
+      "step": 5573
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 4.323892977235624,
+      "learning_rate": 7.517480040497572e-10,
+      "loss": 0.0181,
+      "step": 5574
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 1.8751831593629684,
+      "learning_rate": 7.024678999900669e-10,
+      "loss": 0.0309,
+      "step": 5575
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 2.7347866857145986,
+      "learning_rate": 6.548581541593324e-10,
+      "loss": 0.0186,
+      "step": 5576
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 2.8926365897897406,
+      "learning_rate": 6.08918782464829e-10,
+      "loss": 0.0178,
+      "step": 5577
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 4.6764673593214505,
+      "learning_rate": 5.646498002553902e-10,
+      "loss": 0.0231,
+      "step": 5578
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 5.802111064150024,
+      "learning_rate": 5.220512223219621e-10,
+      "loss": 0.0308,
+      "step": 5579
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 2.0047609631552525,
+      "learning_rate": 4.81123062898714e-10,
+      "loss": 0.0163,
+      "step": 5580
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 1.0730242231401868,
+      "learning_rate": 4.4186533565915293e-10,
+      "loss": 0.0146,
+      "step": 5581
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 3.672437423453078,
+      "learning_rate": 4.042780537205637e-10,
+      "loss": 0.0192,
+      "step": 5582
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 4.753438723651524,
+      "learning_rate": 3.6836122964178934e-10,
+      "loss": 0.0247,
+      "step": 5583
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 2.127300877016818,
+      "learning_rate": 3.341148754232304e-10,
+      "loss": 0.0155,
+      "step": 5584
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 7.34220557033025,
+      "learning_rate": 3.015390025068454e-10,
+      "loss": 0.039,
+      "step": 5585
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 6.750258237012312,
+      "learning_rate": 2.706336217767058e-10,
+      "loss": 0.0253,
+      "step": 5586
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 2.231439477264196,
+      "learning_rate": 2.4139874355955105e-10,
+      "loss": 0.0146,
+      "step": 5587
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 3.31797560363478,
+      "learning_rate": 2.138343776231233e-10,
+      "loss": 0.0244,
+      "step": 5588
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 4.782531101487246,
+      "learning_rate": 1.8794053317672255e-10,
+      "loss": 0.0175,
+      "step": 5589
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 1.6196649621284065,
+      "learning_rate": 1.6371721887287196e-10,
+      "loss": 0.0192,
+      "step": 5590
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 1.720724803964426,
+      "learning_rate": 1.4116444280398711e-10,
+      "loss": 0.0169,
+      "step": 5591
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 3.9654605728386154,
+      "learning_rate": 1.2028221250570683e-10,
+      "loss": 0.015,
+      "step": 5592
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 3.692236086799763,
+      "learning_rate": 1.0107053495522767e-10,
+      "loss": 0.0205,
+      "step": 5593
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 2.5848778676255066,
+      "learning_rate": 8.35294165718592e-11,
+      "loss": 0.027,
+      "step": 5594
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 5.861472760760072,
+      "learning_rate": 6.765886321646874e-11,
+      "loss": 0.021,
+      "step": 5595
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 6.288974718765686,
+      "learning_rate": 5.345888019092638e-11,
+      "loss": 0.0381,
+      "step": 5596
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 3.6870045895899226,
+      "learning_rate": 4.092947224032529e-11,
+      "loss": 0.0194,
+      "step": 5597
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 7.961768982598845,
+      "learning_rate": 3.007064355076139e-11,
+      "loss": 0.0204,
+      "step": 5598
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 2.888016420410043,
+      "learning_rate": 2.088239775044354e-11,
+      "loss": 0.0248,
+      "step": 5599
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 6.1509337395528,
+      "learning_rate": 1.3364737909138392e-11,
+      "loss": 0.025,
+      "step": 5600
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 3.4536548965699705,
+      "learning_rate": 7.517666539280654e-12,
+      "loss": 0.039,
+      "step": 5601
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.643426850528451,
+      "learning_rate": 3.3411855937526273e-12,
+      "loss": 0.0276,
+      "step": 5602
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 3.699342004217049,
+      "learning_rate": 8.352964681046516e-13,
+      "loss": 0.0218,
+      "step": 5603
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.6873280750994422,
+      "learning_rate": 0.0,
+      "loss": 0.0218,
+      "step": 5604
+    },
+    {
+      "epoch": 4.0,
+      "step": 5604,
+      "total_flos": 750239094767616.0,
+      "train_loss": 0.1176493737971587,
+      "train_runtime": 15635.0784,
+      "train_samples_per_second": 22.932,
+      "train_steps_per_second": 0.358
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 5604,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 24000,
+  "total_flos": 750239094767616.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}